Bronica Medium Format Lenses - Depth of Field Tables
Related Sites:
DOF Calculator [5/2002]
F/Calc DOF and Photo Math Calculator [02/00]
Hyperfocal Distance Formula
Links Page (archive Excel DOF calculator..)
|
Subject: Response to Depth of Field with MF |
|
You've discovered what, to me, is the most serious disadvantage of moving
from
35mm to MF. Lenses of a given focal length have the same depth of field no
matter what size film is put behind them. In 35mm, a 20mm lens at f/22 can
keep everything in focus from a foot or foot and a half to infinity. The
approximately equivalent lens on the Pentax 6x7 (45mm) shows depth of
field at
f/22 as about 3.5 feet to infinity (and this is calculated using a larger
"circle of confusion" than with the 35mm lens). |
| Chris Patti (Medium Format Digest Date
1997-09-16) |
Use the table below to see what Chris Patti
means..
List of Selected Bronica Lenses:
| Bronica Lens | 35mm equiv.
largest
square |
|---|
| Small Bayonet Mount |
|---|
| 40 mm F4 Auto-Nikkor | =17 mm |
| 50 mm F3.5 Auto-Nikkor | =22 mm |
| 75 mm F2.8 Auto-Nikkor | =32 mm | |
| 135 mm F3.5 Auto-Nikkor | =58 mm |
| 200 mm F4 Auto-Nikkor | =84 mm |
| 100 mm F2.8 Auto-Zenzanon | =43 mm |
| 150 mm F3.5 Auto-Zenzanon | =65 mm |
| Large Bayonet Mount |
|---|
| 105 mm F3.5 Auto-Nikkor | =46 mm |
| 300 mm F4.5 Auto-Zenzanon | =130 mm |
400 mm Tele-Nikkor
with focusing mount | =173
mm |
600 mm Tele-Nikkor
with focusing mount | =260
mm |
800 mm Tele-Nikkor
with focusing mount | =346
mm |
1200 mm Tele-Nikkor
with focusing mount | =520
mm |
More Format
Comparison Information
| Circles of Confusion Diameter |
|---|
| 35mm | .03mm |
| 645 | .05mm |
| 6x6 | .06mm |
| 6x7 | .065mm |
Compare 35mm with 6x6!
Depth-of-Field Calculator - NS 2.0 MSIE 3.0 or later
Required
To find the near, far and hyperfocal distance for a specific lens, enter
the focal length (mm), aperture (f/number) and
the object distance (whole or decimal number).
The JavaScript program will calculate the near, far, depth and hyperfocal
distances.
The unit of measure for distances is feet. This
can be overridden by selecting meters. Your MSIE and later Netscape
browser will print the tables, others must print screen to the clipboard
and print from some other application.
The formula
used for this table.
You may use this code freely for any non-commercial purpose, written
authorization for all other uses.
Please direct any questions, suggestions or comments via Email to: [email protected]
[Editor's Note: Be sure to check out Mr. Gillett's 35mm Photography
pages!]
Related Postings
From: Bob Buckles [email protected]
Newsgroups: rec.photo.misc
Subject: Determining the Depth of Field
Date: Fri, 27 Mar 1998
The subject of this column is based partly on an article that appeared
last Summer in one of the popular camera magazines and depth of focus lens
charts that appear on the web. Many times we find ourselves in the field
and don't have the information at our finger tips. Here's an easy way to
determine the depth of focus in the field.
When photographing landscapes, we are often faced with a decision as
to what will be "in-focus" and what will have to be sacrificed to
"out-of -focus." There's a simple formula that can be used in the field
to ensure a distant object (infinity) will be sharp and at the same
time, let you know how much in front of infinity will also be in sharp
focus. After you select the focal length of the lens you will be using,
you can determine your point of focus (at F/11) with the following
formula: FD = FL divided by 10 squared, where FD is your focus distance
and FL is the focal length of the lens you are using to photograph the
scene. Example: Using a 50mm lens, 50/10 X 50/10 or 5 X 5 = 25 feet. Set
the distance scale on the lens at 25 feet. With the lens aperture at
F/11, we can be assured infinity will be in-focus. To determine how much
foreground will be in-focus simply divide 25 by 2 = 12.5 feet. In other
words, with a 50mm lens at F/11, focused at 25 feet, everything from
12.5 feet to infinity will be in focus. Most auto-focus lens today
have very few markings to help you set the distances manually. I
suggest you view the scene, select a spot about 25 feet away and
auto-focus on this spot. Then switch off the camera's auto focus
feature and take the picture.
Should the scene require more foreground in focus, increase the depth
of field by stopping the lens down to a smaller aperture. For F/16,
multiply 25 (the distance we determined in the above example) by 2/3.
The new focusing distance is 17 feet. With a 50mm lens focused at 17
feet and F/16, now everything from 8.5 feet to infinity is in focus.
More depth of field is required? Stop down another stop to F/22 and
multiply 17 again by 2/3. The new distance is now 11 feet. Set the
50mm lens at 11 feet and everything from 5.5 feet to infinity will be in
focus.
If all of this seems a little too complicated, the table below should
help. I suggest you cut out this table, laminate it, and carry it in
your camera bag. An inexpensive way to laminate is to buy clear, self
sticking shelf paper and put it on both sides of the object to be
laminated, leaving a 1/8 " border. This works great for protecting
important papers and booklets such as your camera manual.
F/11 F/16 F/22 F/32
28mm 8' 6' 4' 3'
35mm 11' 8' 5' 4'
50mm 25' 17' 11' 8'
70mm 49' 33' 22' 15'
100mm 100' 67' 45' 30'
135mm 182' 121' 81' 54'
200mm 400' 267' 178' 119'
500mm 2500' 1667' 1111' 741'
From: [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: 645 vs. 35mm depth of field question
Date: Mon, 02 Mar 1998
The same conclusion was reached in an earlier discussion with a more
technical and rigorous explanation by Ben Weiner:
From: [email protected] (Ben Weiner)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: MF depth of field
Date: 18 Dec 1996
Garry Lee [email protected] writes:
>The point remains that the DOF is the same, view for view in any format.
>For the same sharpness, the Circle of Confusion is bigger in a bigger
>format BUT that is because the negative is bigger and needs less
>enlargement.
>It is a MYTH that larger formats have less DOF.
>They don't.
With all due respect to the participants in this discussion, it is
a confusing subject, but the above is simply not true. If you have
ever struggled to get enough DOF with a reasonable f-stop while
using a "normal" length lens on 4x5 or 8x10 you will know that
larger formats DO have less DOF.
Let's look at hyperfocal distance as an indicator of DOF.
There is a simple formula for hyperfocal distance:
h = F^2 / (N*c)
h = hyperfocal distance
F = focal length
N = f-number
c = diameter of circle of confusion
This formula comes from the Lens FAQ which David Jacobson posts
to rec.photo.moderated and can also be found in any number of books.
It is really simple to derive from similar triangles, but it
is difficult to show this in a text medium, without pictures.
OK. Suppose we compare 35mm, and 6x7, which is about twice as big
in linear size. (No arguments over "4x as big in area", please!)
Take a 50mm lens on 35mm, and a 100mm lens on 6x7. If you stand in
the same place with the two cameras, these give roughly the same
"normal" perspective and angle of view. I want to compare how
the two formats do when taking the SAME picture.
Let's suppose my subject matter extends from 5 meters to infinity.
To get it all in sharp focus, I need to set the lens so it
has a hyperfocal distance of 10 meters. I want to know what
f-stop I need. I can solve the above formula for N, the f-number:
N = F^2 / (h*c)
BUT I need to know the circle of confusion. Let's take c = 0.025 mm
(1/100 inch) for 35mm. Since I only need to enlarge the 6x7 neg
half as much, I can use a twice-as-big circle-of-confusion for
6x7; c = 0.05 mm.
OK, plug in the numbers (don't forget to convert h from meters to mm):
For 35mm, N = (50*50) / (10,000 * 0.025). N = 10, i.e. use f/10 (or f/11).
For 6x7, N = (100*100) / (10,000 * 0.05). N = 20, i.e. use f/20 (or f/22).
See? You need to stop down two extra stops, from f/11 to f/22, to get
the same depth of field.
Why is this? Basically, it's because of the F^2 in the formula.
The hyperfocal distance gets larger - and the depth of field gets
smaller - by a factor of 4 when you double the film size and
lens size. The circle of confusion also gets larger, but only
by a factor of 2. There's a factor of 2 left over. In order
to get back the depth of field you've lost, you have to make
the f-number go down by a factor of 2 (which is 2 f-stops).
Another way to look at it is this: out-of-focus areas are blurred
because of the physical diameter of the aperture. That's why
stopping down any lens (smaller diameter aperture) gives more depth
of field.
If you stand in the same place and photograph the same scene with
two lenses that give the same angle of view, you will get the same
DOF if the lenses have the same aperture diameter - physical
diameter in mm, (NOT the same f/number).
The 50mm lens at f/10 has an aperture of 5mm.
The 100mm lens at f/10 has an aperture of 10mm, hence less DOF.
To get the same DOF as the smaller lens, you need an aperture
of 5mm, hence 100m lens at f/20.
If you don't believe me, try it. If you try taking a picture with a
35mm camera and normal lens at f/8 or so, you will get respectable
DOF. With a 6x7, normal lens, f/8, you will get less DOF. With a
4x5 or 8x10 camera and its normal lens at f/8, you will get so
little DOF it will make your head spin. (Large format photographers
rarely use apertures as fast as f/8 for this reason, but sometimes
it can be put to creative use.)
I have spoken.
{end quote by Ben Weiner}
"Charles Petzold" wrote:
>
> Let's pretend that you own a lens that you can attach to either your 35mm
> camera or your MF camera. You place both cameras on tripods side by side
> pointed at the same scene. As you switch the lens between the two cameras,
> what happens to the depth of field?
>
> Answer: The DOF obviously remains the same.
>
> HOWEVER, the problem is that the particular scene you're shooting is
> rendered in the *same size* on the 35mm negative and the MF negative.
> You're wasting part of the larger MF negative. So what do you do? You have
> two choices:
>
> You could switch to a longer lens on the MF camera, and the depth of field
> decreases.
>
> Or, you could move closer to your subject with the MF camera, and the depth
> of field also decreases.
>
> So, what you're observing is normal and natural and an unavoidable
> consequence of using a larger film format.
>
> -- Charles Petzold
>
From: Tim Forcer [email protected]
Newsgroups: rec.photo.misc
Subject: Re: Mirror lenses
Date: Wed, 08 Apr 1998
Ron Higgs wrote:
>
>DOF depends on focal length and aperture, not whether it
>is a mirror or non-mirror lens.
This is true as written, but ignores the VERY significant point that a
cat DOESN'T have a circular aperture. Have a look at the front element,
and you'll see a black disc in the middle. That's the back of the
secondary mirror element, and means that the aperture is an annulus,
with an effective diameter greater than that of focal length divided by
numerical aperture (since numerical aperture relates light-gathering
area, not any linear dimension of the lens, to focal length). Thus, for
a given numerical aperture, with focal length and all other factors
identical, the diameter of the circle of confusion for a given
out-of-focus point source with a mirror lens is LARGER than for an
all-refracting lens (the mirror lens' "circle" of confusion will be an
annulus). In other words, the depth of field will be smaller.
So, in practical terms, DOF *DOES* depend on lens construction.
Doughnuts, anyone?
Tim Forcer [email protected]
The University of Southampton, UK
[Ed. Note: the Excel spreadsheet for DOF calculation cited below is also
archived and linked at Links Page]
From: "Michael K. Davis" [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: DOF Calculation
Date: 9 Jun 1998
Tyson Kopfer [email protected] wrote:
: Hey all,
: Schneider lists some greats charts on DOF. Doesn't print out nice and
: small like the aforementioned spreadsheet, but a great reference
: nonethless. They tables list DOF for 6x9, 4x5, 5x7, and 8x10 and a
: zillion lenses for each format. Check it out at:
: http://www.schneideroptics.com/large/depth/depthof.htm
I took the figures they give for Near and Far sharps for various
combinations of focal length, format and focus distance and did the math
backwards to determine what diameter circles of confusion Schneider has
chosen as permissible (when the image is enlarged to a 10 inch diagonal
print to be viewed at a distance of 10 inches -- the standard criteria.)
Every combination I picked fell between 1/150th and 1/155th of an inch.
Quoting page 131 of "Basic Photographic Materiels and Processes" by
Stroebel, Compton, Current, and Zakia (c1990 Focal Press): "Permissible
circles of confusion are generally specified for a viewing distance of 10
inches, and 1/100 inch is commonly sited as an appropriate value for the
diameter. A study involving a small sample of cameras designed for
advanced amateurs and professional photographers revealed that values
ranging from 1/70 to 1/200 inch were used -- approximately a 3:1 ratio."
It's somewhat subjective, but I like 1/175 inch -- toward the more
critical end of the range used by manufacturers. The rotating-disk Depth
of Field calculators published by Kodak in their Photoguides use a
generous value of 1/100 inch. Not nearly as conservative as Schneider.
Not only do most depth of field calculators deprive you of the ability to
select your own permissible diameter for circles of confusion, they also
make no effort to account for cropping to aspect ratios that are different
than that of the original format. In other words, when calculating depth
of field, the diagonal of that portion of the image that will be lifted to
produce the final print should be used, not the full image diagonal.
I have posted an Excel 5.0+ spreadsheet in rec.photo.misc under the
subject: Excel-based DOF calculator with a twist
It encourages you to measure the actual image area produced by the
camerase you own, enter those dimensions (or take the ones given) and then
it calculates diagaonals for maximum crops to get popular aspect ratios
like 1:1, 4:5, 5:7 and 11:14. You can even specify a unique aspect ratio.
Then it asks for focal length and your choice of permissible diameter
for circles of confusion and then any focus distance you want to specify.
So, the resulting DOF chart is specific to all the above variables!
Mike
--
/---------------------\
Michael K. Davis
[email protected]
From: "Michael K. Davis" [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: DOF Calculation
Date: 9 Jun 1998
David Payne [email protected] wrote:
: Art,
: There is a wonderful DOF calculator on a web page at the address listed below.
: It works with any format and focal length. The author also allows folks to use
: it for non-comercial purposes. I downloaded the page and now I can use it
: anytime for a quick calculation. I printed out a few key combinations for lenses
: I own, laminated them with plastic and threw them into my camera bag.
: http://www.smu.edu/~rmonagha/brondof.html
: Dave
This DOF calculator used the following fixed values for circle of
confusion diameter:
Quoting:
Circle of Confusion Diameter = .03mm for 35mm, .05mm for 645, .06mm for
6x6, .065mm for 6x7, .075mm for 6x9, .15mm for 4x5, .20mm for 5x7, .30mm
for 8x10
These circle of confusion diameters are at the film plane for each format.
Let's divide each of these into the full image diagonal for each format:
35mm 43.27mm / .03mm = 1442.33
645 69.70mm / .05mm = 1394
6x6 79.20mm / .06mm = 1320
6x7 89.25mm / .065mm = 1373.08
6x9 102.63mm / .075mm = 1368.4
4x5 158.18mm / .15mm = 1054.53
5x7 214.07mm / .20mm = 1070.35
8x10 320.82mm / .30mm = 1069.4
The calculated values are each the ratio of a given format's full image
diagonal to the selected permissible diameter for circles of confusion for
that format. By dividing each by 10, we can get the denominator of the
fraction 1/n that describes the diameters of circles of confusion in a
10-inch diagonal print viewed at a distance of 10 inches (the accepted
standard for specifying circles of confusion.) So, dividing each by ten,
and taking the reciprocal, we can get the diameters that will result in a
10-inch diagonal print:
35mm 1/144 inch
645 1/139 inch
6x6 1/132 inch
6x7 1/137 inch
6x9 1/137 inch
4x5 1/106 inch
5x7 1/107 inch
8x10 1/107 inch
OK, now that all the work is done, we can see that first, this depth of
field calculator intrinsicly exhibits a bias that permits larger circles
of confusion in the final print for large format cameras than medium
format, than small format. Personally, if I had to use a DOF calculator
that produces inconsistent diameters for circles of confusion in the
finaly print, from one format to the next, I would want the bias to be
reversed. The more critical choice of 1/144 inch (at the print) for 35mm
will drive the user to stop down more than would be necessary to limit
circles of confusion to 1/107. This in turn will increase the
probability of inducing visible diffraction effects. This bias might also
lull large format users into thinking that movements need not be employed
to increase depth of field when a generous depth of field scale says it
isn't necessary.
For a point of reference in choosing a diameter for circles of confusion,
a fixed value of 1/100 inch (in a 10-inch diagonal print) is used by Kodak
in their rotating-disk DOF calculators found in their Photoguides. This
is consistent for all formats. I prefer 1/175 inch for all formats. This
literally permits the print to be viewed with a 75% reduction in viewing
distance to get the same illusion of depth of field.
This becomes especially important when using wide-angle lenses. Why?
The best perspective is had when viewing a print at
Viewing Distance = Focal Length * Negative-to-Print Magnification
A 50mm lens on a 35mm camera, generating a 10 inch diagonal print with a
4:5 aspect ratio, having a negative-to-print magnification ratio of 6.61
(254 mm / 38.42mm), should be viewed at a distance of:
50mm * 6.61 = 330.56mm = 13.01 inches
For the 50mm focal length, a 13 inch viewing distance is greater than the
more critical 10 inch viewing distance for which I have subjectively
decided to use 1/175 inch circles of confusion. For any degree of
magnification from the 24x30mm useful area of the 35mm format (for prints
with a 4:5 aspect ratio) it is not until focal lengths under 1.5 inches
(38mm) are used that the so-called best viewing distance would bring the
viewer inside the 10-inch distance for a 10-inch diagonal print. this
would jeopardize the apparent sharpness of our 1/175 inch circles of
confusion in our 10-inch print and is why the final print size, degree of
magnification and viewing distances should be considered when selecting
a diameter for circles of confusion to be used with depth of field
calculations. Viewing the results is what it's all about after all. The
variables involved CAN be managed.
Of course, you can always erect a fence in front of your gallery walls to
keep everyone at great distances.
Mike
--
/---------------------\
Michael K. Davis
[email protected]
From: Bob Wheeler [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: DOF Calculation
Date: Wed, 10 Jun 1998
Like many other things in photography, one should test such things for
oneself.
Using the eye's resolution limit on a print at a given distance is
interesting as a hint to what one should require of the film resolution,
but it is the film resolution that is primary. If one picks a film
resolution, say 20 lines per millimeter, then one can ask questions about
the camera settings that produce it, for example, what is the effect of
diffraction?
If one tests, as I have done, the film resolution as a function of f-stop
and depth of focus (which maps into depth of field on the other side of
the lens), one will find that there is a competition between the two
parameters with respect to resolution: as depth of focus increases the
resolution decreases, and it is necessary to decrease the size of the lens
opening to compensate (larger f-stop). My tests indicate that 20 lines per
millimeter is well approximated by using a circle of confusion of 1/16 mm
= 0.0625 mm = 1/400 in. My testing indicates that 1/175 in produces about
13 lines per millimeter. (This constant works for most of the large format
lenses that I have tested, but I am sure there are lenses for which it
should be larger or smaller, and it is possible that for some lenses, a
more complicated correction will be needed.)
--
Bob Wheeler --- (Reply to: [email protected])
ECHIP, Inc.
From: [email protected] (Lee McFearin)
Newsgroups: rec.photo.equipment.large-format
Subject: Re: DOF Calculation
Date: 10 Jun 1998
Here is an attempt at a DOF Chart. I'd appreciate any feedback
as I work on a program to produce the Table for various ranges.
This table uses the distance the back (or front) standard must travel
to bring the image into focus. To use this table:
1. Focus on the farthest point. The distance between the lens and the
film plane is the Minimum Focus Distance. (The two standards
will be closest)
2. Focus on the nearest point. The difference between this and the #1
distance is the Distance between focus points.
3. Choose a Circle of Confusion and look it up in the COC column.
4. Scan Across the table to the appropriate Distance between focus
points. Remember this column!
5. Scan down the the Minimum Focus Distance column until you find
the value of the minimum focus distance.
6. The "Aparure Size" columns are aligned with the "Distance between
focus points" columns. Scan Across to the same column you chose in
number 4. This is the Apature size (in mm) you must use. For the
F number, use focal_len/Apature (or another table).
Note, Given any three of COC, Min Focus Distance, Distance Between Focus
Points, or Apature Size this table can find the fourth!
Note also, the optimal focus point can be found given a different table.
However, it was usually very close to the center of the Min and Max
focus points.
The equations I used were:
ds and dl are the distances from the lens to the film plane when the
camera is focused on the nearest and farthest points to be in the DOF.
c is the circle of confusion
a is the apature size (a = f/A : f= focal length, A = F number)
d is the optimal distance between the lens and film plane for both the
near and far points to be equally blurry.
a = c(ds+dl)/(dl-ds) and d= dl(1- (dl-ds)/ds+dl))
lee
COC(mm)| Distance between focus points (mm)
-------------------------------------------------------------
0.0250 |0.50 0.59 0.71 0.84 1.00 1.19 1.41
0.0354 |0.71 0.84 1.00 1.19 1.41 1.68 2.00
0.0500 |1.00 1.19 1.41 1.68 2.00 2.38 2.83
0.0707 |1.41 1.68 2.00 2.38 2.83 3.36 4.00
0.1000 |2.00 2.38 2.83 3.36 4.00 4.76 5.66
0.1414 |2.83 3.36 4.00 4.76 5.66 6.73 8.00
0.2000 |4.00 4.76 5.66 6.73 8.00 9.51 11.31
0.2828 |5.66 6.73 8.00 9.51 11.31 13.45 16.00
0.4000 |8.00 9.51 11.31 13.45 16.00 19.03 22.63
_______|______________________________________________________
|
Min Foc|
Dist | Apature Size (mm)
|
40mm |4.10 3.46 2.93 2.48 2.10 1.78 1.51
50mm |5.10 4.30 3.64 3.07 2.60 2.20 1.87
60mm |6.10 5.15 4.34 3.67 3.10 2.62 2.22
70mm |7.10 5.99 5.05 4.26 3.60 3.04 2.57
80mm |8.10 6.83 5.76 4.86 4.10 3.46 2.93
90mm |9.10 7.67 6.46 5.45 4.60 3.88 3.28
100mm |10.10 8.51 7.17 6.05 5.10 4.30 3.64
110mm |11.10 9.35 7.88 6.64 5.60 4.72 3.99
120mm |12.10 10.19 8.59 7.24 6.10 5.15 4.34
130mm |13.10 11.03 9.29 7.83 6.60 5.57 4.70
140mm |14.10 11.87 10.00 8.42 7.10 5.99 5.05
150mm |15.10 12.71 10.71 9.02 7.60 6.41 5.40
160mm |16.10 13.55 11.41 9.61 8.10 6.83 5.76
170mm |17.10 14.40 12.12 10.21 8.60 7.25 6.11
180mm |18.10 15.24 12.83 10.80 9.10 7.67 6.46
190mm |19.10 16.08 13.54 11.40 9.60 8.09 6.82
200mm |20.10 16.92 14.24 11.99 10.10 8.51 7.17
210mm |21.10 17.76 14.95 12.59 10.60 8.93 7.52
220mm |22.10 18.60 15.66 13.18 11.10 9.35 7.88
230mm |23.10 19.44 16.36 13.78 11.60 9.77 8.23
240mm |24.10 20.28 17.07 14.37 12.10 10.19 8.59
250mm |25.10 21.12 17.78 14.97 12.60 10.61 8.94
260mm |26.10 21.96 18.48 15.56 13.10 11.03 9.29
270mm |27.10 22.80 19.19 16.15 13.60 11.45 9.65
280mm |28.10 23.65 19.90 16.75 14.10 11.87 10.00
290mm |29.10 24.49 20.61 17.34 14.60 12.29 10.35
300mm |30.10 25.33 21.31 17.94 15.10 12.71 10.71
310mm |31.10 26.17 22.02 18.53 15.60 13.13 11.06
320mm |32.10 27.01 22.73 19.13 16.10 13.55 11.41
330mm |33.10 27.85 23.43 19.72 16.60 13.97 11.77
340mm |34.10 28.69 24.14 20.32 17.10 14.40 12.12
350mm |35.10 29.53 24.85 20.91 17.60 14.82 12.47
360mm |36.10 30.37 25.56 21.51 18.10 15.24 12.83
370mm |37.10 31.21 26.26 22.10 18.60 15.66 13.18
380mm |38.10 32.05 26.97 22.69 19.10 16.08 13.54
390mm |39.10 32.89 27.68 23.29 19.60 16.50 13.89
rec.photo.equipment.large-format
From: "Michael K. Davis" [email protected]
[1] Re: Image Circles/Different Formats ??????
Date: Fri Jun 19 20:42:01 CDT 1998
Hi!
Jon Wells [email protected] wrote:
[snip]
: 3. I have a 4x5 press camera and a separate 6x7 back for it. Assuming
: no change in lens, when the 6x7 back is used, the image size on the
: negative is the same, but the field of view is cropped. To obtain the
: same field of view, I need to step back twice as far, which results in
: an image size that is one-half. I could then enlarge negative 200% and
: thus have an image size and field of view that is the same as the 4x5.
: If I did this, however, wouldn't the photo taken with the 6x7
: necessarily be less detailed because, among other reasons, the
: object-to-lens distance was greater so less detail made it to the
: object-to-lens distance was greater so less detail made it to the
: negative?
:
: I appreciate your comments very much. Jon
Others have done a fine job answering your question, but I wanted to add
something relevant. When you double the distance between the camera and
the object on which you are focusing, you will quadruple the depth of
field. Backing up to increase depth of field is a great trade off. When
you crop the negative to achieve the perspective had before increasing the
object distance, the loss in sharpness due to cropping will be linear,
not exponential like the gain in depth of field. Of course, you may not
have needed that extra gain in depth of field and would have preferred to
avoid increased magnification of the grain structure.
Mike Davis
--
/---------------------\
Michael K. Davis
[email protected]
rec.photo.equipment.large-format
From: Bob Wheeler [email protected]
[1] Re: Small Format DoF Advantage is Squelched
Date: Thu Jul 09 12:01:20 CDT 1998
Chris Ellinger wrote:
> "Michael K. Davis" [email protected] wrote:
>
> > In the table below, the third column (Near Sharp at f/22) was calculated
> > using a maximum permissible diameter for Circles of Confusion of 1/175th
> > of an inch. The fourth column (Largest Aperture with Visible Diffraction)
> > was calculated with aerial disk diameters of 1/175th of an inch, also.
>
> Should the same diameter for Circle of Confusion be used for all formats? Some
> assume the diameter for CoC to be based on a standard print size, and the
> magnification required to produce that print from each format, with a larger
> diameter allowed for larger formats. Wouldn't this also apply to the aerial
> disk diameter?
>
> Chris Ellinger
> Ann Arbor, MI
Yes, that is correct. When you double the format, all parameters,
including c, m ust be doubled to hold DOF constant. Those who claim small
formats are better, usual ly err in making comparisons with the f-stop
constant for both formats. The resul t is that the larger format is
penalized with too large an aperture. The f-stop mu st be doubled if the
format is doubled -- this is why large format cameras range do wn to f64,
and 35mm only to f16 or f22. These f-stops produce approximately the sam e
minimum (physical) aperture size for the two formats, which means that the
diffraction is the same for both.
--
Bob Wheeler --- (Reply to: [email protected])
ECHIP, Inc.
From: [email protected] (Richard Knoppow)
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Small Format DoF Advantage is Squelched
Date: Fri, 10 Jul 1998
[email protected] (Chris Ellinger) wrote:
>"Michael K. Davis" [email protected] wrote:
>
>> In the table below, the third column (Near Sharp at f/22) was calculated
>> using a maximum permissible diameter for Circles of Confusion of 1/175th
>> of an inch. The fourth column (Largest Aperture with Visible Diffraction)
>> was calculated with aerial disk diameters of 1/175th of an inch, also.
>
>Should the same diameter for Circle of Confusion be used for all formats? Some
>assume the diameter for CoC to be based on a standard print size, and the
>magnification required to produce that print from each format, with a larger
>diameter allowed for larger formats. Wouldn't this also apply to the aerial
>disk diameter?
>
>Chris Ellinger
>Ann Arbor, MI
>
Depth of field is really an optical illusion. It works on the
premise that the eye has limited resolution itself. So, as long as
the images of objects at various distances remain within this limit of
percieved resolution they will appear sharp.
The circle of confusion is the size of the blur spot from a point
source object. Obviously, the size of the circle of confusion
acceptable on the negative must take into account the size of the
final image. In fact, to be correct, it must take into account the
size of the image as it appears on the retina of the eye. That is,
one must really include both the size of the print and the expected
viewing distance when calculating depth of field.
It should be obvious that for a given print size a smaller negative
must have a smaller circle of confusion. It must be smaller by
exactly the magnification that the negative requires to make the
print.
Depth of field charts (at least Kodak's) are calculated for a
certain fraction of the focal length. These charts assume a typical
magnification of the final image. This is perfectly fine for lenses
or "normal" focal length but may not work for wide-angle or long
lenses, since the magnification of the image isn't any longer what the
table was calculated for. e.g. the chart for a 50mm lens on a 35mm
camera will not be correct for a 50mm very wide angle lens used for a
4x5" camera.
Because the conditions for depth of field are often not clearly
stated DOF has become one of the most misunderstood and confusing of
all photographic subjects.
---
Richard Knoppow
Los Angeles, Ca.
[email protected]
From: [email protected] (Photoinfo)
Newsgroups: rec.photo.misc
Subject: Depth of Field Info
Date: 24 Jul 1998 02:20:11 GMT
Web page provides an overview of depth of field, discusses common
misperceptions, and describes how to use DOF scales, hyperfocal cards.
Visit http://members.aol.com/photoinfo/dof.html
Bob McCabe
From: "Michael K. Davis" [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Small Format DoF Advantage is Squelched
Date: 7 Jul 1998
I have been playing math games again and discovered something I have never
come across before. It is surely not an unprecedented finding, but it was
news to me.
Everybody laments that an 8x10 has less Depth of Field than a 4x5, than a
6x7, etc., assuming they are using equivalent focal lengths, but get this:
The achievable diffraction-limited Near Sharp distances are IDENTICAL for
all formats, given that the ratio of focal length to image diagonal is the
same from one format to the next. In other words, if several formats are
using focal lengths that are equivalent in their ratio to the format
diagonals, the effects of diffraction will limit each format to a unique
minimum aperture at which diffraction becomes visible (you knew that) AND
(here comes the news) it turns out that if you calculate the Depth of
Field for each focal length/image diagonal pair AT THOSE UNIQUE APERTURES,
you'll find that ALL the formats can achieve the SAME Near Sharp (without
movements and at different f/stops, course).
It is late and that reads like a big wad of cotton is in my mouth, but
here goes an example that should illuminate what I'm celebrating.
In the table below, the third column (Near Sharp at f/22) was calculated
using a maximum permissible diameter for Circles of Confusion of 1/175th
of an inch. The fourth column (Largest Aperture with Visible Diffraction)
was calculated with aerial disk diameters of 1/175th of an inch, also.
This reduces to the equation:
Format Diagonal in mm / 2.36920501777 = Aperture where diffraction becomes
visible. Focal length is not a variable for this calculation.
Format, Focal Near Sharp Largest Near Sharp at
Using 4:5 Length Distance Aperture Largest Aperture
Aspect Ratio (mm) at f/22 With Visible With Visible
Diagonal (feet) Diffraction Diffraction
(f/stop) (feet)
35mm 35.3 4.2 16.2 5.7
6x7cm 80.0 9.6 36.8 5.7
4x5in 141.2 16.9 64.9 5.7
8x10in 285.3 34.2 131.1 5.7
11x14in 406.2 48.7 186.6 5.7
First, notice that at f/22, the Near Sharps are much closer for the
smaller formats. (The Far Sharps are all nominally at infinity.) So you
can see that 8x10 has half the Depth of Field enjoyed by 4x5, with a
resulting Near Sharp that's twice as far from the camera.
But, also notice that the last column calculates the Near Sharp distances
for each format at each format's largest aperture with visible
diffraction. The Near Sharps at these apertures all work out to be
exactly the same! For these focal lengths, they are all 5.7 feet.
And notice that thanks to diminished diffraction, the larger formats can
bring their Near Sharps toward that had by 35mm, just by stopping down to
apertures where small formats should not follow.
The *finding* is that diffraction limits them all to the SAME Near Sharp.
Large Format can get just as much Depth of Field as small format because
diffraction is getting out of the way exactly in proportion to the loss of
Depth of Field.
I hope nobody is laughing. It illustrates the practical benefit of the
fact that the impact of diffraction is diminished linearly just as the
Depth of Field diminishes with increase in format diagonal.
So, to avoid diffraction, the astute 35mm photographer stops down no
further than f/16, and the depth of field promised at apertures smaller
than f/16 can never actually be enjoyed. With a focal length of 35.3mm,
the 35mm format can not enjoy a Near Sharp closer than 5.7 feet. BUT,
this figure holds true for EVERY format using equivalent focal lengths, at
the diffraction-limited minimum apertures for each format. They can all
achieve a Near Sharp of 5.7 feet when stopped down to their respective
diffraction-limited minimum apertures. Obviously, the larger formats will
need longer exposures to achive this Depth of Field (independent of
movements) and the 11x14 format would be hard pressed to find a lens with
f/180.
So, in summary, I contend that even without their movements, large-format
cameras CAN achieve the exact SAME Depth of Field had by the smaller
formats (with longer exposures.)
Mike
--
/---------------------\
Michael K. Davis
[email protected]
From: [email protected] (Richard Knoppow)
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Small Format DoF Advantage is Squelched
Date: Wed, 08 Jul 1998
Depth of field is confusing because it can be stated for several
conditions. What you have discovered is the principle that for equal
image magnification the depth of field depends on the physical size of
the stop, regardless of focal length. So, if the same subject is
photographed from the same distance with a 50mm lens at f/2 (25mm
aperture) and with a 300mm lens at f/12 (again 25mm) and the image
from the 50mm negative magnified to the same size as that from the
300mm negative, the two will have exactly the same DOF.
Since diffraction depends on both the size of the hole and pupil
magnification it becomes constant with f/stop regardless of FL. So,
when the image from the 50mm lens at f/2 is magnified to the size of
that from the 300mm lens at f/12 the resolution of both images will be
identical.
In other words, when image magnification is taken into account the
seeming advantage of the larger f/stop required by the shorter FL lens
disappears. However, it will require a lot less light for a given
depth of field with the 50mm lens.
Which is to say you are right.
---
Richard Knoppow
Los Angeles, Ca.
[email protected]
From: "Michael K. Davis" [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Small Format DoF Advantage is Squelched
Date: 10 Jul 1998
Hi!
Chris Ellinger [email protected] wrote:
: "Michael K. Davis" [email protected] wrote:
: > In the table below, the third column (Near Sharp at f/22) was calculated
: > using a maximum permissible diameter for Circles of Confusion of 1/175th
: > of an inch. The fourth column (Largest Aperture with Visible Diffraction)
: > was calculated with aerial disk diameters of 1/175th of an inch, also.
: Should the same diameter for Circle of Confusion be used for all formats? Some
: assume the diameter for CoC to be based on a standard print size, and the
: magnification required to produce that print from each format, with a larger
: diameter allowed for larger formats. Wouldn't this also apply to the aerial
: disk diameter?
: Chris Ellinger
: Ann Arbor, MI
You are absolutely justified in being concerned about the magnification of
image degrading critters when negatives are printed. The calculations
I've posted regarding both circles of confusion and diffraction effects
have not overlooked that important point. Let me answer the first issue,
magnification of circles of confusion, by just quoting a question I
received by e-mail two days ago and the reply I sent. I'm going to let
the author of the question be anonymous.
Anonymous wrote:
>my question relates to the use of the same circle of confusion (1/175th)
>for all formats - I think this is the source of your unity effect. In
>general, everybody insists on using a different circle of confusion for
>each format related to some perceived enlargement effect factor (as far
>as I can tell, anyway - seems arbitrary to me ;-) If you factor in
>different circles of confusion or enlargement factors, then you would see
>near/far differences between formats. the linkage between circle of
>confusion and in-focus acceptance criteria is direct if buried in the math?
>
Mike Davis wrote:
Actually, compensating for the degree of magnification that will occur is
built into the depth of field calculations by way of the variable that
reports the format diagonal. The selection of a maximum permissable
diameter for cirlces of confusion in my depth of field calculator for
example, allows the user to specify the diameter that will occur in the
print itself after enlargement or reduction to a 10-inch diagonal print to
be viewed at a distance of 10 inches. (To be specific, the 10-inch viewing
distance is measured from the eye to any equidistant corner of the print
being viewed, not along a line right angle to the plane of the print.)
Once a diameter is selected, it is appropriate for any format [because you
are selecting the diameter seen on the print, not at the negative] and it
doesn't just work for 10-inch prints. Having made a fuss about this
seemingly arbitrary 10-inch diagonal, let me add that if you make the
print larger or smaller, the circles of confusion will increase or
decrease proportionately BUT, as long as the viewing distances remain
equal to the print diagonals, the illusion of depth of field will remain
absolutely constant at all magnifications. As the print gets larger, so
do the circles of confusion and so does the viewing distance.
[The use of a 10-inch diagonal print to be viewed at a distance of 10
inches is truly a standard for which discussions of image degradation have
used for years. When Kodak selected 1/100th inch diameter circles of
confusion for their rotating disk depth of field calculators, they did so
with the idea that for any format, whether a reduction or an enlargement
is needed to get to a 10-inch diagonal print, the resulting circles of
confusion on THAT size print will not exceed 1/100th inch if the
photographer works within the confines of the near and far sharps given by
their rotating disk calculator. Some print size had to be selected for
folks like us to reference. I can say to you that I consider 1/175th-inch
circles of confusion to be the maximum permissible diameter on a 10-inch
diagonal print viewed at 10 inches and you can say that you are less
conservative and have chosen 1/125th-inch circles of confusion on a
10-inch diagonal print viewed at 10 inches. We both understand that we
mean this to be the case no matter what format we are using and the best
part of this seemingly arbitrary standard is that it is not at all
limiting. Follow this. If you decide to make a 40-inch diagonal print
from that same negative, you will indeed enlarge the circles of confusion
by a factor of 4, but if the viewing distance is ALSO increased, the
illusion of depth of field will NOT change one bit. So, as long as the
viewing distance equals the print diagonal, it doesn't matter WHAT
magnification we use, right? We talk about 10-inch diagonal prints just
so that when I say 1/175th and you say 1/125th we know we are on the same
curve!
Again, as the prints get larger, so do the circles of confusion, but so
does the viewing distance.]
That's why it's important to consider decreasing the diameter of circles
of confusion when you know in advance that viewing distances will be
closer than the print diagonal. Similarly, you can be less aggressive
when you are certain that viewing distances will be some multiple of the
print diagonal. It's a linear relationship, so if your favorite
permissable diameter is a very conservative 1/200th of an inch at the
print and you know the print will most likely be viewed at four times the
diagonal, by all means, enjoy the increase in shutter speed had by
entering a value of 1/50th inch in the depth of field calculator. [You'll
be telling the calculator that you know in advance you can tolerate the
larger circles of confusion at the print.]
Another point to consider is that to achieve the correct perspective for a
given focal length, the best viewing distance is equal to Focal Length *
Negative-to-Print Magnification. All you have to do is stick your nose to
a print that was shot with a really short focal length (relative to the
format diagonal) to know in your heart that this is true. Super wide
shots look great close up. Using this formula to determine viewing
distance, any focal length shorter than the image diagonal will result in
a viewing distance that is less than the print diagonal. Now you have got
to increase your depth of field to compensate, by specifying a smaller
diameter for the permissible circles of confusion. (The age old myth that
wide-angle lenses have more depth of field is excruciatingly false in
light of the need to force additional depth of field to compensate for
appropriate viewing distances that are less than the print diagonal!)
Unfortunately, I only know of one calculator that allows you to supply the
circle of confusion diameter as a variable. I'm patting myself on the
back, excuse me, but while I'm at it, I have to add that no other depth of
field calculator encourages you to specify the exact diagonal of only that
portion of the image that will be enlarged to make the final print, a
critical oversight in calculators that assume you will be using the entire
format diagonal. [The calculator is a MS-Excel 5.0+ spreadsheet,
MDOFcal2.XLS, that is not available on a server, but I did
post it once before as a uuencoded pkzip file.]
END OF QUOTE
Now to the subject of diffraction and how I've accounted for
negative-to-print magnification.
The diameter (mm) of diffraction disks is calculated as
0.00135383 * f-number
To set the size of the diffraction disks EQUAL to what I consider a
tolerable diameter (1/175th inch) for circles of confusion for any format,
once enlarged or reduced to a 10-inch diagonal print to be viewed at 10
inches, I just have equate to 1 the quotient had when circles of confusion
diameter is divided by diffraction disk diameter, then reduce.
1 = (Format Diagonal mm / 1750) / (0.00135383 * f-number)
or
f-number = Format Diagonal mm / 2.36920501777
The above formula does account for the different magnifications that will
result from using different formats.
To wrap up my answer to your concern about magnification of diffraction
effects, I'll quote another e-mail reply I gave in response to a different
person's queries regarding this thread:
Anonymous #2 wrote:
>Mike,
>One question, if f/16 is the minimum size for 35mm, what did you
>determine to be the minimum size for Medium format and Large format.
Mike Davis wrote:
To find the largest aperture at which the effects of diffraction are just
visible, use this formula [This yields airy disks no larger than 1/175th
inch on a 10-inch diagonal print made from any format.]
Format Diagonal in mm / 2.369 = f/stop where diffraction becomes
visible.
Remember that when measuring the format diagonal, you have to consider
whether you are going to enlarge the entire image. If you intend to crop,
then you should use the diagonal of the cropped area.
So, to answer your question, here are the f/stops at which diffraction
just starts to become visible for several formats, assuming that for each
format, a 4:5 aspect ratio print (8x10, 16x10, etc.) is going to be
cropped from the total image area. [So, in the table below, and in the
original post at the top of this thread, I have decided to calculate the
largest apertures where diffraction effects are still barely visible using
format diagonals that are NOT the full image diagonals, but rather, the
diagonals of that portion of the entire image area which can be cropped to
produce a 4:5 aspect ratio print.]
Format, Largest
Using 4:5 Aperture
Aspect Ratio With Visible
Diagonal Diffraction
35mm f/16.2 (f/16) -- f/11 would show no diffraction at all
6x7cm f/36.8 (f/32) -- f/22 would show no diffraction at all
4x5in f/64.9 (f/64) -- f/45 would show no diffraction at all
8x10in f/131.1 (f/128) -- f/90 would show no diffraction at all
11x14in f/186.6 (f/180) - f/128 would show no diffraction at all
The punchline of my article [top of this thread] was that because of
degradation due to diffraction, there is no point trying to get depth of
field by exceeding these f/stops (by exceeding f/16 for 35mm, for
example). In examining that truth, I discovered that even though large
formats have to use smaller apertures to get the same depth of field
enjoyed by smaller formats, when you run each format out to its smallest
diffraction-limited aperture, you get exactly the SAME depth of field
across all formats. So, diffraction is every bit as rough on small
formats as depth of field is on the large formats. The relationship is
perfectly linear. The net outcome is that if you use the
diffraction-limited apertures for each format, you'll get the same depth
of field for every format.
The large formats can achieve the same depth of field as the small formats
(without using camera movements to reposition the plane of focus) but
their exposures have to be longer to do so. When the large format
photographer then employs movements to enhance depth of field, one real
result is that he can increase his shutter speeds because he doesn't have
to use as small an aperture to get the same depth of field had when no
swings or tilts were used. Now he's ahead of the smaller formats in how
much depth of field he can muster.
END OF QUOTE
Whew! I've beaten it to death, huh? Die you dead horse, DIE!!!
I really want to be corrected if I've made any boo-boos in my thinking,
here. I've always felt the BEST way to learn is to try to explain your
understanding of something. BEST gets BETTER still if someone
in the audience can offer correction!
Hurt me! I want to know the TRUTH! ZERO MYSTERIES! :-)
Mike
--
/---------------------\
Michael K. Davis
[email protected]
From: "Michael K. Davis" [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: seek info on depth of field tables
Date: 9 Aug 1998 18:34:45 GMT
steven T koontz [email protected] wrote:
: mike maloney, with denver o'toole wrote:
: >
: > What is a good source and a good type of depth of field table to use
: > in the field? I have a Rodenstock 210 Apo-S lens.
: There is a great DOF chart at schneiders web site. Has all formats and
: lens lengths..
: --
: steve's photography & Z car stuff @ http://www.mindspring.com/~skoontz
: [email protected]
I took the figures they give for Near and Far sharps for various
combinations of focal length, format and focus distance and did the math
backwards to determine what diameter circles of confusion Schneider has
chosen as permissible (when the image is enlarged to a 10 inch diagonal
print to be viewed at a distance of 10 inches -- the standard criteria.)
Every combination I picked fell between 1/150th and 1/155th of an inch.
Quoting page 131 of "Basic Photographic Materiels and Processes" by
Stroebel, Compton, Current, and Zakia (c1990 Focal Press): "Permissible
circles of confusion are generally specified for a viewing distance of 10
inches, and 1/100 inch is commonly sited as an appropriate value for the
diameter. A study involving a small sample of cameras designed for
advanced amateurs and professional photographers revealed that values
ranging from 1/70 to 1/200 inch were used -- approximately a 3:1 ratio."
It's somewhat subjective, but I like 1/175 inch -- toward the more
critical end of the range used by manufacturers. The rotating-disk Depth
of Field calculators published by Kodak in their Photoguides use a
critical end of the range used by manufacturers. The rotating-disk Depth
of Field calculators published by Kodak in their Photoguides use a
generous value of 1/100 inch. Not nearly as conservative as Schneider.
To answer the question:
: > What is a good source and a good type of depth of field table to use
: > in the field? I have a Rodenstock 210 Apo-S lens.
I would say print your own and make sure it takes into account not only
the focal length and format diagonal, but also YOUR PERSONAL PREFERENCE
for a maximum permissible diameter for circles of confusion.
Kang Myong-jin once wrote in this newsgroup:
>Btw, I use the DOF chart only as a starting point. I've found that I need
>to stop down more than the DOF chart that I use, depending on the film.
Many people have abandonded the use of DoF calculators completely because
they believe their experience has proven the math can't be trusted. Well
it just isn't so.
There's a web page that I just read yesterday, but can't find at the
moment, where the author strongly advocates that for landscapes, you
should set your focus at infinity and then select an aperture that will
give you an acceptable near sharp. He has even posted an image of a
Newfoundland village taken from a hill where there's a cannon in the
foreground, a village near infinity and hills and ocean at infinity where
the near sharp is well beyond the foreground cannon, but he argues that
for his tastes, this is allowable (the cannon is at 30 feet from his 90mm
lens!) because everything else, out to infinity, is razor sharp and that
is much more important to him. He says he is tired of using DoF
calculations that encourage focusing at the hyperfocal distance because
this has so often resulted in less than crystal clear backgrounds.
I hear him loud and clear but I honestly believe he is throwing away
useful depth of field by focusing at infinity whenever infinity is part of
his subject. There is twice as much depth of field between the point of
best focus and the far sharp as there is between the near sharp and the
point of focus. So he's throwing away 2/3's of his depth of field because
he never again wants to see infinity out of focus.
I want to encourage him to come back to the math and trust it. All he
needs is control over the one variable that MOST depth of field
calculators deprive you of using -- the ability to specify the maximum
permissible diameter for circles of confusion.
Not only do most depth of field calculators deprive you of the ability to
select your own permissible diameter for circles of confusion, they also
make no effort to account for cropping to aspect ratios that are different
than that of the original format. In other words, when calculating depth
of field, the diagonal of that portion of the image that will be lifted to
produce the final print should be used, not the full image diagonal. If
you are routinely cropping your 35mm slides to make 8x10, 11x14 or 16x20
prints (4:5 aspect ratio), then your format diagonal is not that of the
full 24x36mm image area. The nominal 4:5 crop uses only a 24x30mm portion
of the full image and that means the depth of field calculation should use
a format diagonal of 38.42mm not 43.27mm. Here's the impact of using the
wrong diagonal for a 100 mm lens at f/8 seeking a maximum Circle of
Confusion of 1/175 inch after magnification to a 10-inch diagonal print:
Diagonal Near Sharp Hyperfocal Distance Far Sharp
43.27 mm 82.93 feet 165.86 feet Infinity
38.42 mm 93.40 feet 186.80 feet Infinity
That's a ten foot difference to the Near Sharp, just due to the difference
in magnification when cropping to a 4:5 aspect ratio. If we use Kodak's
spinning-disk depth of field calculators that have fixed the maximum
permissible diameter for Circles of Confusion at a sloppy 1/100th inch
(for magnification to a 10-inch diagonal print) and which assume no
cropping will be done, we would be told our Near Sharp, using same
example, is 47.39 !! That's nearly half the distance to the 82.93 foot
Near Sharp we calculated when accounting for a 4:5 crop and selected a CoC
of 1/175 inch! So is it any wonder that people are disgusted with DoF
calculators?
Here's one last source of disappointment -- viewing distances. I like to
use 1/175 inch as the maximum diameter for CoC on a 10-inch diagonal print
to be viewed at ten inches. This works out to be OK for any size print as
long as it is viewed at a distance equal to or greater than the print
diagonal. But, if I know in advance that the print will be viewed at a
distance less than the print diagonal, I have to DECREASE the CoC diameter
proportionately. I have a lot of prints hanging in a track-lighted
hallway in my home and the width of the hallway forces viewing distances
less than 36 inches and in practice, encourages viewing disances of 5 to
20 inches. This kind of scrutiny demands 8x10 or smaller prints even from
6x7 format, but I do have several 16x20's in that hallway that can take
the punishment.
I have, on occaision, specified CoC's as small as 1/350 inch (one half the
diameter of 1/175 inch) for 16x20 prints to be viewed at a distance equal
to half their 25-inch diagonal. Sticking with the above example (35mm
format, cropped to 24x30, using 100mm lens at f/8), this runs my near
sharp out to 186.80 feet with the Far Sharp at Infinity, but Kodak's DoF
calculator would still say 47.39 feet! The Kodak calculator would have me
set my focus index at a distance that is twice this near sharp, or 94.78
feet and under the circumstances I've just described (any size print
viewed at a distance equal to half its diagonal), this means the REAL Near
Sharp would be 75.60 feet and the REAL Far Sharp would be at 127.0 feet
NOT infinity! Again, no wonder people don't trust DoF calculators and
it's no wonder that fellow is tired of seeing out of focus backgrounds!
I have again posted an Excel 5.0+ spreadsheet in rec.photo.misc under the
subject: Excel-based DOF calculator with a twist
It encourages you to measure the actual image area produced by the
cameras you own, enter those dimensions (or take the ones given) and then
it calculates diagaonals for maximum crops to get popular aspect ratios
like 1:1, 4:5, 5:7 and 11:14. You can even specify a unique aspect ratio.
Then it asks for focal length and your choice of permissible diamter for
circles of confusion and then any focus distance you want to specify. The
resulting DOF chart is specific to all the above variables. It also
calculates the aperture at which Airy disks start becoming visible due to
diffraction, the aperture at which the quest for depth of field should be
achieved by other methods (increasing subject distance, using tilts, etc.)
Mike
From: "Michael K. Davis" [email protected]
Newsgroups: rec.photo.misc
Subject: Excel-based DOF calculator with a twist
Date: 9 Aug 1998
The following is a uuencoded zip file containing a MS-Excel Spreadsheet
called MDoFCalc.xls (zip format). This is a
Depth of
Field Calculator offering three unique advantages over others I've seen:
1) It allows the user to specify the REAL format diagonal produced by
his or her camera or even the diagonal of that portion of the full image
area that will be used after cropping, and it provides a sub-calculator
that allows this diagonal to be easily determined for any aspect ratio
print that might be produced by cropping from any format up to 11x14 inch.
2) It allows the user to specify the maximum diameter for circles of
confusion permissible after magnification to a 10-inch diagonal print (or
any other size print to be viewed at a distance equal to the print
diagonal.) This automatically accomodates the difference in magnification
had from one format to another, but more importantly, it lets the user
decide just how aggressive the depth of field calculations need to be so
that factors like print viewing distance can be compensated.
3) It calculates the aperture at which Airy disks start becoming visible
due to the effects of diffraction, the aperture at which the quest for
depth of field should be abandoned to other methods, like the use of
camera movements.
If you do not have the software to uudecode and/or unzip the following
uuencoded zip file, just request the Excel file by e-mail and I'll send it
to you as a MIME attachment.
Mike Davis
-----------
Errata:
The MDoFCalc.xls spreadsheet has a huge *typo* in Note 2.
On rows 139 and 141, the words
*image circles*
should be replaced with
*Circles of Confusion*.
Unforgiveable. Must get more sleep...
You can change it yourself by doing a Tools, Protection, Unprotect then
edit as directed and do a Tools, Protection, Protect and a File, Save.
Or, you can just ask me to e-mail you a corrected version.
Mike Dais
--
/---------------------\
Michael K. Davis
[email protected]
Date: Mon, 14 Sep 1998
From: Peter Klosky [email protected]
To: [email protected], [email protected]
Subject: Re: 120 Makro Planar
>> 120mm on a MF is equivalent to about 70mm on the 135 format.
>The 120 is within 2 degrees horizontally of an 85 on a 35 camera: 26
to 24: so make it equivalent to an 80. The 135 is about equivalent to
an 85: 23 to 24. The 150 is about equal to a 100 on 35mm: 21 to 20.
Here are some details from my view angle calculator, which was posted a
while back. Btw, the best way to rough out these calculations is to use
the rule of thumb that a 50mm on a 135 is very close to an 80mm on MF.
Applying the ratio of 50/80 or 80/50 will yield a result within 1%, as
shown by the table below, which is even more accurate.
The 120 MF is closer to 27 degrees than 26. A 75mm on a 135 is almost
exactly 27.
The 135 MF is, as you point out, very close to an 85mm on a 135.
The 150 MF is, like you say, about a 100mm on a 135, or about 95mm, to be a
little more precise.
frame, focal length in mm, diagonal, width and height in degrees.
57.00 by 57.00 frame. 80 53.48 39.22 39.22
57.00 by 57.00 frame. 100 43.90 31.82 31.82
57.00 by 57.00 frame. 120 37.13 26.72 26.72
57.00 by 57.00 frame. 135 33.25 23.84 23.84
57.00 by 57.00 frame. 150 30.08 21.52 21.52
24.00 by 36.00 frame. 50 46.79 39.60 26.99
24.00 by 36.00 frame. 70 34.35 28.84 19.46
24.00 by 36.00 frame. 75 32.18 26.99 18.18
24.00 by 36.00 frame. 80 30.26 25.36 17.06
24.00 by 36.00 frame. 85 28.56 23.91 16.07
24.00 by 36.00 frame. 95 25.66 21.46 14.40
24.00 by 36.00 frame. 100 24.41 20.41 13.69
24.00 by 36.00 frame. 105 23.28 19.46 13.04
24.00 by 36.00 frame. 135 18.21 15.19 10.16
Another effect is that the 135 shooter may be interested in making 8x10
prints, in which case he loses some of his view angle, adding to his
confusion in the field. In that case, he is best to mark his mirror or
screen with fine line tape and calculate his view angle on a 24 x 30
frame. I'll leave interpretation of this table as an exercise to the
reader.
24.00 by 30.00 frame. 50 42.03 33.40 26.99
24.00 by 30.00 frame. 70 30.69 24.19 19.46
24.00 by 30.00 frame. 75 28.73 22.62 18.18
24.00 by 30.00 frame. 80 27.00 21.24 17.06
24.00 by 30.00 frame. 85 25.47 20.02 16.07
24.00 by 30.00 frame. 95 22.86 17.95 14.40
24.00 by 30.00 frame. 100 21.75 17.06 13.69
24.00 by 30.00 frame. 105 20.73 16.26 13.04
24.00 by 30.00 frame. 135 16.20 12.68 10.16
----------------------------------------------------------------------
From: Bob Buckles [email protected]
Newsgroups: rec.photo.technique.misc
Subject: Re: DOF and bulb mode
Date: Tue, 20 Oct 1998
If you want to check the DOF and don't have a DOF preview on your camera
do this: Manually set the lens at the F/stop you want to use at, focus
your lens, and then, very carefully, release the lens release button and
rotate the lens (about 1/4 turn) as though you were removing it. The
lens will stop down to the F/stop you have set and you can view the
subject and see the DOF.
Good Luck,
Bob
Date: Sat, 21 Nov 1998
From: [email protected]
To: [email protected]
Subject: Re: [Rollei] Rolleinar DOF Table
In a message dated 98-11-20 08:28:34 EST, [email protected] (Ferdinand
Stutterheim) writes:
> I will try to put the table in a decent format this weekend.
> Today I have trouble with MSOE in my office. I will have to sort that
> out first. My mail is inaccessable at the moment.
===========================================
I uploaded a scan of Pearlman's DOF table for a 75mm lens with Rolleinars
to http://homepages.infoseek.com/~rbender/Dofnar.jpg
R. J. Bender ( A Nikon, Mamiya and Rollei user. ) [email protected] or
[email protected]
From Medium Format Digest:
From: Paul Roark [email protected]
Subject: Response to medium format DOF
Date: 1999-01-12
Contrary to popular wisdom, 35 mm has no advantage regarding DOF in
actual picture-taking situations. In fact, as a shooter of both in
the outdoors, Id give the advantage to MF.
It is true that a 35 mm has more optical DOF than the equivalent MF
lens at the same aperture. For example, the formula for the
hyperfocal distance is often stated as follows: H = (F x F)/(f x d),
where H is the hyperfocal distance, F is the lens focal length, f is
the aperture, and d is the circle of confusion. The circle of
confusion for a 35 mm is often assumed to be .03 mm. Since the MF
frame is 1.86 times larger, use .056 for its d value. If you work
through the numbers, youll find the H for f8 for a 50 mm (normal 35
mm) lens is 10.4 meters; for an 80 (for MF) it is 14.3 meters. Of
course, since the 80 is slightly wider angle, the advantage for 35 mm
is even more. For total DOF, figure that the closest point of
"sharp" focus is = the hyperfocal distance. (Of course, even at f64
there is only one true plane of focus; were just talking about how
much blur were willing to accept.)
However, the lens DOF is not the issue. The question is the final
sharpness of the image. Here, recall that the system resolution is
affected by each of its components. These include, most
importantly, diffraction limit of the lens, film sharpness and grain,
and, related to these, film size.
Because of lens diffraction and the smaller acceptable circle of
confusion of 35 mm, MF has at least a one f-stop advantage. On the
film front, youll find that 100 ASA film in an MF is much sharper
and less grainy than 25 ASA film in a 35 mm. So, there is a good 2
stop advantage for MF.
The bottom line Ive found is that, as a practical matter, MF has
more DOF than 35 mm if you are working to a specific final picture
sharpness standard. The net advantage of more film area seems to
outweigh the optical DOF advantage of 35 mm.
From Medium Format Digest:
From: Warren Kato [email protected]
Subject: Response to medium format DOF
Date: 1999-01-12
Numbers aside (and I am not disputing anyone's numbers here) as a
practical matter, I seem to be more disappointed with the extreme wide
angle shots of MF versus 35mm. As a point in comparison, my P645 35mm has
approximately the same angle of view as my Zuiko 21/3.5. The 21 at f 16
is in focus from about 1.5 foot to infinity according to the scale on the
camera although it poops out at the infinity end a bit. However the P645
35mm at f 22 per the DOF scale only focuses from 3 feet to infinity. In
practice the DOF is seems to be even shallower than this, and, of course,
I expect higher sharpness overall from MF. With one focusing at 3 feet and
the other at 1.5 feet, with the same angle of view, the
foreground/background comparison will be more accentuated with the 35mm
system and the foregound objects will be twice as large. I can get a
4X5/tilt & shift look with the 21mm that I can't get with MF. At 80mm
(MF) versus 50mm (35 mm), there isn't such a noticeable problem.
From: "Willis B. Boyce" [email protected]
Newsgroups: rec.photo.film+labs
Subject: Re: Delta 3200 in 120: Results!
Date: Thu, 17 Dec 1998
Joe,
I think that there are some subtle flaws in your reasoning.
Your assumption about the DOF is incorrect. Comparing a 50mm at f/1.4 lens
and an equivalent medium format lens, say 80mm, at f/4, the 80mm lens will
have almost exactly twice as much depth of field as the 50mm. Here are my
figures:
50mm f/1.4 @ 2m -> 0.11m DOF (assuming 35mm)
80mm f/2.8 @ 2m -> 0.15m DOF (assuming 6x6)
80mm f/4.0 @ 2m -> 0.22m DOF (assuming 6x6)
The format counts because larger formats tolerate larger circles of
confusion. You lose DOF when you go from 35mm to medium format because of
the increased magnification, but some of that loss is won back because the
larger format needs to be enlarged less, and so larger circles of confusion
on the negative will still be acceptably small on the print. I've used
the 1/1730 standard as per the Lens FAQ. If you tend to enlarge 35mm-sized
areas of your 6x6 negs and consider the extra film area to be compositional
slop space, then you should base your DOF calculations on the maximum
acceptable COC for the 35mm format, which in turns means that you can't
trust the DOF tables that came with your medium format lens; half the
numbers for a good estimate. Isn't this fun?
Anyway, DOF aside, there are other reasons why you might want to use Delta
3200 in 120. You might want a nice, grainy picture and also the convenience
of being able to shoot a Polaroid beforehand. You might want to switch in
midroll between Delta 3200 and a different type of film. You might want to
only carry one camera with you and still be able to shoot with ISO 3200
film. You might want to be able to use a waist level finder. You might
want to use a TLR because you're shooting through a really dark filter and
don't want to take it off and put it on all the time. You might like the
square format. You might like being able to work with big medium format
negatives rather than fiddling around with tiny 35mm negs. You might like
using old folders and Rolleis and don't like having to use a 35mm camera
when the light gets dim. :-)
I suppose that my point is that Delta 3200 in 120 format is just like any
film in that it is useful for certain things. It's up to you to decide
whether or not it's worth using.
Regards,
Willis
Joe Berenbaum wrote
>"Willis B. Boyce" [email protected] wrote:
>
>At 22:38 16/12/98 -0000, you wrote:
>>>After reading that, my question is- would you have any reason now to
>>>shoot Delta 3200 in medium format rather than Tri-X (or equivalent) in
>>>35mm?
>>Sure. Three stops. This means that I can shoot with available light,
>>without a tripod, which is exactly what I did.
>>Willis
>
>Hi- thanks for the response. Let me explain what is bothering me
>though- if I get out a light meter and take readings in a low light
>situation, I can use a 35mm camera with an f1.4 lens wide open and 400
>ISO film and get a shot at at handholdable shutter speed, say 1/125.
>If I pick up a medium format camera- I have several and for this kind
>of thing I would be using a folder or a TLR with a standard lens of
>around f3.5 which I would again use wide open, and I use a film rated
>at ISO 3200, the actual advantage is (quickly consults analog light
>meter) -nothing! Well to be precise, about 1/3 of a stop. If I point
>my LunaPro at a nearby wall now with the ISO set at 400 I get EV 8,
>f1/4 at 1/125. If I change the ISO to 3200, I get EV 11, f4 at 1/125.
>Thus with the equipment choices of Rolleiflex 3.5F with a Planar or
>Bessa II with f3.5 Heliar I don't get any real advantage using Delta
>3200 in 120 over using Tri-x in a 35mm camera, and the 35mm camera
>would probably have a more recently designed lens into the bargain
>with a more usable maxiumum aperture. This is using a Leica M6 to
>compare fairly with medium format cameras that don't have mirror
>shock. Now of course if I used Rolleiflex 2.8F I would get a one-stop
>advantage. But it still isn't such a major advantage over the 35mm
>equivalent.
>
>My reason for using medium format would be principally to get less
>grain that an equivalent 35mm negative. If it doesn't actually GIVE me
>less grain because I can use a 3-stop faster lens in 35mm on 3-stop
>slower film with the same shutter speed and get approximately
>equivalent depth of field as well, then what advantage is there in the
>end, to using ISO 3200 film in medium format? Do you see what I mean?
>If you get different figures I would appreciate hearing about it, but
>I think this all stands up.
>
>Don't get me wrong, I really WANT there to be an advantage. I love
>medium format folders and Rolleis. I want there to be a reason to use
>them with Delta 3200. I just ordered a couple of rolls and will try it
>out soon.
>
>Whaddya think?
>
>Joe B. (Please remove ".gov" for email.)
From: "Michael K. Davis" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Circle of Confusion
Date: 27 Dec 1998
Hi!
EDGY01 [email protected] wrote:
: In the formula for Depth of Field one of the variables is the "permissible
: diameter of the circle of confusion." In today's modern optics (with
: substantial correction), what is the diameter that a lens manufacturer uses
: when they lay down the depth of field marks on their lenses? In other words,
: what has become known as an "acceptable range of sharpness?" (Fifty years ago
: we were talking a circle of a diameter of 1/250th of an inch). Any definitive
: information out there today?
:
: [email protected]
Elsewhere in this thread, Franicis Miniter shares a lot of good
information including the often overlooked discussion of how viewing
distance influences depth of field. Let me answer your two questions
breifly, then I'll give you my long version.
The diameter that lens manufacturers use varies from one manufacturer to
the next, but most run between 1/100th inch (very generous -- likely to
dissappoint you even at viewing distances equal to the print diagonal) to
1/300th inch (pretty conservative -- definitely yielding an illusion
of sharpness in images viewed at distances equal to or greater than the
print diagonal.) The fact remains, however, that for a given focal length
and format diagonal (after cropping!) a print of given magnification,
viewed at a distance equal to its diagonal, will have exactly the same
depth of field for a lens made by one manufacturer as that of another.
The math rules -- depth of field can not be engineered into a lens.
Schneider makes large format lenses of great quality. Their published
depth of field tables use a maximum permissible diameter for circles of
confusion equal to 1/152 inch. Mamiya's Mamiya 7 rangefinder has perhaps
the most expensive lenses made for medium format and they are arguably the
sharpest. The depth of field scales engraved on the Mamiya 7 lenses are
based on a Circle of Confusion maximum diameter of 1/145 inch.
So there are two values for you -- one from perhaps the finest lens makers
in large and medium format respectively, if that means anything, but let
me explain why their choices aren't right for all situations. The long
version...
Quoting page 131 of "Basic Photographic Materiels and Processes" by
Stroebel, Compton, Current, and Zakia (c1990 Focal Press): "Permissible
circles of confusion are generally specified for a viewing distance of 10
inches, and 1/100 inch is commonly sited as an appropriate value for the
diameter. A study involving a small sample of cameras designed for
advanced amateurs and professional photographers revealed that values
ranging from 1/70 to 1/200 inch were used -- approximately a 3:1 ratio."
It's somewhat subjective, but I like 1/175 inch -- toward the more
critical end of the range used by manufacturers. The rotating-disk Depth
of Field calculators published by Kodak in their Photoguides use a
generous value of 1/100 inch. Not nearly as conservative as Schneider.
I would say print your own table and make sure it takes into account not
only the focal length and format diagonal, but also YOUR PERSONAL
PREFERENCE for a maximum permissible diameter for circles of confusion.
Many people have abandonded the use of DoF calculators completely because
they believe their experience has proven the math can't be trusted. Well
it just isn't so.
There's a web page that I read once, where the author strongly advocates
that for landscapes, you should set your focus at infinity and then select
an aperture that will give you an acceptable near sharp. He has even
posted an image of a Newfoundland village taken from a hill where there's
a cannon in the foreground, a village near infinity and hills and ocean at
infinity where the near sharp is well beyond the foreground cannon, but he
argues that for his tastes, this is allowable (the cannon is at 30 feet
from his 90mm lens!) because everything else, out to infinity, is razor
sharp and that is much more important to him. He says he is tired of
using DoF calculations that encourage focusing at the hyperfocal distance
because this has so often resulted in less than crystal clear backgrounds.
I hear him loud and clear but I honestly believe he is throwing away
useful depth of field by focusing at infinity whenever infinity is part of
his subject. There is twice as much depth of field between the point of
best focus and the far sharp as there is between the near sharp and the
point of focus. So he's throwing away 2/3's of his depth of field because
he never again wants to see infinity out of focus.
I want to encourage him to come back to the math and trust it. All he
needs is control over the one variable that MOST depth of field
calculators deprive you of using -- the ability to specify the maximum
permissible diameter for circles of confusion.
Not only do most depth of field calculators deprive you of the ability to
select your own permissible diameter for circles of confusion, they also
make no effort to account for cropping to aspect ratios that are different
than that of the original format. In other words, when calculating depth
of field, the diagonal of that portion of the image that will be lifted to
produce the final print should be used, not the full image diagonal. If
you are routinely cropping your 35mm slides to make 8x10, 11x14 or 16x20
prints (4:5 aspect ratio), then your format diagonal is not that of the
full 24x36mm image area. The nominal 4:5 crop uses only a 24x30mm portion
of the full image and that means the depth of field calculation should use
a format diagonal of 38.42mm not 43.27mm. Here's the impact of using the
wrong diagonal for a 100 mm lens at f/8 seeking a maximum Circle of
Confusion of 1/175 inch after magnification to a 10-inch diagonal print:
Diagonal Near Sharp Hyperfocal Distance Far Sharp
43.27 mm 82.93 feet 165.86 feet Infinity
38.42 mm 93.40 feet 186.80 feet Infinity
That's a ten foot difference to the Near Sharp, just due to the difference
in magnification when cropping to a 4:5 aspect ratio. If we use Kodak's
spinning-disk depth of field calculators that have fixed the maximum
permissible diameter for Circles of Confusion at a sloppy 1/100th inch
(for magnification to a 10-inch diagonal print) and which assume no
cropping will be done, we would be told our Near Sharp, using same
example, is 47.39 !! That's nearly half the distance to the 82.93 foot
Near Sharp we calculated when accounting for a 4:5 crop and selected a CoC
of 1/175 inch! So is it any wonder that people are disgusted with DoF
calculators?
Here's one last source of disappointment -- viewing distances. I like to
use 1/175 inch as the maximum diameter for CoC on a 10-inch diagonal print
to be viewed at ten inches. This works out to be OK for any size print as
long as it is viewed at a distance equal to or greater than the print
diagonal. But, if I know in advance that the print will be viewed at a
distance less than the print diagonal, I have to DECREASE the CoC diameter
proportionately. I have a lot of prints hanging in a track-lighted
hallway in my home and the width of the hallway forces viewing distances
less than 36 inches and in practice, encourages viewing disances of 5 to
20 inches. This kind of scrutiny demands 8x10 or smaller prints even from
6x7 format, but I do have several 16x20's in that hallway that can take
the punishment.
I have, on occaision, specified CoC's as small as 1/350 inch (one half the
diameter of 1/175 inch) for 16x20 prints to be viewed at a distance equal
to half their 25-inch diagonal. Sticking with the above example (35mm
format, cropped to 24x30, using 100mm lens at f/8), this runs my near
sharp out to 186.80 feet with the Far Sharp at Infinity, but Kodak's DoF
calculator would still say 47.39 feet! The Kodak calculator would have me
set my focus index at a distance that is twice this near sharp, or 94.78
feet and under the circumstances I've just described (any size print
viewed at a distance equal to half its diagonal), this means the REAL Near
Sharp would be 75.60 feet and the REAL Far Sharp would be at 127.0 feet
NOT infinity! Again, no wonder people don't trust DoF calculators and
it's no wonder that fellow is tired of seeing out of focus backgrounds!
An Excel-based DOF calculator with a twist can be downloaded from this
address:
http://home.sol.no/~gjon/mdofcalc.xls
It encourages you to measure the actual image area produced by the
cameras you own, enter those dimensions (or take the ones given) and then
it calculates diagaonals for maximum crops to get popular aspect ratios
like 1:1, 4:5, 5:7 and 11:14. You can even specify a unique aspect ratio.
Then it asks for focal length and your choice of permissible diamter for
circles of confusion and then any focus distance you want to specify. The
resulting DOF chart is specific to all the above variables. It also
calculates the aperture at which Airy disks start becoming visible due to
diffraction, the aperture at which the quest for depth of field should be
achieved by other methods (increasing subject distance, using tilts, etc.)
Mike Davis
--
/---------------------\
Michael K. Davis
[email protected]
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Circle of Confusion
Date: Sun, 27 Dec 1998
On 27 Dec 1998 02:37:30 GMT, "Michael K. Davis"
[email protected] wrote:
...
>The diameter that lens manufacturers use varies from one manufacturer to
>the next, but most run between 1/100th inch (very generous -- likely to
>dissappoint you even at viewing distances equal to the print diagonal) to
>1/300th inch (pretty conservative -- definitely yielding an illusion
>of sharpness in images viewed at distances equal to or greater than the
>print diagonal.) The fact remains, however, that for a given focal length
>and format diagonal (after cropping!) a print of given magnification,
>viewed at a distance equal to its diagonal, will have exactly the same
>depth of field for a lens made by one manufacturer as that of another.
>The math rules -- depth of field can not be engineered into a lens.
Ummm, what about the poor lens vs. the really superb...? ;-)
Upon close examination of the images, the sharper lens may
appear to have less DOF, since there could be a greater
visible difference in sharpness between the in-focus image,
and the limits of DOF - as determined by the selected disk
diameter, assuming that the disk diameter is not unusually
small... (I know you have specified the relative viewing
distance - but that is often violated...! ;-)
>There's a web page that I read once, where the author strongly advocates
>that for landscapes, you should set your focus at infinity and then select
>an aperture that will give you an acceptable near sharp. He has even
>posted an image of a Newfoundland village taken from a hill where there's
>a cannon in the foreground, a village near infinity and hills and ocean at
>infinity where the near sharp is well beyond the foreground cannon, but he
>argues that for his tastes, this is allowable (the cannon is at 30 feet
...
Yes - using most DOF guides, infinity images look
insufficiently sharp (I counter this by "fudging"
focus toward infinity some, and using a somewhat
smaller lens stop). The web page author does throw
away a lot of useful DOF by focusing at infinity,
but I do find that equal sharpness at the near and
far limits does generally look worse than unequal
sharpness favoring the distant focus somewhat
(even if the near focus sharpness is compromised).
This may be due to the differences in the scale
of the detail...
I prefer very distant subjects either to be very
sharply rendered, or to be very unsharply rendered
(they look very bad to me if they look "almost in
focus, but not quite"...;-).
Thanks for a nice post!
David Ruether
From: "Michael K. Davis" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Circle of Confusion
Date: 28 Dec 1998
Neuman - Ruether [email protected] wrote:
>On 27 Dec 1998 02:37:30 GMT, "Michael K. Davis"
>>The fact remains, however, that for a given focal length
>>and format diagonal (after cropping!) a print of given magnification,
>>viewed at a distance equal to its diagonal, will have exactly the same
>>depth of field for a lens made by one manufacturer as that of another.
>>The math rules -- depth of field can not be engineered into a lens.
>
>Ummm, what about the poor lens vs. the really superb...? ;-)
>Upon close examination of the images, the sharper lens may
>appear to have less DOF, since there could be a greater
>visible difference in sharpness between the in-focus image,
>and the limits of DOF - as determined by the selected disk
>diameter, assuming that the disk diameter is not unusually
>small... (I know you have specified the relative viewing
>distance - but that is often violated...! ;-)
Well said! This is especially true when the resolving power tapers
significantly from the center to the corners. There's a Mamiya 7 user
forum at www.mamiya.com where a few threads discuss exactly this. Those
lenses have such incredible resolving power and contrast that the plane of
sharpest focus makes everything else look bad. Several Mamiya 7 users
admit to focusing at infinity because they don't want ANY bokeh in the
background -- it hurts too much with those lenses.
>>I want to encourage him to come back to the math and trust it. All he
>>needs is control over the one variable that MOST depth of field
>>calculators deprive you of using -- the ability to specif
>
>Yes - using most DOF guides, infinity images look
>insufficiently sharp (I counter this by "fudging"
>focus toward infinity some, and using a somewhat
>smaller lens stop). The web page author does throw
>away a lot of useful DOF by focusing at infinity,
>but I do find that equal sharpness at the near and
>far limits does generally look worse than unequal
>sharpness favoring the distant focus somewhat
>(even if the near focus sharpness is compromised).
>This may be due to the differences in the scale
>of the detail...
>I prefer very distant subjects either to be very
>sharply rendered, or to be very unsharply rendered
>(they look very bad to me if they look "almost in
>focus, but not quite"...;-).
This is where I say it's better to decrease the diameter of your maximum
permissible circle of confusion until focusing at the hyperfocal distance
yields a sufficiently sharp infinity. If you use my Excel spreadsheet DoF
calculator, making the CoC diameter smaller will push both the hyperfocal
distance and the near sharp (1/2 hyperfocal) further away from the camera,
all the while keeping the far sharp at infinity. When you hit upon a CoC
diameter that yields a satisfactory infinity, you can print a table that
never fails you -- at least for that lens, at whatever magnification and
viewing distance you were shooting for. It's an absolute pity to throw
away depth of field for lack of knowing exactly how far short of infinity
you can focus.
Hate carrying DoF tables? Even if you were to just use the DoF scales
engraved on your lens, always indexing your focus as if you were using the
next wider aperture than the one you're really using, you'll push the near
sharp and the hyperfocal distance further out and thus, shrink the CoC's
in half at infinity. If that doesn't do it for you, move on -- index your
focus against the next wider aperture, until you're satisfied. Once you
find an offset that's pleasing, you'll never be guessing thereafter.
>Thanks for a nice post!
You're welcome -- I appreciate the feedback!
Mike Davis
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Circle of Confusion
Date: Tue, 29 Dec 1998
On 28 Dec 1998 04:25:20 GMT, "Michael K. Davis"
[email protected] wrote:
>Neuman - Ruether wrote:
[most deleted...]
>Hate carrying DoF tables? Even if you were to just use the DoF scales
>engraved on your lens, always indexing your focus as if you were using the
>next wider aperture than the one you're really using, you'll push the near
>sharp and the hyperfocal distance further out and thus, shrink the CoC's
>in half at infinity. If that doesn't do it for you, move on -- index your
>focus against the next wider aperture, until you're satisfied. Once you
>find an offset that's pleasing, you'll never be guessing thereafter.
Yes, that is essentially what I do (use the DOF indicators
for the next wider stop, but then I sometimes use a stop
smaller to cover the near end better [and infinity more
conservatively {assuming I have enough "stopping-room"
before nasty diffraction sets in...;-}]) - and it works
well! (And doesn't throw away all that nice near-side DOF,
which is what happens with simply setting the lens at
infinity-focus...)
David Ruether
[email protected]
[email protected]
http://www.fcinet.com/ruether
From: "Michael K. Davis" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Circle of Confusion
Date: 29 Dec 1998
I found the page!
Michael K. Davis [email protected] wrote:
: Above, you say that I was referring to Harold Merklinger's web page when I
: wrote of a web site that advocates focusing at infinity. It's a simple
: mistake, so no hard feelings, but I can't remember where I saw that nor
: who the author was and never said it was Merklinger, but I do remember A
: web page photo with out of focus cannons in the foreground accompanied by
: text that advocates focusing at infinity. The closest I ever came to
: indentifying the web site was:
:
: >There's a web page that I read once, where the author strongly advocates
: >that for landscapes, you should set your focus at infinity and then
: >select an aperture that will give you an acceptable near sharp.
:
: Is that Merklinger? If a Merlinger page has the picture I described and
: you've put it together that I'm quoting him, then you're right -- I AM
: misquoting him, indeed. Do you have the address to the page with that
: photo? I've read a lot of Merklinger's work and know that he's too savvy
: to be as inflexible as the guy I'm thinking of.
:
: Mike
I went and looked at the address you gave for Merklinger's pages and found
the cannon picture here:
http://www.trenholm.org/hmmerk/DOFR.html
You were right, it was Merklinger's page, but I'm not so sure I misquoted
him when I said he advocates focusing at infinity. Let me quote the text
that accompanies the photo:
Caption, Figure 3: This scene, the village of Placentia in Newfoundland,
was taken with a 90 mm lens set at f/8 and focused at infinity. As
explained in the text, to have focused at the hyperfocal distance would
have seriously degraded the image of the village, while making negligible
improvement to the foreground. [snip] The foreground is admittedly not
tack-sharp. But I can recognize the cannon, the gravel, the grass and the
trees. Had I focused at the hyperfocal distance the telephone poles in the
village would have been almost erased, and windows in buildings would have
been just blurs.
Well, from this example alone, we can see that Merklinger is willing to
throw away some depth of field to make sure his distant subjects are razor
sharp. There's no room for debate here. I'm dissappointed in his having
made this decision even once, because to me, it seems like a departure
from his proven mastery of the subject. He has not stated anywhere in the
text that he purposely sought a soft foreground. To the contrary, he
laments it not being tack-sharp and states:
>... there is really only one way to make the foreground sharper, and
>still have a sharp view of the village. That is to use a smaller lens
>opening. In this example, I could have reduced the spot size to 5.5 mm
>by using f/16. Resolution of the village, however, would have been
>slightly degraded by diffraction effects.
On this I'd like to comment how pleased I am to see someone else make
mention of the diffraction 35mm suffers at f/16. I have been scoffed a
few times for saying that 35mm users should avoid going to apertures
smaller than f/11, so hurray for that bit of wisdom! But, when he says he
wouldn't want to use f/16, he forgets to mention that the photo was taken
at f/8, so he could have had a sharper foreground at f/11 and still left
his focus at infinity, but back to that strategy: He is promoting the
abandonment of focusing at the hyperfocal distance in favor of focusing at
infinity. That's the claim I made when I didn't know who I was quoting
and that's the claim I'm making now. Let me quote Merklinger again, from
the same page (above the photo):
>What bothers me even more is the old story about maximizing depth-of-field
>by focusing at the hyperfocal distance. If you follow that advice you will
>be guaranteed that scenes in the distance will never be resolved any
>better than mediocre. You will have sealed in that "minimum acceptable
>standard".
Sounds like I didn't misquote him in my original post.
When talking about depth of field scales on lenses, he also says:
>Thus the maximum permissible "circle of confusion" is 1/30 mm in diameter
>for a 35 mm camera with its 50 mm standard lens. The problem is that
>today's films and lenses are capable of achieving a resolution standard
>at least five times as stringent, and maybe more. But if one enters that
>revised standard into the formulae, depth of field just about disappears.
>And that doesn't square with experience either.
I'll make the same arguement I made before. He needs to CHOOSE a SMALLER
maximum permissible diameter for Circles of Confusion and recalculate his
depth of field tables. He CAN trust the math!!
In discussing his photo, he says:
>The hyperfocal distance for a 90 mm lens at f/8 is 106 feet.
He says this as if the CoC is carved in stone. By the way the max. CoC
would be about 1/138th-inch in a 10-inch diagonal print to yield a
hyperfocal distance of 106-feet for a 90mm lens at f/8 with 35mm format.
No wonder he's disappointed. Most people would not find that to be sharp
unless their viewing distance was greater than the print diagonal by half
again as much, at least. If he likes to stick his nose to his prints,
he'll have to select a CoC of at least 1/300th-inch.
Calculating new hyperfocal distances for each of his lenses at each of
their f/stops, with experimentation, he could quickly find a CoC diameter
that is satisfying AND THEN his hyperfocal distance (WHICH HE SAYS
SHOULDN'T BE USED) would move further away from the camera, but it would
still fall well short of Infinity. INFINITY itself would be SATISFYINGLY
SHARP, AND BEST OF ALL, he would have left some depth of field to work for
him in the foreground instead of throwing it away by FOCUSING AT INFINITY!
When he focuses at Infinity, two thirds of his depth of field is pushed
out beyond infinity -- a total waste! He could have much sharper
foregrounds and still preserve a degree of sharpness at Infinity that
SUITS HIS PERSONAL CHOICE.
I MUST disagree with his final advice on the subject, which includes:
>The general rule for scenic photographs, where one wishes to maximize the
>depth of field, is as follows. Set the focus at the distance of the most
>distant object. [Infinity?] Then set the lens opening to the size of the
>smallest object to be resolved in the foreground. No calculations
>needed! [Those stones in the foreground are about an inch in diameter, so
>... gee! I can shoot at f/4 with my 90 mm lens and everything will be
>sharp as a tack!! I like this no calculations stuff!]
and
>As a general rule, lens aperture diameters of 3 to 5 mm are often the
>optimum for scenic purposes. [Let's see... 90mm divided by 3mm is about
>f/32 -- That ought to do the trick! Ooops! What's this? Diffraction
>kicks in at f/16?] Blades of grass and small branches are resolved, and
>resolution of distant objects is about equal to that allowed by the
>majority of modern films and lenses. This rule applies to all formats
>[except the ones I use in my cameras.]
Do the math!
http://home.sol.no/~gjon/mdofcalc.xls
or
http://home.sol.no/~gjon/mdofcal2.xls (should be available by 1/7/99)
A special note to Harold Merklinger: I honestly admire the tremeondous
body of work you've done and I understand your perspective completely,
even though I disagree. I too much prefer a soft foreground to a weak
background when shooting landscapes. Although I can not apologize for the
technical content of my argument, I do acknowledge my abrasive style and
beg your tolerance. You deserve a less provocative debate.
Mike Davis
--
/---------------------\
Michael K. Davis
[email protected]
From: Anders Blomqvist [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Circle of Confusion
Date: Tue, 29 Dec 1998
"Michael K. Davis" wrote:
> I found the page!
>[snip]
> I went and looked at the address you gave for Merklinger's pages and found
> the cannon picture here:
>
> http://www.trenholm.org/hmmerk/DOFR.html
>
> You were right, it was Merklinger's page, but I'm not so sure I misquoted
> him when I said he advocates focusing at infinity. Let me quote the text
> that accompanies the photo:
>
> Caption, Figure 3: This scene, the village of Placentia in Newfoundland,
> was taken with a 90 mm lens set at f/8 and focused at infinity. As
> explained in the text, to have focused at the hyperfocal distance would
> have seriously degraded the image of the village, while making negligible
> improvement to the foreground. [snip] The foreground is admittedly not
> tack-sharp. But I can recognize the cannon, the gravel, the grass and the
> trees. Had I focused at the hyperfocal distance the telephone poles in the
> village would have been almost erased, and windows in buildings would have
> been just blurs.
Yes, I can understand you think he always advocates focusing at infinity
if you merely look at the simplified version on
http://www.trenholm.org/hmmerk/DOFR.html. However, I claimed that he does not
_always_ promote focusing at infinity (or most distant subject), as you
can find in his more detailed Shutterbug articles (especially SHBG04.pdf
"Adjusting Depth of Field Part IV" on
http://www.trenholm.org/hmmerk/HMbook14.html). I agree that he certainly
doesn't approve of hyperfocal distance calculations.
[snip]
> I'll make the same arguement I made before. He needs to CHOOSE a SMALLER
> maximum permissible diameter for Circles of Confusion and recalculate his
> depth of field tables. He CAN trust the math!!
Well, it looks like he has done his math (SHBG01.pdf, "Adjusting Depth of
Field"), and claims he wants at least five times smaller CoC to be pleased
with his photos. This if not very often achievable, so he has developed
his alternative approach to DOF. This approach is, by the way, also based
on math. I think your approaches are somewhat different: he wants his
distant details to be tack sharp when viewing the picture at a close
distance, and is willing to sacrifice near-end sharpness, you are talking
about a uniform sharpness when viewed from _normal_ viewing distances. I
personally don't think either of you are wrong.
[snip]
> I MUST disagree with his final advice on the subject, which includes:
>
> >The general rule for scenic photographs, where one wishes to maximize the
> >depth of field, is as follows. Set the focus at the distance of the most
> >distant object. [Infinity?] Then set the lens opening to the size of the
> >smallest object to be resolved in the foreground. No calculations
> >needed! [Those stones in the foreground are about an inch in diameter, so
> >... gee! I can shoot at f/4 with my 90 mm lens and everything will be
> >sharp as a tack!! I like this no calculations stuff!]
Yes, this is his simplified general rule for tack-sharp distant objects
and less sharp foreground. However, he also has more detailed calculations
available for more technically advanced photographers (you?) in his
articles.
[snip]
> A special note to Harold Merklinger: [snip] Although I can not apologize for the
> technical content of my argument, I do acknowledge my abrasive style and
> beg your tolerance. You deserve a less provocative debate.
Who doesn't?
--
Anders Blomqvist, Singapore
From: Bob Wheeler [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Depth of field
Date: Thu, 31 Dec 1998
Tom Rittenhouse wrote:
> Somehow my post didn't go out correctly. Hopefully this
> will. Glad I saved it before posting.
>
> Bob Wheeler ([email protected]) wrote:
> :
> : DOF = (2NcM)/(1-K),
> :
> : where M=(1+m)/m^2, and K= (Nc)/(mf),
> :
> : with
> :
> : N: f-stop,
> : c: diameter of the circle of confusion
> : m: magnification -- image size divided by subject
> : size
> : f: focal length.
> :
>
> Well, with your formula you don't need to consider the format.
Yes you do. If for 35 mm you have
N=16,c=0.025,m=0.01,f=50, then for 4x5 you have
N=64,c=.1,m=0.04, and f=200. All parameters must
be multiplied by the same factor. If you choose
not to scale N, then DOF decreases with increasing
format. If you choose to scale it, then DOF
increases slightly with increasing format. I
uploaded an Acrobat file with a couple of graphs
that may help
ftp://www.echip.com/ftp/public/photo.
As far as c is concerned, it is usually chosen as
a film parameter that makes the print resolution
equal to the discrimination of the human eye. The
normal eye can resolve about 5 lines per
millimeter, l/mm, in a print held at the closest
distance of clear vision -- about 25 cm. For an
8x10 print, this means that one needs 10 l/mm on
the film, or that the diameter of the circle of
confusion should be the reciprocal, or 0.1 mm.
> Using m = image size on print / size of subject
> thus c becomes a constant .1mm(?)
> (BTW .1mm = about .004in that is 2.5 times smaller than I mentioned)
> So, f, N and m are the only changes.
> lets make m = .1 (70in tall person, 7 high on print)
> Now, f, N are the only variables.
>
> (2xNx.1x((1+.1)/.1x.1)) / ((1-(8x.1)/(.1xf))
>
> f-8/100mm: (2x8x.1((1+.1)/.1x.1))/((1-(8x.1)/(.1x100))
> (1.6((1.1)/.01))/(1-(.8)/(10))
> (1.6x110)/(1-.08)
> 176/.92
> 191.3mm
>
> f-8/50mm: (2x8x.1x((1+.1)/.1x.1)) / ((1-(8x.1)/.1x50))
> (1.6((1.1)/.01)) / (1-(.8)/(5))
> (1.6x110) / (1-.16)
> 176/.84
> 209.5mm
>
> Compare: 191.3/209.5 = .91 = 91%, or
> 9% difference
>
> As I said, "similar DOF".
>
> --tom
>
--
Bob Wheeler --- (Reply to: [email protected])
ECHIP, Inc.
rec.photo.misc
From: "Mick McCarron" [email protected]
[1] Re: Depth of Field Charts
Date: Fri Feb 12 09:37:12 CST 1999
Rick Davis wrote
>I'm looking for Depth of Field Charts for my 300mm f5.6 lens (8x10). I have
>a nice set of charts made by Polar Productions in 1992 for a 190mm lens I
>had at the time. They show critical use and general use respectively. Is
>there someone out there who can help me locate similar charts. I need them
>to be laminateable, if possible. Thanks.
If you have EXCEL or a Psion organiser there is a spreadsheet available on
this site,
http://www.pocketinfo.org/leisure_mis.html
With this you can calculate your own.
Mick
From: "Michael K. Davis" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: DoF Tables for Mamiya 7
Date: 24 Mar 1999
Hi!
S. Gareth Ingram [email protected] wrote:
: Just for arguements sake - I'm not too keen on Dof tables - they give a
: false sense of security - they try to establish a clear cut rule where
: none is possible. What is acceptable as sharp to one person is not
: acceptable to another. You will be dissapointed if you expect Dof tables
: to be the answer.
Your argument is perfectly reasonable in light of the fact that nearly
EVERY depth of field calculator available, provides NO WAY TO MODIFY
that subjective variable you are referencing! Depth of field tables are
no more or less likely to be dissapointing than the engraved scales we
find on our lenses. Did you notice I provided two sets of tables for the
Mamiya 7 lenses? One set matches their lens barrel DoF scales, the
other is twice as aggressive. The math CAN be trusted and if you want to
do critical work with a rangefinder, it MUST be trusted. You have no
choice if you want to avoid rolling the dice every time you make an
exposure.
The wholesale rejection of depth of field scales and tables stems from the
simple lack of exposure to using a calculator that treats Circle of
Confusion diameter as a user-specifiable VARIABLE instead of as a
CONSTANT. When you are permitted what is apparently the LUXURY of
deciding for yourself how sharp you want your prints to be at a
given viewing distance, you can begin to take control of depth of
field. Most DoF calculators don't even document what constant they use
for maximum permissible diameter of circles of confusion. Those few that
do, specify separate values for each format (for CoC diameter at the film
plane, before enlargement), which makes it really tough to go shopping for
a calculator that suits your tastes if you work in more than one format.
Every calculator I have ever seen also treats format diagonal as a
constant! Do you print all of your images full frame? No wonder we are
disillusioned with DoF calculators! When you crop, you are decreasing the
diagonal and thus increasing the magnfication required to reach a given
print size - diminishing the resulting depth of field! Hello!
Every depth of field calculator should provide the following features:
1) User-specifiable maximum permissible diameter for Circles of Confusion
at the Near and Far Sharps: This should be expressed in units common to
all formats! Not at the film plane, but rather AFTER MAGNIFICATION to a
reference print size, to be viewed at a reference distance. This is the
ONLY way to compare CoC diameters from one format to another or for
different crops from a single format! Hello! THE CONVENTION IS:
Specify the maximum permissible diameter YOU PERSONALLY ARE WILLING TO
TOLERATE in a 10-inch diagonal print to be viewed at a distance of
10-inches. The beauty of this reference standard is that once YOU CHOOSE
a diameter you find acceptable for YOUR Near and Far Sharps, you will find
it works FOR ANY SIZE PRINT viewed at a distance equal to its diagonal!
AND: It will work for ANY FORMAT, with ANY CROPPING you impose!
CONSISTENT, REPEATABLE RESULTS! (Quoting page 131 of "Basic Photographic
Materiels and Processes" by Stroebel, Compton, Current, and Zakia (c1990
Focal Press): "Permissible circles of confusion are generally specified
for a viewing distance of 10 inches, and 1/100 inch is commonly sited as
an appropriate value for the diameter. A study involving a small sample
of cameras designed for advanced amateurs and professional photographers
revealed that values ranging from 1/70 to 1/200 inch were used --
approximately a 3:1 ratio.") In light of this, consider that one set of
tables I generated for the Mamiya 7 lenses uses 1/145-inch to match
MAMIYA'S choice as exhibited by their engraved lens scales and the other
set of tables uses 1/290-inch -- a value I challenge you to find lacking
for most situations. (It is three times as agressive as the value Kodak
uses in their Photoguide spinning disk calculators, 1/100-inch!) The
impact of using so aggressive a value is simple -- your near sharps will
be further away from the camera. If you can't back away or go up (as
Ansel often did, climbing to the rooftop platform of his International
Harvester) to get your foreground subjects far enough away to fall within
the near sharps that result with such an agressive value, then you'll have
to suffer a compromise (or use movements to reposition the plane of
focus.)
Once you find a value that YOU find acceptable, the only time
you'll have to change the CoC variable is when you anticipate viewing
distances that won't be equal to the print diagonal. In fact, if you know
in advance that a print will be displayed such that viewers are prevented
from coming closer than twice the print diagonal, you can cut your CoC
diameter in half and enjoy the same illusion of depth of field by opening
up two full stops, with a resulting shutter speed that's 4 times as fast!
If you know in advance that viewers will be able to examine the print at
1/2 its diagonal, all you have to do is decrease the maximum permissible
diameter for CoC's to 1/2 of your preferred value for the reference print
size and viewing distance. If you like 1/200-inch for prints viewed at a
distance equal to their diagonal, you'll have to calculate tables using
1/400-inch for this project. The resulting DoF tables will push your focus
point and your Near Sharp further away from the camera than they were at a
less agressive CoC diameter. You CAN control the variables if you have a
calculator that PERMITS you that control!
2) User-specifiable format diagonals: If you crop a 35mm slide to produce
an 4:5 aspect ratio print (4x5, 8x10, 16x20, etc.), at best your format
diagonal has just gone from 43.3 mm to 38.4mm. For a 50mm lens at f/16,
limiting the CoC diameters to 1/175-inch at the Near and Far Sharps,
focusing at the hyperfocal distances in both cases, your Far Sharp will be
Infinity for both, but where the full-frame Near Sharp is at 10.37 feet,
the 4:5 ratio crop's Near Sharp will be at 11.67 feet. Hello! Format
diagonal should be treated as a VARIABLE, not a CONSTANT. All you need is
a calculator that allows YOU to specify the format diagonal, after
cropping, and wouldn't it be nice if it had a front end that helped you to
calculate that diagonal for various aspect ratio crops? You CAN take
control of depth of field if you have a calculator that PERMITS it.
3) A depth of field calculator should use the variables already provided
to calculate the aperture at which the effects of diffraction will become
visible: What good is it to pursue Depth of Field while letting
diffraction in the back door?! Hello! If diffraction's Airy disks are
larger than your circle of confusion diameters, why make them larger still
by trying to get smaller cirles of confusion? Smaller aperture increases
depth of field as it increases diffraction. Depth of field isn't the only
factor affecting image quality. If it were, we wouldn't need lenses, we
would all be shooting with pinhole cameras -- awesome depth of field,
awesome diffraction.
Here is a calculator that provides all these features. It's a MS-Excel
5+ spreadsheet. I'm told it works on Macs as well as Windows machines:
http://home.sol.no/~gjon/mdofcal2.xls
For those who can't make use of such a calculator to generate tables for
use in the field, just try this. No math required: If you don't like the
results you are getting with your lens barrel DoF scales, just index the
scales at a wider aperture than the aperture you are actually using to make
the exposure. For example, if you are shooting at f/16, imagine that you
are using f/8 when examining the Depth of Field scale on your lens. This
will double the aggressiveness of your lens manufacturer's choice for
maximum permissible Circle of Confusion diameter. If you are shooting at
f/11, index the Near and Far Sharps against f/5.6. You might find that
it's not necessary to double your manufacturer's choice of CoC diameter.
It might be acceptable to shoot at f/16 but index at f/11, only one stop
wider, instead of two. Experiment, find the offset that is acceptable to YOU
and remember that if the viewing distance will be cut in half you have to
index two stops wider. If you will be cropping 35mm to produce a 4:5
aspect ratio print, index one stop wider than you would for full frame.
Think.
Mike Davis
--
rec.photo.equipment.medium-format
From: "Patrick Bartek" [email protected]
[1] Re: Medium format recommendations
Date: Thu Mar 25 05:34:28 CST 1999
Regarding Re: Medium format recommendations, SPECTRUM wrote:
> >Last but not least, don't forget that as you move to larger film sizes, you
> >will have less depth of field when working at larger magnifications compared to
> >35mm systems.
>
>
> I really wish someone would take the time to clarify this as I
> think it's one of those things that just keeps getting repeated ad
> nauseum.
> IMO, a 6X7 will produce a similar appearance of DoF in an 8X10
> as will a 35mm at a lower aperture. It's all a result of the final
> magnification and for some reason this seems to be consistently
> overlooked. I realize that the average DOF charts take into account
> the variations in the CoC but the end results of most DoF calculations
> don't seem to concur with my observations. What gives ?
Okay, here's the last and true words on this depth of field thing:
"Depth of field varies according to the focal length and f-number of
the lens, the distance to the subject, and the acceptable degree of
sharpness (circle of confusion)." [ref: M&M Photo Lab Index, 1971;
page 9-21, Photographic Optics and Filter Data: "Depth of Field of
Camera Lenses"]
Notice that film format is not a part of the DOF equation. So, a
300mm lens at f8 focused at 25 feet with the same circle of
confusion has the same depth of field whether its a 35mm lens
on 35 camera or an 8x10 lens on a 8x10 or vice versa. The image
circles projected are indeed different for each lens, but the depth of
field is the same (REALLY!), even though a 300 on a 35 is a telephoto,
but is a normal on an 8x10.
However, note this: the circle of confusion chosen for 35mm lenses is
smaller than for 8x10 lenses, since 35mm negs must be enlarged more to
gain useful print sizes. So, in the real world, with this difference
taken into account, a 300 on 35 will have a smaller DOF, than a 300 on
8x10.
--
Patrick Bartek
NoLife Polymath Group
[email protected]
From: [email protected] (Andreas Nicklass)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: DOF with a close-up lens
Date: Thu, 8 Apr 1999
Lim Meng Shi [email protected] wrote:
> Does anyone know how to calculate the DOF for a close-up lens
> attachment?
Okay, if nobody else volunteers to answer this, I'll try.
It's actually quite easy because we are in the macro region
and the only parameters that matter are the magnification M
and the aperture number F.
dof = 2 c F / M^2
c is the circle of confusion you will tolerate.
For example with M = 0.4, F = 16 and c = 0.025 mm DOF will
be 5mm (1/5 inch).
Andreas
From: [email protected] (Thomas Bantel)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: AF blues Re: Are manufacturers right...nobody cares about
DOF???
Date: 27 May 1999
>Yeah, I guess you are right. Since I am only paying 50% more for these
>new AF lenses with the same optical designs, mostly, as my old MF lenses,
>I guess I shouldn't begrudge Nik/Can/Min/Oly/Pent/Con/Yas/Sig/Tam/Tok...
>saving a few dollars by leaving off those distance scales and DOF tables.
>A savings of $7 is $7 in their pockets, and they probably need the money...
Don't get me wrong. I didn't mean it *that* way. Of course every lens
should have a distance scale and (where feasible) DOF markings. But,
as a matter of fact, lots of todays lenses don't have them. And there
are lots of people stuck with such lenses. All I wanted to say is, they
still can use hyperfocal focusing and it is not overly hard, even without
a distance scale. Most of my lenses have a distance scale and some even
have the DOF marks ;-) But, I don't use them all that often. Why? First,
I have found that distance scales are unreliable with most of my zooms.
They are varifocals and no real zooms, resulting in different distances
on the scale for different focal lengths and same real subject distance.
Second, the DOF markings are too optimistic for my taste anyway.
When I'm out, I use as hyperfocal distance simply
50mm lens: h = 100 meters / aperture number
24mm lens: h = 23 meters / aperture number
Really easy to calculate, no computer needed ;-) Often I estimate the
distance even when the lens has a reliable distance scale. I look around for
something in an appropriate distance and focus on it. That way I don't even
have to take the camera down to look at the distance scale. When I'm off by
10 percent with my estimate distance, it's no big deal. I'll never notice
it in the final picture.
Of course, if you don't use hyperfocal focusing but want a DOF not including
infinity, things get more complicated. But, fortunately, my cameras have the
DEP mode which does the calculations for me. I stop down one stop from that,
to compensate for the (IMHO) optimistic values the program gives.
>and hey, if they drop the distance scales, that's not a problem either, I
>can carry a 100 meter tape and just run it on out there whenever I'm
>using one of my long lenses - I need the exercise - sure hope those
>critters don't bolt when they see me jumping thru the bushes...
>
>yep, I'm sure happy I can contribute all these monies to "upgrade" my
>lenses (if not my optics designs) and help out the economy by giving up
>these expensive features like distance scales and DOF markings and stuff.
>
>heh heh heh - grins bobm
They already drop the distance scales and I sure don't like it. All I
wanted to say is, even without these tools it is doable - and IMHO quite
easy - to use hyperfocal focusing. Otherwise some lenses would be *really*
unusable IMHO. As it is, they are, hmm, inconvenient :-)
Thomas Bantel
From: "NEDSNAKE" [email protected]
Subject: DOF
Date: Sat, 5 Jun 1999
Newsgroups: rec.photo.equipment.35mm
A system I have used for DOF is to focus 1/3 of the distance to the subject
and stop down as far as possible or practical. This system works well in
low light situations. Give it a try, see if it works for you.
Mike
From Rollei Mailing List:
Date: Thu, 24 Feb 2000
From: Bob Shell [email protected]
Subject: Re: [Rollei] Low light TLR performance
Huh???
Final image size is only part of the picture.
Allen is right in what he says. The rule of thumb was always that for
a given angle of coverage you would need to stop down two additional
stops in medium format to get the same depth of field as 35 mm if the
final print sizes were the same. I learned this early in my days
working for a portrait/wedding studio.
Bob
...
From Leica Mailing List:
Date: Sun, 2 Apr 2000
From: "Henning J. Wulff" [email protected]
Subject: Re: [Leica] Photographig Photographers
.......
>BTW, the CompuServe Photo Forum was supposed to open with free access to
>general internet users. I'll have a look at www.compuserve.com to see
>whether this has happened.
>
>--Mitch
Well, I haven't been on Compuserve for over 5 years, but the boredom
factor was pretty high there at the time. Maybe things have changed.
Here, people sometimes feel DOF is boring, and neck strap length is at
least something new, if not for everyone. If you want maximum sharpness at
infinity, set your lens at infinity. If you want 'critical' sharpness
between a closer object and infinity, set your lens to 2 or 3 stops
smaller
than the DOF scale on your camera recommends and use the hyperfocal
setting. These thing are not fixed, just as the criteria of 'critical
sharpness' is not fixed. I usually feel satisfied in 35mm work if I set my
lens to 2 stops smaller than the DOF indeces dictate. If I need more, I
haul out a 4x5 and play with that. Sometimes you just can't get enough DOF
and you have to make a decision as to what's important, and focus exactly
on that. Maybe you have to open up to maximum aperture to properly isolate
the item to make sure everyone knows you don't want DOF. All the above
'rules' that you mentioned are applicable under some circumstances. None
are universally applicable.
* Henning J. Wulff
/|\ Wulff Photography & Design
/###\ mailto:[email protected]
|[ ]| http://www.archiphoto.com
From Rollei Mailing List:
Date: Mon, 27 Sep 1999
From: [email protected]
Subject: Re: [Rollei] OT : Depth Of Field, etc...
About Depth of Field and other basic formulae : a reminder, the good
technical article in :
http://www.graflex.org/lenses/photographic-lenses-tutorial.html
--
Emmanuel BIGLER
[email protected]
<
Date: Fri, 11 Jun 1999
From: [email protected] (Bill Welch)
Newsgroups: rec.photo.technique.nature
Subject: Re: More depth of field
Mr. Burian is correct to say the depth of field is less with longer focal
lengths "from the same shooting distance". A lot of people don't
understand that distinction.
However, the depth of field is the same for different focal lengths, when
the subject is reproduced, on the film, at the same image size! A 180mm
macro and a 100mm macro will have the same depth of field, at the same
aperture and same image size.
All things being equal, a longer focal length lens will have less depth of
field, than a short focal length lens; when they are both focused the same
distance from the subject, and at the same aperture.
That's not really too relavant, because the goal of most macro work is to
produce an image at a specific size. When photographing a caterpillar
with a 100mm macro lens, the photographer usually gets as close as
necessary to produce the desired image size on the film. With a 180mm
macro, the same thing happens, just at nearly double the working distance.
At the same aperture and image size the depth of field should be the same.
William Welch
Photo website at:
http://siscom.net/~wwelch/
Date: Wed, 23 Jun 1999
From: "B. Buckles" [email protected]
Newsgroups: rec.photo.technique.nature
Subject: Re: Hyperfocal Focus with AF lenses?
Before you throw all your lenses away, try this" (unless you have Pentax
lenses, then throw them my way) So, for the 35 mm camera, take the focal
length you are shooting at and divide that by 10. Then square this
number. That's the distance, in feet, to focus your lens at with the
f-stop set at f/11 to have infinity in focus. This is called the
hyperfocal distance. One half this distance in front will also be in
focus. Example: 50 mm lens set at f/11. 50/10=5. 5X5=25'. Focus at 25'
and everything from 1/2 of 25' to infinity is in focus. If you need more
DOF, use the number you calculated for f/11 and multiply that number by
2/3 or 18'. Now set your lens at f/16 and focus at 18'. Everything from
18/2 or 9' to infinity is in focus. For f/22 is another 2/3 the f/16
distance. (18 X 2/3 = 12') Going the other direction, multiply the
hyperfocal distance by 3/2. Make your self a small chart with these
numbers and tape it to the back of your camera body as I have.
Good Luck,
Bob
From: "Michael K. Davis" [email protected]
Newsgroups: aus.photo,rec.photo.equipment.35mm
Date: 11 Jul 2000
Subject: Re: depth of field and redeye
A useful formula is:
Min. Safe Distance Between Lens and Flash = Focal Length * 4
This was published in Popular Photagraphy about a year ago, but not in
such a simple format - they had a table that showed how the probability of
avoiding red eye improved with an increase in the ratio of flash-lens
distance to focal length. The above formula is my interpretation of that
table for a 90% probability of avoiding red-eye.
Even if a flash is ten feet above the lens, with a short focal length,
there exists the possibility that the subject's eyes will be looking at a
point between flash and lens which is at precisely the right angle to
reflect the light of the flash off the retinas into the lens. If we can
assume the subject will be looking at the lens, Popular Photography's
article indicated there is a 90% chance of avoiding red-eye by displacing
the flash to a distance that is four times the focal length. Achieving a
100% probability of red-eye avoidance is nearly impossible with a specular
(non-diffuse) point source. I chose to remember that 90% could be
achieved with a 4x displacement. I don't recall the probabilities they
published for other ratios of lens-flash separation to focal length.
Mike Davis
photos33 [email protected] wrote:
: I once saw a mathematical formula's for depth of field and redeye can
: anybody help?
: The redeye one was something like.
: A x B = C
: "A" ....distance from lens to flash
: "B".....?????
: "C".....safe working distance for red not to occur.
:
: Is there a formula for working out the depth of field for a particular focal
: length. For example a 50mm lens using f8 focused at 5 metres
: regards Gary
[Ed. note: reminder tip re: DOF in teleconverters...]
From Hasselblad Mailing List:
Date: Thu, 27 Jul 2000
From: "Q.G. de Bakker" [email protected]
Subject: Re: Teleconverter questions
Paolo Pignatelli wrote:
> I have a 250 5.6 Sonnar, and was considering purchasing a teleconverter for
> it. First question, any recommendations as to which one? I see that there is
> a 1.4X, which would be okay, but I would rather have a 2X.
> Second question, while I realize that I will lose speed, will I also get the
> DOF increase that normally accompanies the smaller f stop? My application
> is principally landscape photography.
If you would rather have a 2x converter, you should of course get a 2x
converter. Try to find the original Zeiss Mutar 2x used. Or else the
Hasselblad 2xE must do.
You will *not* get any increase in DOF. You'll only lose light as if the
aperture was closed two extra stops, you don't really get a smaller
f-stop. What happens is that the image is made by your f/5.6 250 mm lens,
and *after* it is created, it is magnified by the converter. This means
spreading out the light over a larger are, which is why you lose 2 stops.
But it also means the unsharpness in the image is enlarged as well, so you
will *lose* DOF. In fact you will get DOF as if you used a f/5.6 (!) 500
mm lens.
From: [email protected] (PBurian)
Newsgroups: rec.photo.equipment.35mm
Date: 28 Jul 2000 12:55:36 GMT
Subject: Re: Another question about telephoto lens
>Not all telephoto lenses are designed to stop focusing at infinity. You
>still have to focus them, even at objects ordinarily considered at infinity.
Right. Many telephoto lenses will focus beyond infinity. If you manually
set them for the furthest possible distance, your photos of distant
subjects may not be critically sharp.
Peter Burian
[ed. note: thanks to Mr. Michael Gudzinowicz for sharing these
pointers...]
From: [email protected] (Michael Gudzinowicz)
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Circle of confusion
Date: 18 Aug 2000
"Peter De Smidt" [email protected] wrote:
>I'd like to work out hyperfocal distances for my lenses. In order to do
>that, though, I need to know what most people think is an acceptable circle
>of confusion. Stroebel says that 1/250" is appropriate for 4x5 negatives.
>Do all of you agree? I'd also like to know what you suggest for 5x7 and
>8x10 sized negatives. Any insight on this subject would be appreciated.
Stroebel's value is approximately 0.1 mm, which is just beyond the limit
of sharpness for the human eye, so it would work for critical contact
prints. For 8x10 prints, the value in the print would be 2 x 0. 1 mm = 0.2
mm, which is close to the value of 0.25 mm which is used as an upper limit
to define "just acceptable" sharpness in a print.
I prefer to use a value of 0.125 mm in the print which is at the limit of
vision. The circle of confusion for the print is divided by the print
magnification to give that required in the negative. For a 4x print,
a value of 0.125 / 4 = 0.03125 mm may be used.
From Panoramic Mailing List:
Date: Tue, 24 Oct 2000
From: "M. Denis Hill" [email protected]
Subject: RE: Image sharpness and focus distance.
A reasonable definition is, "Acceptable Circle of Confusion is the size of
the largest circle which the eye cannot distinguish from a dot. In 35mm
format cameras, a 0.03mm diameter circle of confusion is considered
acceptable. It is used to calculate depth-of-field or depth of focus."
See a detailed discussion at http://artzonephoto.com/artzone/zlensthe.htm,
and more at http://www.minoxlab.com/Don_Krehbiel/mpl/dkdof.htm.
M. Denis Hill
Area 360 Communications
http://www.area360.com
From LEica Mailing List;
Date: Thu, 09 Nov 2000
From: Jim Brick [email protected]
Subject: [Leica] Re: Re: DOF
After you click "plot", wait. It can take a minute.
Have fun.
http://www.dof.pcraft.com/dof-frames.cgi
Jim
Date: Thu, 09 Nov 2000
From Leica Mailing List;
From: Jim Brick [email protected]
Subject: [Leica] Re: Re: DOF
This one is also fun.
http://www.photozone.de/depth.htm
Jim
[Ed. note: a great review of DOF concepts by Mr. Brick, a noted
photographer and photobook author and engineer involved in digital camera
design...]
From Leica Mailing List:
Date: Wed, 07 Feb 2001
From: Jim Brick [email protected]
Subject: [Leica] RE: DOF
>> Depth of field is a factor of image size and relative aperture. The film
>> format/ lens combination affects the image size. A 45mm lens in on a 4x5
>> camera will record an object "smaller" in the final print than a 45mm on a
>> 35mm camera. You could enlarge the 4x5 negative more than normally is done
>> and print an identical image size (assuming you shot both cameras at the
>> same subject distance) and achieve identical depth of field (and COC) but
>> this is not normal practice.
Here is a good DOF site. After hitting PLOT, don't be impatient as it
takes
some time to calculate.
http://www.dof.pcraft.com/dof-frames.cgi
Set up 75mm lens, f/1.4, .5 feet, and you will see that at f/1.4, 2', DOF
is minuscule. 1/4" at best. The thickness of a tulip petal.
For a given COC, a 50mm lens 10 feet away from the subject has the EXACT
SAME DOF as a 100mm lens 20 feet away from the subject. A 50mm lens at
f/1.4 focused at 10 feet has from 9'7" to 10'5" in focus. A 100mm lens at
f/1.4 focused at 20 feet (exact same image size on the film as the 50mm
example), the DOF is from 19'7" to 20'5". The depth of field is EXACTLY
the same. EXACTLY 10 INCHES IN BOTH CASES for the given COC.
DOF has EVERYTHING to do with IMAGE SIZE, which is a combination of the
focal length you are using AND how far away you are from the subject.
If the "image size" on the film is the same, the DOF is the same for a
given f/stop, REGARDLESS OF FOCAL LENGTH and REGARDLESS OF THE FILM SIZE.
This is a simple fact.
*** MORE ***
Depth Of Field is a function of f/stop and image size. An 80mm lens on a
4X5 has precisely the same DOF as an 80mm lens on a Hasselblad,
which has exactly the same DOF as an 80mm lens on a Leica, in every
situation.
One does not give less or more DOF than the other.
Take two Hasselblads, one with an 80mm lens, the other with a 160mm lens.
Take the 160mm Hasselblad exactly twice the distance from the subject as
the 80mm Hasselblad is (this will make the photographed image - its size -
on the film exactly the same for both cameras) and the DOF in the image in
both cameras, at exactly the same f/stops, will be exactly the same.
Conversely, if you photograph with the 80 and 160 from exactly the same
distance and f/stop, the DOF in the 160 image will be considerable less
than the DOF in the 80 image. The 160 image size will be twice as large as
the 80 image as well.
One manufacturers 80mm lens has the same optical characteristics as any
other manufacturers 80mm lens. Or 100mm lenses, or 150mm lenses. It's not
a function of manufacturer or film size, it is the function of image size
on the film and f/stop. It is the law of optical physics.
An 80mm lens on a 35mm camera has the same DOF as an 80mm lens on a 6x6
camera which has the same DOF as an 80mm lens on a 4x5 camera.
The 6x6 and 4x5 80mm has more "coverage" than the 35mm format, but the
image size (the size of what it is you are photographing, as recorded on
the film) is identical, therefore the DOF is identical.
What f/stop you use and where you focus in your scene will determine how
much is in focus. From the exact point of focus, DOF extends 1/3 forward
(toward the camera) and 2/3 back (away from the camera). If you photograph
a persons face, at wide open (f/2.8 or f/3.5) and focus on their eyes,
there is a good chance that the end of their nose will be out of focus.
If, however, you focus on the tip of their nose, their eyes will be in
focus. Except for perhaps Pinocchio. As you stop down, more comes into
acceptible focus, increasing 1/3 front and 2/3 back.
Jim
From Leica Mailing List:
Date: Sun, 11 Feb 2001
From: "Jacques Bilinski" [email protected]
Subject: Re: [Leica] DOF: Erwin
> Erwin has a nice page on DOF issues:
> http://www.imx.nl/photosite/technical/DoF/DoF.html
>
> I quote out of context to pique your interest:
Erwin states:
So the calculated depth of field is eqaul on both sides. The welllknown
rule that the total length of permissible sharpness extends 1/3 in front
and
2/3 beyond the image plane/plane of focus is derived from the fact that
many
enlarged pictures are not viewed from the proper distance that is required
when the perspective of the lens and at the moment of the picture is the
same as that from the viewing distance. Many pictures are viewed from a
distance that is too short for the proper perspective.
So you are looking at this photograph. The depth of field is 1/3 in 2/3
out.
You now move the photograph away from your face to its proper viewing
distance for perspective. The depth of field is now 1/2 in 1/2 out. This
seems highly counter intuitive to me.
From Photography Teachers Mailing List:
Date: Mon, 12 Feb 2001
From: "Allen, Bob" [email protected]
Subject: RE: Teaching depth of field
Greetings.
I too find that many folks have a difficult time understanding depth of
field, aperture, shutter speed, exposure, and reciprocity. I'm pleased
that David opened up this line of discussion so we can see what each other
has come up with. I'm primarily a close-up shooter (bugs & flowers, I'm a
biologist not an artist) and to me this is the most important topic to
master. Keep in mind that my students are also biologists and that many
need high quality photos of their research species, so close-up
photography is vital. Thus the need for a firm grasp of dof.
Here is my contribution, from simple tricks to snazzy illustrations that
I've posted in the Files section at the PhotographyTeachers website at:
http://groups.yahoo.com/group/PhotographyTeachers/files/Presentations/Depth-of-Field/
Use whatever you may find helpful. I posted them as jpeg files, exported
from my original Adobe Illustrator files which are of higher quality and
file size. You may want to download all files, then print this note, and
use them together.
The files are:
dof.jpg - DOF examples, one-page handout
dofb.jpg - DOF examples, simplified version of dof.jpg
dof.ppt - DOF examples, with dofb.jpg on a Microsoft PowerPoint
presentation (requires PowerPoint 2001/98 for Mac or 2000/97 for Windows).
exp1.jpg - Reciprocity Calculator, example #1
exp2.jpg - Reciprocity Calculator, example #2
exp3.jpg - Reciprocity Calculator, example #3
exp4.jpg - Reciprocity Calculator, example #4
exp-recip.jpg - Reciprocity Calculator, with examples #1-4 and text.
Basically, I want to give them the concepts of 3D depth of vision and
convince them to *want* to learn to control it in order to make beautiful
images with their cameras. I am a strong believer in working from general
"black box" concepts then moving on to the details, not details first.
Your results may vary.
First, you must have already discussed what shutter speed and aperture
are. I have a handful of diaphragms and enlarging lenses that I pass out
to class and walk them through the apertures - wide to narrow, then back
again. Tactile learners need this, as so most other folks. Someone always
asks "why" apertures are important, so I do several demos. I've meant to
attach a bright flashlight to one of these things, but have yet to do so.
It would make a nice demo of light recording onto film. Someday I'll get
to it. Back to depth of field...
Without a camera, I ask them to look at a nearby object, focus their eyes,
and notice that the background is blurry and out of focus. Then I ask them
to keep their head still and focus on the background, noticing how the
foreground goes blurry and out of focus. This is quite easy using the old
Cub Scout "Floating Sausage" trick. With each hand, make a fist, then
extend your index finger as if to point at something. Hold your hands up
in front of your face with fingers pointing at each other, about 6 inches
in front of your eyes. Focus on the background right behind the fingers,
*not* the fingers. Your binocular vision crosses at this point, making it
appear as if there is a sausage (your fingers) floating in front of your
face. Now ask them to focus on their fingers and note how the background
goes blurry. Yes, you can do the same thing with a single finger held
straight up at arm's length, but the sausage trick is more fun. The point
is to make them understand that they can focus on one object and produce
an out-of-focus background, then refocus on the background and produce an
out-of-focus foreground. Tell them that cameras can do the same thing, but
not without their help.
Now introduce the idea of something being in focus from near them, to
farther away from them. I borrow a student for this one. Position the
student in front of the class, preferably with a distracting background at
a distance behind them. Point out that you have a camera and are taking a
portrait of that person. Ask what type of focus you should use: person in
focus, background in focus, or both person and background in focus? Ask
for predictions for what the photo will look like with each choice. Then I
give them a handout, Depth of Field and Aperture (posted in the Files
section), that illustrates this concept with sample distances.
We all look at the handout and discuss how the optical phenomenon of
narrowing of the opening works to deepen the depth of field. (I'm not
interested in the physics of this, nor do I discuss it). The handout has a
camera, apertures, subject, point of critical focus, near-limit dof,
far-limit dof, and infinity marked on it. I ask them to follow the
apertures on paper with the physical diaphragms and lenses. We discuss
this and I show slides that are paired: one with deep dof and another of
the same shot with narrow dof, then flip back and forth again to show
differences. Looking at the handout helps firm this up in their minds.
Now the question of how to control dof comes up. We discuss correct
exposure (a different topic, so not included here) and how it relates to
dof. From our demos and discussions comes the idea that wide open = less
dof, stopped down = more dof. Instead of a lengthy discussion of how dof
and exposure are related, I give them a handout, Correct Exposure &
Reciprocity (uploaded to the Files section.) The handout illustrates what
I call the Reciprocity Calculator (or "Reciprocity See-Saw"). I made a
large version of this with a file folder, cut-out text, and a brass tack.
There is a list of shutter speed values and a list of aperture values
arranged vertically, plus a two-headed arrow and a fulcrum. For now, tell
them to use 1/125 at f11. Place the arrow flat on the page, parallel with
the top or bottom of the paper. Place the fulcrum just below the center of
the arrow to simulate a see-saw. Then lay the shutter speed list on the
left side of the arrow, and the apertures on the right. Slide the shutter
speed list up or down so that the left arrowhead points at the example
shutter speed (1/125) and the aperture side up or down so that the right
arrowhead points at the example aperture (f11). From here on out, do not
move the lists (students want to tape them down, but don't let them).
This assumes that correct exposure is 1/125 at f11. But what if we want to
increase depth of field? From earlier discussion and demo, they know that
they need to stop down to a smaller aperture. Have them move the right
arrowhead upward so that it points at the next aperture value, f16. If the
center of their see-saw is held at the same point, they will see that the
left side (shutter speeds) now point to a slower shutter speed of 1/60.
Tell them that this is equivalent to the earlier exposure setting of 1/125
at f11 but that it gives them more dof. Do it again to get f22 at 1/30,
then go then return to the original exposure of 1/125 at f11 and go the
other way to get a faster shutter speed (1/250 at f8). I intend to video
this in action and post it as an mpeg file, but haven't gotten to it yet.
This is also a good time to extoll the virtues of a steady tripod in those
situations when they need dof and thus longer shutter speeds.
-Bob Allen
...
[Ed. note: thanks to Dave Mason for sharing these notes and
observations!]
Date: Sun, 1 Jul 2001
From: [email protected]
To: [email protected]
Subject: How much sharpness is enough?
The formula for computing hyperfocal distances offers surprising insights
into the choice of 35mm, medium or large film format.
Assume you need everything from 2 meters to infinity in focus. You have
a 35mm with 40mm lens, a 6x6 cm with 80mm and a 4x5 inch with 160mm lens.
Because of the 4:2:1 enlaragement ratio you can accept a .001" circle of
confusion on a 35mm negative, .002" on 6x6 or .004" on 4x5.
For sharpness from two meters to infinity you need a four meter
hyperfocal distance and must stop down to f/16 with 40mm on 35mm, f/32
with 80mm on 6x6, or f/64 with 160mm on 4x5. Assuming all three lenses
are then diffraction-limited, lens resolution would be in a 4:2:1 ratio
and becomes unity with enlargement to a constant print size! This
probably follows from the fact that the three lens apertures are all
identical, so a diffraction-limited lens of a given absolute size
resolves a constant amount of detail, no matter what size image is
projected.
If larger negatives still yield more satisfying end results, either (1)
depth of field requirements were not severe enough to require
diffraction-limited apertures, or (2) film was a limiting factor such
that better results were obtainable simply by projecting the same image
onto a larger piece of film.
- Dave Mason
[Ed. note: in my reply to Dave, I noted that too often, film is limiting
in real world contrast shots - http://www.smu.edu/~rmonagha/mf/film.html
for film lpmm resolution limits (often 50 lpmm or so for color film!).]
Date: 19 Aug 2000
From: [email protected] (Michael Gudzinowicz)
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Circle of confusion
"Q.G. de Bakker" [email protected] wrote:
>FrederickL wrote:
>
>> I think the acceptable circle of confusion will depend on the size print
>you
>> wish to make. If you are just making contact prints you can tolerate a
>larger
>> coc than you would if you were making mural enlargements.
>
>The size of print alone will tell you nothing at all. Viewing distance must
>be taken into account as well. A mural seen from 100 meters can get away
>with a much larger COC than a 8x10 viewed at arms length.
>So what you would need to consider is the angle of view, which is a function
>of both size and distance.
coc_in_mm = print_view-distance_in_mm / (print_magnification * factor)
factor:
1000 for just acceptable sharpness - like 35 mm 8x10 prints viewed from
10" using typical lens dof scales - around 4 lpmm in the print 2000 for
critically sharp results - "contact print" quality
Examples:
coc for an 8x10 "acceptable" print from 35 mm viewed at 250 mm =
250/(8 * 1000) = 1/32 mm
coc for critically sharp 8x10 print from 4x5 viewed at 250 mm =
250/(2 * 2000) = 1/16 mm
coc for critically sharp 30x40 print from 8x10 viewed at 1000 mm =
1000/(4 * 2000) = 1/8 mm
coc for critically sharp 30x40 print from 8x10 viewed at 250 mm =
250/(4 * 2000) = 1/32 mm
coc for critically sharp 100x150 print from 35 mm viewed at 5,000 mm =
5,000/(100 * 2000) = 1/40 mm
From Nikon MF Mailing List:
Date: Thu, 26 Jul 2001
From: Rick Housh [email protected]
Subject: DOF technical/hyperfocus-"t-stop"/MTF
Riccardo Polini wrote:
>Just to conclude, I would like to inform you that the Italian magazine
>TuttiFotografi (which performs MTF tests in their own lab) usually gives,
>besides MTF charts, a kind of DOF when the lenses are focused at infinity
>and at each aperture. The DOF is the minimum distance a subject will be
>focused at when the focus ring is set at infinity.
Which brings to mind another slightly different concept, the hyperfocal
distance for a lens at a given aperture. The hyperfocal distance is the
focus point, at which anything between half the distance to that focus
point and infinity is in focus. I can't recall this has been mentioned yet
in this discussion. This is very useful for a situation where you must
fire off a lot of shots without time to focus precisely.
You can do it with a DOF chart, but it is especially easy if you have
useful DOF markings on the lens. Set the infinity mark on the focusing
scale opposite the DOF mark for the furthest distance for the f-stop in
use, rather than on the dot used for sharpest focus. Then the number on
the focusing scale opposite the DOF mark for the closest distance for the
f-stop in use will be the nearest distance which will be in
focus. Everything between infinity and that nearest point should be in
focus, and the "sharpest focus" dot then points at the hyperfocal
distance.
For example, on the 35mm f/2.0 lens I happen to have on my desk, if I
align the infinity mark with the yellow mark representing the far distance
of focus at f/11, the near distance yellow mark points to about 6.5 feet
and the point of sharpest focus point at about 13 feet. In other words,
at F/11 the DOF of this lens will allow any object between infinity and
6.5 feet to be in acceptable focus when the lens is focused at 13 feet.
I use this hyperfocal principle all the time.
If you use a chart for this purpose, be careful. The theoretical.
calculated depth-of-field at any given aperture may not match the markings
on the lens' focus scale, which can be because the light transmission loss
of any particular lens may affect its effective aperture, and the
manufacturers adjust the f-stop markings on the lens to compensate. In
other words, the f-stop markings on the lens may not match the theoretical
calculations for that lens. This effect is minor with most lenses, but
some, such as "mirror lenses", can have a transmission factor as low as
about 75%, so their effective f-stop will not be the same as the
calculated f-stop. This, compensated, f-stop (called "t-stop" in some
quarters) may accurately represent the light value, but the depth-of-field
calculation is still based on the theoretical aperture.
Riccardo, do you think the results of the TuttiFotografi depth-of-field
results are affected by the fact that the effective f-stops for different
lenses are adjusted to account for transmission factors, as well as the
effect of MTF's? I couldn't tell from what you said.
With respect to MTF charts, here is an interesting article criticizing the
use of MTF charts to represent a lens's resolution and "sharpness",
because of interpretive differences. The author argues that the views of
experienced, professional lens users may be of more value than MTF values.
He points out that many manufacturer's MTF charts are based on designs
that are so critical the manufacturers cannot achieve the design goals in
actual manufacturing practice, and a less ambitious MTF design goal may
actually result in better practical lenses. He points out how Zeiss bases
their MTF charts on actual measurements of production lenses, as
TuttiFotografi seems to. Still, it depends on the measuring technique.
Interesting stuff, most of the proofs beyond my mathematical grasp:
http://www.normankoren.com/Tutorials/MTF2.html#Zeiss
- Rick Housh -
From Nikon MF Mailing List;
Date: Thu, 26 Jul 2001
From: [email protected]
Subject: Re: DOF technical/hyperfocus-"t-stop"/MTF
Rick Housh [email protected] wrote:
> Riccardo, do you think the results of the TuttiFotografi depth-of-field
> results are affected by the fact that the effective f-stops for different
> lenses are adjusted to account for transmission factors, as well as the
> effect of MTF's? I couldn't tell from what you said.
I think they calculate DOF from MTF charts. Probably it's not meaningful
per se. The numbers they obtain are just indicative of a different quality
of lenses.
Anyway, I too "firmly" believe that a keen, experienced photographer (not
necessarily a professional) is more reliable than a MTF chart.
Moreover, mathematical formulas giving DOF as a function of aperture,
focal length, subject-to-film distance, reproduction ratio and so on DO
NOT (and cannot) take into account the "quality" of a lens. So I think
that lenses can exhibit different DOF, despite the same aperture, focal
length, etc... It's interesting to note what Bjorn Rorslett writes in his
Web site about the different DOF behaviour of AI 25-50/4 and AF 20-35/2.8
D in landscape photography. The former is less sharp in the focus plane
but is sharper in the out-of-focus area. I used both the lenses to take
landscape pictures and I confirm that what Bjorn noted is TRUE! He prefers
the DOF behaviour of the 25-50/4 and I couldn't agree more. In conclusion,
the same aperture, focal length and subject-to-distance value DO NOT
ensure the same quality of the out-of-focus areas, i.e. what I have
defined as DOF behaviour.
Regards
Riccardo
Date: Mon, 2 Jul 2001
From: "annqlee" [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: LF DOF
John,
I have written a Javascript
http://carcassi.eng.uci.edu/Eportfolio/technical/diffraction/diffraction.htm
If you have Internet Explorer, I have also just written a Javascript to
compare views between formats.
http://carcassi.eng.uci.edu/Eportfolio/film.htm
Ann
[Ed. note: long sold by the time you read this, but cited here as a
reference work]
From Nikon mailing list;
Date: Tue, 31 Jul 2001
From: John Wall [email protected]
Subject: Depth of Field: The standard book
All this duscussion of depth of field prompts me to say that I have a
copy of the classic book on DOF that I would like to sell.
Alfred Blaker's classic APPLIED DEPTH OF FIELD (1985), is in soft cover
but spiral bound. In 280 pages, Blaker tells you everything you ever
wanted to know about depth of field. There are loads of diagrams and
formulas. There are 150 pages of tables showing DOF for different lenses
in "Near-Object Photography" and high magnification photography.
This is the standard reference in the field. It covers all formats -- 35
mm, MF, and view cameras. It covers everything. There is far more
information here than I will ever need. My copy is in Mint condition
(one slightly bent corner on the front cover). This is a hard-to-find
but classic item. Amazon.com lists this book for $60.00 + shipping (if
they have it) and I will send my copy to you postage paid for $50.00.
Or make me a reasonable offer!
Best,
John Wall
(NikonJohn)
From Nikon MF Mailing List;
Date: Tue, 31 Jul 2001
From: "shepherdlen" [email protected]
Subject: Re: Depth of Field: The standard book
From: John Wall [email protected]
snipped
>I would like to sell.
> Alfred Blaker's classic APPLIED DEPTH OF FIELD (1985),
> There are 150 pages of tables showing DOF for different lenses
> in "Near-Object Photography" and high magnification photography.
Hi,
IMO there is only one problem with this book.
The depth of field formulae for magnified images on page 263 makes no
allowance for effective focal length or effective aperture being changed
by extension.
I have checked out the formulae given by Kodak and many others, both
mathematically and in field tests and found this book to use a formulae
which does not agree with the others and gives wrong answers.
The book also ignores that from the early 1970's many lenses became
asymmetrical and even "correct" close up formulae has to be adjusted for
the pupilary exit factor.
The tables look good - it is such a pity many of them give wrong
answers.
Len Shepherd.
From Nikon MF Mailing List;
Date: Sun, 5 Aug 2001
From: "Jack Jansen" [email protected]
Subject: RE: Enhanced DOF
My optometrist has been doing this with film and an MF camera for quite a
while. He described an example of increasing dynamic range and depth of
field in the same picture: the foreground of the scene (in Hawaii) was in
shadow while farther out the scene was brightly lit. He exposed a frame
focused for the foreground, with the lens relatively open, and exposed a
second frame focused for farther out with the lens relatively stopped
down.
He merged the two pictures with Photoshop. He also uses a technique
for increasing resolution: he takes a four- or five-frame panorama and
merges the frames with Photoshop so that the effective resolution is four
or five times greater than would be obtained with the same MF camera and
film with the necessary greater enlargement.
He showed me an example: his brother is on the opposite side of their
hotrod from him, leaning with his forearms on the fender; the hood is up
and the chrome and brass hardware in the engine compartment is gleaming.
While his brother remained still, the optometrist shot four verticals
across the scene. He merged them with Photoshop and made a 20x30
horizontal with which I was very much impressed.
Jack J
...
From Nikon MF Mailing List;
Date: Mon, 6 Aug 2001
From: "Hansen, Lars Holst" [email protected]
Subject: RE: Re: Depth of Field
See also http://www.luminous-landscape.com/blended_exposures.htm
Best regards,
Lars
Date: 6 Aug 2001
From: [email protected] (John Stafford)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: hyperfocal/DOF is useless in MF cuz' Re: Depth of Field
[email protected] (Robert Monaghan) wrote
> the problem with all this theoretical DOF discussion is that many medium
> format cameras don't keep film truly flat at the film plane, [...]
Which reminds me of our mutual interest in Veriwide cameras, Bob. I
had the Plaubel for some time, and now use the Brooks. The later has a
'longer' back, which lays the film quite flat. And reverting to an
earlier thread where someone criticized the Hasselblad for its
occassional 'wrinkled' (wavy) film - indeed, tis true sometimes,
especially when a frame is left in place for a day or so before
shooting so just-in-case, I waste that frame and use the next. All in
all, however, I rarely see anyone shooting wide-open in MF where the
imperfect film flatness becomes most obvious.
As an aside, has anyone compared their camera's DOF (or hyperfocal)
indicator (printed on the lens) with various hyperfocal DOF charts? It
might be interesesting to see if some manufacturers use pessimistic or
optimistic metrics.
Date: Thu, 9 Aug 2001
From: [email protected] (Mark Anderson)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: different mfgers CoC and DOF ranges 1/70th to 1/200th inch..
Robert Monaghan wrote:
> from http://www.smu.edu/~rmonagha/brondof.html from noted DOF expert
> Michael Davis (of DOF spreadsheet fame) - quote:
>
> Quoting page 131 of "Basic Photographic Materiels and Processes" by
> Stroebel, Compton, Current, and Zakia (c1990 Focal Press): "... A study
> involving a small sample of cameras designed for advanced amateurs and
> professional photographers revealed that values ranging from 1/70 to 1/200
> inch were used -- approximately a 3:1 ratio."
The scales on my 35 mm camera lenses (Canon) are based on a 1/1000"
COC. The scale on my 6x9 cm folder (Zeiss) is based on a 1/250" COC.
--
Mark Anderson
DBA Riparia www.teleport.com/~andermar/
Date: Sun, 29 Jul 2001
From: [email protected] (Mark Anderson)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Depth of Field - 35mm vs MF
[email protected] wrote:
> No. A given focal length will have the same depth of field in any format so
> a 180mm will give you the same DOF in medium format as a 180mm in 35mm.
> Of course you will have a wider angle of view with medium format.
Yes, but! If you're defining DOF by a certain circle of confusion on
the negative, you're right. But what matters is acceptable sharpness on
the FINAL PRINT! Since an 8x10 print is less of an enlargement from
medium format, than it is from 35 mm., you can get the same degree of
sharpness, at a specific print size (covering a specific field of view),
using a larger circle of confusion in a larger format, than you'd use
with a smaller format. So.... what matters is final image size, not
focal length.
The circle is usually defined as 1/1000" (apx. .025 mm) in the 35 mm
format. The DOF markers on my 105 mm Tessar on my 6x9 folder are based
on COC of 1/250", and 1/100-1/200" may be used for LF photography.
So, you've got a choice, make prints from medium and large format no
bigger than from 35 mm, and experience equivalent DOF on the print using
equivalent F.L.'s (based on coverage) in different formats, OR... notice
that the DOF is "less" with the longer lenses used with larger formats
because you're now making bigger prints, wherein you need to hold the
COC smaller to maintain an acceptable level of sharpness.
--
Mark Anderson
DBA Riparia www.teleport.com/~andermar/
Date: Wed, 26 Sep 2001
From: ralph fuerbringer <[email protected]>
To: [email protected]
Subject: Re: Shift for Scheimpflug? Yes or No?
the statement that you gain two stops in light (dept of field) with a lens
1/2 the focal length is in error.It's much better. as maitani the designer
of the olympys om 1-4 states in his book: a 50mm lens at f2 has the same
depth of field as 100mm lens at f8. dont bother to argue the point. it is
fact. i didnt believe it myself but if you have a 50 and 100 lens in your
system with distance scales you can verify it. just like the fact that a 2x
telextenter loses two stops in transmission it retains the dept of field of
the original lens, a 4 stop gain in dept of field over using a lens of twice
the focal lenght. unfortunately if you want the brokeh of the longer lenses
you will be disappointed and confused without this info. i lost bets on both
these facts. ralph
Date: Thu, 04 Oct 2001
Subject: Re: SV: [HUG] Hasselblad 50mm cf focus accuracy
From: [email protected]>
To: [email protected]>
Hi,
Not unusual with W/A lenses in my experience, but there is an explanation
(kind of...).
Contrary to popular belief, the tolerance for distance discrepancies BEHIND
a lens is smaller the shorter that lens is (with some assumptions that I'll
skip here). Since depth of field is greater for shorter lenses, this seems
counterintuitive, but can be inferred from the much smaller pitch on the
focussing helicoid on a w/a lens. Try refocussing from infinity to, say, 2
m on a 50, an 80, and a 250, and you see that the front of the lens moves
forward much more on the longer lenses.
Thus, a small, fixed error may go unnoticed with most of your lenses, but
show up only with your shortest one; the important thing is that the
focussing screen and the film agree, not that the scale, or infinity stop
does!
What you can do is a simple test: with a tripod, and with your shortest lens
wide open, take a few pics of a distant, detailed building or tree. Focus
with the screen, take your first pic. Now turn to the infinity stop, note
the difference on the focussing scale, take another pic, and then turn the
focus past true focus on the screen, and the same difference you just noted
away from infinity (you may need to mark the rings with some tape and a
fine-tipped marker). After processing, look at the film with a strong
loupe: if the first exposure is sharpest, and the other two are about
similar to each other, then all is well.
Good luck
Per
> From [email protected]>
> Datum: Thu, 4 Oct 2001
> Till: [email protected]>
> =C4mne: SV: [HUG] Hasselblad 50mm cf focus accuracy
>=20
> Hi Chris
> I have noticed the same thing.
> ....
> My other
> lenses (80, 150 and 250) with or without 2x converter focuses infinity
> correctly. And so did the old 500 I got rid of some years ago. How the SWC
> behaves is a little more tricky to tell but as far as I have seen with a
> screen on it it appears to focus infinity at the inf. spot. That is not an
> ideal distance for that lens but that is another topic............
> Ulf
> ----- Ursprungligt meddelande -----
> Fr=E5n: "Christopher Gonzaga" [email protected]>
>> I just got back my recollimated 50mm lens from Panorama Camera Center NY. I
>> took the lens up to the rooftop to test it for infinity focus and it still
>> says that buildings that are 1-2 miles away are in focus before infinity.
>> Is this normal. My 80mm lens focusses the same buildings at infinity.
>> ....
>> Do others who have
>> the same lens have the same experience?
>> Thanks,
>> Chris
From: "Peter Klosky" [email protected]>
To: [email protected]>
Subject: Re: SV: [HUG] Hasselblad 50mm cf focus accuracy
Date: Thu, 4 Oct 2001
> the important thing is that the focussing screen and the film agree, not that the scale, or infinity stop does ...
My experience and technique does not agree with this statement. When I use my 50mm lens, I do not screen focus. I focus by scale.
What I am doing is photos of people. I can estimate that 2m is waist up, 3m is full length, 5m is a group, etc. This way, I can
work quickly in a dark room, without error-prone hunting that screen focusing offers.
To me, saying that the lens scale being off has no impact is similar to what I heard from a technician with a local firm once at a
camera show, regarding a 35mm body with/meter: "It doesn't matter if the meter reading you get with the camera is correct, as long
as the actual shutter speeds are longer/shorter than marked, as it relates to the amount that the metering is off. You'll get the
correct exposure on the film, if the two are off in such a way as to self-cancel." The obvious problem with these lines of thinking
is that many of us expect to be able to take a meter reading with a hand-held meter or measure the focus distance using estimation
or a tape measure, and are upset to hear of service folks that refuse to properly line up each piece, to read properly and be fully
interchangeable. Field techniques in practice and bench theories do not always agree, and this is two of these cases. This is
probably a reason that Ansel Adam's well-received technique suggests calibrating many of the aspects of the meter, camera, film and
development for actual results, instead of relying on calibrations that may be off.
Peter
From: [email protected]
Subject: Re: [Rollei] slight OT, DOF in 6x6, for a 80 vs. 35mm/50
To: [email protected]
Date: Fri, 9 Nov 2001
> I may be crazy, but it seems to be there's significantly more shooting a
> 50 mm lens on a 35 mm camera at about four feet than there is shooting
> the Rollei...I think part of the shock is due to my tending to forget
> that an 80 is an 80, even if it is the "standard" lens...
The question of compared depth-of-field (DOF) issues in MF-6X6 vs.
35mm is often debated and the conclusions are not always clear. In the
RUG-FAQ I have put some pointers to detailed web articles trying to
give the most comprehensive overview of this DOF question, namely
All you need to know about Depth Of Field on Bob Wheeler's web site
from the most simple...
http://www.bobwheeler.com/photo/DOFExpo2.pdf
...to the most complex including shift and tilt
http://www.bobwheeler.com/photo/ViewCam.pdf
When you say "an 80 is an 80" you are right in terms of focal length,
but as pointed by Mark Rabiner you should also take into account the
final magnification of the image and negligible influence of grain in
MF.
So, yes, common experience shows that you get much less depth of field
in MF vs. 35mm, (arguing against would be not sound reasonable
;-);-)), but the comparison may not be fair taking into account the
global increase in final printed image quality.
Now a well-know (but definitely non-intuitive) rule of macro-
photography states that if you compare (this holds even for far
distant objects!) images taken with *any* focal length, the depth of
field for a given f/stop will be the same as long as the
*magnification* between the objet and the image remains the same.
So the "loss" in DOF with a 80 (whatever 80, in MF or not) vs. a 50
when shooting, say, at 4 feet is quantitatively explained according to
this rule by the different object/image magnification, being 8/5
bigger with the 80. But, yes, who cares for the explanation ?;-);-)
less DOF, this is what you'll get!!!
--
Emmanuel BIGLER
[email protected]>
From: [email protected]
Subject: Re: [Rollei] slight OT, DOF in 6x6, for a 80 vs. 35mm/50
To: [email protected]
Date: Fri, 9 Nov 2001
E.B.:
> > So the "loss" in DOF ...is qualitatively explained according to
> > this rule by the different object/image magnification, being 8/5
> > bigger with the 80....
Siu Fai:
> A rule of thumb I have read is to stop down accordingly. For example DOF of
> 2.8/80 in 6x6 is similar to 1.4/50 in 35mm. If you're used to use f5.6 with
> 35mm (4 stops) then you'll need f11 in 6x6.
Yes the rule is easy to apply in practise, and the explanation is
actually so simple that I can't resist writing it here ;-);-)
Assume that you want to keep the hyperfocal distance with your 80mm
lens in MF identical to the one you are used to with a 50mm in 135
film. Since (for reasonably far-distant objects) the hyperfocal
distance "H" determines the far- and near- limits of sharpness for a
given object-lens distance, if "H" is the same in both cases then you
are all set.
A well-known formula defines H as H=f*f/(N*c) where f=focal length,
N=the f-stop Number, and c=the diameter of the circle of least
confusion. So you can rewrite the formula as H=(f/c)*(f/N).
Assume that you scale the reference value of c proportionally to the
film format size or focal length (for example c=30 microns in 135,
c=30*8/5 ~= 50 microns in 6x6, both reasonable values, close to what
is considered as standard) so f/c does not change ; now if you want H
to be the same you simply need (f/N) to be the same in both cases. For
a 8/5 ratio this yields N(6x6)=(8/5)*N(135)=1.6*N(135). So if I start
from a f/1.4-50mm lens in 135, I have to stop down to 1.4*1.6=2.2,
i.e. f/2.2 with a 80mm in 6x6, or 1.5 f/stop, slightly less stringent
than Siu's "2.8" (2 f-stops) according to his "rule of thumb".
--
Emmanuel BIGLER
[email protected]>
From: [email protected]
Date: Fri, 9 Nov 2001
Subject: [Rollei] DOF Indices
To: [email protected]
At the recent photo show at the Javits Center in NYC I engaged a
knowledgeable Swede at the H********* counter in a discussion of DOF and the
index markings on their lenses.
He declared that the markings are theoretical, not very precise and overly
generous. You're not going to get what the indices indicate in your
photograph.
Use the indices for the NEXT WIDER APERTURE than you are setting for your
exposure. OR do as he always does: use the indices for that f-stop which is
2 stops wider than you are using for your exposure in order to get all you
expect/need in focus.
I would infer that the same applies to Rollei lens markings as well.
Vincent L. Gookin
[email protected]
From: [email protected]
Subject: Re: [Rollei] DOF Indices, on R-TLR and other lenses
To: [email protected]
Date: Mon, 12 Nov 2001
From Vincent L. Gookin, about DOF scales:
> ...I engaged a knowledgeable Swede at the H********* counter in a
> discussion of DOF and the index markings on their lenses.
> He declared that the markings are theoretical, not very precise
Definitely, I would not say that. The engravings are as precise as any
good analog scale. And something theoretical can be as precise as you
want if the model is correct and input values are correct !!! But
don't forget that, whatever model you are using, "garbage in, garbage
out." ;-)
For all serious lens manufacturers, DOF scales engravings are
*precisely* designed according to the most simple law : H=3Df*f/(N*c),
and this model is not questionable except for macro work.
**BUT**:
> ...and overly generous. You're not going to get what the indices
> indicate in your photograph.
... you have to decide on a realistic value for "c" !!! On the H******
planar CF 100, the engravings are coherent with an hyperfocal value H
= 6 metres @ f/22. So if you extract the value of the corresponding
circle of confusion "c", you eventually find
*** c = 75 microns *** !!!!
i.e. an "acceptable" sharpness limit of about 6-7 line pairs/mm !!! for
a lens capable of "passing" 100 lp/mm in good conditions, "overly
generous" is an understatement...
> Use the indices for the NEXT WIDER APERTURE than you are setting for
> your exposure. OR do as he always does: use the indices for that
> f-stop which is 2 stops wider than you are using for your exposure
> in order to get all you expect/need in focus.
My understanding is that c=3D75 microns is simply not a realistic value
in practice!!!. Changing the reading by 2 f-stops means dividing the
value of "c" by 2, i.e taking c=3D37.5 microns. This sounds more
coherent with actual performance of Z**-H** lenses, capable on the
whole MF frame of performances in sharpness quite similar to a good
35mm lens on 35mm format. I've also checked on a 50mm Z**-H** distagon
C (with the Synchro-Compur, moving red DOF indices) and found a circle
"c" of about 50 microns (H=3D2,2m @f/22) identical to the one Rollei
uses on R-TLRs (see below).
Other values for those interested
lens focal length H(metres)@f/22 c=3Df*f/(H*N) (in microns)
H********* (read from W***i's book)
planar 80 4.9 60 microns
on a 35mm Voigtl=E4nder Bessamatic
skoparex 35 1.8 31 microns
septon 50 3.6 31 microns
super-dynarex 135 18. 46 microns
on a Rollei 35 SE
sonnar 40 3. 24 microns
> I would infer that the same applies to Rollei lens markings as well.
Rolleiflex T
tessar 75 5. 51 microns
---------------------------------
A first remark is that a great fantasy seems to be behind the choice
of the "c" value.
Another remark is that the DOF scale on a R-TLR focus knob is
reasonable, being based on a circle c=3D50 microns. However this means
10lp/mm, and for a lens capable of passing a minimum of 60-80 lp/mm (I
hope I do not open a passionat RUG thread here ;-);-), again a
sharpness criterion of 10 lp/mm is, at least, "generous".
Good'ol Voigtl=E4nder lenses are based on a reasonable c=3D30 microns for
35mm format ; except for the telephoto : probably they cheated a
little, did not want to frighten the users with an "impossible to use"
lens "afflicted" by excessively narrow DOF limits... ;-);-)
Now the strange 75 microns for the H**** planar CF 100 is for me a
real puzzle in contradiction with a legend I used to believe in (until
today ;-) "Z*** used to design its lens DOF scales with tighter
sharpness tolerances tha other manufacturers". A last remark to be
confirmed is that manufacturers seem to increase the value of "c" for
longer focal lengths even if used for the same image format.
Well eventually most (not all) modern manufacturers of 35mm auto-focus
lenses have definitely solved the question : they ignore the formula
H=3Df*f/(N*c) and do not engrave anything at all... not even a distance
scale. So the customer gets what he deserves and cannot complain being
"cheated" on non-existent DOF or distance engravings... ;-);-)
--
Emmanuel BIGLER
[email protected]>
Date: Tue, 13 Nov 2001
To: [email protected]
From: Richard Knoppow [email protected]>
Subject: Re: [Rollei] DOF Indices, on R-TLR and other lenses
you wrote:
> From Vincent L. Gookin, about DOF scales:
>
>> ...I engaged a knowledgeable Swede at the H********* counter in a
>> discussion of DOF and the index markings on their lenses.
>
>> He declared that the markings are theoretical, not very precise
>
>Definitely, I would not say that. The engravings are as precise as any
>good analog scale. And something theoretical can be as precise as you
>want if the model is correct and input values are correct !!! But
>don't forget that, whatever model you are using, "garbage in, garbage
>out." ;-)
>
>For all serious lens manufacturers, DOF scales engravings are
>*precisely* designed according to the most simple law : H=3Df*f/(N*c),
>and this model is not questionable except for macro work.
>
>**BUT**:
>
>> ...and overly generous. You're not going to get what the indices
>> indicate in your photograph.
>
>... you have to decide on a realistic value for "c" !!! On the H******
>planar CF 100, the engravings are coherent with an hyperfocal value H
>=3D 6 metres @ f/22. So if you extract the value of the corresponding
>circle of confusion "c", you eventually find
>
>*** c 75 microns *** !!!!
>
>i.e. an "acceptable" sharpness limit of about 6-7 line pairs/mm !!! for
>a lens capable of "passing" 100 lp/mm in good conditions, "overly
>generous" is an understatement...
>
>> Use the indices for the NEXT WIDER APERTURE than you are setting for
>> your exposure. OR do as he always does: use the indices for that
>> f-stop which is 2 stops wider than you are using for your exposure
>> in order to get all you expect/need in focus.=20
>
>My understanding is that c=75 microns is simply not a realistic value
>in practice!!!. Changing the reading by 2 f-stops means dividing the
>value of "c" by 2, i.e taking c=37.5 microns. This sounds more
>coherent with actual performance of Z**-H** lenses, capable on the
>whole MF frame of performances in sharpness quite similar to a good
>35mm lens on 35mm format. I've also checked on a 50mm Z**-H** distagon
>C (with the Synchro-Compur, moving red DOF indices) and found a circle
>"c" of about 50 microns (H=2,2m @f/22) identical to the one Rollei
>uses on R-TLRs (see below).
>
>
>Other values for those interested=20
>
>lens focal length H(metres)@f/22 c=3Df*f/(H*N) (in microns)
>
>H********* (read from W***i's book)
>planar 80 4.9 60 microns
>on a 35mm Voigtlaender Bessamatic
>skoparex 35 1.8 31 microns
>septon 50 3.6 31 microns
>super-dynarex 135 18. 46 microns
>on a Rollei 35 SE
>sonnar 40 3. 24 microns
>
>> I would infer that the same applies to Rollei lens markings as well.
>Rolleiflex T
>tessar 75 5. 51 microns
>
>---------------------------------
>
>A first remark is that a great fantasy seems to be behind the choice
>of the "c" value.
>
>Another remark is that the DOF scale on a R-TLR focus knob is
>reasonable, being based on a circle c=50 microns. However this means
>10lp/mm, and for a lens capable of passing a minimum of 60-80 lp/mm (I
>hope I do not open a passionat RUG thread here ;-);-), again a
>sharpness criterion of 10 lp/mm is, at least, "generous".
>
>Good'ol Voigtl=E4nder lenses are based on a reasonable c=30 microns for
>35mm format ; except for the telephoto : probably they cheated a
>little, did not want to frighten the users with an "impossible to use"
>lens "afflicted" by excessively narrow DOF limits... ;-);-)
>
>Now the strange 75 microns for the H**** planar CF 100 is for me a
>real puzzle in contradiction with a legend I used to believe in (until
>today ;-) "Z*** used to design its lens DOF scales with tighter
>sharpness tolerances tha other manufacturers". A last remark to be
>confirmed is that manufacturers seem to increase the value of "c" for
>longer focal lengths even if used for the same image format.
>
>Well eventually most (not all) modern manufacturers of 35mm auto-focus
>lenses have definitely solved the question : they ignore the formula
>H=3Df*f/(N*c) and do not engrave anything at all... not even a distance
>scale. So the customer gets what he deserves and cannot complain being
>"cheated" on non-existent DOF or distance engravings... ;-);-)
>
>--
>Emmanuel BIGLER
>[email protected]>
>
FWIW, Kodak states that the depth of field charts published for their
large format lenses is based on a circle of confusion equal to 2 min of arc
or approximately 1/1750 of the focal length.
I don't know what criteria was used for the depth of field scales on
their focusing lens. I rather think it depended on the application and
assumed magnification of the final image. Kingslake may say something about
this in his old book on photgraphic lenses.
In any case depth of field is an optical illusion and can be calulated in
a couple of ways depending on whether one chooses a fixed circle of
confusion (as makes sense for a camera with interchangible lenses, or a
fraction of the focal length. DOF seems to generate more confusion than
almost any other subject in photography probably because it can be defined
in more than one way.
----
Richard Knoppow
Los Angeles, CA, USA
[email protected]
From: [email protected]
Subject: Re: [Rollei] OT DOF indices, circle = 1/1750
To: [email protected]
Date: Tue, 13 Nov 2001
From Richard Knoppow:
> FWIW, Kodak states that the depth of field charts published for
> their large format lenses is based on a circle of confusion equal to
> 2 min of arc or approximately 1/1750 of the focal length.
I like the simplicity and (apparent) universality of this definition
for the acceptable circle of confusion for DOF scales very much, since
it is exactly twice the angular resolution limit of the average naked
eye (1 minute of arc), and thus *potentially* as universal and less
questionable as was the definition of the metre (equal to 1/40,000,000
of the Earth circumference, blue, white, red flag again: personally
instead of 1/1750 I would have chosen 1/1789 ;-);)
More seriously, it means that if the final image is observed at the
"standard" viewing distance (i.e a viewing distance equal to the focal
length times the enlarging factor) we'll get exactly 4 "eye-pixels"
per "image-pixel" in the zones of the image where the acceptable
limits of sharpness are reached, hopefully a small fraction of the
whole image only, since we usually want the main subject to be the
sharpest.
Probably the "1/1750-th" rule should apply only to "standard" lenses
covering a "standard" angle of about 54 degrees; wide angle or
telephoto lenses requiring probably a more detailed analysis.
So let's put this definition to work: here is what we get
focal length circle c=f/1750 | focal length circle c=f/1750
f(mm) (in microns) | f(mm) (in microns)
35 20 | 75 43
40 23 | 100 57
50 29 | 150 86
The values of c=24 microns for the Rollei 35 (f=40) and c=30 microns
for the Bessamatic (f=50, and many others 35 mm cameras) fit very well
within this "1/1750-th" rule. The value of c=50 microns for the R-TLR
is slightly larger than a "suggested" c=43, but clearly a value of
c=75 microns as chosen on some Z**-H** CF lenses, surprisingly, does
not match the rule at all.
-- Emmanuel BIGLER [email protected]>
Date: Tue, 13 Nov 2001
To: [email protected]
From: Richard Knoppow [email protected]>
Subject: Re: [Rollei] Re: DOF Indices, on R-TLR and other lenses
you wrote:
>Richard Knoppow wrote:
>
>> FWIW, Kodak states that the depth of field charts published for their
>> large format lenses is based on a circle of confusion equal to 2 min of arc
>> or approximately 1/1750 of the focal length.
>> I don't know what criteria was used for the depth of field scales on
>> their focusing lens. I rather think it depended on the application and
>> assumed magnification of the final image. Kingslake may say something about
>> this in his old book on photgraphic lenses.
>
>
>Just to put the Kodak value in context, the DOF tables in my Rollei TLR
>manuals are calculated on f/1400 for 66 and f/2000 for Rolleikin, so the
>Kodak tolerances are stringent indeed...
>
>
>
>Eric Goldstein
>
>
Here is some more from _Lenses in Photography_ 1st ed., Rudolph
Kingslake, 1951, Garden City Books
He gives the following values for Kodak depth of field markings or charts
on various types of lenses:
8-mm cine 1/2000" (0.012mm)
16-mm cine 1/1000" (0.025mm)
24mm x36mm 1/500" (0.50mm)
Small folding cameras 1/200" (0.127mm)
Large cameras, critical 2min of arc (f/1700)
Large cameras, liberal f/1000
The discussion in the text indicates these are based on assumed typical
image magnification and viewing conditions. Kingslake has a very full
discussion of depth of field in this book with illustrations. He points out
that for cameras, like motion picture cameras, with interchangible lenses,
the depth of field must be calculated from a constant circle of confusion
rather than from one which is a fraction of the focal length.
There are some funny constants with depth of field. For instance, where
pictures are taken from the same viewpoint with two different focal length
lenses, and the images magnified to the same size, the DOF is constant with
the physical size of the diaphragm opening. That is, a half inch hole,
which is f/48 for a 12" lens and f/4 for a 2" lens, will have the same
depth of field when the image from the 2" lens is magnified to be the same
size as the other.
When the image size is adjusted to be identical by moving the viewpoint
the DOF will be constant with f/stop. i.e., if I place set up a camera with
a 12" lens and a camera with a 2" lens so that the actual images are the
same size the 12" lens will obviously be much further away from the
subject. At the same f/stop the depth of field of the two will be the same.
There was a second edition of Kingslake's book. I think I have it but
have a lot of books stored away. Both editions are reasonably easy to find
used and worth it. Very clear explanations of optics applying to cameras
intended for working photographers.
----
Richard Knoppow
Los Angeles, CA, USA
[email protected]
From: "Leonard Evens" [email protected]>
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Depth of Field - 35mm vs MF
Date: Mon, 26 Nov 2001
"Lear"
[email protected]> wrote:
* On Sat, 24 Nov 2001
* "Michael K. Davis"
> [email protected]> wrote:
>
>>Math rocks!
>
> I haven't follow the thread, but I believe that an Xmm DOF lens is the
> same at all formats, so a 50mm lens will have the same DOF regardless of
> format, although coverage will be different (as long as the image circle
> is big enough). so its logical that a 50mm lens in 35 and a 80mm in 66
> will have more or less the same angle of view, but since the latter is
> almost double the length of the former, it will have far less DOF, and
> that is the apparent �less DOF�in MF.
>
> Am I wrong? or is all that math needed to understand this simple idea?
I'm sure others who hate math will disagree, but the math really is
necessary. Moreover, your analysis doesn't avoid math, it just uses it
less precisely or perhaps even incorrectly. How in the world can one
discuss the relation of focal length, which is a number, to depth of field,
another number, without math? You are just trying to minimize the
amount of math in order to (over) simplify. For each kind of
problem, you have to use the math which is appropriate to that problem,
no more and certainly no less. Do you avoid the use of math to balance
your checking account? Ordinary arithmetic is sufficient, but if you
precluded the use of subtraction, you would only make it harder.
Calculus on the other hand is not (usually) necessary to handle one's
financial affairs.
Let me try my standard lecture on depth of field.
First the term depth of field is a vague imprecise term unless you
specify what you mean. There are of course formulas for what will be in
focus when you focus at a specific distance and specify a specific
criterion for sharpness (the circle of confusion). But these hardly
avoid the use of math. A simpler approach is to look at the hyperfocal
distance H. This is the distance such that when you focus at it,
everything from infinity down to half that distance will be in focus.
If you focus at another distance, there are simple formulas for what will
be in focus which depend on the hyperfocal distance. I include those
below in an appendix, but don't look if it offends you.
The point about the hyperfocal distance H is that the smaller it is, the
more will be in focus at any given distance and the larger it is the less
will be in focus. If you want a lot of depth of field, you need a small
hyperfocal distance.
How does the hyperfocal distance relate to focal length and format?
Here is the formula I find most useful. It applies to relatively
distant objects. Let f be the focal length in mm, E the degree of
enlargement for an acceptable 8 x 10 print, and N the f-number. U will
be a fudge factor which includes your personal specification for
UNsharpness. In my case, using fairly standard assumptions, U = 77.4
if I want the hyperfocal distance in feet, and U = 254 if I want it in
meters. You might want a SMALLER number if you are more demanding.
Then the hyperfocal distance is given by
H = f^2 * E/(N * U)
(This can be in either feet or meters by an appropriate change of scale
incorporated in U.)
To use this with a calculator, enter the focal length in mm, press the
squaring key, multiply it by the enlargement factor, divide the result by
the f-number and then divide that result by the UNsharpness factor.
The important point here is that as you increase the degree of
enlargement, if you want to maintain the same angle of view, you have to
decrease the focal length to compensate. If you keep the product
constant, you don't change the angle of view for the final image.
Thus, an 80mm lens with an
enlargement factor of 4 would be equivalent to a 40 mm lens with an
enlargement factor of 8. Enlargement factors are determined by the
format of the negative frame but also the amount of cropping. The
enlargement factor of course is relative to the final size image. I've
taken it to be 8 x 10. (But see appendix 2 below for more on that.)
Notice however that there is still one more factor of f, the focal
length involved. That means if you change the enlargement factor and
the focal length so as to maintain angle of view, the hyperfocal distance
changes by the ratio of the focal lengths. So doubling focal length
and correspondingly halving focal length doubles the hyperfocal distance.
Roughly you could say then that it halves the depth of field, although
precise estimates requires the formulas given in the appendix 1.
Note this really depends on the math. The formula could have been
H = f * E/(N * U)
where the factor f was not squared. In that case, the hyperfocal
distance would remain unchanged.
I don't see any way to understand this without the formula or one
equivalent to it. Words alone can't substitute for math to specify
precise numerical relationships.
Note also that the formula
H = f^2 * E/(N * U)
also tells you what happens to the hyperfocal distance if you keep the
same amount of enlargement (i.e., fix the format) but change the focal
length. Here there is a much more dramatic change because H depends on
the square of the focal length. Thus for 35mm enlarged the same amount
doubling the focal length increases H by a factor of four and decreases
the depth of field accordingly.
Appendix 1.
Let H be the hyperfocal
distance and let D be the distance at which you focus. For sufficiently
distant objects
Near focus = H*D/(H+D)
Far focus = H*D/(H - D) (at least if D is less than H. When D >= H, the far
focus is at infinity.
Appendix 2.
What should you do if you want a print larger than 8 x 10. I chose
8 x 10 because if one views a print at the closest comfortable distance
(for a normal eye), taken to be 10-12 inches, 8 x 10 is a reasonable
size. Larger prints would normally be viewed from correspondingly
further away, and in principle although they will be enlarged more from
the original negative, depth of field calculations for an 8 x 10 print
should still apply if we assume the greater viewing distance compensates
for the extra enlargment. One could get this from the formula by saying
that if you double the viewing distance, you should be less demanding of
what is in the print so you double U. You also have to double the
amount of enlargement E, but because E is in the numerator and U is in
the denominator, the two factors cancel. If on the other hand, you are
not going to view a print twice as large quite as uncritically, you may
want to make a smaller change in U, so in total there still will be an
increase in H, and you will have overall less depth of field.
> Diego K:
--
Leonard Evens [email protected] 847-491-5537
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
From [email protected]
Date: Tue, 27 Nov 2001
Newsgroups: rec.photo.equipment.medium-format
Subject: why DOF isn't all it should be ;-) Re: Depth of Field
I don't think the DOF math calculations are wrong, just less accurate and
useful than they at first appear. The reasons for this include:
a) if you want to see what the DOF is going to be, hit the stop down DOF
lever and take a look (give your eyes some seconds to adapt if darker..)
besides being direct and more visual than cerebral, you will also catch
all those poles and branches coming out of somebody's head ;-)
b) for most med fmt lenses, which are fixed focal lengths in 99.8%, the
vast majority have mfgers DOF markings. No need to do the math. While
these DOF markings make various assumptions, a bit of experience will
determine if they are liberal or conservative, and you can adjust (I
usually use one stop more conservative markings to be "on safe" side
c) as per past notes on film buckling and flatness, in med fmt you have
more problems with many cameras having " " problems, so you can't rely on
the film plane being precisely where the calculations assume it is. If the
subject at infinity is really important (landscapes..) then using
hyperfocal or DOF settings may result in softer than useful infinity
imagery due to film buckling.
d) all this assumes you know what the magnification factors and hence CoC
and other elements will be. But if I shoot a really great sunset, maybe I
will splurge and have it blow up to 30x40". At that point, in a small
office, I will be most annoyed by the out of focus elements at infinity if
I was using the standard DOF calculations and values for an expected
enlargement to 11x14 or 16x20. So the CoC would have to be varied to
adjust for enlargement size, and not knowing that for sure beforehand, it
generally isn't factored in or allowed for.
e) the math for DOF shifts as you move into macro ranges (1/3rd:2/3rds to
1/2:1/2 etc), and a lot of my shots are of flowers and stuff inbetween ;-)
f) when you do ultraviolet or infrared film, what happens to the DOF? The
shifts in the focal plane are sometimes surprising, depending on distance
g) many lenses have some degrees of curvature, in some cases a whole lot
of field curvature, which can make assumptions of a flat focal plane in
DOF calculations a real source of error and soft focusing defects...
in short, there are lots of reasons why most of us don't use DOF as much
as it might seem we ought to do so, and it isn't because of the math (I'm
an engineer) but the variability and realities of general shooting
experiences...
regards bobm
--
From: "Leonard Evens" [email protected]>
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: why DOF isn't all it should be ;-) Re: Depth of Field
Date: Wed, 28 Nov 2001
"Robert Monaghan" [email protected]> wrote:
> I don't think the DOF math calculations are wrong, just less accurate
> and useful than they at first appear. The reasons for this include:
Of course, that is the case. I keep forgetting to add that caveat
whenever I respond to one of these postings. The formulas are only a
starting point. Usually, what I'm trying to do is make the point that
the formulas do tell you something specific that can't be described in
less explicit terms.
>
> a) if you want to see what the DOF is going to be, hit the stop down DOF
> lever and take a look (give your eyes some seconds to adapt if darker..)
> besides being direct and more visual than cerebral, you will also catch
> all those poles and branches coming out of somebody's head ;-)
I can do that much better now that I've had cataract surgery. But
with my medium format Horseman Technical Camera where I focus on a ground
glass, I sometimes find it hard to see exactly what is there when stopped
way down. It would probably be easier with a larger format camera, but
with a relatively small dim image, one can be surprised when one sees the
final print
>
> b) for most med fmt lenses, which are fixed focal lengths in 99.8%, the
> vast majority have mfgers DOF markings.
My Horseman lenses don't. My old Rolleiflex does, but the markings are
based on assumptions about degree of enlargement and a circle of
confusion that don't seem to apply to me.
> No need to do the math. While
> these DOF markings make various assumptions, a bit of experience will
> determine if they are liberal or conservative, and you can adjust (I
> usually use one stop more conservative markings to be "on safe" side
I'm probably unusual, even for a mathematician, that even at 68 I can do
rough arithmetic in my head. I learned the hyperfocal distances that
apply for the lenses I use, and I can home pretty close to estimating the
near and far focus points by mental calculation. But I admit most
people would need a calculator to find H*D/(H+D) and H*D(H-D).
>
> c) as per past notes on film buckling and flatness, in med fmt you have
> more problems with many cameras having " " problems, so you can't rely
> on the film plane being precisely where the calculations assume it is.
> If the subject at infinity is really important (landscapes..) then using
> hyperfocal or DOF settings may result in softer than useful infinity
> imagery due to film buckling.
I guess I've been lucky with the cameras I use, and it hasn't seemed too
much of a problem. But I did discover that my old Rollei doesn't focus
at infinity when set at infinity. I had to tape my own scale to the
knob.
>
> d) all this assumes you know what the magnification factors and hence
> CoC and other elements will be. But if I shoot a really great sunset,
> maybe I will splurge and have it blow up to 30x40". At that point, in a
> small office, I will be most annoyed by the out of focus elements at
> infinity if I was using the standard DOF calculations and values for an
> expected enlargement to 11x14 or 16x20. So the CoC would have to be
> varied to adjust for enlargement size, and not knowing that for sure
> beforehand, it generally isn't factored in or allowed for.
Agreed.
>
> e) the math for DOF shifts as you move into macro ranges (1/3rd:2/3rds
> to 1/2:1/2 etc), and a lot of my shots are of flowers and stuff
> inbetween ;-)
For closeups I would assume the math was even more important. The
formulas are more complicated and one certainly can't rely on scales.
Certainly one would want to rely a lot on what one saw on the ground
glass.
>
> f) when you do ultraviolet or infrared film, what happens to the DOF?
> The shifts in the focal plane are sometimes surprising, depending on
> distance
I don't know about that, but certainly using one's eyes would be
pointless. There should be detailed information available somewhere
about how to calculate it in such cases since the optics is basically the
same. But presumably one also has to experiment to check that one has
got it right.
>
> g) many lenses have some degrees of curvature, in some cases a whole lot
> of field curvature, which can make assumptions of a flat focal plane in
> DOF calculations a real source of error and soft focusing defects...
Agreed.
>
> in short, there are lots of reasons why most of us don't use DOF as much
> as it might seem we ought to do so, and it isn't because of the math
> (I'm an engineer) but the variability and realities of general shooting
> experiences...
Again, I say that the calculations are just a starting point, and they
have to be adjusted for a variety of reasons based on visual evidence,
experience with one's own equipment, and judgement.
But the usual questions about this involve simplified assumptions about
depth of field, which are just wrong. The usual misconception, for
example, is that depth of field doesn't really depend in any serious way
on focal length, so if one decreases the focal length propotionately to
the format size, keeping the same angle of view, and same final image
size, then one will get the same depth of field. That is patently
false, as anyone who has used a typical digital camera should know,
but it seems some people believe it.
>
> regards bobm
--
Leonard Evens [email protected] 847-491-5537
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
From: [email protected] (Robert Monaghan)
Newsgroups: rec.photo.equipment.medium-format
Subject: focusing tricks Re: why DOF isn't all it should be ;-)
Date: 28 Nov 2001
Here's a possible trick that really helps me, originally from Roger Hicks
of Medium Format Handbook etc. fame; find some of the bar codes on a
bright white/black plastic packaging and cut out. Put on subject and use
to focus. This is as high as contrast can get, so this is as easy as
focusing is going to get too ;-) You can get different sizes, put a bit of
double sided tape on them. The hard part is to remember to take 'em off
before taking the shot - or folks wonder why that tree there in the middle
has a bar code on it???? ;-) ;-)
very handy, esp. if using a split image rangefinder with med fmt and
non-moving subjects (like trees, sculpture)...
for macro, there is also a trick using some mineral oil in the central
spot area (reduces contrast but lots brighter) that some folks like (I
don't) - from an old 1960s pop photography article IIRC and some later
cites...
a fresnel screen is also most helpful, esp with wide angles, see
http://people.smu.edu/rmonagha/bronfresnel.html (cheap too, $2 at Home
Depot)
see also mf/vision.html for some other possible solutions/options
regards bobm
From: "Jim Crawford" [email protected]>
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Hyperfocal distance for a Mamiya TLR
Date: Sat, 8 Sep 2001
Hello folks,
I am late getting to this thread, but would like to throw in my two
cents (or I would like to think my two sense). If the calculation of
the hyperfocal distance is so complicated that even a college math
professor has trouble with it, what chance do we mortals have to get it
right.
In the original post, Andy-J asked if there was a simple way to
determine the hyperfocal distance for his TLR at f8. I would like to
offer a little more practical approach than has been offered - at least
to me it seems so. We spend thousands of dollars on equipment and then
ask questions like what should the hyperfocal distance of a 55mm lens be
at f8.
Kodak did and I guess they still do, publish a small book that measures
about 6X9 inches, costs less than $20, and is cram packed with
information including depth of field for all focal length lenses for all
different camera formats (including MF). It has more info on filters,
flash, fill-in flash, close up exposure, film data, and on and on than
you may ever want to know. It is called the Kodak Professional
Photoguide and can be found at most quality camera stores. Using this
guide you can easily look at one of the three DoF calculators and see
that for 6X6 MF, 50mm lens @ f8, the range of focus (acceptable
sharpness) is from slightly less than 4 feet to infinity and the
hyperfocal distance is approximately 8 feet. For a 55mm lense, one has
to interpolate a little and I estimate the range to be from 5.5 ft to
infinity w/ a hyperfocal of 11 feet. Then, if your purpose is critical,
close down one more stop, adjust the shutter speed appropriately (for
proper exposure), and shoot.
I think this is so much more practical than multiplying the circle of
confusion times the aperture times the conversion factor from inches to
millimeters and back to feet, etc. Folks, Kodak has worked all of this
out for you. A world of photo info is available for less than a fourth
of the price of a decent filter. It is an invaluable tool.
Now maybe I am being overly simplistic, but if your purpose is to take
or make photographs, then use the guide and you will do just fine. If,
however, you are doing lense testing, the professor's approach may be
the better way.
I am not trying to step on anyone's toes here and if I am out of line,
please let me know. But as I read the questions in this and other
groups, I am amazed at not the questions that are asked, but that people
don't take advantage of the excellent, inexpensive resources available
to them. Maybe they just do not know that this information is
available.
I have no ties to Kodak, but have carried this little "encyclopedia" in
my camera bags since I bought it in 1981. If you don't have one, get
one.
I apologize for going on about this for so long, but this photoguide is
so handy that I don't know how any serious photographer would get along
without it. Thanks for listening.
Jim Crawford
Dallas, TX (actually Garland)
"Leonard Evens" [email protected]> wrote
> There has been some discussion of how to set a Mamiya TLR at the
> hyperfocal distance. Since I contributed some misleading advice, I would
> like to give what I hope is some more accurate advice.
>
> If my calculations are correct, when the lens is focused at the
> hyperfocal distance, the displacement of the lens from its infinity
> setting is given very closely by
> h ~ cN
> where c is the diameter of the circle of confusion and N is the
> f-number. (Note this is independent of the focal length.) For
> example, for c about .0025 inches, and f/8, this is about 0.5 mm.
> So for larger apertures, it is going to be very difficult to set the lens
> at the hyperfocal distance accurately using a distance scale unless the
> camera has a gearing mechanism for relating object distance scale to lens
> movement. My Rolleiflex 2.8F does have such a gearing mechism, so it is
> not too hard to set the distance scale at about 42 feet, which is the
> hypefocal distance for an 80 mm lens at f/8. My Mamiya C-3 has a
> distance scale on the side which is in one to one ratio to the actual
> lens movement, so while I can set it very roughly to 42 feet, it is much
> harder to get it at all close by using the distance scale. The only
> practical method to focus at the hyperfocal distance is to choose some
> object at that distance and focus as carefully as possible on it. One
> might use for example an auxilliary rangefinder device for estimating
> distances.
>
> Since with a TLR, you can't stop down the focusing lens, there is
> unfortunately no way to check the depth of field visually at the f-stop
> being used for the exposure.
>
> One thing to keep in mind however, is that focusing at the hyperfocal
> distance is only crucial if the near limit of what is in focus is
> specially critical. If you focus at infinity, everything from infinity
> down to the hyperfocal distance will be in focus. As you move the focus
> down from infinity to the hyperfocal distance, you move the near focus
> down to half the hyperfocal distance. So a rule of thumb might be
> something like this---at least for the Mamiya TLR. Pick the diameter c of
> the CoC that you want to use. Pick the f-number N. Focus at infinity and
> then move back some very small amount that you are sure the lens moves
> less than cN. Finally stop down one stop more for safety.
>
> If one uses this idea and does some tests, it should be possible to
> control focus using hyperfocal technique in this way. It may even be
> possible to put marks on the distance scale for a few standard shooting
> situations, decided upon after testing.
> But given the delicate movements, it is going to be very hard
> to use such markings consistently.
> --
>
> Leonard Evens [email protected] 847-491-5537
> Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
From: eric boxall ericb@abc_fenris99.demon.co.uk>
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Depth of Field - 35mm vs MF
Date: Thu, 29 Nov 2001
Leonard Evens [email protected]> writes
>"someone"
>[email protected]> wrote:
>
[Snipped ...]
>In principle, that would be about right if one made 8 x 10 prints in both
>cases with minimal cropping and one used a 50 mm lens for 35 mm and an 80
>mm lens for 6 x 6. (But the angle of view in the 6 x 6 case would be
>slightly larger.)
>
>But other factors, such as degree of film flatness, could affect the
>result. Medium format film tends not be held as flat as 35 mm film.
>
>One might want to stop down 1 1/2 to 2 stops just to be sure, but of
>course there is nothing like testing one's own equipment against one's
>own standards to be sure.
Sorry Leonard but I just can't go along with that.
If the film isn't flat in the gate, stopping down will make things
worse, not better. This happens because the image-side NA increases as
the object-side NA diminishes. Stopping down reduces the object-side NA.
A further point is object distance; if object distance is large (approx.
infinity) the Depth of Focus becomes very small and film flatness is
very critical ... Depth of Field may well be large but Depth of Focus is
approaching zero. Conversely, for near object distances, Depth of Field
is small ... Depth of Focus becomes progressively less critical.
Two working examples of these phenomena which I have actually seen ...
[1] Zeiss aerial-survey cameras (which work at or near the infinity
focus setting at about f=8 to f=11 and consequently have a very small
object-side NA ... probably less than 0.10) have a vacuum-and-glass
pressure plate system to hold the film precisely in correct register.
[2] Zeiss Photomics II and III (using objectives with a very high
object-side NA of 1.3-1.4) did not have a pressure plate of any sort and
lead the film through the gate emulsion outwards, producing a variation
in flatness across the gate of about 1 to 2mm along the length of the
35mm frame.
I raised this matter with the experts when I was in Oberkocken and got
the explanation which I have paraphrased above.
Best regards,
--
Eric
Email to:beric@fenris99(dot)demon(dot)co(dot)uk
From: "Leonard Evens" [email protected]>
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Depth of Field - 35mm vs MF
Date: Thu, 29 Nov 2001
"eric boxall"
ericb@abc_fenris99.demon.co.uk> wrote:
> Leonard Evens [email protected]> writes
>>One might want to stop down 1 1/2 to 2 stops just to be sure, but of
>>course there is nothing like testing one's own equipment against one's
>>own standards to be sure.
>
>
> Sorry Leonard but I just can't go along with that.
>
> If the film isn't flat in the gate, stopping down will make things
> worse, not better. This happens because the image-side NA increases as
> the object-side NA diminishes. Stopping down reduces the object-side
NA.
Clearly, I have something to learn. One reason I post opinions is to
be corrected. :-)
What is NA?
>
> A further point is object distance; if object distance is large (approx.
> infinity) the Depth of Focus becomes very small and film flatness is
> very critical ... Depth of Field may well be large but Depth of Focus is
> approaching zero. Conversely, for near object distances, Depth of Field
> is small ... Depth of Focus becomes progressively less critical.
My references, e.g. Photographic Optics by Arthur Cox, say that depth of
focus, for distant objects is proportional to the f-number and
independent of focal length.
The actual formula that Cox gives is unclear in my copy because there is
a typo, but I believe it is
C * N * v/ f
where C is the diameter of the circle of confusion, N is the f-number, v
is the lens to image distance, and f is the focal length. For distant
objects, v/f is very close to 1. I do see that depth of focus increases
for close objects, but I don't see why it approaches zero for distant
objects. Perhaps it has something to do with the fact that one can't
use the same considerations to account for depth of field and depth of
focus at the same time? I also don't see why stopping down decreases
depth of focus. Could you explain in more detail?
>
> Two working examples of these phenomena which I have actually seen ...
>
> [1] Zeiss aerial-survey cameras (which work at or near the infinity
> focus setting at about f=8 to f=11 and consequently have a very small
> object-side NA ... probably less than 0.10) have a vacuum-and-glass
> pressure plate system to hold the film precisely in correct register.
>
> [2] Zeiss Photomics II and III (using objectives with a very high
> object-side NA of 1.3-1.4) did not have a pressure plate of any sort and
> lead the film through the gate emulsion outwards, producing a variation
> in flatness across the gate of about 1 to 2mm along the length of the
> 35mm frame.
1-2 mm variation across a 35 mm frame? A rather generous circle of
confusion for 35 mm might yield C = 0.03 mm,
which at f/11 would yield a depth
of focus of .33 mm. So 1-2 mm variation in the film plane would mean
total garbage. Did I calculate something wrong?
>
> I raised this matter with the experts when I was in Oberkocken and got
> the explanation which I have paraphrased above.
>
> Best regards,
--
Leonard Evens [email protected] 847-491-5537
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
From: "Michael K. Davis" [email protected]>
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: different mfgers CoC and DOF ranges 1/70th to 1/200th inch..
Date: 11 Aug 2001
Hi Bob!
I sincerely appreciate your having called me a "DoF expert" - In my not so
humble opinion, I thought I was alone in that assessment! Just kidding...
Anyway, I thought I would jump in to offer the link to the spreadsheet,
graciously hosted by Jon Grepstad of Norway. He just updated the link
yesterday to my latest version, v9.7.2
http://home.online.no/~gjon/mdofcal9.xls
Even if one doesn't like the idea of using printed DoF tables in the
field (I don't), it's quite easy to use them to determine how many f-stops
adjustment you must offset each camera lens DoF scale to achieve the
results you desire, as shown on the tables you've calculated and never
printed.
I can't resist contributing to this thread, since the whole subject is on
how to select the CoC diameter you desire.
My approach is copiously documented in the Notes tabs of mdofcal9.xls, but
I'm compelled to summarize it here. ALL of the following variables should
be considered when selecting (calculating) the maximum permissible CoC
diameter:
Focal length
Format diagonal (after cropping)
Enlargement factor (to reach your desired print size)
System resolution (combined resolving power of lens and film)
Desired resolution in the final print
Viewing distance (how closely will it your print be scrutinized)
Max CoC diameter is the only variable in a DoF calculation that allows YOU
control over the results, so why not do a little homework to see what
value will suit your needs? You CAN trust the math. It works and the
spreadsheet referenced above does the math for you.
Once you calculate tables for the a target print size you like to produce,
that will give satisfactory results at a stated viewing distance, you can
just examine each lens' engraved DoF scale to see what offset is
necessary, in stops, to deliver the same DoF as the tables require. Note
this for use in the field. Once in the field, you accomodate the
possibility that viewing distances will be greater or less than that.
For example, if you decide in advance of shooting an image that you want
the final print to survive a viewing distance that's only half that
specified when calculating the tables, and your DoF table (or DoF scale
offset) indicates a hyperfocal distance of 9 feet, with a corresponding
near sharp of 4.5 feet, one solution would be to refocus at 18 feet and
make sure the closest subject is no closer than 9 feet. This will shrink
your CoC's to half their original diameter, at the near sharp and at
infinity.
Another example, would be a situation where you simply aren't able to get
the DoF you need, in the field, at the time of exposure. If the shutter
speed required does not permit an aperture that would yield the DoF you
need, don't resign yourself to recording unwanted subject motion, just
resign yourself to increasing the ratio of print diagonal to viewing
distance. Print an 11x14 and hang it over the piano, instead of printing
a 24x30 and hanging it in the hallway. The point is, these compromises
can be considered BEFORE exposure, to yield better results than would be
had when it's too late to refocus.
This combination of starting with a solid, reliable DoF reference and
modifying it on the fly gives you complete control over DoF. It CAN be
managed. It CAN be trusted.
For me, the most valuable fruit of this method has been my knowing
precisely when to squelch the quest for depth of field. I consider the
wholesale use of minimum apertures to be one of the biggest mistakes made
in landscape photography. The majority of SF and MF lenses deliver their
best resolution near f/8 or f/11, but many landscapes are shot at the
smallest aperture that camera and subject motion will permit, for sheer
ignorance of how much DoF is enough. The smaller formats are especially
vulnerable to diffraction and it's a shame to see it introduced when a
wider aperture and corresponding faster shutter speed would have yielded
sufficient depth of field.
Mike Davis
From: "Leonard Evens" [email protected]>
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: different mfgers CoC and DOF ranges 1/70th to 1/200th inch..
Date: Sat, 11 Aug 2001
"Michael K. Davis"
[email protected]> wrote:
> Hi Bob!
> I sincerely appreciate your having called me a "DoF expert" - In my not
> so humble opinion, I thought I was alone in that assessment! Just
> kidding... Anyway, I thought I would jump in to offer the link to the
> spreadsheet, graciously hosted by Jon Grepstad of Norway. He just
> updated the link yesterday to my latest version, v9.7.2
> http://home.online.no/~gjon/mdofcal9.xls
I had not looked at this worksheet before because I usually running Linux
and I doubted if the spreadsheets I had available would be able to make full
use of the worksheet. But I just downloaded it and I was surprised to
see that the basic calculator appears to be all there. But all the notes
and examples are missing. I suppose I will have to fire up Windows and
look at it under Excel. :-( :-( But it does look rather impressive.
Of course math whizes like me can just carry the formulas around in our
heads and with a few relevant paramters make the calculations on the fly.
:-) :-). Seriously, though, I tried that a couple of days ago with my
digital camera, just to estimate the hyperfocal distance, and I was
surprised that I didn't come out a couple of orders of magnitude off.
Popular misconceptions aside, mathematicians are not specially good at
mental arithmetic.
As I said, your tool does look rather impressive. Even if one doesn't use it
regularly as a matter of course, just working out a few examples will
lead one to be much more aware of the different factors affecting depth
of field.
--
Leonard Evens [email protected] 847-491-5537
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
From: "Leonard Evens" [email protected]>
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Depth of Field - 35mm vs MF
Date: Sat, 01 Dec 2001
"Q.G. de Bakker"
[email protected]> wrote:
> Brian Ellis wrote:
>
>> The usual standard for circle of confusion size is about 1/100 to 1/200
>
> 1/100 what?
>
>> in
>> the final print. So you could work backwards from there based on the
> degree
>> of enlargement from negative to print (e.g. a 10x enlargement will
>> require
> a
>> circle of confusion of 1/1000 to 1/2000). . Of course the circle of
>> confusion changes as the lens aperture changes - smaller aperture
>> equals smaller circle of confusion, which is why you read that depth of
>> field increases as aperture size decreases - so any depth of field
>> numbers you work out are going to be valid only for one particular
>> aperture anyhow.
>
> What you forgot to mention was that what size the CoC may be allowed to
> be not only depends on degree of enlargement, or aperture size, but also
> on viewing distance.
> If there ever can be a hard and fast criterion, it must be the average
> (there goes being hard and fast) angular resolution of the human eye.
> Details in a print will not appear to be unsharp (which isn't
> necessarily the same as appearing to be sharp) when they are beyond the
> limit of what the eye can see.
>
>
>
As I noted elsewhere, what is considered a maximum acceptable circle of
confusion in the final print does not depend on the aperture. Whether
or not a given point in the image will produce something smaller than
that does depend on the aperture.
As you say, what is considered acceptable in the final print depends on
the viewing conditions, the visual accuity of the viewer, and probably
other considerations. One thing I would like information about is the
extent to which larger prints viewed at correpsondingly greater distances
are equivalent to smaller prints viewed closer up. If one consideres
only angular displacement at the eye, detail in a 16 x 20 viewed at
20-24 inches should look exactly the same as the same detail in an 8 x 10
viewed at 10-12 inches. But in practice, people seem to demand more in
a 16 x 20 even if it is viewed from further away. For example, one
might say that a certain combination of film, format, and lens can
produce an acceptable 8 x 10 but not an acceptable 16 x 20. Can anyone
enlighten me as to the additional factors that lead to such evaluations?
Personally, I hang my 16 x 20 prints on walls, and when I view them, I do
see pretty much what I saw in correpsonding 8 x 10s viewed closer up. Of
course, it I get close, or to be completely obsessive, start looking at
them with a loupe, I can see differences, but I don't think that should
be the criterion for what is acceptable.
--
Leonard Evens [email protected] 847-491-5537
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
From: "Leonard Evens" [email protected]>
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Effect of Format on Depth of Field
Date: Mon, 03 Dec 2001
"Peter Caplow"
[email protected]> wrote:
> One frequently hears that larger formats have less depth of field, due
> to the longer focal length of equivalent lenses.
>
> Suppose one took identical pictures, first with a 35mm camera, and then,
> with an 8x10 view camera, using lenses of equivalent focal lengths, as
> well as identical film, apertures and shutter speeds. If the resulting
> negatives were both printed as 16x20's and compared side by side, would
> the depth of field in the picture taken by the 35mm camera be
> substantially greater, or would it be the same, since the cameras were
> the same distance from the subject?
The brief answer is yes.
Here is the longer answer. First the formats of a 35 mm and an 8 x 10
view camera are different. The first has an aspect ratio of 1.5 and the
second an aspect ratio of( 1.25. So let's assume you crop the 35 mm
image so you only use 24 x 30 mm. That means to produce an 8 x 10
print, you need to enlarge by 8*25.4/24 = 8.47 times. Let's assume you
take the picture from the same point and you have exactly the same part
of the subject in the final picture. To do that you will have to divide
the focal length you use with the 8 x 10 view camera by the enlargement
factor or 8.47. (This assumes the subjects are distant. It could get a
bit more complicated for close up objects.)
Now here is the crucial factor. The depth of field is very roughly
inversely proportional to the degree of enlargement (assuming the same
criterion for sharpness in the final print) and inversely proportional to
the SQUARE of the focal length. Multiplying the first by 8.47 and
dividing the second by 8.47 cancels only part of the effect of the focal
length. So roughly speaking, by dividing the focal length by 8.47, you
increase the depth of field by that amount. Or reversing the roles, the
8 x 10 view camera has about 1/8.47th the depth of field of the 35 mm
camera with these assumptions. You can compensate by multiplying the
f-number by 8.47.
These calculations actually apply to the hyperfocal distance. The actual
depth of field at any focused distance can be calculated from the
hyperfocal distance by simple formulas.
Important caveats!
There are a variety of other factors which can effect perception of
sharpness. One obvious one is grain. An 8 x 10 contact print will be
virtually grain free. Also, diffraction can make everything in the
image fuzzy, but at the same f-stop, the 8 x 10 lens should be less
subject to diffraction limits. Another important issue is film flatness. The
term depth of focus refers to the degree of departure from flatness which
is acceptable. For distant objects this is independent of focal length
and inversely proportional to degree of enlargement needed to produce the
final print. But the calculations of depth of field and depth of focus
are based on the same geometric considerations, so the two quantities are
not independent of one another. Generally, it is easier to keep 35 mm
film flat then to keep 8 x 10 film flat. I have to admit I haven't
worked out myself how the
different factors contribute to the final result, but others
have commented that lack of film flatness can easily invalidate
theoretical calcuations of depth of field. Finally, all the
calculations of depth of field are based on the assumption that the lens
has no field curvature. This is clearly not true for any real lenses,
so the actual results could depend significantly
> Peter Caplow
>
Leonard Evens [email protected] 847-491-5537
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
From: [email protected] (Michael Gudzinowicz)
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Effect of Format on Depth of Field
Date: 3 Dec 2001
Peter Caplow [email protected]>
> One frequently hears that larger formats have less depth of field, due
> to the longer focal length of equivalent lenses.
>
> Suppose one took identical pictures, first with a 35mm camera, and then,
> with an 8x10 view camera, using lenses of equivalent focal lengths, as
> well as identical film, apertures and shutter speeds. If the resulting
> negatives were both printed as 16x20s and compared side by side, would
> the depth of field in the picture taken by the 35mm camera be
> substantially greater, or would it be the same, since the cameras were
> the same distance from the subject?
The quick answer is that 8x10 depth of field would be much less than 35 mm
at the same aperture, view angle, and print size. To get equivalent depth
of field with 8x10 under those circumstances, the aperture required would
be approximately 8X that for 35 mm (35 mm f/8 = 8x10 f/64).
In the following discussion, I'm assuming that film isn't a factor,
but it may be in the case you've constructed since the MTF of film
at the 64 lpmm required for the 35 mm print is much lower than that
at 8 lpmm required for the same print from 8x10.
If you check the formulas for depth of field, you'll observe that for the
same depth of field, the hyperfocal distance (H) must be the same.
One formula is:
H = (focal_length)^2 / (f/stop x circle_of_confusion)
If one compares the 8" and 24 mm dimensions of the 8x10 and 35 mm formats,
the proportion is approximately 8 to 1. For the same view angle on 16x20
prints, 8x10 would require a lens with 8X the focal length of 35 mm.
The enlargment from 35 mm would be 16X, and if one wants to have 4 lpmm in
the print (just acceptable), then 64 lpmm are needed in the negative. That
translates to a circle of confusion of 1/64 mm or 0.015625 mm. For 8x10, the magnification
is 2X, and circle of confusion required 0.125 mm.
You can compare the f/stops required by rearranging the equation above,
and taking the ratio. H is the same for both (same depth of field). The
focal length for 8x10 is 8X that for 35 mm, but the term is squared. The
circle of confusion permitted for 8x10 is 8X that required for 35 mm.
f/stop_8x10 (focal_length_8x10)^2 / (H x circle_of_confusion_8x10)
----------- = ----------------------------------------------------
f/stop_35mm (focal_length_35mm)^2 / (H x circle_of_confusion_35mm)
or substituting for the 8x10 terms one has:
f/stop_8x10 (8 x focal_length_35mm)^2 / (H x 8 x circle_of_confusion_35mm)
----------- = ----------------------------------------------------
f/stop_35mm (focal_length_35mm)^2 / (H x circle_of_confusion_35mm)
which, for the same H reduces to:
f/stop_8x10
----------- = 8
f/stop_35mm
For 4x5, the ratio is 4, and for medium format (6x7), is is around 2.
Therefore, for the same view angle and print size:
35 mm f/8 = MF f/16 = 4x5 f/32 = 8x10 f/64
From: "Q.G. de Bakker" [email protected]>
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: when infinity isn't infinite ;-) Re: Depth of Field - 35mm vs MF
Date: Thu, 6 Dec 2001
Robert Monaghan wrote:
> I have to agree with Leonard (what, again?) that infinity isn't at
> infinity for most camera lenses; the design and mfgers usually use an
> arbitrary value between 100 and 200 times the lens focal length as an
> approximation of infinity, depending on the lens design and focal length.
Lens manufactureres use collimators to set infinity on lenses. So, depending
on the quality of their collimator, infinity should really be at infinity.
> second, [...]
> third, [...]
> and finally [...]
And beyond finally, there are other factors to consider. Like temperature
effects, affecting both lens (that's why so many long lenses do not have an
infinity stop) and body (!!!) focus lengths. And the effect of stopping down
on focal length.
Both effects vary per lenstype and build, and therefore can not be accounted
for in simple formulae. And both can do enough to play havoc with what
formulae and scales would have you expect.
So if you want to know for sure (as far as that is possible) you can't rely
on mathematics, but only use visual inspection, with the lens stopped down
to the taking aperture. Not easy, no.
From: "Michael K. Davis" [email protected]>
Newsgroups: rec.photo.equipment.large-format
Subject: Ann's formula vs. Michael's (was: Confusion over CoC)
Date: Thu, 13 Dec 2001
Ann's formula vs. Michael's (was: Confusion over CoC)
Hello to Peter, Ann, Richard and Michael, who all participated in this
thread I am resurrecting from April of this year.
Let me recap the four-post thread as briefly as possible, embedding my
comments and questions:
******
From: Peter De Smidt ([email protected])
Subject: Confusion over circle of confusion
Newsgroups: rec.photo.equipment.large-format
Date: 2001-04-17
>
>I'm trying to figure out the hyperfocal distance for my various lenses
>and f-stops, the lenses being 135, 203 and 300mm, and the f-stops being
>f22, f32, f45. I would like to use a value for the circle of confusion
>that would give me a visually sharp 16x20 inch print from a 4x5 negative.
[snip]
>Peter
*******
From: annqlee ([email protected])
Subject: Re: Confusion over circle of confusion
Newsgroups: rec.photo.equipment.large-format
Date: 2001-04-17
>
>Hi Peter,
>
>If I want DOF I never use anything wider than f22. I even use f22 when I
>focus near infinity. I would not hesitate to use f45 if I were to enlarge
>4x or even 5x. You test should be at which fstop is the diffraction
>unacceptable to you. It is my belief that you will find that f45 is still
>very good. I don't even use those formulas anymore, if it is kinda far I
>use f32. If it is really far I use f22.
>
>Here are some tests:
>http://carcassi.eng.uci.edu/Eportfolio/technical/diffraction/diffraction.htm
>
>Good Luck,
>
>Ann
Ann,
I had a look at your calculator, at the address you posted above. Nice
job, but...
Supplying the following values:
Print Magnification Factor: 4.21
Focal Length: 43
F-Stop: 22
Circle of Confusion on Film: 0.03115
Object Distance: 2620
It calculated the following Output:
Hyperfocal Distance: 2698.088428425507
Near Distance: 1329.235453229893
Far Distance: 90525.46997047497
Circle of Confusion on Print: .1311415 mm = 3.8126756213708095 lpmm
Diffraction Limit on Print: 15.547397970200823 lpmm=0.032159722222222215
mm
Questions:
What do you mean by lpmm?
Do you read that as "lines per millimeter" or do you read that as "line
pairs per millimeter"?
1st possibility:
If by lpmm you mean "lines per millimeter", the output would suggest that
one can translate CoC diameter to lines per millimeter as follows:
lines per millimeter = 1 / 2c
and thus, you would calculate "line pairs per millimeter" this way:
line pairs per millimeter = 1 / 2c / 2
(since there are two lines in a pair...)
So, if by lpmm you mean "lines per millimeter", a CoC diameter of
0.1311415 mm = 3.8126756213708095 lines per millimeteres, as your
calculator shows -OR- 1.906337810685 line PAIRS per millimeter, which is
had by dividing by 2.
2nd possibility:
If by lpmm you instead mean "line pairs per millimeter", then your formula
is:
line pairs per millimeter = 1 / 2c
In which case, your calculator is stating that a CoC diameter of 0.1311415
mm = 3.8126756213708095 line PAIRS per millimeter.
Punchline: Is it your understanding then that a CoC diameter of
0.1311415mm is equivalent to 1.906337810685 line PAIRS per millimeter or
3.8126756213708095 line PAIRS per millimeter?
Problem: I don't believe either possiblity can be correct. I think
1.906336810685 is too small by a factor of 4 and 3.8126756213708095 is too
small by a factor of 2. So whether you use lpmm to mean "lines per
millimeter" or "line pairs per millimeter" the number of line pairs per
millimeter works out to be too small using the numbers I get with your
calculator. On what do I base this observation? Well, for starters,
Michael Gudzinowicz's comments, two posts down, in the same thread - which
no one else seems to have caught. More below...
******
From: Richard Knoppow ([email protected])
Subject: Re: Confusion over circle of confusion
Newsgroups: rec.photo.equipment.large-format
Date: 2001-04-18
[snip]
>This is a neat calculator with all the math hidden, worth looking
>at. Po girl seems to be good at several things.
Richard: This thread was graced by three "heavyweights" of which you are
perhaps the most respected, and I don't say that lightly. With respect, I
would like to ask if you had considered the relationship of CoC diameter
to "lpmm" Ann's calculator exhibits when you gave it your blessings. I
realize that you by no means implied that you had given it a thorough
going over, but what say ye now?
******
From: Michael Gudzinowicz ([email protected])
Subject: Re: Confusion over circle of confusion
Newsgroups: rec.photo.equipment.large-format
Date: 2001-04-18
[snip]
>At the distance where the eye focuses best on near objects (around 10" or
>250 mm), one should easily be able to resolve 4 lpmm. Most people can see
>6 lpmm, and 8 lpmm usually has "texture" rather than even appearance. If
>you want to use 4 lpmm in the print, which is commonly used for 8x10
>prints from 35 mm, then the circle of confusion for a 4X will be
>1 /(4 x 4) = 1/16 mm.
>For the 8X enlargement from 35 mm to 8x10, it's 1/32 mm.
Here, Michael uses the same nomenclature (lpmm) and clearly states that to
achieve 4 lpmm (whatever that is) with a 4x enlargement factor, we must
seek an on-film maximum CoC diameter of 1/16 mm, which is 0.0625 mm.
Problem: Ann's calculator, using the formula:
lpmm (whatever that is) = 1 / 2c
would equate 0.0625 mm CoC's to 8 lpmm, NOT 4 lpmm (be that line PAIRS per
millimeter, or simply lines per millimeter).
>
>If the print viewing distance is not 10", one can accomodate the change
>in viewing distance with the following formula:
>
> (viewing_distance_in_mm)
>coc = --------------------------------------------
> (print_res_lpmm) (magnification) (250 mm)
Let's convert Michael's formula to solve for CoC when viewing distance
actually equals 250mm. This gives us:
1
coc = --------------------------------
(print_res_lpmm) (magnification)
Solving for print_res_lpmm, we get:
1
print_res_lpmm = --------------------------
coc * magnification
Turning to the example input I used with Ann's calculator, where my CoC is
0.03115 and my magnification is 4.21, we can use this derrivative of
Michael's formula to calculate:
print_res_lpp = 1 / (0.03115 * 4.21) = 7.62535 lpmm, which is exactly
TWICE the resolution Ann's calculator came up with for my example:
Circle of Confusion on Print: .1311415 mm = 3.8126756213708095 lpmm
This is because Ann's formula for print resolution is effectively:
1
print_res_lpmm = -----------------------------
coc * magnification * 2
Unfortunately, even if we try to reconcile the difference between
Michael's formula and Ann's formula with contrary definitions for lpmm, we
can't get any closer than a factor of 2, as discussed above.
[snip]
For the record, I always write lp/mm to mean line pairs per millimeter.
I seldom have occaision to use an acronym for lines per millimeter, but it
would be lpmm.
******
So: Whether Richard and Michael realized at the time, or not, Ann's
calcuations disagree with Michael's.
And: Since Ann's Diffraction Limit on Print is calculated the same way,
the Diffraction calculations are impacted as well.
Just to be clear, looking at these two outputs again:
Circle of Confusion on Print: .1311415 mm = 3.8126756213708095 lpmm
Diffraction Limit on Print: 15.547397970200823 lpmm=0.032159722222222215
mm
I find the diameter specified in mm for Circle of Confusion on Print to be
absolutely correct. The problem lies in the conversion to lpmm.
I heartily invite all three experts (and I don't use that word
facetiously!) to reconcile this by answering these two questions:
1) For each of you, what does lpmm mean?
For the record, I always write lp/mm to mean line pairs per millimeter.
I seldom have occaision to use an acronym for lines per millimeter, but it
would be lpmm.
2) Which of these two formulas is correct:
1
print_res_lpmm = --------------------------
coc * magnification
-OR-
1
print_res_lpmm = -----------------------------
coc * magnification * 2
It gets more interesting still: Above, I said that I absolutely agree
with Ann's calculated Circle of Confusion on Print for my stated inputs
(0.1311415 mm), but I can not begin to figure out how you are calculating
the Diffraction Limit on Print. I can see that you are using the same
formula to equate diffraction's disk diameter to lpmm, but I don't
understand how you calculate either diffraction value in the first place.
So, to Ann: What formula are you using for Diffraction Limit on Print?
And: Wouldn't it be best to consider your Diffraction Limit to be the
same diameter as your CoC diameter? (Why stop down further if doing so
will force the diameter of diffraction's Airy disks to exceed your chosen
CoC diameter?)
To avoid visibile diffraction, I calculate the f-stop at which
diffraction's Airy disk diameters equal my maximum permissible CoC
diameters using this formula:
Smallest usable f-Stop = Max. CoC Diameter on film (mm) / 0.00135383
Using this formula, for a maximum CoC diameter of 0.03115 on-film, I would
not stop down below f/23.0 (just a tad beyond f/22). This will keep my
Airy disks smaller than 0.03115 also.
Going on the information Ann provides in the captions of her sample
prints, I ran this data through her calculator:
Print Magnification Factor: 5
Focal Length: 210
F-Stop: 64
Circle of Confusion on Film (here, exactly equal to film_diagonal / 1000):
0.15367
Object Distance: 4484
The calculated Output was:
HperFocal Distance: 4484.040476345415
Near Distance: 2242.0101190406826
Far Distance: 496745377.81597006
Circle of Confusion on Print: .76835 mm = .6507451031430989 lpmm
Diffraction Limit on Print: 4.5 lpmm=.11111111111111112 mm
If, as your calculator recommends, we specified a CoC diameter
equal to the 4x5 film diagonal divided by 1000, my formula for the f-stop
at which diffraction's Airy disks would reach that diameter calculates:
Smallest usable f-Stop = 0.15367 / 0.00135383 = f/113.5
f/64 is almost two stops away from f/113.5 ! I submit that the reason you
are content to use f/64 with 4x5 is because even at f/64, diffraction's
Airy disks aren't doing nearly as much damage as what I would describe as
the *enormous* CoC diameters you are allowing at your Near and Far sharps.
Michael's formula would recommend that if we want to achieve 4 lpmm in a
5x print, we'll have to limit our on-film CoC diameters to:
1/(4*5) = 1/20 mm = 0.05mm on-film which is less than 1/3 the diameter we
would get by dividing the film diagonal by 1000, which gave us 0.15367mm
for the calculations above.
If I were to use the CoC diameter Michael's formula recommends, 0.05mm
on-film, my formula for smallest usable f-stop would calculate:
Smallest usable f-stop = 0.05 / 0.00135383 = f/36.9 instead of f/113.5
In fact, using Michael's formula for calculating Print resolution, the
format_diagonal / 1000 choice for CoC diameter would result in a Print
resolution of only 1.30 lpmm in a 5x print, instead of 4 lpmm. Thus,
Michael's formula, contrasted against the statement Ann wrote at the
bottom of her calculator page ("Conclusion, f64 was fine at 5x
enlargement. I would not hesitate to use it next time if DOF is needed.")
says that Ann's tolerance for the diffraction seen with f/64 in a 5x print
from 4x5 goes hand in hand with her tolerance for CoC's that equate to
only 1.30 lpmm at the print.
I look forward to your answers.
Thank you,
Mike Davis
--
From: [email protected] (Michael Gudzinowicz)
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Ann's formula vs. Michael's (was: Confusion over CoC)
Date: 14 Dec 2001
"Michael K. Davis" [email protected]> wrote:
> Ann's formula vs. Michael's (was: Confusion over CoC)
I've included the post below, but moved a few sections to the beginning
for clarity.
Re: Ann's post:
> Questions:
> What do you mean by lpmm?
> Do you read that as "lines per millimeter" or do you read that as "line
> pairs per millimeter"?
"Lines per mm", "line pairs per mm" and "cycles per mm" are all
equivalent terms in this context, and have been for decades in
publications. "Lines per mm" may be abbreviated as "l/mm" or to
avoid confusion in programs and formulas, as "lpmm. The preferred
form for "lines per inch" is "lpi" rather than "l/i" which resembles
"1/i".
The inverse is the circle of confusion for defocus. The number
"2" isn't part of the formula. If one wishes to consider a target
of 32 lpmm, the coc is 1/32 mm. Each black bar is 1/64 mm in
width with a white space of 1/64 mm. If the center of the coc is
placed in the center of the black bar, its perimeter extends
halfway into the neighboring white spaces. Therefore the coc's
on adjacent black bars do not overlap which is defined as
"complete resolution".
With respect to Ann's math, the formulas are defined in the
script which can be viewed with an html editor. She used the
traditional formula for the hyperfocal distance, and simple
approximation formulas for dof.
The Rayleigh limit has been modified to include a 0.9 factor
(there are all sorts of modifications around with other names and
derivations; 0.9 looks like it dropped out of an OTF limit) and
she used another factor of two to indicate complete
non-overlapping resolution rather than the placement of one
maximum over the first Airy ring. With a source of light
compromising the entire visible spectrum, the absolute value of
the Bessel function has to be integrated over all wavelengths
which removes the rings. The result resembles the diameter of the
disk for green light. Since there's no way one will resolve
details with a LF lens and sheet film, that isn't a bad choice to
use with the defocus coc which is evenly illumination over the
cross section. (The traditional Rayleigh formula works better
with film resolution in lens tests; see my old discussions with
Bob Atkins for curve fits, etc.)
My post:
I use the following simple formula to estimate the resolution
required in the negative. The magnification, of course, is the
linear negative to print magnification. No enlarging losses are
assumed.
> (viewing_distance_in_mm)
> >coc = --------------------------------------------
> > (print_res_lpmm) (magnification) (250 mm)
The result may be used in the calculation of hyperfocal distance.
If you check Ann's formula, you'll note that she didn't use print
magnification to modify the coc used for dof calculations. It was
just used in the estimate of the diffraction effect at the focal
plane.
The expression of coc as diagonal/1000 is somewhat archaic, and
actually assumes a diagonal of 250 mm at 4 lpmm or a coc of 0.25
mm. For an 8x10, it is a little soft. Other dominators which have
been used are 1500 (6 l/mm) and 2000 (8 lpmm).
There are other ways of examining sharpness which give results
similar to the Rayleigh approximation.
These approaches are simplistic for LF photography, since the
effect of diffraction may approximate that of defocus. When they
are equal, coc chosen to define dof will only be seen at the
focal plane, and dof at that level defined by the coc is zero.
A more realistic calculator would use the sum of diffraction and
defocus circles to define the boundaries of sharp focus and dof
limits. Since diffraction effects change with f/stop and image
distance from the lens, the derivation and practical use is a
little more complicated. Also, the optimum position of the focal
plane with respect to near and far limits shifts with f/stop
unlike the simple cases discussed in this newsgroup.
Rest of original post:
---------------------------------------------------------------------
> Hello to Peter, Ann, Richard and Michael, who all participated in this
> thread I am resurrecting from April of this year.
>
> Let me recap the four-post thread as briefly as possible, embedding my
> comments and questions:
>
> ******
> From: Peter De Smidt ([email protected])
> Subject: Confusion over circle of confusion
> Newsgroups: rec.photo.equipment.large-format
* Date: 2001-04-17
* > >
> >I'm trying to figure out the hyperfocal distance for my various lenses
> >and f-stops, the lenses being 135, 203 and 300mm, and the f-stops being
> >f22, f32, f45. I would like to use a value for the circle of confusion
> >that would give me a visually sharp 16x20 inch print from a 4x5 negative.
>
> [snip]
>
> >Peter
>
> *******
>
> From: annqlee ([email protected])
> Subject: Re: Confusion over circle of confusion
> Newsgroups: rec.photo.equipment.large-format
> Date: 2001-04-17
> >
> >Hi Peter,
> >
> >If I want DOF I never use anything wider than f22. I even use f22 when I
> >focus near infinity. I would not hesitate to use f45 if I were to enlarge
> >4x or even 5x. You test should be at which fstop is the diffraction
> >unacceptable to you. It is my belief that you will find that f45 is still
> >very good. I don't even use those formulas anymore, if it is kinda far I
> >use f32. If it is really far I use f22.
> >
> >Here are some tests:
> >http://carcassi.eng.uci.edu/Eportfolio/technical/diffraction/diffraction.htm
> >
> >Good Luck,
> >
> >Ann
>
> Ann,
>
> I had a look at your calculator, at the address you posted above. Nice
> job, but...
>
> Supplying the following values:
>
> Print Magnification Factor: 4.21
> Focal Length: 43
> F-Stop: 22
> Circle of Confusion on Film: 0.03115
> Object Distance: 2620
>
> It calculated the following Output:
>
> Hyperfocal Distance: 2698.088428425507
> Near Distance: 1329.235453229893
> Far Distance: 90525.46997047497
> Circle of Confusion on Print: .1311415 mm = 3.8126756213708095 lpmm
> Diffraction Limit on Print: 15.547397970200823 lpmm=0.032159722222222215
> mm
>
> Questions:
> What do you mean by lpmm?
> Do you read that as "lines per millimeter" or do you read that as "line
> pairs per millimeter"?
>
> 1st possibility:
>
> If by lpmm you mean "lines per millimeter", the output would suggest that
> one can translate CoC diameter to lines per millimeter as follows:
>
> lines per millimeter = 1 / 2c
>
> and thus, you would calculate "line pairs per millimeter" this way:
>
> line pairs per millimeter = 1 / 2c / 2
>
> (since there are two lines in a pair...)
>
> So, if by lpmm you mean "lines per millimeter", a CoC diameter of
> 0.1311415 mm = 3.8126756213708095 lines per millimeteres, as your
> calculator shows -OR- 1.906337810685 line PAIRS per millimeter, which is
> had by dividing by 2.
>
>
> 2nd possibility:
>
> If by lpmm you instead mean "line pairs per millimeter", then your formula
> is:
>
> line pairs per millimeter = 1 / 2c
>
> In which case, your calculator is stating that a CoC diameter of 0.1311415
> mm = 3.8126756213708095 line PAIRS per millimeter.
>
>
> Punchline: Is it your understanding then that a CoC diameter of
> 0.1311415mm is equivalent to 1.906337810685 line PAIRS per millimeter or
> 3.8126756213708095 line PAIRS per millimeter?
>
> Problem: I don't believe either possiblity can be correct. I think
> 1.906336810685 is too small by a factor of 4 and 3.8126756213708095 is too
> small by a factor of 2. So whether you use lpmm to mean "lines per
> millimeter" or "line pairs per millimeter" the number of line pairs per
> millimeter works out to be too small using the numbers I get with your
> calculator. On what do I base this observation? Well, for starters,
> Michael Gudzinowicz's comments, two posts down, in the same thread - which
> no one else seems to have caught. More below...
>
> ******
>
> From: Richard Knoppow ([email protected])
> Subject: Re: Confusion over circle of confusion
> Newsgroups: rec.photo.equipment.large-format
> Date: 2001-04-18 17:44:34 PST
>
> [snip]
>
> >This is a neat calculator with all the math hidden, worth looking
> >at. Po girl seems to be good at several things.
>
> Richard: This thread was graced by three "heavyweights" of which you are
> perhaps the most respected, and I don't say that lightly. With respect, I
> would like to ask if you had considered the relationship of CoC diameter
> to "lpmm" Ann's calculator exhibits when you gave it your blessings. I
> realize that you by no means implied that you had given it a thorough
> going over, but what say ye now?
>
> ******
>
> From: Michael Gudzinowicz ([email protected])
> Subject: Re: Confusion over circle of confusion
> Newsgroups: rec.photo.equipment.large-format
> Date: 2001-04-18 04:40:08 PST
>
> [snip]
>
> >At the distance where the eye focuses best on near objects (around 10" or
> >250 mm), one should easily be able to resolve 4 lpmm. Most people can see
> >6 lpmm, and 8 lpmm usually has "texture" rather than even appearance. If
> >you want to use 4 lpmm in the print, which is commonly used for 8x10
> >prints from 35 mm, then the circle of confusion for a 4X will be
> >1 /(4 x 4) = 1/16 mm.
> >For the 8X enlargement from 35 mm to 8x10, it's 1/32 mm.
>
> Here, Michael uses the same nomenclature (lpmm) and clearly states that to
> achieve 4 lpmm (whatever that is) with a 4x enlargement factor, we must
> seek an on-film maximum CoC diameter of 1/16 mm, which is 0.0625 mm.
> Problem: Ann's calculator, using the formula:
>
> lpmm (whatever that is) = 1 / 2c
>
> would equate 0.0625 mm CoC's to 8 lpmm, NOT 4 lpmm (be that line PAIRS per
> millimeter, or simply lines per millimeter).
>
>
>
> >
> >If the print viewing distance is not 10", one can accomodate the change
> >in viewing distance with the following formula:
> >
> > (viewing_distance_in_mm)
> >coc = --------------------------------------------
> > (print_res_lpmm) (magnification) (250 mm)
>
>
> Let's convert Michael's formula to solve for CoC when viewing distance
> actually equals 250mm. This gives us:
>
> 1
> coc = --------------------------------
> (print_res_lpmm) (magnification)
>
>
> Solving for print_res_lpmm, we get:
>
>
> 1
> print_res_lpmm = --------------------------
> coc * magnification
>
>
> Turning to the example input I used with Ann's calculator, where my CoC is
> 0.03115 and my magnification is 4.21, we can use this derrivative of
> Michael's formula to calculate:
>
> print_res_lpp = 1 / (0.03115 * 4.21) = 7.62535 lpmm, which is exactly
> TWICE the resolution Ann's calculator came up with for my example:
>
> Circle of Confusion on Print: .1311415 mm = 3.8126756213708095 lpmm
>
> This is because Ann's formula for print resolution is effectively:
>
> 1
> print_res_lpmm = -----------------------------
> coc * magnification * 2
>
> Unfortunately, even if we try to reconcile the difference between
> Michael's formula and Ann's formula with contrary definitions for lpmm, we
> can't get any closer than a factor of 2, as discussed above.
>
> [snip]
>
> For the record, I always write lp/mm to mean line pairs per millimeter.
> I seldom have occaision to use an acronym for lines per millimeter, but it
> would be lpmm.
>
> ******
>
> So: Whether Richard and Michael realized at the time, or not, Ann's
> calcuations disagree with Michael's.
>
> And: Since Ann's Diffraction Limit on Print is calculated the same way,
> the Diffraction calculations are impacted as well.
>
> Just to be clear, looking at these two outputs again:
>
> Circle of Confusion on Print: .1311415 mm = 3.8126756213708095 lpmm
> Diffraction Limit on Print: 15.547397970200823 lpmm=0.032159722222222215
> mm
>
> I find the diameter specified in mm for Circle of Confusion on Print to be
> absolutely correct. The problem lies in the conversion to lpmm.
>
> I heartily invite all three experts (and I don't use that word
> facetiously!) to reconcile this by answering these two questions:
>
> 1) For each of you, what does lpmm mean?
>
>
> For the record, I always write lp/mm to mean line pairs per millimeter.
> I seldom have occaision to use an acronym for lines per millimeter, but it
> would be lpmm.
>
> 2) Which of these two formulas is correct:
>
> 1
> print_res_lpmm = --------------------------
> coc * magnification
>
> -OR-
>
> 1
> print_res_lpmm = -----------------------------
> coc * magnification * 2
>
>
>
> It gets more interesting still: Above, I said that I absolutely agree
> with Ann's calculated Circle of Confusion on Print for my stated inputs
> (0.1311415 mm), but I can not begin to figure out how you are calculating
> the Diffraction Limit on Print. I can see that you are using the same
> formula to equate diffraction's disk diameter to lpmm, but I don't
> understand how you calculate either diffraction value in the first place.
>
> So, to Ann: What formula are you using for Diffraction Limit on Print?
>
> And: Wouldn't it be best to consider your Diffraction Limit to be the
> same diameter as your CoC diameter? (Why stop down further if doing so
> will force the diameter of diffraction's Airy disks to exceed your chosen
> CoC diameter?)
>
> To avoid visibile diffraction, I calculate the f-stop at which
> diffraction's Airy disk diameters equal my maximum permissible CoC
> diameters using this formula:
>
> Smallest usable f-Stop = Max. CoC Diameter on film (mm) / 0.00135383
>
> Using this formula, for a maximum CoC diameter of 0.03115 on-film, I would
> not stop down below f/23.0 (just a tad beyond f/22). This will keep my
> Airy disks smaller than 0.03115 also.
>
> Going on the information Ann provides in the captions of her sample
> prints, I ran this data through her calculator:
>
> Print Magnification Factor: 5
> Focal Length: 210
> F-Stop: 64
> Circle of Confusion on Film (here, exactly equal to film_diagonal / 1000):
> 0.15367
> Object Distance: 4484
>
> The calculated Output was:
>
> HperFocal Distance: 4484.040476345415
> Near Distance: 2242.0101190406826
> Far Distance: 496745377.81597006
> Circle of Confusion on Print: .76835 mm = .6507451031430989 lpmm
> Diffraction Limit on Print: 4.5 lpmm=.11111111111111112 mm
>
> If, as your calculator recommends, we specified a CoC diameter
> equal to the 4x5 film diagonal divided by 1000, my formula for the f-stop
> at which diffraction's Airy disks would reach that diameter calculates:
>
> Smallest usable f-Stop = 0.15367 / 0.00135383 = f/113.5
>
> f/64 is almost two stops away from f/113.5 ! I submit that the reason you
> are content to use f/64 with 4x5 is because even at f/64, diffraction's
> Airy disks aren't doing nearly as much damage as what I would describe as
> the *enormous* CoC diameters you are allowing at your Near and Far sharps.
>
> Michael's formula would recommend that if we want to achieve 4 lpmm in a
> 5x print, we'll have to limit our on-film CoC diameters to:
>
> 1/(4*5) = 1/20 mm = 0.05mm on-film which is less than 1/3 the diameter we
> would get by dividing the film diagonal by 1000, which gave us 0.15367mm
> for the calculations above.
>
> If I were to use the CoC diameter Michael's formula recommends, 0.05mm
> on-film, my formula for smallest usable f-stop would calculate:
>
> Smallest usable f-stop = 0.05 / 0.00135383 = f/36.9 instead of f/113.5
>
> In fact, using Michael's formula for calculating Print resolution, the
> format_diagonal / 1000 choice for CoC diameter would result in a Print
> resolution of only 1.30 lpmm in a 5x print, instead of 4 lpmm. Thus,
> Michael's formula, contrasted against the statement Ann wrote at the
> bottom of her calculator page ("Conclusion, f64 was fine at 5x
> enlargement. I would not hesitate to use it next time if DOF is needed.")
> says that Ann's tolerance for the diffraction seen with f/64 in a 5x print
> from 4x5 goes hand in hand with her tolerance for CoC's that equate to
> only 1.30 lpmm at the print.
>
> I look forward to your answers.
>
> Thank you,
>
> Mike Davis
From: "Michael K. Davis" [email protected]>
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Ann's formula vs. Michael's (was: Confusion over CoC)
Date: Sat, 15 Dec 2001
Hi Ann!
ann lee [email protected]> wrote:
: Hi Michael,
: I did not mean to offend you.
Oh, don't be silly! :-)
: I appreciate mathematical applications. I am
: not good with math or numbers though.
I'm no mathemetician, but judging by your calculator, like you, I enjoy
putting together tools that accomplish something - even though I honestly
struggle to implement formulas I've read about into spreadhseets and such.
The learning curve for my HP48G+ calculator was nearly more than I could
manage and to this day, all I know how to do on it is the stuff I
programmed into it months ago, with tremendous difficulty. (But I
documented it all so I can reprogram easily if the batteries go dead!)
: With all due and utmost respect, many times in the field, you do not know
: what distances your objects are at, in fact you don't know what is your
: effective f-stop. In conjuction, you really don't know your depth of focus.
: Just like my research, I can not predict within 15% the concentrations of
: pollutants in the atmosphere, but I can basically tell you what will happen
: if the temperature increase, more NOx emissions etc. Same thing with DOF, I
: believe the formulas we are using are very simplistic, however I do know
: that reducing aperature increases DOF from the math. But I did not know if
: it is ok for me to shoot at f45,f64,f90,f128? Also, while it is popular
: convention to say the eye resolves 8lpmm at 10", but that doesn't mean much
: to me, is 4lpmm ok? 2lpmm ok? Honestly, I have no idea even till today
: because I have never measured the lpmm.
Still, the intent of your calculator is very appealing and I KNOW it
genuinely has practical merit. I say this from the position of having
used the formulas in my HP48G+, in the field, for nearly every shot I
take, for almost a year now, with great success - all the empirical
evidence one could want. (More on this in a moment...)
Many people have found my use of a calculator to be laughable, not to
mention obcessive. Indeed, I was "laughed out of town" by anti-math
zealots on a stereography list I subscribe to, after posting an article
where I detailed every last step I go through to setup a shot (including
my use of a laser rangefinder to measure near and far point distances and
later, to find a target subject on which to focus, that is precisely at
the calculated best distance to place the plane of focus). I've since
redeemed myself with these folks...
I use twin Mamiya 7 II's to produce 50x50mm stereo pairs from two 6x7
chromes. The total setup time runs about 18 to 20 minutes, including the
measurements, the calculations for optimum aperture, focus distance and
something called base separation (the distance between the two camera
lenses, which varies with the ratio of Far to Near point distances, the FL
of the taking lens and the FL of the stereo viewer.) Some of that time is
consumed by more conventional tasks, of course, like metering (L-608) and
just physically assembling the rig, but because my end product in the
stereo viewer is a pair of mounted original chromes (dupes don't cut it),
I also spend a lot of time fussing with various in-camera contrast control
techniques, including pre-exposure - to elevate shadows, for which I spent
a great deal of time and energy to perfect, and yes, I use math to make
decisions about whether pre-exposure should be applied and precisely how
much pre-exposure is necessary to compress the metered luminance range of
the scene into the film's latitude. The only thing not done in my camera
is the E-6 processing, done by a commercial lab. What you see in my
stereo viewer is the film as it came from my cameras, except for E6
processing.
I have worked very hard to identify every link in the chain affecting
image quality which I might have some ability to influence and I've
learned that taking it to the Nth degree the way I do requires a lot of
patience in the field. Eric Boxall, who posts to rec.photo.*, once
described what I do as being more like photogrammetry than photography.
So, I know that I am the eccentric among centrics, but in the interest of
convincing the rest of the world that math, even the crude, oversimplified
formulas we use, has a place in photography, even in the field, let me
share some "rave reviews" I've enjoyed in recent months:
******
>From Paul Talbot of www.rmm3d.com, posted in the Photo-3D list a couple of
days after we first met at a stereography convention:
"A definite highlight of the day was having the opportunity to meet
Photo-3D 'math maven' Mike Davis and to view his medium format 3D slides.
Mike D. was just introduced to stereo photography late last summer, so he
is very new to 3D. The images he showed in his SaturnSlide viewer were so
stunning that they grabbed firm hold of my eyeballs and refused to let
them go! I could scarcely stand to hand the viewer to the next person
waiting in line!
In about 5 years of being involved with stereo photography I have seen
many thousands of stereo slides: from having visited several clubs,
viewing 'Hall of Fame' and other PSA loaner sets, SSA folio images, NSA
stereo theater shows, medium format folios and expos, and private
exchanges. Viewing Mike's photos on Saturday, I thought I had slipped
briefly into a 'parallel universe' where the quality of 3D imaging exceeds
anything I had ever imagined possible.
During the three hour drive home this morning my head was still spinning
in wonder at the quality of those chromes. What am I to do now? Rent a
dumpster to toss my own thousands of 3D photos and start from scratch?
(I won't, but the thought did enter my mind!)
Mike uses twin Mamiya M7 IIs, and has given some detailed info on his
techniques on the MF3D list. Mike is very humble about the 'artistic'
skill that he adds to his highly refined technical ability, but I can say
his photos show he has a very fine photographic eye to complement his near
total mastery of the science of photography.
Note: None of this is meant to suggest that Mike's techniques are
appropriate in all circumstances (nor does he apply all of them to all his
own shots). Point-and-shoot away for family photos and spontaneous
images, by all means! But if you've ever been dissatisfied with the
technical aspects of any of your images (as I certainly have with mine),
you may be as excited as I am to know there is hope to achieve the quality
you seek!"
******
I readily confess that I succumb to pride when I read those words, but
they actually contain a message much more important to me than the
hollowness of self-glorification: Those words say that a mathematical
approach works!
I feel at times that I'm waging a one-man crusade for the "realtime" use
of math in the field, because most people think I've taken it way too far;
that I've left the planet. I'm not doing any one thing that's all that
innovative or unprecedented, but I am doing everything I CAN to deliver
extremely high quality images, literally everything - I don't cut any
corners. Anyone can do what I'm doing and I just want to spread the word.
I can't share my stereo viewer with this audience, so let me suffer you
with some more reviews of my work, again, for the sake of my crusade -
I'll put some ice on my swollen head...
******
>From Bruce Rosenberger ([email protected]), who saw my work for the first time
at the National Stereography Association convention, in Buffalo, this past
July:
"I suspect that anyone that had the great fortune to see Mike's slides
will never again question his system for taking pictures. It will take a
while for all his ideas to sink into my Pennsylvania Dutch brain, but his
pictures will stay with me for a long time!"
******
>From Boris Starosta (www.dynamicsymmetry.com):
"The most amazing things I saw at the convention:
Imax Stereo Theatre show of views from the Space Station. [snip]
Zebra Imaging's full color full parallax CG holograms. [snip]
Mike Davis' MF stereoviews. Astonishing beauty and perfect execution -
the man is an artist."
******
And lastly, from David W. Kesner ([email protected]):
"Mike has given us the numbers and numbers don't lie. He has also
given us the product and the product is beyond questioning.
As someone who has seen the products of Mike's formulas let me say
that you can't argue with perfection. I do not use that word lightly
either. His images are the most easily viewed and visually stunning I
have ever seen. They just plain don't get any better.
For the record these are 6x6 images taken with single or twin Mamiya
7's."
******
Punchline:
1) Using math in the field can improve image quality.
2) Using math in the field requires more patience than brains.
3) Using math in the field is appealing if you have a passion for
quality.
OK, I'll get off that soapbox...
: I was not aware that the circle of confusion actually covers one black and
: two halves of a white. That seems weird.
It does sound strange.
: I don't know the diffraction limit, I only know its theoretical limit and I
: multiplied that by 0.9 because I figure that will be better than 1.0 which
: is impossible for the wavelengths of most scences. I don't think anyone can
: give you a number for the generic scene. But I am guessing that my 0.9 is as
: good as any.
: So confession time, I have never used that calculator page. 3 lashes with
: wet noodles.
I'm really moved by your honesty, but for the sake of my "crusade" I'm
crushed. That's OK, I've just issued a stay of execution against the wet
noodle lashing. :-)
: I am not abandoning ship, I still hold my one CoC=1black line or white line.
: I still hold that 0.9*theoretical diffraction limit is as good enough. I
: still hold that you have to try it for yourself, not about measuring DOF,
: but above how far can you stop down before it is objectionable to you. But I
: do know that Michael G et al. are more educated in this matter, so I guess I
: will shut my webpage and redirect it to his.
Be not defeated, but rather encouraged - to either stand your ground or
fix it, whichever brings you the most peace. Shutting it down would be a
shame.
: I hoped I have cleared assumptions of my erroneous calculator.
You have, and again, I'm very impressed by your candor and humility.
Mike Davis
--
From: [email protected] (Michael Gudzinowicz)
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Ann's formula vs. Michael's (was: Confusion over CoC)
Date: 15 Dec 2001
"Michael K. Davis" [email protected]> wrote:
The posts are getting a little long. In this reply, I've included
Ann's formulas below if you can't view at the script.
> : The Rayleigh limit has been modified to include a 0.9 factor
> : (there are all sorts of modifications around with other names and
> : derivations; 0.9 looks like it dropped out of an OTF limit) and
> : she used another factor of two to indicate complete
> : non-overlapping resolution rather than the placement of one
> : maximum over the first Airy ring. With a source of light
> : compromising the entire visible spectrum, the absolute value of
> : the Bessel function has to be integrated over all wavelengths
> : which removes the rings. The result resembles the diameter of the
> : disk for green light. Since there's no way one will resolve
> : details with a LF lens and sheet film, that isn't a bad choice to
> : use with the defocus coc which is evenly illumination over the
> : cross section. (The traditional Rayleigh formula works better
> : with film resolution in lens tests; see my old discussions with
> : Bob Atkins for curve fits, etc.)
>
> Your interpretation of her math is excruciating in the absence of the
> formula she used. What is it?
Input Definitions:
Print Magnification Factor="mag"
Focal Length="fl"
F-Stop="fs"
Circle of Confusion on Film: (~film_diagonal/1000)="coc"
Object Distance : (same units as focal length)="o"
Output Definitions:
HyperFocal Distance="hf"
Near Distance="near"
Far Distance="far"
Circle of Confusion on Print:="coc2"
Diffraction Limit on Print:="diff"
Calculations and formatting:
hf=fl*fl/(coc*fs)
near=hf*o/(hf+o)
far=hf*o/(hf-o)
if far0 then far="Inf"
coc2=coc * mag+ " mm = " +0.5 / (coc * mag) + " lpmm";
diff= 1600 / fs * 0.9 / mag + " lpmm=" +0.5 / (1600 / fs * 0.9) * mag + " mm";
Date: Sat, 11 Aug 2001
From: [email protected] (Mark Anderson)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: different mfgers CoC and DOF ranges 1/70th to 1/200th inch..
Robert Monaghan [email protected] wrote:
> in short, if you rely on the DOF of your lenses, you may be using a
> liberal or a conservative or middle of the road CoC, but most folks can't
> tell you what the CoC is on their given lens(es), and most of us use
> experience to decide how liberal or conservative to be when using such DOF
> markings and tables and scales and so on ;-)
Experience quickly showed me that I was not satisfied with the DOF based
on my Zeiss lens markings. (SuperIkonta C, 105 mm 6x9 cm). Determining
that it was based on a COC of .004" rather than the .001" of my 35mm
format lenses was useful. Since I don't determine in advance what my
print size or typical viewing distance is going to be, (prints are
usually 8x10 or 11x14 and personally I at least initially view the print
up really close), I figure that generally I want medium format
enlargements and detail to be limited by grain, not by COC. Therefore,
if at all possible, if I use the lens markings of DOF, I use the
markings for two stops smaller than I'm using. The distance that gives
a COC of .001" at f/11, will give a COC of .004" at f/5.6. IOW, if you
use the next more conservative DOF marking on your lens, you'll halve
your COC.
BTW, there are certainly times that having the COC far smaller than
needed at normal viewing distances becomes useful.
Two examples.
1) Using a 10x magnifier on a roughly 8x10 albumen contact print from
the 1880's allowed me to read a minute port name on a boat so that I
could identify where the shot was taken.
2) Using the same magnifier on a glossy 2 1/4 by 4 1/4 contact snapshot
from the 1940's, allowed me to read a sign that gave me information to
eventually track down my two lost uncles.
--
Mark Anderson
DBA Riparia www.teleport.com/~andermar/
From: "Michael K. Davis" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Not Just Another DoF Calculor! v9
Date: 23 Jan 2001
Have a look at this MS-Excel 5.0+ Spreadsheet (PC or Mac) that allows you
to create Depth of Field Tables taking every possible variable into
account - even the resolving power of the eye:
http://home.online.no/~gjon/mdofcal9.xls
If you prefer that every print tolerate examination at a viewing distance
of only 25 cm (9.8 inches), rather than at a distance equal to its print
diagonal, a goal set by many DoF Calculators, you can make use of the
system resolution related features to recommend a CoC diameter that meets
that goal while preserving as much DoF as possible. This feature requires
that you specify system resolution (the combined resolving power of your
lens and film) and the resolving power of your eye - which is treated as
the print resolution you want to achieve (i.e. 5 lpmm) after
magnification.
Having previously specified the format diagonal (after cropping), the
spreadsheet calculates the maximum print dimensions for various aspect
ratios (i.e. 1:1, 2:3, 4:5, 5:7, and 11:14), or for a user-specified
aspect ratio, that will deliver your specified resolution in the print,
given the system resolution. You then specify the actual print size you
intend to make, using the recommended dimensions that will guarantee your
desired print resolution after magnification (assuming the specified
system resolution is correct) OR you can just enter in any print
dimensions you desire.
The resulting after-magnification resolution at the print is then
calculated - it will be equal to your desired print resolution if you
entered one of the recommended dimensions, a higher lpmm if your
dimensions are smaller (less magnification) or a lower lpmm if your
dimensions are larger.
The spreadsheet then calculates and suggests a limit for the smallest
Maximum Permissible CoC Diameter you should specify to avoid going
smaller than can be resolved by the eye (i.e. 5 lpmm). Going smaller
would be a waste of DoF. It similarly calculates and suggests a limit for
the smallest Maximum Permissible CoC Diameter you should specify to avoid
going smaller than the system can resolve at the print. Again, going
smaller would be a waste of DoF. You are instructed to consider the
greater of the two diameters (the lesser of the two denominators, where
the Maximum Permissible CoC Diameter is expressed as 1/n inch) to be the
smallest useful Maximum Permissible CoC Diameter (the greatest useful
denominator) at the Near and Far sharps.
For example, when the magnification ratio is really excessive, as for a
20x24 print from 35mm, the system resolution at the print would be 2.18
lpmm (assuming a system resolution at the film of 45 lpmm and a human eye
resolving power of 5 lpmm). At only 2.18 lpmm in the print, there is no
need to force Max CoC diameters down to 5 lpmm, so the second
recommendation above would take precedence over any concern for matching
what the eye can resolve.
When the magnification ratio is smaller than what the system resolution
can actually deliver, as for a 4x5 print from 35mm, the system resolution
at the print would be 10.63 lpmm (again assuming a system resolution at
the film of 45 lpmm). At 10.63 lpmm in the print, we now have no need to
force Max CoC diameters beyond 5.00 lpmm - the maximum resolution your eye
can resolve at a 25 cm viewing distance. In this situation, the first
recommendation above would take precedence over any concern for matching
CoC diameter to the print resolution. Again, we have optimized our DoF -
in fact, for every situation where the magnification ratio is small, like
an 8x10 from 6x7cm, the resulting DoF tables are very generous compared to
those calculated for the same lens/format combination at larger
magnification ratios.
If you want to make a really big print, where the magnification ratio
exceeds the ratio of system resolution to eye resolution, go ahead - this
spreadsheet will recommend the appropriate Maximum CoC Diameter AND it
will recommend the minimum viewing distance for that size print, the
distance which simulates a 5 lpmm resolution when viewed at only 25cm.
That's not all... The effects of diffraction increase as we stop down to
increase Depth of Field. This spreadsheet calculates the aperture at
which the effects of diffraction will become visible (where the diameter
of diffraction's Airy disks equal your chosen Maximum CoC Diameter.) If
for example, this is calculated to occur at f/20, you will know that it is
pointless to seek additional depth of field by stopping down from f/16 to
f22 or f/32. To do so would increase DoF at the expense of diffraction.
When you print the resulting Depth of Field tables, headers document the
intended print dimensions, the chosen resolution at the print, and the
recommended minimum viewing distance. By calculating optimized tables
for various print sizes, where not one inch of acceptable depth of field
has been lost to ignorance of the variables, you'll be able to accurately
predict acceptable print dimensions in the field.
As a quick example, consider that a 16x20-inch print, capable of
tolerating examination at 25cm (9.8 inches), made from 6x7cm format,
enjoys a Near Sharp of 2.1 meters (6.88 feet), at f/16 with a 43mm lens,
when focused at the hyperfocal distance of 4.19 meters (13.76 feet).
Any subject closer than 2.1 meters simply won't be found acceptably sharp
in a 16x20 print viewed at a distance of 25cm. Want more Depth of Field?
Take a look at your DoF table for the next smaller print dimensions you
have optimized. If that's an 11x14 print, you'll find that a print of
that size will survive examination at 25cm with the Near Sharp moved
closer to the camera - 1.47 meters (4.81 feet) at f/16 with the same 43mm
lens focused at the new hyperfocal distance of 2.94 meters (9.63 feet).
This is possible because of the reduced magnfication ratio - the CoC
diameters at the film can be larger for an 11x14 print than for a 16x20.
We're not done yet! This spreadsheed also includes calculators for
increasing Depth of Field by decreasing focal length (without moving) or
by increasing subject distance (without changing focal length). The
accompanying notes explain simple methods for doing this in the field.
Enjoy!
Mike Davis
from contax mailing list:
Subject: RE: [Contax] The Digital Way
Date: Thu, 10 Jan 2002
From: "Cousineau, Bernard" [email protected]
> What you write applies to depth of _focus_, but definitely not to depth of
> _field_. For depth of field you need to know the focus (object) distance as
> well. Taking your lenses as an example and focussing both to 50 feet; the
> 200mm at f/4 would have a DOF of about 5.4 feet, but the 100mm at f/2 would
> have a DOF of about 9.2 feet.
This is true if you don't enlarge the 100 mm image to the same size as the
200 mm image. If, however, you want the final images to be the same size,
you have to enlarge the 100 mm neg (slide) twice as much, and you will end
up with roughly the same depth of field as the 200. For example: Let's say
that I want to have an 8x10 inch portrait of my girlfriend, and I have 3
cameras at my disposal: a 6x7 with a 200 mm f:4, a 35mm with a 100 mm f:2
and a digicam with a 50 mm f:1. Let's say, for the sake of argument (and
to simplify the math), that each format is 1/4 the size (1/2 of each
linear dimension) of the next size up (so 48x64mm for 6x7, 24x32mm for 35,
12x16mm for digital, after cropping). In this case, the depth of field of
each portrait will be the same, in theory, once they are enlarged to 8x10
inches.
These back-of-the-envelope calculations break down when you factor-in that
"all things are not equal." Grain, for instance, will be very different in
the 3 photos, which can contribute significantly to the impression of
sharpness.
An other interesting monkey wrench in this calculations is that two lenses
of the same focal length and aperture don't necessarily have the same
depth of field! There is an interesting thread on photo.net (entitled "CFI
250 Superachromat vs CFE 350 Tele-superachromat") in which Kornelius J.
Fleischer of Zeiss explains that the Zeiss Superachromat 5.6/250 displays
noticeably less depth of field than the Zeiss Sonnar 5.6/250...
Even with the same lens/film/technique/magnification, depth of field can
vary considerably depending on subject contrast (high contrast subjects
will hold some detail farther from the plane of focus than low-contrast
subjects).
This means that all bets are off if you want to compare two different
imaging technologies using different optical systems and printing
techniques in terms of depth of field. The general trend that you need a
much faster lens to get a given amount of depth of field on a smaller
sensor/film still holds though.
Bernard
from contax mailing list:
Date: Fri, 11 Jan 2002
From: muchan [email protected]
Subject: Re: [Contax] The Digital Way
"Cousineau, Bernard" wrote:
> An other interesting monkey wrench in this calculations is that two lenses of
> the same focal length and aperture don't necessarily have the same depth of
> field! There is an interesting thread on photo.net (entitled "CFI 250
> Superachromat vs CFE 350 Tele-superachromat") in which Kornelius J. Fleischer
> of Zeiss explains that the Zeiss Superachromat 5.6/250 displays noticeably
> less depth of field than the Zeiss Sonnar 5.6/250...
Maybe I'm still in the center of CoC again, but I think this difference is
not hard to understand... if a lens has more spherical abberation, for
example, than another lens of the same focal length, a point-light source
would make bigger circle on film, doesn't it? So, the Sonnar wide open
could have more DOF than Tessar wide open... (???)
muchan (don't know anything about Superchromat, though...)
from contax mailing list:
From: "Ginny" [email protected]
Subject: RE: [Contax] The Digital Way
Date: Thu, 10 Jan 2002
Luminous-Landscape (a great website) has a good discussion on COF and DOF
and he discusses what Michael has just brought up about the subject
remaining at the same size with any lens you get the same DOF.
http://www.luminous-landscape.com/dof.htm
Ginny
From: Frank Loeffel [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Arca-Swiss depth of field device
Date: Tue, 05 Feb 2002
harvey weiss wrote:
> An Arca-Swiss depth of field indicator is illustrated in a photo on
> p.29 of N.McGrath, Photographing Buildings (2nd ed, 1993).
> I would appreciate any information concerning its usefulness and its
> availability.
Here I go spreading misinformation on the LF group again.
It's called Brainbox and has been discontinued for many years.
There's a discussion on it on
http://hv.greenspun.com/bboard/q-and-a-fetch-msg.tcl?msg_id=0031UM
If anyone see one for sale I'd like to know.
Frank Loeffel
From: "Leonard Evens" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: DOF "overrated"?
Date: Sun, 17 Mar 2002
"Mxsmanic" [email protected] wrote:
> "Leonard Evens" [email protected] wrote ...
>
>> Of course the focal length of the human optical system is pretty short,
>> but I have no idea what a reasonable "circle of confusion" might be for
>> the retina.
>
> About 30 seconds of arc, under ideal conditions, and many times more in
> less than ideal conditions.
If my calculations are correct, that comes out to about .0032 mm on the
retina. (I'm using a focal length of about 22 mm for the eye.) I'm not
sure what the relative aperture is for the eye, but again if my
calculations are correct, at f/22, this would yield a hyperfocal distance
of about 22 feet and at f/44 about half of that. If the eye the latter
aperture were accurate and the eye actually focuses at about 10 feet
instead of at infinity, then according to the standard formulas,
everything from 5 feet to infinity would appear in focus.
Of course, there is no particular reason to believe that the standard
analysis of depth of field applies to the eye, and the estimates for
aperture may be way off. But it is interesting that this calculation
seems consistent with my personal observations. So maybe the orders of
magnitude aren't far off. Maybe I should add that with an interocular
lens inserted during cataract surgery, the focal length of my eye's
opical system is probably significantly greater than 22 mm since before
that I had quite a lot of myopia. It is probably something more like 26
mm, and that would increase the hyperfocal distance and decrease the
depth of field.
--
Leonard Evens [email protected]
From: "res0ajxi" [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: DOF
Date: Sat, 26 Jan 2002
After reading Dr. Harold Merklinger's book, 'Focusing the View Camera', I am
convinced that the Scheimphlug Principle alone is not sufficient for
focusing. Also, he shows depth of field and has charts for various lens
tilts. The book was very illuminating.
I have built a calculator/table generator to using Dr. Merklinger's book as
a basis. It's located at http://www.MrCalculator.com/focus.html.
Larry
From: "Michael K. Davis" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Formula for diffraction, was: Re: DOF "overrated"?
Date: Sun, 7 Apr 2002
Hi Bogdan!
Bogdan Karasek [email protected] wrote:
: Hi,
[snip]
: I just bought an APO-Nikkor 305mm/f9-f128. I was wondering if
: there was a way of calculating beforehand if diffraction sets in
: at f128??? What effects should I be looking for on the negative
: (8x10)? Since the APO-Nikkor is a process lens, would it handle
: diffraction differently from a regular photo lens of similar
: focal length?
: Thanks for any help that may be forthcoming.
: Regards,
: Bogdan
Quoting David Jacobsen's Lens Tutorial at Photo.net:
http://www.photo.net/learn/optics/lensTutorial
***
Approximating the aperture-to-film distance as f and making use of the
fact that the aperture has diameter f/N, it follows directly that the
diameter of the first zero of the diffraction pattern is 2.43934*N*lambda.
Applying this in a normal photographic situation is difficult, since the
light contains a whole spectrum of colors. We really need to integrate
over the visible spectrum. The eye has maximum sensitivity around 555 nm,
in the yellow green. If, for simplicity, we take 555 nm as the wavelength,
the diameter of the first zero, in mm, comes out to be 0.00135383 N.
As was mentioned above, the normally accepted circle of confusion for
depth of field is .03 mm, but .03/0.00135383 = 22.1594, so we can see that
at f/22 the diameter of the first zero of the diffraction pattern is as
large is the acceptable circle of confusion.
***
Thus, to answer your question Bogdan, let's consider this formula:
N at which diffraction becomes visible
= Maximum CoC diameter / 0.00135383
Given this formula, your question becomes: "What Circle of Confusion
diameter do I want to use in my Depth of Field calculations?"
Here's the answer:
Maximum Permissible On-film CoC Diameter
= 1 / Desired Print Resolution / Enlargement Factor
We're almost there. The question now becomes: "What is my desired Print
Resolution?"
Most people would say somewhere between 5 and 10 lp/mm if the final print
is to survive scrutiny at a viewing distance of 10 inches.
To accommodate potential film flatness problems associated with a sheet
film as large as 8x10, we should probably calculate DoF with a fairly
aggressive demand for small CoC's, so let's go for a print resolution of
at least 7.5 lp/mm. Let's plug that into the last formula I gave you
above and assume for this exercise that you will be making 16x20 prints
where the enlargement factor is 2:
Maximum Permissible On-film CoC Diameter = 1 / 7.5 / 2 = 0.06667 mm
Now, we can use this CoC diameter in our DoF calculations, but let's also
use it in our diffraction calculations.
Going back to the first formula I provided, above:
N at which diffraction becomes visible = Maximum CoC diameter / 0.001353
N at which diffraction becomes visible = 0.06667 / 0.001353 = 49.27
So: As you stop down toward f/45 from wider apertures, CoC diameters will
be decreasing in size (more DoF) as diffraction's Airy disk diameters are
increasing in size. But if you go just a little bit beyond f/45 on your
way to f/64, at f/49.27, diffraction's Airy Disks will reach the same size
as your chosen maximum permissible CoC diameter. At this point you would
not want to stop down any further (not if you want to achieve a resolution
of 7.5 lp/mm in a 16x20 print from 8x10 format that will be viewed at a
distance of 10 inches.)
The fact that your lens is a true APO design helps it avoid chromatic
aberration but does not overcome the physics of diffraction. To make use
of f/64, f/90 or f/128 without causing visible diffraction, you will have
to decrease the enlargement factor or increase the viewing distance or
both (relative to the criteria given in the scenario above.) An 8x10
contact print viewed at 10 inches will deliver the equivalent of 7.5 lp/mm
Airy disks at f/100 (twice f/49.27, rounding up), for example. So you
could shoot at f/90 without concern for a viewing distance of 10 inches as
long as your enlargement factor is only 1 (an 8x10 contact print). You
could use f/128 if you had an enlargement factor of 1 and also forced a
minimum viewing distance of 20 inches (or an enlargement factor of 2 with
a minimum viewing distance of 40 inches, etc.)
You were wise to ask about using f/128 with 8x10 - they don't go together
very well - using that aperture with the 4x5 format would be even worse --
Mike Davis
Date: Wed, 06 Feb 2002
From: george day [email protected]
To: [email protected]
Subject: [HUG] Handy online DOF calculator
Folks,
I've been horsing around, drawing up some DOF tables for my LF gear, and
came across this Web site:
http://www.dof.pcraft.com/dof.cgi
If it's accurate, this is extremely cool!
Date: Thu, 7 Feb 2002
From: Austin Franklin [email protected]
To: [email protected]
Subject: RE: [HUG] Re: Handy online DOF calculator
My understanding is Leica uses 0.025 and others use 0.033 for 35mm. The COC
varies per format as follows:
35 mm 0.025 mm COC
35 mm 0.033 mm COC
APS 0.025 mm COC
6x45 cm 0.05 mm COC
6x6 cm 0.06 mm COC
6x7 cm 0.065 mm COC
6x9 cm 0.095 mm COC
5x4 inch 0.15 mm COC
10x8 inch 0.3 mm COC
Here is yet another COC calculator:
http://www.darkroom.com/Images/DOFCalc.html
Austin
....
Date: Thu, 07 Feb 2002
From: Mike Kirwan [email protected]
To: [email protected]
Subject: RE: [HUG] Handy online DOF calculator
I use one that is free and runs on my Palm Pilot, very handy for field work.
Comes with an excellent 40 page user guide. It is called "vade mecum" and
written by Robert E. Wheeler and can be downloaded from
http://www.bobwheeler.com/photo/Software/software.html
Mike
From: "Leonard Evens" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: DOF "overrated"?
Date: Tue, 02 Apr 2002
"John Stafford" [email protected] wrote:
> "Leonard Evens" [email protected] wrote
>> [...]
>> The diameter of the Airy disc is given by C*l*N
>>
>> where l is the wavelength of the light source, N is the f-number of
>> the lens opening, and C is a proportionality factor depending on the
>> units.
>
> Does the color (wavelength) of the subject have a _significant_ effect
> upon _everyday_ photography? For example, the differences of a blue
> subject to a red one... discounting for the moment the problem of focus
> planes.
Wavelength is usually measured in angstroms. Visible light ranges from
7000 angstroms for red light down to 4000 or 3500 angstroms for violet.
I'm not sure what aspects of every day photography you are concerned
with. The Airy disc would have about twice the diameter for red light
as for violet light for the same f-stop. But the eye is not equally
sensitive to light of different wavelengths as noted before.
Any normal illumination won't be monochromatic, that is consist of light
of one wavelength. It will be a mixture of different wavelengths. The
dominant trend of the light is described by the term "color temperature".
In real life situations the perceived color
of subjects is not a simple function of the wavelength of the light they
reflect. The human visual system makes adjustments so that scenes
viewed under very different lighting still look more or less the same.
That is called color constancy. So color as perceived by the human
visual system is not at all the same thing as wavelength.
The same is not true for photosensing systems such as film or sensor
arrays in digital cameras. A picture taken under incandescent lighting
on film designed for daylight will have a strong reddish cast, for
example. There are films designed for different kinds of illumination,
and digital cameras usually allow one to set the white balance to
deal with this.
These days, most artificial lighting is done with flash which
is designed to mimic daylight lighting. But critical photographers
certaintly have to be aware of the character of the illumination of the
scene as well as how their films or sensor arrays respond..
Even for viewing color prints, despite color constancy, the
color temperature of the illumination can make a difference in human
perception of the colors. When I did color darkroom work, I used
special flourescent bulbs with daylight color balance for judging the
color of my prints. Similarly, the color temperature of a monitor can
make a subtle difference in how colors on the monitor are perceived.
--
Leonard Evens [email protected] 847-491-5537
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
From: [email protected] (Ross Bagley)
Date: Tue, 02 Apr 2002
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: DOF "overrated"?
...
>Does the color (wavelength) of the subject have a _significant_ effect
>upon _everyday_ photography? For example, the differences of a blue
>subject to a red one... discounting for the moment the problem of
>focus planes.
Well, there are several lenses that have the appelation "apochromatic"
or APO, presumably because other similar designs are not designed with
the goal of eliminating chromatic abberation.
Chromatic abberation is another lens artifact where the different
wavelengths of light follow different paths through the optics
(exactly the same way different wavelengths of light follow different
paths through a prism). That prismatic separation of colors occurs in
the off-axis portions of the lens and the amount of abberation differs
with the materials chosen for each lens element.
Older wide aperture lenses tend to display this artifact better,
probably due to changes in glass chemistry and modern multi-coatings,
but enlarging lens manufacturers will still uniformly recommend their
APO lenses over their non-APO lenses for color enlargement.
Regards,
Ross
-- Ross Bagley http://rossbagley.com/rba
From camera makers mailing list:
From: Dave Schneider [email protected]
To: [email protected]
Subject: RE: [Cameramakers] standard movement on camera
Date: Fri, 5 Apr 2002
Murray,
The way I would see it, without those movements you don't really have a view
camera, just a big, clunky, hard to carry around camera that uses big film.
The movements are the main (perhaps only) reason I use a view camera. The
control of focus plane and perspective is the main attraction. If you are
not able to employ any movements then a medium format camera would be easier
to use and produce nearly equal results.
Many times people discuss the great depth of field available with a view
camera. This is a misconception since the DOF is actually less with a normal
lens on a 4x5 camera than you would have with a smaller format. What a view
camera does allow, which is often confused with DOF, is the ability to
adjust the plane of focus. In any other camera the plane of focus is
parallel to the film. With a view camera employing front tilt the plane of
focus could run from your toes to the top of the light post down the street.
Now the depth of field, that is the sharp focus on either side of that
plane, may be quite small. Small enough that while the top of the light post
could be in sharp focus the no parking sign tacked to it 6 feet above the
ground may be very out of focus.
I would consider front tilt, swing and rise to be essential. There are many
good camera designs that have these movements and no, or very limited, rear
movements. The rear movements are almost a luxury in the sense that you can
accomplish everything with front only movements if you are willing to move
the camera around on the tripod after you made the front movements. If you
are doing landscapes and similar things where only limited movements are
used a front movement only design would not be very limiting. But don't go
to all the work of building a camera without any movements.
Dave
From: "Q.G. de Bakker" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Kowa 250/5.6 vignetting at close focus
Date: Tue, 26 Mar 2002
Robert Monaghan wrote:
> re: digital manipulation -
> yep, I've been impressed by the PBS specials, such as the Empire of the
> Bugs (?) where DOF went from bug level to distant fields etc. Amazing!
> I suspect we will see some amazing microscopy shots now with digitally
> enhanced imaging - esp. if C. Zeiss adds this technology to the 'scopes
> http://www.cdm-optics.com/site/pr/zeiss-cdm.html Even layering of multi
> images in photoshop is a potential solution to some DOF issues...
"Amazing" DOF (like in this Empire of Bugs thing) can be achieved with quite
down to earth ancient techmology.
Pick a lens, any lens, and point it at anything of interest. Then, put a
tiny object, like a bug, in the image space behind that lens. Next, pick a
camera and macro lens, and focus it on the bug. Then move the first lens
back and forth until the image it produces is picked up by the camera lens
and you get both the bug and the background in focus simultaneously. Kind of
back projection, but without a screen.
From rollei mailing list:
Date: Fri, 03 May 2002
From: Bob Shell [email protected]
Subject: Re: [Rollei] Re: Rolleiflex pics
Dave Wyman at [email protected] wrote:
> "Q.G. de Bakker" [email protected] wrote:
>
>> I'd like to know if depth of field would have been the same if I'd used
> 35mm gear and enlarged
>> the image to the same size as the Rollei photograph. Does anyone know the
> answer?
>
> The answer is yes.
>
> That's good to know. The subject of DOF is complicated. But less so for me,
> now.
>
> Dave
Good to know, but wrong. If you shoot two photos, one on 35mm and one on
medium format, and shoot both at f/8, the depth of field will not be the
same on the same sized print. The 35mm shot will have more depth of field.
To get the same depth of field, the rule of thumb is to stop down two stops
more on the medium format. So to get about the same depth of field you
would shoot the 35mm at f/8 and the medium format at f/16.
Bob
From rollei mailing list:
Date: Fri, 3 May 2002
From: [email protected]
Subject: Re: [Rollei] Re: slight OT DOF in MF vs. 35mm
From Bob Shell
> > ...The 35mm shot will have more depth of field....the rule of
> > thumb is to stop down two stops more on the medium format.
From P. Tempel:
> ...Assuming 50mm lens on both 35mm and 6x6, they should have the
> same DOF, no?
I think that both of you are right in a sense but in fact DOF is not
decided *only* by the focal length : it is eventually decided by
*your* sharpness criteria. So you do what you want according to the
final examination of the print ; if the print is to be enlarged 8x,
plus examined with a loupe ;-);-) you'll need of course more stringent
DOF rules.
Between a 50mm lens used on a 35mm camera and a wide-angle 50mm in 6x6
you'll probably not enlarge to the same degree of magnification. So
the conditions under which you examine the final image decide of what
is the acceptable sharpness and thus what is the reasonable value for
the diameter of the circle of confusion "c", the hyperfocal distance
on which all DOF scales are based being equal to H = (f*f)/(N*c) (N =
numerical aperture). A good but stringent rule would be to say : I
want professional images and a significant increase in quality vs.
35mm cameras with my expensive medium format system. Consequently I'll
apply the same values in 6x6 for "c" as I use in 35mm. This makes
sense and yields Philippe's conclusions (H and DOF would not change
for the same focal length whatever the format wight be) but does not
correspond to conventionally engraved DOF scales, even on professional
MF cameras.
Usually DOF scales are engraved by manufacturers on the base of a set
of purely conventional, indicative, DOF circles. Common values for "c"
are 24-35 microns in 35mm cameras, about 50 microns in 6x6cm (51
microns is what I get from my R-TLR scales). But Hasseblad uses an
"impossible" "unrealistic" value of 75 microns on some CF lenses !!
(this was discussed before). So you are free to work with a 30 micron
value for "c" but this is not what is engraved on the barrel of a
80/2.8 planar, for example.
Even more : as a paradox if you take in 6x6 a bigger and less
stringent value for "c" (because the final print will be enlarged
less), this yields a smaller value for H in 6x6 when the focal length
"f" and aperture "N" are the same! so, paradoxically, a *smaller*
value for H implies that DOF would *increase* in 6x6 !
Bob's excellent rule of 2 f-stops (exactly the opposite conclusion !!)
is in fact the most reasonable practical, usable trade-off that
actually takes into account the fact that you do not enlarge the print
as much in 6x6 as in 35mm, but that you also demand an excellent final
image quality. But as far as I understand this rule holds for
comparison of DOF between two lenses of the same angular coverage but
different focal lengths, e.g. a 50mm lens on 135 film compared to DOF
you get with a 80 in 6x6. In this case the 2 f-stop IMHO is the most
realistic and can be explained easily from the hyperfocal formula.
--
Emmanuel BIGLER
[email protected]
From Rollei Mailing List:
Date: Fri, 03 May 2002
From: "John A. Lind" [email protected]
Subject: Re: [Rollei] Re: Rolleiflex pics
There are numerous errors by several folks in this thread. I worked out
the numbers for all this a while ago.
Alas, I'm at home during lunch and will post a more definitive response
later this evening. Two points now though . . .
(1a) CoC diameter, at whatever arbitrary diameter one defines it for
necessary print sharpness to appear "in focus" increases in linear
proportion to increases in linear film format dimensions. Considered
alone, this also *increases* the DoF in linear direct proportion. But
wait, there's more.
(1b) To maintain the same perspective for an equivalent image, lens focal
length must also increase in linear proportion to increases in linear film
format dimensions. Considered alone, an increase in focal length
*decreases* the DoF in inverse proportion to the square of the focal length.
(1c) The net result of (1a) and (1b) is a linear decrease in DoF.
(2) From (1a) to (1c) above, DoF for same _angle_of_view_ (or perspective)
decreases with increases in film format size if the same level of
"sharpness" is demanded in the same size prints from each. To compensate
and achieve the same DoF from the same perspective (AoV), one must stop
down the lens more if film size increases. However, comparing 35mm small
format to "medium format" requires caution. What size medium format: 645,
6x6, 6x7, 6x9, etc.? They're all different film frame sizes. One must
stop down about two stops from 35mm small format when using 6x9 medium
format. 6x7 is about 1.5 stops, and 645 or 6x6 are approximately one stop.
"More at 11" . . .
-- John
...
From rollei mailing list:
Date: Fri, 03 May 2002
From: Bob Shell [email protected]
Subject: Re: [Rollei] Re: Rolleiflex pics
Philippe Tempel at [email protected] wrote:
> Are you talking both lenses the same focal length or
> lenses "normal" for each format? Assuming 50mm lens
> on both 35mm and 6x6, they should have the same DOF,
> no? The only problem is the wider angle of veiw on
> the 6x6 camera (makes you move in closer the the
> subject to get the same angle of view and if you're
> closer it will make foreground objects appear larger).
I'm assuming lenses with roughly the same angle of view.
But in your example of a 50mm lens uses on both formats at the same f/stop,
different degrees of enlargement to produce an 8 X 10 would have an effect.
Perhaps small.
Bob
From: "Q.G. de Bakker" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: DOF "overrated"?
Date: Tue, 19 Mar 2002
[email protected] wrote:
> Then why do Seagulls have visible diffraction at f22 and Lubitels do not?
Because Seagulls have better lenses to begin with, so the degradation due to
diffraction is noticeable, while the Lubitel's lens is always worse than
diffraction can make it be.
> And then why do some lenses stop down to f45 and others don't? I would
> think how far a lens can stop down is a result of the design process as
> well as physics.
Yes. Stopping down (a bit) usually decreases the effect of aberrations (a
lens design issue) while at the same time diffraction starts degrading the
maximum theoretically achievable resolution. But as long as the effects of
aberrations are worse, they, not diffraction, limit the performance of the
lens. So if stopping down reduces the design flaws enough image quality can
actually improve.
But at some point there is no more improvement, and with further stopping
down only the increasing diffraction will be present. When this happens is
indeed a matter of design (which by the way, is not separate from physics).
> Understandably meniscus lenses will have different
> characteristics than three element lenses but it appears that there is a
> range of differences even between three element designs. And I would also
> think that there is a reason why a lens designer would design a lens to
> stop down to f45, otherwise all lenses would stop down to pretty much the
> same aperture.
In a great many lenses the minimum aperture depends on the maximum aperture.
Many mounts have a set number of stops, resulting in f/5.6 lenses having a
minimum aperture one stop smaller than f/4 lenses, etc.
In wide angle lenses however minimum aperture usually is restricted.
From rollei mailing list:
Date: Sat, 04 May 2002
From: "John A. Lind" [email protected]
Subject: Re: [Rollei] Re: Rolleiflex Pics
Dave Wyman wrote:
>Somehow I not happy with the assertion that it's necessary to stop down
>two f/stops to achieve the same DOF between 35mm and MF. I'm not happy,
>not because I disagree, but because I still don't understand it yet, at
>least under the conditions I mentioned.
>
>My original conditions called for two images of the same subject, with the
>same area in view, enlarged to the same size, one with 35mm film, one with
>MF film.
Dave,
With a 35mm fitted with a 50mm lens and a 645 or 6x6 fitted with an 80mm
lens, it's about 1 f-stop, *not*, repeat *not* two f-stops. I'm writing up
the detailed posting about it all today. Couldn't find my notes from
several years ago. Needed to derive all the equations again and test it
all in a spreadsheet again . . . just to make certain if I were to assert
something that varied from statements made by some members of the list, it
was based on *fact* and not anecdotal information or heuristics.
BTW, one cannot simply use the hyperfocal equation be accurate, although it
is an approximation and it should explain why the DoF changes.
H = f^2/(A*c), where:
H = hyperfocal distance
f = focal length
A = lens aperture
c = maximum allowable circle of confusion diameter on the film
Solving for lens aperture:
A = (f^2)/(H*c)
Both "f" and "c" increase in direct proportion to increased linear
dimensions of film size for the same apparent DoF in a print. Thus, if you
increase linear dimensions of the film by some value "x" it produces:
A = [(x*f)^2]/(H*x*c)
A = [(x^2)*(f^2)]/[x*(H*c)]
A = x*[(f^2)/(H*c)]
In order to maintain the same hyperfocal distance, aperture number must
increase by "x" with a linear dimension increase of "x" in the film
size. To maintain the same exact, apparent front and rear boundaries for
DoF in the print, the aperture increase is slightly different from this
value, and that is what I'll post later to show in practical application,
645 and 6x6 format requires stopping down approximately one f-stop compared
to 35mm small format (**NOT** TWO).
-- John
From rollei mailing list:
Date: Sat, 4 May 2002
From: [email protected]
Subject: Re: [Rollei] Re: Rolleiflex Pics
Dave--
If you use the same lens on 35mm and on 21/4 the depth of field will be
the same. The lens is dum, and doesn't know or care what format is used.
Bigger, smaller, the image is the same size. Only one is on bigger film and
showing more of the scene. Formate and size of the print, the image is the
same. So the DOF is the same. Don't be taken in by the idea that 35mm gives
you more DOF tha MF-film. a 50mm on a Blad and the 50 on 35mm are the same.
From Rollei Mailing List:
Date: Sat, 04 May 2002
From: "John A. Lind" [email protected]
Subject: Re: [Rollei] Re: slight OT DOF in MF vs. 35mm
Bob Shell wrote:
>[email protected] at [email protected] wrote:
>
> > Bob's excellent rule of 2 f-stops (exactly the opposite conclusion !!)
[snip]
> > But as far as I understand this rule holds for
> > comparison of DOF between two lenses of the same angular coverage but
> > different focal lengths, e.g. a 50mm lens on 135 film compared to DOF
> > you get with a 80 in 6x6. In this case the 2 f-stop IMHO is the most
> > realistic and can be explained easily from the hyperfocal formula.
>Absolutely right.
>
>Bob
Absolutely incorrect!!
It's approximately ***ONE*** f-stop!!!! Furthermore, it's not the
hperfocal formula alone, although it's an element of it.
For the same DoF using the same critical focus distance between 35mm and a
645, or 6x6 or 6x7 using as much of the frame as possible when making a
print, the following are how many stops the 80mm lenses on 645 or 6x6, and
90mm lenses on 6x7 medium format cameras must be stopped down farther than
a 50mm lenses on 35mm small format cameras to have the same depth of field.
f-Stops to Stop Down
Critical Focus = 15 Feet
645 6x6 6x7
Print Size 80mm 80mm 90mm
------------------------------------
3.5 x 5 inch 1.31 1.31 1.37
5 x 7 inch 1.26 1.26 1.32
8 x 10 inch 1.15 0.93 0.99
11 x 14 inch 1.15 0.98 1.05
Depth of Field (DoF) is based on how a human perceives what is apparently
in focus. Technically, anything not on the plane in space at the critical
focus distance is not in focus. A point in space only forms a point on
film when it's in the plane at critical focus distance. Those points in
space not on the plane at critical focus distance form a spot on the film.
However, this spot at the focal plane must be sufficiently large (i.e., out
of focus) before a human can perceive it is an out of focus spot. This
phenomenon is directly related to the acuity limits of the human
eye. These limits create a zone in front and behind the critical focus
distance in which everything appears to be in focus because the spots are
so small they cannot be differentiated from points. This zone is the
DoF. In practical application a photographer uses this effect when
focusing a camera through the viewing or taking lens using a matte focusing
screen (on a technical field camera, TLR or SLR).
Since DoF is based on human eye acuity limits and perception, any exact
value used is subjective. It is based on the average densities of rods,
L-cones, M-cones and S-cones in the retina across the human population,
resulting in an average acuity of 1 minute of arc (i.e., 1/60th of a
degree). Some people will have greater acuity, and some will have less acuity.
In practical photographic application the goal is determining apparent DoF
in a print. The "average" human acuity is confounded by print size and the
distance at which it is viewed. One of the generally accepted standards is
a 12 inch viewing distance for a 3.5 x 5 inch print. Another (currently
used by Kodak) is a 14 inch viewing distance for a 4 x 6 inch print. Using
the human average acuity of 1 minute arc at a viewing distance of 12 inches
the threshold diameter for a spot to be perceived as a spot versus a point is:
Cp = 2*12*tan[(1/60)/2 deg.]
Cp ~= .0035 inches, or ~0.089 mm
For "Kodak's" 4 x 6 inch print at a 14 inch viewing distance:
Cp = 2*12*tan[(1/60)/2 deg.]
Cp ~= .0041 inches, or ~0.103 mm
The two are in general agreement being approximately 1/1000th of the short
dimension of the print (relevant to printing 35mm small format film).
What, then, is the largest diameter circle of confusion "c" allowable on
film? For 35mm small format, it would be approximately 1/1000th of the
short dimension of the film frame:
24mm/1000 = 0.024mm
This is in general agreement with a commonly used value of 0.025mm for the
maximum circle of confusion diameter used to define DoF on 35mm
film. Other values used are 0.030mm and 0.033mm. I won't get into exactly
*what* the allegedly "correct" value for "c" should be as it's somewhat
subjective, and therefore an *exact* value is somewhat arbitrary. As will
be seen it's not relevant to how many more stops a "standard" lens for
medium format must be stopped down to achieve the same DoF as with a
"standard" lens for 35mm small format.
It should be obvious now that a larger film format can allow a larger
diameter circle of confusion before a point is perceived as a spot (and
therefore out of focus) on the same size print, and that this relationship
is directly proportional to the relative linear dimensions of the film
frame. Following are the linear dimensions of 35mm small format and common
medium format film frames (Kodak's sizes date to when they created 120):
35mm: 24mm x 36mm
645: 2.25 x 1.75 in. (Kodak), or
56mm x 41.5mm (Mamiya)
6x6: 2.25 x 2.25 inches, or
56mm x 56mm (Hasselblad)
6x7: 2.25 x 2.75 in. (Kodak), or
56mm x 69.5mm (Mamiya)
In making an "edge to edge" print, each film format requires magnification
that results in some cropping along one dimension to fill the other
dimension completely. Following are the approximate magnifications
required of each film format to fill common print sizes using the Mamiya
and Hassleblad film frame sizes:
Film Format Magnification by Print Size
Print size 35mm 645 6x6 6x7
3.5x5 in: 3.70X 2.27X 2.27X 1.83X
5x7 in: 5.29X 3.18X 3.18X 2.56X
8x10 in: 8.47X 4.90X 4.54X 3.65X
11x14 in: 11.64X 6.73X 6.35X 5.12X
The 4 x 6 inch print is omitted because it is exclusive to 35mm small
format and this is about making the approximately the same images from
different film formats. There may be some lab, somewhere that will make a
4x6 print from a medium format film frame. I haven't found one yet. They
will make 3.5x5 from both formats. Other print sizes, such as 4x5 and
16x20 have the same aspect ratio as an 8x10 and therefore do not need to be
listed. These print sizes represent unique aspect ratios and they differ
from the film frame aspect ratios.
In maintaining the apparent DoF in a print, the maximum circle of confusion
diameter "c" can be allowed to increase from that for 35mm with the ratio
of its magnification to the magnification of the larger film format being used:
Max. Circle of Confusion Increase
from 35mm Small Format
Film Format
Print size 645 6x6 6x7
3.5x5 in: 1.63X 1.63X 2.03X
5x7 in: 1.67X 1.67X 2.07X
8x10 in: 1.73X 1.87X 2.32X
11x14 in: 1.73X 1.83X 2.28X
Because DoF is also related to lens focal length, the following are
considered the commonly used "standard" lenses used for each format and a
factor for the increase in focal length compared to 35mm small format is
also given:
Format Focal Length Increase
35mm: 50mm 1.0X
645: 80mm 1.6X
6x6: 80mm 1.6X
6x7: 90mm 1.8X
These lengths are used to keep this as a practical application. An exact
equivalent length for each format based on print aspect ratio could be
used, but would produce focal lengths that cannot be found for these film
formats. A 645 varies from 78mm to 86mm, depending on print size, for
exactly the same perspective. A 6x6 would use a 78mm regardless of print
size. A 6x7 would use a 97mm lens. The common "standard" length was
chosen as it represents what is most commonly available from every camera
maker for each film format.
Terms used in the DoF equations:
H = Hyperfocal Distance
f = Focal Length
A = Lens Aperture
c = max. Circle of Confusion Diameter
S = Critical Focus Distance
Sn = Near DoF Boundary Distance
Sf = Far DoF Boundary Distance
The common equations used for DoF calculations are:
H = (f^2)/(A*c)
Sn = (H * S)/[H + (S - f)]
Sf = (H * S)/[H - (S - f)]
The presence of the lens focal length in Sn and Sf are the reason the
equation for hyperfocal distance *cannot* be used for finding how many
f-stops to stop down the lens for the same DoF in a larger film
format. Only one of the two DoF Boundary Distance equations need be
used. Both give the same result.
Solving for Hyperfocal Distance in the equation for Sn:
H = (S - f)/[(S/Sn) - 1]
This equation can be equated to the one for hyperfocal distance:
(f^2)/(A*c) = (S - f)/[(S/Sn) - 1]
Solving this for lens aperture gives:
A = (f^2)*[(S/Sn) - 1]/[c*(S - f)]
Factors for increase in both focal length and in maximum circle of
confusion diameter for each film format and each print size can now be
performed to find the factor by which aperture f-number must be increased
to maintain the same DoF. Example for 35mm to 645 for 11x14 print size:
A = [(1.6f)^2]*[(S/Sn) - 1]/[(1.73c)*(S - 1.6f)]
A ~= 1.6 * {(f^2)*[(S/Sn) - 1]/[c*(S - f)]}
The exact factor changes the third significant digit slightly as critical
focus distance is varied. Approximate factors by which f-number increases
for film format and print size at critical focus distances from 5 to 50 feet:
Film Format
Print size 645 6x6 6x7
3.5x5 in: 1.6X 1.6X 1.6X
5x7 in: 1.5X 1.5X 1.6X
8x10 in: 1.5X 1.4X 1.4X
11x14 in: 1.5X 1.4X 1.4X
To find the number f-stops from some f-number multiplier "M":
f-stops = log(M^2)/log(2)
Either Common or Naperian Logarithms may be used.
Crunching the numbers for a critical focus distance of 15 feet gives the
following:
f-Stops to Stop Down for Same DoF as 35mm
Critical Focus = 15 Feet
645 6x6 6x7
Print Size 80mm 80mm 90mm
------------------------------------
3.5 x 5 inch 1.31 1.31 1.37
5 x 7 inch 1.26 1.26 1.32
8 x 10 inch 1.15 0.93 0.99
11 x 14 inch 1.15 0.98 1.05
Now . . . I'll state it again . . .
It's NOT two f-stops!!!!
-- John
From Rollei Mailing List:
Date: Sun, 05 May 2002
From: Bob Shell [email protected]
Subject: Re: [Rollei] Re: Rolleiflex Pics
John A. Lind at [email protected] wrote:
> With a 35mm fitted with a 50mm lens and a 645 or 6x6 fitted with an 80mm
> lens, it's about 1 f-stop, *not*, repeat *not* two f-stops.
Sorry, John, but it is two stops. Tried and proved quite a few years ago,
and a standard rule of thumb since. You can spout equations for days, but
if you actually make the prints you will see that it requires two stops to
get them to look the same. QED.
Bob
From Rollei Mailing List:
Date: Sun, 5 May 2002
From: Austin Franklin [email protected]
Subject: RE: [Rollei] Re: slight OT DOF in MF vs. 35mm
Here's an on-line depth of field calculator:
http://www.darkroom.com/MiscDocs/DOFCalc.html
It's a Unix server, so the URL is case sensitive.
From Rollei Mailing List:
Date: Sun, 05 May 2002
From: "John A. Lind" [email protected]
Subject: RE: [Rollei] Re: slight OT DOF in MF vs. 35mm
Austin Franklin wrote:
>Here's an on-line depth of field calculator:
>
>http://www.darkroom.com/MiscDocs/DOFCalc.html
>
>It's a Unix server, so the URL is case sensitive.
Yes, there are a number of these around. Look at the CoC values
though. These should be adjustable and my "complaint" with them is most do
not allow setting them to desired values. A 3.5x5 print made from as much
film frame as possible (minimum cropping) at 12" viewing distance and 1
minute arc visual acuity requires approximate CoC values (on film) for each
film format as follows (rounded to nearest 0.005mm):
3.5 x 5 Inch Print CoC
35mm 0.025mm
645 0.041mm
6x6 0.041mm
6x7 0.051mm
If 0.025mm CoC is used for 35mm film with a 11x14 print, the values for
other film formats shift because the aspect ratio of the print is different
and requires comparatively different enlargement ratios:
11 x 14 Inch Print CoC
35mm 0.025mm
645 0.043mm
6x6 0.046mm
6x7 0.057mm
IOW, the CoC values for larger formats tend to be too large compared to
that used for 35mm. Why? They don't take into account the true
magnification required to make a print for each common print size. They
are based more on the length of the film frame diagonal and the ratios of
them between film formats.
These CoC values are also presumptive that larger prints will be viewed at
greater distances, in direct proportion to linear size dimensions (e.g. a
16x20 is viewed from twice as far as an 8x10). In practice this isn't
true. Large prints (8x10 to 16x20) made for juried exhibition will be
examined at about 12-14 inches regardless of size. A casual viewer
browsing through the gallery may back off to about 18 inches. If
"accepted" CoC values are used blindly for critical DoF control, most
notably for using hyperfocal focusing or with close macro work, the
apparent DoF in the print will be noticeably shallower than was
"calculated." I have seen larger prints in which it became obvious the
photographer used hyperfocal focusing for maximum DoF without considering
the size of print that would be made (using. The very distant background
looks slightly fuzzy as a result. For 35mm and 645 destined for large
prints I tighten up the CoC to bring it more in line with print size and
anticipated viewing distance based on 1 minute arc of visual acuity.
If the model in the link is used for different film formats at the same
distance, even a one stop stop-down from 35mm to the medium format sizes
produces a larger depth of field. This is the first clue that the CoC
values for the various medium format film sizes are too large.
-- John
From rollei mailing list:
Date: Sun, 5 May 2002
From: Austin Franklin [email protected]
Subject: RE: [Rollei] Re: slight OT DOF in MF vs. 35mm
John,
> Austin Franklin wrote:
> >Here's an on-line depth of field calculator:
> >
> >http://www.darkroom.com/MiscDocs/DOFCalc.html
> >
> >It's a Unix server, so the URL is case sensitive.
>
> Yes, there are a number of these around. Look at the CoC values
> though. These should be adjustable and my "complaint" with them
> is most do
> not allow setting them to desired values.
Well, whether they "should" or not is questionable. These are based on a
"standardized" viewing distance and an average visual acuity. Most people
don't understand how and why the CoC is derived, and prefer to just have
default values.
I am not disagreeing that it might be useful to some few people to be able
to vary the CoC either directly, or by varying some variables in an equation
that derives CoC, but I do believe that number is very few.
> If the model in the link is used for different film formats at the same
> distance, even a one stop stop-down from 35mm to the medium format sizes
> produces a larger depth of field. This is the first clue that the CoC
> values for the various medium format film sizes are too large.
I don't understand what you are saying here. If you take a 35mm/CoC 0.025
using 50/f2.8, you will get near the same DOF with a 6x6/CoC 0.060 with
80f2.8. I don't understand what your one stop-down issue is here? The two
SHOULD be the same, that's the point of the CoC.
Austin
From rollei mailing list:
Date: Mon, 06 May 2002
From: "John A. Lind" [email protected]
Subject: RE: [Rollei] Re: slight OT DOF in MF vs. 35mm
Austin Franklin wrote:
>John,
>
>I don't understand what you are saying here. If you take a 35mm/CoC 0.025
>using 50/f2.8, you will get near the same DOF with a 6x6/CoC 0.060 with
>80f2.8. I don't understand what your one stop-down issue is here? The two
>SHOULD be the same, that's the point of the CoC.
No . . . the DOF for a 6x6 using an 80mm lens should be shallower compared
to a 50mm lens on a 35mm camera, if focus distance and lens aperture are
held constant. (Bob Shell and I agree on this, but disagree by how
much.) The CoC can increase, but it can only do so at the same rate as the
linear dimensions of the film increase. This is my problem with the
model. It uses 0.025mm for 35mm format and over twice that for a 6x6. The
6x6 film frame dimensions are less than twice that for 35mm. The relevant
dimensions in making common size prints from the two are 36mm versus 56mm,
which is about 1.56X larger, not the 2.4X larger implied by going from
0.025mm to 0.060mm for the CoC. The CoC for the 6x6 should be about
0.040mm if 0.025mm is used for 35mm format.
The focal length must also increase at the same rate as the linear
dimensions of the film. If either of the equations for near limit or far
limit of the DOF are solved for aperture though, a key factor is "f^2/c"
and the rest of the equation is multiplied by this. I will use the example
of two common technical field camera sizes to simplify the math:
A 4x5 camera has a 6 inch lens; call this 6" focal length "f". For the
same perspective (field of view) with an 8x10 camera it must must have a 12
inch lens because the linear dimensions of the sheet film have doubled and
it becomes "2f". If the circle of confusion diameter for the 4x5 is "c",
it can can also double from "2c" with the 8x10. In maintaining the same
near and far DOF boundaries and solving for lens aperture number, you end
up with this factor being:
(f^2)/c for the 4x5, and
[(2f)^2]/(2c) for the 8x10
Simplifying, it becomes:
4(f^2)/(2c) and
2(f^2/c)
The aperture number doubles from this factor alone, which is 2 f-stops
(e.g., if it was f/5.6 for the 4x5 it becomes f/11 for the 8x10). In going
from a 4x5 to an 8x10 field technical camera, at focusing distances longer
than about 10 feet, the lens aperture must be stopped down by two stops to
maintain essentially the same DOF. The rest of the equation does affect
the exact result, but it's very little at focusing distances greater than
about 10 times the lens focal length and the "f^2/c" it's multiplied by
becomes the overwhelming effect.
-- John
From Rollei Mailing List:
Date: Mon, 06 May 2002
From: Bob Shell [email protected]
Subject: Re: : RE: [Rollei] Re: slight OT DOF in MF vs. 35mm
[email protected] at [email protected] wrote:
> Thus even if the formula yields about 1 f-stop, 2 f-stops IMHO are
> more realistic for professional image quality.
This is what I originally said that sparked such a debate.
One thing I always tell my students when they begin to get balled up in
trying to figure things out with math is the simple but often forgotten two
word maxim, "Try it". Rather than debating it based on math and philosophy
as the Greeks did, this is a much more pragmatic approach.
If you actually try it you will find that two stops is what works.
Bob
From rollei mailing list:
Date: Mon, 6 May 2002
From: [email protected]
Subject: Re: : RE: [Rollei] Re: slight OT DOF in MF vs. 35mm
From Dave:
> I have been following this topic...with great interest....Could the
> group as a whole help me to understand better the physics, why I
> picked on John's mailing is quite simple.
Dave. Everything is contained in 3 simple formulae, valid for
far-distant objects.
1) H = f*f/(N*c) H = hyperfocal distance, f focal length N numerical
aperture, c CoC value. H determines **everything**.
2) 1/p1 = 1/p + 1/H p1 = near limit of sharpness, p nominal focusing
setting
3) 1/p2 = 1/p - 1/H p2 = near limit of sharpness
You do not need a powerful computer to do this nor any spreadsheet or
whatever.
The only possible controversy is on the choice of the value for the
Coc "c" and how you compare things between various formats. If you
consider that "c" is scaled like "f", then H is proportional to f/N
i.e. H scales like f at a given numerical aperture N. More generally
to keep H constant between formats you get a formula (see above where
it comes from)
N2 = N1 * (f2*f1)^2 * (c1/c2)
this contains everything and there is little to argue except putting
figures in this formula.
Example. Take c1 = 33 microns in 35mm photography, C2=50 microns in
6x6cm. Now consider a f1=43mm lens in 135 (exact format diagonal) with
a CoC of 33 microns, and a F2=79mm in 6x6 (diagonal of a 56x56 frame)
with a CoC of 50 microns. Then if you want to keep H constant you
should scale N by 1.51 i.e. slightly more than 1.4 (1 f-stop).
Now in medium format you demand ultimate image quality. So there is
little use of sharpness criteria defined before WWII for amateur
cameras. Do not forget that a value of 50 microns for a CoC
corresponds to a resolution limit of 10 line pairs per mm. All readers
know that their MF lens can resolve at least 50 lp per mm. Zeiss
claims 100 lp/mm for lens+(good film) combinations with modern MF
lenses. So there are plenty of good reasons to stop down more than
what is given by the convenional DOF formula.
Thus even if the formula yields about 1 f-stop, 2 f-stops IMHO are
more realistic for professional image quality.
The physics behind is simply geometrical optics. Three limits for the
above mentioned formulae :
i) the value for the CoC should be in any case bigger than the
diffraction spot, roughly equal to N microns whatever the film format
migh be. In other words using DOF formulae in 6x6 at f/128 is a
nonsense.
ii) for far-distant objects p, p1, and p2 being measured from the
lens, you shoud use slightly more complex formulae, valid also in
macro :
4) 1/p1 = 1/p + 1/H*(1-f/p) p1 = near limit of sharpness, p nominal
focusing setting
5) 1/p2 = 1/p - 1/H*(1-f/p) p2 = near limit of sharpness
This holds even in macro and yields the value of 4*N*c for the total
DOF range (p2-p1) at 1:1 (2f-2f) magnification ratio.
iii) above mentioned formulae are derived from a symmetrical or
quasi-symmetrical lens formula. For an asymmetric retrofocus or
telephoto lens where the pupil magnification factor is different from
unity, DOF formulae are more complex but as far as the Hyperfocal
distance is concerned there is little change ; even in a 50 mm
distagon with a pupil magnification factor of 1.8, when H is much
larger than f, H=f*f/(N*c) will be valid in most cases.
-- Emmanuel BIGLER [email protected]
From rollei mailing list:
Date: Mon, 06 May 2002
From: "John A. Lind" [email protected]
Subject: RE: [Rollei] Re: slight OT DOF in MF vs. 35mm
Austin Franklin wrote:
>John,
>
>Understood, but that meets YOUR requirements, it is arbitrary for other
>people's needs though. The "common" CoC is based on the diagonal
>measurement, and I believe is far more applicable for what most people do.
>Another issue is when projecting rectangular images, you have to have a
>screen that accommodates both horizontal and vertical, unless you can
>guarantee only one orientation for all your images.
Absolutely. All screens I've used are square. That means both horizontal
and vertical images (from the same film format) are equally magnified.
>6x6 is a VERY common format, and makes for very common print sizes, at least
>with the labs I have used, and the album books that I have printed for.
>BTW, paper for digital printing follows standard printer paper sizes, like 8
>1/2 x 11, 11 x 17, 13 x 19 etc. So what's considered "common" is not so
>common any more. I don't believe most people print TO the entire edges of
>the paper, and also do crop for non-square formats, at least in one
>dimension.
Agreed about 6x6 being a common film format. Perhaps I wasn't clear on
this. My point was most common rectangular print aspect ratios are not the
same as most common film format aspect ratios. Those with their own
darkrooms can make custom print sizes at will. Those who do not have their
own darkrooms, and this is an enormous number, must use common print sizes
offered by a lab or pay an exhorbitant cost for the lab to make custom
print sizes.
The original question was about how to get the *same* DOF from medium
format as from 35mm small format. The analysis to answer this *must*
compare apples to apples between film formats, and not apples to oranges,
tangerines or grapefruit, or somehow apply a different standard of quality
to one and not the rest. This means using the same print dimensions, the
same viewing distance, the same presumed viewer visual acuity, the same
critical focus distance, and as close to the same field of view as is
practical using common focal lengths available for each film format.
Determining the DOF for a specific lens aperture and critical focus
distance must follow the following steps to define the maximum acceptable
film CoC constant in the DOF equations *before* entering in lens focal
length, lens aperture, critical focus distance and grinding the numbers out
for the near and far DOF limits:
(1) Defining the dimensions and a viewing distance for at print or screen
projection. This is purely arbitrary based on what is intended for the
image being made.
(2) A scientifically based assumption is made about viewer visual
acuity. This is an angle to make it independent of distance. The value I
use (1 second of arc, or 1/60th of a degree) is the accepted one published
in optometry and opthamology textbooks. It is based on the average density
of rods, L-cones, M-cones and S-cones in the retina of the human eye across
the human population, presumes the eye can focus properly, and presumes
20/20 vision with aberrations or distortions either absent or corrected.
(3) Simple trigonometry is used with the viewing distance and average
visual acuity limit in (1) and (2) to determine the maximum acceptable CoC
diameter on the print or projection at the threshold of a viewer being able
to discern the it's a spot and not a pinpoint.
(4) Magnitude of magnification required to produce the print or projection
is calculated for the desired film format (i.e. image frame size on
film). The maximum acceptable diameter CoC in the print or projection is
divided by this magnification to find the maximum acceptable diameter CoC
on film.
Once these four steps are performed for a specific film format, print size
and viewing distance, the DOF for a specific focus distance, focal length
and lens aperture can be calculated. Using models with hard-wired values
for the "film" CoC makes presumptions about the combination of print size
and viewing distance that may or may not be applicable to the individual
using the model! This is why I built my own DOF models that allow defining
print (or projection) size, viewing distance and film format dimensions to
calculate the *film* CoC to be used. I enter those values before entering
focal length, aperture and critical focus distance. I found the values
hard-wired into other models inappropriate and grossly presumptive.
Exemplia gratia:
Have prints made using three print sizes offered by a lab I use in Austin,
Texas: 11x11, 11x14 and 11x16 inches. Make edge-to-edge prints in each of
these sizes with minimum cropping of a 6x6 film frame and a 35mm film frame:
11x11:
35mm mag. is 11.64X
6x6 mag. is 4.99X
11X14:
35mm mag. is 11.64X (mag. unchanged; print DOF unchanged)
6x6 mag. is 6.35X (mag. increases; print DOF decreases)
11X16:
35mm mag. is 11.64X (mag. unchanged; print DOF still unchanged)
6x6 mag. is 7.26X (mag. increases more; print DOF decreases more)
There are six prints. In all three from the 35mm frame, the apparent DOF
in the prints will look the same. In all three from 6x6 frame, the
apparent DOF in the prints will look different in each one.
>Personally, for framed prints, I just print my squares at the largest square
>the paper I am using can handle, and matte to the edge of the image. For
>unframed prints, I just print the square with an equal border on the top and
>two sides. No fuss, no muss.
I'm hoping from the above you recognize a 0.060mm film CoC value may or may
not be applicable to what you are doing. Some other value(s) may be for
the print sizes you are generating and what you anticipate the average
viewing distances will be for them.
>Obviously, your "requirements" are different than most...
I don't believe having 11x14 prints made from 35mm and 645 chromes, and
projecting chromes to a square screen is uncommon. Numerous 35mm chromes
are printed to 11x16, but the magnification is the same as for an 11x14
print. I will concede that producing slide shows is an uncommon activity,
but using a square screen for it is common practice.
>And, I don't know about you, but either I am shooting full open, where I
>want minimum DOF, or I am capturing a range, that I am conscious of, and
>use DOF
>accordingly...35mm or MF...I always use ~1 more stop than what the lense
>lists as the DOF I need. I've never had a problem with this working
>perfectly.
I reverse engineered the DOF markings on my 35mm and MF lenses to calibrate
them to the DOF accuracy desired and was horrified at what I discovered. I
also stop down ~1 stop more because all are approximately one stop off from
what I require for print sizes, screen projection and their intended
viewing distances.
-- John
from minolta mailing list:
Date: Fri, 12 Apr 2002
From: "twm47099" [email protected]
Subject: File "CoC vs Distance.xls" uploaded was Re: DOF, function of focal length only?
I've uploaded an Excel spreadsheet to the files section of this
group. It plots the Circle of Confusion (CoC) vs distance from the
camera. It is located at:
http://groups.yahoo.com/group/Minolta/files/CoC%20v%20distance.xls
You enter the Focal Length, F-stop, and focusing distance
It gives you:
o Magnification (in ratio and x) based on FL and focus distance
o Hyperfocal Distance (and multiples of HFD) based on FL and F-stop
It plots the CoC vs distance along with:
o the CoC at infinity,
o diameter of the diffraction disk (based on F-stop)
o HFD and 2* HFD
It also shows the maximum print enlargement ratio based on the CoC
that will result in 5 lpmm resolution on the PRINT.
The notes sheet provides explanation of the spreadsheet, a smplified
technical basis for it, and the sources and web links that I used.
Let me know if it works for you (it did for me) or if you have any
questions. I'd appreciate any comments, good or bad (including "too
much time on his hands").
Tom
From: [email protected] (Richard Knoppow)
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Focusing and Depth of Field concerns
Date: Fri, 31 May 2002
"Robert J. Wood" wrote:
>Greetings,
>
>I recently worked my way through the focus tutorial, formulas, etc. on
>the large format page:
>
>http://www.ai.sri.com/~luong/photography/lf/fstop.html#result2
>
>There are a few things I don't understand.
>
>
>1) First of all, how devoted should one be to these formulas? Should I
>always assume that, if enlarging from 4x5 to 11x14, my circle of
>confusion should be 0.071mm? I'd love to hear of peoples experiences
>with these approaches.
>
>2) The author says that for "close ups" you use the formula:
>
> N = f/v * D/(2c) = 1/(1+M) * D/(2c)
>
> to find f:stop and for regular shots, you use:
>
> N = D/(2c)
>
>What defines "close up"? WHen do you stop using one formula and begin
>using another? Ive got a Wisner 4x5 Tech. Field and am using a 135mm
>Rodesntock Apo-Syronar-S. When trying to keep both infinity in focus as
>well as an object about 6 feet away, I get drastically different
>f:numbers from the two formulas--176 from the "non close-up" formula,
>and 47 from the "close up" formula. So obviously, an object 6 feet away
>would call for the "close up" formula.
>
>3.) What is the best way to integrate lense tilt into these
>considerations? Can anyone direct me to a good book/source? To this
>point, all of my knowledge was gained either from the Ansel Adams books
>or from very random places on teh web. I'd love ANY comments on these
>things.
>
>Regards,
>rw
Depth of field seems to generate more confusion than any other
subject in photography.
Don't take the formulas as fixed. DOF is an optical illusion.
In principle only an infinitessimally thin plane on the object side
is in "perfect" focus on the film, that is, a point is reproduced as a
point. In practice, this isn't quite so because no lens is perfect,
but its close enough for our purposes.
At any distance on either side of this plane the image will no
longer be focused sharply. The question is how much out of focus can
it be before the eye interprets it as blured. Research on vision has
determined some values for visual acuity. In fact, visual acuity
varies depending on several factors but for purposes of calculating
DOF a value of about half the maximum acuity is assumed. This comes
out to something like 5 lines per millimeter for an image viewed at
one foot, and of course, scales with distance. The size of the
allowable "circle of confusion" is based on whatever value we choose
for the maximum blur spot the eye will still see as a sharp point.
Most DOF charts assume about the 5 line/mm value above for acuity and
calculate the blur spot size allowable on the negative from this.
The size of the blur spot (or circle of confusion) is dependant
partly on the size of the stop. The smaller the stop the smaller the
angle of the cone of light coming to a focus. The narrower the cone of
light the less it will change with distance, so the variation in
distance on either side of perfect focus that gives a given size of
blur spot become greater, i.e., the DOF becomes greater.
There are complicating issues, namely bluring caused by lens
aberrations and bluring caused by diffraction at the stop, but these
are of small importance compared to just the f/stop size.
For very close work the _efffective_ size of the f/stop is smaller
than its marked value. That is because it is further from the image
while the image is being magnified. In fact, the f/stop of a lens
which is focused by moving the entire lens, is the marked value only
at infinity focus. The difference is negligible down to perhaps five
times the focal length of the lens. If you must compensate exposure
because of bellows draw your DOF calculations will also be off.
When the film is tilted with relation to the lens the sharply
focused plane in the object space is also tilted. So, you can focus a
tilted surface sharply by tilting the film to match. Calculating the
DOF exactly for tilts gets complicated but if you need to you can
calculate the DOF for the near and far parts of the plane just as you
would for the entire image at that distance.
I hope this is helpful. DOF is one of those things which is easier
to explain by demonstration and with drawings and waving your hands:-)
---
Richard Knoppow
Los Angeles, CA, USA.
[email protected]
From: [email protected] (Bill Hilton)
Newsgroups: rec.photo.equipment.medium-format
Date: 10 Jun 2002
Subject: Re: Depth of Feild
>From: Robert Feinman [email protected]
>As concerns bigger prints, presumably they will be viewed from further
>away so once again the degree of magnification becomes a factor.
Nothing to keep people from looking closely at details in a big print ... I'm
talking 16x20" to 30x40" here, not billboards.
I see it all the time at art fairs ... guys with 35 mm making 16x20" prints or
larger, and you can always see the grain if you look carefully, even from 2-3
feet away. Guys with medium format are printing out to 20x24" with no grain,
guys with 4x5" cameras are printing even larger. I personally always take a
closer look at landscapes in particular to see the details and if it's grainy
then it always look cheap to me. Tastes vary ... I agree that if you stay far
enough back it's not noticeable (grin).
Three exhibits I saw in the past few years come to mind ... Richard Avedon had
some massive prints (maybe 8 ft tall) and you could get right up to them and
still see the details clearly. But he was using fine grained film in an 8x10"
view camera. Annie Leibovitz had large prints (maybe 4x6 ft) at the same
museum (I think it was her "Women" exhibit) and even from 3-4 ft they were far
too grainy, and she was using 6x7 cm.
One other example, a recent "Arizona Highways" 75th anniversary exhibit, which
had photos from three key contributors over the years, Ansel Adams, David
Muench and Jack Dykinga. All used mostly 4x5" (Ansel of course used several
other formats too over his many years) and you could stick your nose within
inches of the prints and still see clear grainless details (I think the largest
prints were 30x40" or so). No way you could hold that detail with 35 mm above
about 11x14" print size.
That's why I worry about getting the same type circle of confusion size with
medium format as with 35 mm, so I can keep fine detail when I blow it up
10-12x, because I know someone will be looking for the detail, even in large
prints.
Bill
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Depth of Feild
Date: Tue, 11 Jun 2002
Robert Feinman wrote:
> My normal lens (50mm) on a 35mm camera shows a hyperfocal distance of
> about 15feet at f16.
> My normal lens (105mm) on my 6x7 camera gives about 40 feet at f16.
> I assume from this that both are calibrated using the same criteria,
> such as a circle of confusion of .001".
> I'm wondering if this is fair considering that I only need to enlarge my
> 6x7 images about 1/2 as much.
> Any opionions?
>
It is anyone's guess what assumptions are made in calculating the
hyperfocal distance for a lens used with a given format unless those
assumptions are explicitly stated.
The basic formula for the hyperfocal distance is
(focal_length) squared divided by
the product of f-number with the diameter of the circle of confusion in
the film plane.
What is an acceptable circle of confusion in the film plane is based on
what would be considered acceptable in the final print and the degree of
enlargement. One criterion that is often used is that a disc of
diameter 0.25 mm in a print viewed at 10-12 inches is not
distinguishable from a point. But some people think this is too large
and prefer a smaller circle of confusion. If you enlarge about four
times, as would be typical for an 8 x 10 print from a 6 x 7 cm negative,
you have to divide by four to get 0.0625 mm for the diameter of the
circle of confusion in the film plane. If you enlarge about eight
times as would be typical for a 24 x 36 mm negative, your circle of
confusion in the film plane should be about half that.
If you are more demanding than the 0.25 criterion in the final
print---for whatever reason---then you have to adjust your figures
accordingly. Basing the criterion on an 8 x 10 print viewed at 12
inches assumes that larger prints are viewed proportionately further
away and hence can tolerate proportionately larger circles of confusion.
But some people dispute that.
With a 105 mm lens and a 0.0625 mm circle of confusion in the film
plane, the hyperfocal distance at f/16 is
105^2/(16*0.0625) mm = 11025 mm = 36.17 feet.
That is pretty close to the 40 feet you quote.
For the 24 x 36 mm format, if we take the film plane circle of confusion
to be half 0.0625 or .03125 mm, then for a 50 mm lens at f/16, the
hyperfocal distance is
50^2/(16*0.03125) mm = 5000 mm = 16.4 feet
which is pretty close to the 15 feet you quote.
I find the following formula helpful in calculating the hyperfocal distance
H = f^2 * E/(N * 77.4)
where f is the focal length in mm, E is the degree of enlargement, N is
the f-number, and 77.4 is a fudge factor which incorporates the circle
of confusion in the print and the need to convert from mm to feet.
If you want to be more demanding about the circle of confusion in the
print (or other final image), pick a smaller fudge factor.
One additional point to keep in mind is that as the f-number is
increased, the effect of diffraction is enhanced. This is based on the
size of the Airy disc in the film plane, and that depends just on the
f-number. If you enlarge more, you enlarge the Airy disc more so it may
affect sharpness in the final print. At f/16 for 8 times enlargement,
diffraction may begin to affect the sharpness of the image. But at 4
times enlargement, you still have some way to go. Generally, larger
format lenses can be stopped down more without diffraction degrading the
final image because the film image is enlarged less.
--
Leonard Evens [email protected] 847-491-5537
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: I think I'm going digital
Date: Mon, 22 Jul 2002
John wrote:
> "roland.rashleigh-berry"
> [email protected] wrote:
>
>
>>And if you use your 35mm lenses on the front of these cameras then
>>not only are you losing the advantage of depth of field that a digital
>>cameras can give you
>
>
> You actually get more DoF with larger cameras in the same
> size print. That's why 8X10 contact prints look so sharp when in
> fact their lenses have a slightly lower resolution. They get more
> information into the same size print.
Don't I wish. I recently got a 4 x 5 Toho view camera, which I'm very
happy with. But compared to my Olympus 3040 digital camera, increased
depth of field is not its strong point. For the same angle of view, the
same size final print, and the same depth of field, you have to multiply
the f-number by the ratio of magnifications needed to produce that
print. For an 8 x 10 print, the magnifications are respectively about 2
for 4 x 5 and about 40 for the Olympus. So for the same depth of field
I get at f/4 with the Olympus, I would need to go to f/80 with the 4 x
5. Of course, even if the lens stopped down that far, I would
encounter serious problems with acceptable shutter speeds since the
rules for exposure are the same for both formats. One of the first
things that drew my attention when using the 4 x 5, although I knew
about it in theory, was the difficulty of stopping even moderate motion
while maintaining a large depth of field.
Of course the 4 x 5 image has other advantages over the digital image.
What is in focus certainly appears very sharp, and tonal variations are
smoother. also with a view camera I can, within limits, adjust the
plane of focus, so that everything of interest appears in focus even
with relatively small depth of field. But increased depth of field over
smaller formats is not one of its benefits.
>
> Regards
>
> John S. Douglas, Photographer
> http://www.darkroompro.com
--
Leonard Evens [email protected]
From hasselblad mailing list:
Date: Thu, 01 Aug 2002
From: Jim Brick [email protected]
Subject: RE: [HUG] Teleconverter or Extension Tube?
Frank Filippone wrote:
>PS: Use the extension tubes for the easiest solution...... a teleconverter
>will limit your DOF even more...as the FL doubles.....
>
>Frank Filippone
Frank, you of all people should know that DOF is dependent ONLY on image
size. If you frame the image EXACTLY the same on your ground glass (film),
the DOF is identical REGARDLESS of the focal length of the lens or if you
used extension tubes or not.
So no matter what you use to crop in closer, if the image size on the GG is
the same across all methods, the DOF will be exactly the same.
Providing that you use the same f/stop in all cases.
:)
Jim
From hasselblad mailing list:
Date: Thu, 01 Aug 2002
From: Jim Brick [email protected]
Subject: RE: [HUG] Teleconverter or Extension Tube?
Bob Boggio wrote:
>Using a Teleconverter also requires a larger f stop due to the reduced light
>transmission. So DOF is affected by a Teleconverter.
>
>Bob
That's not true either. A 2x teleconverter makes an f/4 lens into an f/8
lens. If your exposure was 1/60 at f/8, you simply set the lens on f/4 and
your exposure is still 1/60 at f/8. The exposure on the film is the same.
If you were using f/5.6, you can set the lens to f/4 and set the shutter to
1/30 and your exposure will be the same and your DOF will be GREATER. DOF
is not effected by any lens or any attachment. It is based upon image size
and f/stop.
If the image "size" remains the same and the "f/stop" remains the same, the
DOF is the same across all methods of forming the image.
Jim
From hasselblad mailing list:
Date: Thu, 1 Aug 2002
From: Godfrey DiGiorgi [email protected]
Subject: RE: [HUG] Teleconverter or Extension Tube?
omigod, a DoF battle on the HUG!!!
DoF is dependent upon aperture (as distinct from f/number) and image
magnification. Given the same lens opening and subject magnification, DoF
is the same for all lenses. f/numbers normalize lens opening to focal
length for purposes of transmission, but the lens opening at the same
f/number with different focal lengths winds up giving you different
physical apertures and thus different DoF characteristics.
(This is why small format cameras with shorter focal length lenses have
greater DoF than medium format cameras. For instance, a Minox 8x11 camera
with a fixed f/3.5 aperture 15mm lens has about the same angle of view as
a Hassy with 80mm lens, but the Minox hyperfocal distance is 12' covering
an in-focus range from 6' to infinity. To get the same aperture size on
the Hassy with 80mm FL, you'd need to stop down to f/19 and then you'd
have the same DoF.)
Godfrey
...
From Hasselblad Mailing List:
Date: Thu, 1 Aug 2002
From: fritz olenberger [email protected]
Subject: Re: [HUG] Teleconverter or Extension Tube?
DOF may depend, in the general case, on aperture (as distinct from
f/number), but I don't think this dependance is linear. In fact, when the
object distance is small with respect to the hyperfocal distance (usually
the case for head shots), the DOF is linear with respect to f/number (as
distinct from aperture) as follows:
front depth = rear depth = Ne*c/M^2
where
Ne = effective f-number (corrected for bellows factor)
c=diameter of the largest acceptable circle of confusion, and
M=magnification.
Where am I wrong?
-Fritz
From hasselblad mailing list:
Date: Thu, 01 Aug 2002
From: Jim Brick [email protected]
Subject: Re: [HUG] Teleconverter or Extension Tube?
fritz olenberger wrote:
>Where am I wrong?
>-Fritz
Maybe you're not...
But read this:
http://www.clearsightusa.com/depthofield.html
Jim
From kiev88 mailing list:
Date: Wed, 12 Jun 2002
From: Carlos Alvarez [email protected]
Subject: Re: DOF calculator !!!
olivier said something to the effect of:
>hi everybody .... on this link, you can have a DOF calculator ....
>
>http://www.shuttercity.com/DOF.cfm
By the way, if anybody uses a handheld computer (PocketPC), you can get a
free DOF calculator and exposure calculator from Handango.com.
--
Carlos Alvarez
From: "David J. Littleboy" [email protected]
Newsgroups: rec.photo.equipment.35mm,rec.photo.equipment.medium-format
Subject: Re: A theoretical question for you...
Date: Thu, 13 Jun 2002
"Joseph S. Wisniewski" [email protected] wrote:
> Godfrey DiGiorgi wrote:
> >
> > My gut feeling is that you would see images of equal resolution (assuming
> > sensors and supporting firmware are equal in quality) but you would see
> > differences in the DOF dictated by the different focal lengths required to
> > achieve the same field coverage.
>
> Your gut feeling is a little off. The difference in focal length is
> accompanied by a difference in the size of the circles of confusion. You
> get exactly the same picture from each camera, provided you're not
> diffraction limited.
You've obviously not been using consumer digital {g}. It turns out that DOF
is radically different if your sensor size is a factor of 4 (linearly)
different. For the same size print and print CoF, smaller sensors result in
greater DOF. For size differences less than a factor of 2, the difference is
on the order of a single f-stop, so you'd almost never notice it.
David J. Littleboy
Tokyo, Japan
Date: Fri, 23 Aug 2002
From: Robert Feinman [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Focusing and Depth of Field concerns
I noticed at the recent Walker Evans exhibit at the Met Museum in NYC
that the prints hanging on the wall (about 8x10 size) were consistently
viewed by people at about 10 inches. They were hanging in a series of
rooms about 20x20 feet and the viewers went around the room peering
closely at each print. Perhaps the small size had something to do with
it. Once one person got up close they blocked the view of anyone
standing further back so almost everyone circled the room close to each
print.
So much for hanging in a gallery as a way to control viewing distance!
--
Robert D Feinman
[email protected]
Landscapes, Cityscapes, Panoramic Photographs: http://robertdfeinman.com
From: "Michael K. Davis" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: DoF - The Merklinger Method
Date: Tue, 13 Aug 2002
Just a followup induced by a reference that's been made to this thread.
My MDoF993s.xls and CoCCal21.xls spreadsheets can now be found on the
Tools page at:
http://www.accessz.com
Thanks!
Mike Davis
From: "Ralph W. Lambrecht" [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Focusing and Depth of Field concerns
Date: Wed, 28 Aug 2002
Henry Dreyfuss conducted a study for the US Army in the 1960s and determined
that the minimum separable visual angle is about 1 minute of arc or 0.017
degrees. However, he stated that some eyes are twice as good. The study also
shows that the result is largely influenced by the lighting condition and
contrast. Two bright stars against a dark sky, for example, can be separated
down to 0.05 seconds of arc or an impressive 0.000014 degrees.
Using this data for photography, most Depth of Field tables assume a viewing
angle of 2 minutes of arc. Example CoC = 0.022mm for 35mm film. Half that
diameter would be more realistic for critical consumers. Here are some examples
of critical CoC for common film formats.
24x36 0.011
6x4.5 0.020
6x6 0.021
6x7 0.024
6x9 0.026
4x5 0.044
5x7 0.056
8x10 0.089
11x14 0.126
These values assume an uncropped print viewed from a distance equal or larger
than the diagonal of the print. For more information see:
'The Measure of Man'
� 1967 by Henry Dreyfuss
and
'On the Psychophysical Function'
� 1975 by H. L. Resnikoff
Ralph W. Lambrecht
"Q.G. de Bakker" wrote:
> Leonard Evens wrote:
>
> > You offer very good advice, but let me quibble a bit.
> >
> > The old rule of thumb was that the largest disc that the eye can't
> > distinguish from a point at 10 inches is 0.01 inches or about 0.25 mm,
> > which comes out to 4 lp/mm, using your formula. Certainly some people
> > can see better than that, perhaps being able to see as well as the
> > equivalent of 8 lp/mm, as you say.
>
> Let me quibble a bit too.
> If the largest disc that the eye can't distinguish from a point is (about)
> 0.25 mm, you would get 4 of them (single (!) points; stretch them in one
> direction and you will have lines) in a mm. That are only 2 pairs (2 lp/mm),
> not 4.
> No? ;-)
Date: Tue, 27 Aug 2002
From: Alan Davenport [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Focusing and Depth of Field concerns
"Brian Ellis" [email protected] wrote:
>I use the focusing and depth of field system described in a Photo Techniques
>magazine some years back, which I think is discussed in Tuan's article on
>his home page about focusing the view camera. Sorry I don't have the cite
>handy but if you're interested let me know and I'll find it. Basically the
>system involves focusing on the near, focusing on the far, seeing the
>distance that the lens travels between the two points, and then selecting
>the optimum aperture based on that distance and the tables set forth in the
>article. I think the article (actually it's two articles together) talks
>about the circle of confusion size but it's been so long since I read it I'm
>not sure.
The article you're referring to sounds like Paul Hansma's. He
takes a slightly different approach than the second of the articles
(Stephen Peterson's.) Hansma claims that his method give an
optimal aperture given the near/far focus spread, taking into
account both aperture and diffraction. I have tried using his
method and gotten good results; I can't say whether it might
have been possible to get better...
Here are the links to the 4 pages:
http://www.largeformatphotography.info/articles/hansma-dof.1.gif
http://www.largeformatphotography.info/articles/hansma-dof.2.gif
http://www.largeformatphotography.info/articles/hansma-dof.3.gif
http://www.largeformatphotography.info/articles/hansma-dof.4.gif
From: "roland.rashleigh-berry" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Medium Format vs. 35mm
Date: Sat, 7 Sep 2002
"Jeff Haddock" [email protected] wrote..
> I have long wanted to use medium format for my photography but have
> heard conflicting things about it. I am hoping they are simply the
> ramblings of one who is married to the 35mm format and simply does not
> have the will nor the means to try anything else.
>
> Before I begin, I think it may be best if I explain what I photograph
> and how. To begin with, handheld use is of great importance to me as
> much of my work is of urban settings (think Manhattan) but is of
> landscapes at the same time. Think exposures of the side of a building
> for texture, etc. For this reason, I cannot use a tripod in a lot of
> situations, both for practicality in a crowded island and because I
> want to be as inconspicuous as possible. I shoot black and white print
> film, usually enlarging anywhere between 8x10 to about 11x14. I keep
> it this way because I want to teeter on the liquid quality afforded to
> LF users printing contact prints.
>
> That being said, I have heard people stating that there is
> considerable depreciation of depth of field when shooting MF compared
> to 35mm. Does this statement have any merit? If so, why is this the
> case? I have a hunch it may be because most expect to enlarge MF negs
> more than 35mm negs and in doing such the lens shake and relative
> focus become more pronounced.
There is much less depth of field for medium format. It shouldn't come as a
surprise.You are effectively using 35mm telephoto lenses as standard lenses
and I am sure you are aware of telephoto lenses having a more limited depth
of field. To compensate you will have to stop down. But in so doing you lose
out when you have limited light as you describe, This results in longer
exposure times and so some of your shots might be blurred if you hand-hold.
Having said that, lets say shooting conditions are two stops off sunlight,
then at f11 with 100 speed film you will expose at 1/50th sec . You should
be able to get away with that. Or use a faster film. At f11 you will have
good depth of field.
> Beyond this supposed disadvantage, are there any other, real or
> perceived, that I should be aware of? In other words, is it, or why is
> it not, simply the exact same thing with more information because
> there is more negative to expose? Thanks in advance to all who
> participate.
Depth of field is the big problem with medium format. Depending on what you
shoot, this may ruin what you are trying to do. If your subjects are
typically close up and you want the background to be in focus as well then
you might find yourself sticking close to f11 or f16 with medium format,
unless you go in for fast film along with its inherent graininess. And then,
of course, you lose out on the advantages medium format can give you in
terms of extra detail and texture. I think maybe you should stick with 35mm
for this.
From: John Stafford
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Medium Format vs. 35mm
Date: Sat, 07 Sep 2002
roland.rashleigh-berry at [email protected] wrote
> There is much less depth of field for medium format. It shouldn't come as a
> surprise.You are effectively using 35mm telephoto lenses as standard lenses
> and I am sure you are aware of telephoto lenses having a more limited depth
> of field. [...]
How many times are we going to travel down this tired old road? One still
has to consider viewing distance, degree of enlargement, and the normalized
lens for the aspect ratio. So, while strictly speaking a 'normal' lens for
6x6ccm is a long lens (not telephoto) for 2.4x3.6cm, the outcome is not the
same as the same lens on the larger format. It is about outcomes viewed.
If you are grain sniffing, then it'sabout viewing distance. If you want to
consider the rarified air of mathematics, it's about math. Experiencing an
image is a different story.
So, to the original poster - take pictures. Don't get mirred in this virtual
paradigm. It has very little to do with what you want to do. I hope!
From: "David J. Littleboy" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Medium Format vs. 35mm
Date: Sun, 8 Sep 2002
"Leonard Evens" [email protected] wrote:
> Sherman wrote:
> >
> > Jef,
> > The depth of field is the same for the same focal length lens. That is an
> > 80mm lens on a medium format camera will have the same depth of field as an
> > 80mm lens on a 35mm camera at the same aperture.
>
> There may be some interpretation under which this statement is true, but
> with the most common meaning, it is false.
Right so far.
> Depth of field is dependent
> on the criterion for sharpness in the final image, and subject to that
> to the degree of enlargement. You would typically enlarge a 35 mm
> negative at least twice as much as a medium format negative. So, for
> example, the hyperfocal distance for an 80 mm lens and medium format
> would be roughly half that for an 80 mm lens and 35 mm format, if the
> final image has the same size. So in fact the medium format combination
> would have more depth of field, roughly the same amount as the 35 mm
> combination stopped down two additional stops. Of course I'm not sure
> how much such a comparison is worth.
It's a starting point, but, for the theory at least, you should compare
images taken with the same angle of view enlarged to the same size. (I.e.
photographs that are compositionally identical.)
> 80 mm is a normal focal length for
> medium format while it is a relatively long lens for 35 mm. It is not
> entirely surprising that a long lens with one format gives you less
> depth of field than a normal lens with a different format, but of course
> one has to do the mathematics to see what is actually true.
http://www.wrotniak.net/photo/dof/ shows the math. Although it's concerned
with digital cameras, it holds for any camera.
The bottom line seems to be that the DOF is inversely proportional to the
(linear) image (sensor) size, and that doubling the image size requires
stopping down one f-stop to get the same DOF.
The problem, as others have pointed out, is that if you are making 8x10s
from 35mm, you are probably making 16x20s from 6x7, and people who make
prints tend to look closely at what they've made and thus find that their
standards end up being those of the grain sniffers. So the mathematically
and scientifically and logically incorrect claim that an 80 mm lens in MF
has the same DOF as an 80mm lens in 35mm ends up being about right in
practice.
ce.
[off-the-wall-rant]
I haven't tested this yet, but I suspect that for landscape work, the best
hyperfocal distance may infinity. I suspect that if you can detect any loss
of sharpness at infinity when you defocus from infinity to your hyperfocal
distance, you (if you're a grain sniffer) will decide that that loss of
sharpness is unacceptable...
[/off-the-wall-rant]
David J. Littleboy
[email protected]
Tokyo, Japan
From: "David J. Littleboy" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Medium Format vs. 35mm
Date: Tue, 10 Sep 2002
"maf" [email protected] wrote:
>
> But "depth of field," that part of the SCENE (in front of the camera) being
> photographed which is in focus, cannot be compensated for by the differing
> amounts of enlargement between the formats. For a given angle of view (such
> as a "normal" lens used in both formats), and a given f-stop, the smaller
> format (35mm) will have a better depth of field because the 35mm format
> "normal" lens has a shorter focal length (about 50mm) than the MF normal
> lens (75mm-105mm depending on frame size). Period.
You've got the right answer, but you don't have the math to show it. DoF
goes down with the _square_ of the focal length, but goes _up_ linearly with
the format size. (Under the normal definition of DoF, which is a given
circle of confusion on prints of the same size.)
> The difference in the
> enlargement factor between 35mm and MF cannot make up for the difference in
> depth of field.
That's theoretically correct, but you haven't shown that. See:
http://www.wrotniak.net/photo/dof/ for the math.
As a _practical_ issue, since the difference between 35mm and 645 is about
1/2 stop, and the difference between 35mm and 6x9 is a bit over one f-stop,
the difference isn't all that much. The 60mm lens on my GS645S coughs up
surprising DoF on occasion. (Surprising because I haven't done careful tests
to see what I get and usually don't record exposure information.)
David J. Littleboy
Tokyo, Japan
From: [email protected] (DFleming)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Medium Format vs. 35mm
Date: 13 Sep 2002
[email protected] (RD) wrote...
> Anyway, if someone can tell me how the DOF scales are calculated and
> what they mean in the real world, I'll accept this for what it is.
The theory underlying depth of field scales is described here:
http://www.dofmaster.com/theory.html
The DOFMaster program was designed to emulate the depth of field
scales on lenses. You can use it to match the scale on a lens by
varying the circle of confusion in the program. For example, I can
replicate the scales on my Fuji 50mm and 100mm lenses (for 35mm
format) by using a circle of confusion of 0.029 mm. To match the
scales on my Canon lenses I use a circle of confusion of 0.031 mm.
Don
Date: Sat, 21 Sep 2002
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.medium
Subject: Re: Medium Format vs. 35mm
Mike wrote:
> Question.
> I have come across a DOF calculator. No where does the calculator ask for
> print size or viewing distance.
> If print size or viewing distance are all important for calculating DOF on
> film why are they missing from the calculator?
> What am I missing?
> In case you are interested.
> http://www.silverlight.co.uk/resources/dof_calc.html
Because the author has made certain assumptions about the size of the
final print, the distance it will be viewed at, and the ability of the
viewer to resolve points at that distance. If he didn't make these
assumptions personally, they are built into values he got for a crucial
parameter, the circle of confusion, from other sources.
I've worked backwards to calculate what he assumes the diameter of the
circle of confusion in the film plane is. It comes out as follows
35 mm: .033 mm
6 x 6: .06 mm
4 x 5: .15 mm
If one multiplies these by plausible enlargement factors (8, 4, and 2)
one comes up with about the same value for the diameter of the circle of
confusion in an 8 x 10 print: .25 mm in the first two cases and .30 mm
for 4 x 5. Since I don't know exactly what assumptions he made about
how much the negative would be enlarged and what other factors he may
have taken into consideration, I can't say why there would be a
difference for 4 x 5.
But note they are all pretty close, and towards the upper end of what is
usually consider acceptable. .25 mm in an 8 x 10 print viewed at about
10 inches is roughly equivalent to being able to resolve 4 lp/mm at that
distance. Many people can do better than that, and some authorities
choose 8 lp/mm as a more rigorous standard. Had he used such a
standard, his hyperfocal distance calculations would have come out twice
as large.
Note that the guy who made this calculator is not doing anything
mysterious or specially complicated. He is using the same formulas the
rest of us are using, and those formulas require putting in appropriate
parameters. The fact that you found it a snazzy web site doesn't make
more reliable than what I tell you using a simple calculator or even
doing it by hand.
If you do a little searching, you will find web sites which require you
to put additional information in before calculating the depth of field
for you.
--
Leonard Evens [email protected]
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Almost nothing anyone says about DOF is always true.
Date: Sun, 22 Sep 2002
One finds these unending debates about depth of field. This is
surprising because the concept is based on fairly simple optics which
has been understood since early in the 20th century if not earlier.
There are standard formulas for calculating the near and far limits of
what will be perceived in focus. (See Jacobson at
www.photo.net/learn/optics/lensTutorial
for a fairly complete discussion of these formuals.)
These formulas apply under simplifying assumptions and they ignore
various other factors such as film flatness and diffraction. But they
are a good starting point for estimating depth of field, and in many
circumstances they are pretty accurate.
Unfotunately, depth of field is not a qualitative thing which can easily
be described in words. It is governed by mathematical formulas, and
while those formulas are not all that complicated as such things go,
they do encompass complexity that it is impossible to state completely
in words. So almost everything you hear people say about depth of field
is wrong in some circumstances. I would like to give some examples.
First, the general formula is too difficult to use in practice, so it is
usually replaced by different approximations, one which applies to
"distant" objects and the other to closeups. They give rather different
kinds of answers, so, if one were to apply the usual formulas for
distant objects to closeups, the answers would be way off. In addition,
the definition of "closeup" depends on focal length; it is often taken
as less than ten times the focal length. So for my normal 150 mm lens
for 4 x 5 format, distant means further away than 1.5 meters, for my
digital camera set at 10 mm focal length, "distant" means further away
than 0.1 meters. Even at maxium zoom, the focal length is 21.5 mm, so
distant means further away than 0.215 meters. In effect virtually
everything is distant as far as my digital camera is concerned.
Another issue which leads to confusion is subject distance. If one
fixes subject distance, f-number, format, degree of enlargement, and
the criterion for sharpness in the final print, then both the near and
far depth of field ranges increase as the focal length decreases. But
some people have argued that we should consider the size of the image of
the principal subject in the image plane as a standard. If one fixes
the distance to the subject, then of course a shorter focal length leads
to a smaller subject image. If one wants to maintain the size, one must
move proportionately closer. (The standard example of this would be
portraiture.) If one does that, then if the focal length is smaller,
there is actually less near depth field, and the total depth of field
often doesn't change very much. In fact, Jacobson makes exactly this
point at the above website and his mathematics confirms it under certain
assumptions. And it certainly contradicts the other oversimplification
that depth of field always increases if you decrease the focal length.
Unfortunately, those certain assumptions don't always apply. For
example, the principal subject could be at the hyperfocal distance for a
short focal length lens. You could then change to a longer focal length
lens and move far enough away to maintain the same principal subject
image size, keeping everything else such as f-number the same.
Depending on the actual numbers, it can certainly happen that the
subject will be significantly short of the hyperfocal distance for the
longer lens. In that case, the far depth of field for the short focal
length lens will be infinite, but the far depth of field for the long
focal length lens well short of that. So the assertion that depth of
field doesn't change if you keep subject magnification constant also has
to be qualified and depends on the exact circumstances.
Note also that it is a misconception that all photographs have a
principal subject in one plane. Something pretty close to that is true
for portraiture, but it is often not true for scenic photos. If there
are many principal subjects, then moving closer or futher has a more
complex effect. There may not be any particular subject, the image size
of which you want to fix.
The most common source of confusion about depth of field is the belief
that it is an innate characteristic of lenses or at worst lenses and
formats. This is probably a consequence of lens manufacturers putting
depth of field scales on their lenses. But anyone who has studied these
formulas in a discussion of photographic optics will realize that they
require certain inputs. These are focal length, f-number, and coc
(diameter of the maximal allowable circle of confusion in the film
plane). It is the latter which causes problems. It is not given
objectively for all time and all circumstances. Someone has to choose
it. There are a variety of ways of doing that, but they all come down
to making an assumption about the maximal allowable circle of confusion
in the final viewed image, often taken to be a print. That in turn
depends on assumptions about the visual acuity of a typical human eye.
Since visual acuity depends on angles and not absolute distances, that
means the distance from the eye to the print also plays a role. But
many people find that idea very upsetting. It seems too subjective.
But of course it is perfectly objective, as is everything else about
depth of field calculations, which should be distinguished from other
factors affecting perception of sharpness, some of which are subjective.
It just depends on more factors than are apparent to someone used to
relying on lens scales.
Another related idea which sometimes creeps into these discussions is
that something is either in focus or out of focus and that is
objectively determined independent of anything else. In fact, under the
usual simplifying assumptions, there is only one plane where everything
is exactly in focus. Depth of field depends on how far one is willing
to depart from exact focus and still see the details in focus. That is
the reason for the use of a circle of confusion which is the theoretical
image of a point source. This varies in principle from a point for a
source in the plane of exact focus to a disc of increasingly larger size
as one departs from that plane. When it gets too large, it will no
longer be seen as a point in the final image, viewed at a specific
distance by a specific viewer.
Of course, for a real lens, there is no plane of exact focus since
because of diffraction all points are imaged as discs, which are
smallest at what is considered exact focus. Also the surface of best
focus is not actually a plane, but for good lenses, it doesn't depart
very much from a plane.
Having spent some time thinking about this, I've concluded that when all
is said and done, depth of field calculations are what they are based on
the formulas, and in the end that is what one has to rely on.
Qualitative statements may apply in certain circumstances, but they
don't apply universally in all practical situations.
For reference, the formulas which apply for distant objects are
Hyperfocal distance:
H = f^2/(N*C)
where f is the focal length, N is the f-number and c is the diameter of
the acceptable circle of confusion in the film plane. If CP is the
diameter of the largest acceptable circle of confusion in the print
(which depends, as noted above, on the distance it is viewed as well as
the acuity of the eye viewing it), and E is the enlargement factor to
make the print from the given film image, then
H = f^2*E/(N*CP)
I've found the following formula which choose a high end estimate for CP
and also converts from mm to feet useful
H = f^2*E/(75*N)
f here is measured in mm. E you have to pick yourself. If you are more
finicky, you would want a smaller number than 75.
Depth of field limits:
Once one has determined the hyperfocal distance, and one specifies the
distance D focused on, the near and far limits for depth of field are
given by
D_near = D*H/(D+H)
D_far = D*H/(D-H)
Depth of field ranges:
If one wants instead the ranges which are in focus measuring forward and
backward from the point of focus, they are given by
R_near = D^2/(D+H)
R_far = D^2/(D-H)
Although I'm a mathematician, I'm not all that great at mental
arithmetic. But once I calculate and record the hyperfocal distances
for specific f-numbers for each of my lenses, I can estimate these
numbers in my head. My 11 year old grandson can do a lot better. But
certainly the simplest possible calculator is up to the task. I got a
free one from Office Depot which is solar powered and I can carry it in
my back pocket with my wallet.
The formulas for closeups are very different. I don't have much
experience with closeup photography so I won't conjecture about how
useful they may be.
--
Leonard Evens [email protected]
From: "Q.G. de Bakker" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Almost nothing anyone says about DOF is always true.
Date: Sun, 22 Sep 2002
Leonard Evens wrote:
> [...]
>
> The formulas for closeups are very different. I don't have much
> experience with closeup photography so I won't conjecture about how
> useful they may be.
And that's the most interesting bit. Why are the formulae for DOF "very
different" when subject distance are small?
Is there a fundamental change in the nature of DOF happening, for some
unknown reason, at some undetermined point on the way leading into the world
of close-up photography? Not very likely. ;-)
Or do the formulae (the ones you give included) all lack a certain parameter
that increases when subject distances get smaller, that is responsible for
the "very different" results not 'predicted' by what then would be classical
but inherently flawed formulae?
And can those classic formulae then be trusted to give correct results in
the non-close up world? If so, why? By accident? How do we know? If there is
such a parameter, why not put it in the "classic formulae too, and make them
universal?
From: "David J. Littleboy" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Medium Format vs. 35mm
Date: Mon, 23 Sep 2002
"MovieStuff" [email protected] wrote:
> Not so odd considering that this is a forum about medium format
> cameras discussing medium format cameras compared to 35mm cameras and
> my reference above was about medium format cameras; all of which have
> depth of field markings on them, apparently for no good reason it
> would seem.
They have DOF markings that are calculated assuming a certain print size. It
is well known that MF camera DOF markings are usually quite wrong (in
particular, overestimating DOF by one or two stops) for the print sizes most
current MF users are interested in.
http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=000ROw
http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=000uRr
http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=000GkE
These make it very clear that DOF markings are quite arbitrary and
untrustworthy, even in 35mm cameras.
"The fact that different manufucturers use different magical numbers
provides a hint that these numbers are fairly arbitrary. Nikon even used
different numbers for their different ranges of lenses!"
> Beyond that, you can bet that while your large format
> camera lenses may not have depth of field markings on them, there is
> guaranteed to be a depth of field chart somewhere for any one of your
> lenses, again provided by the manufacturer for no good reason,
> according to the popular theory being promoted on this thread. So
> what's your point?
The point, of course, is that DOF is a property of the print, not the lens.
And that degree of enlargement (and individual sharpness requirements)
affect DOF.
David J. Littleboy
Tokyo, Japan
Date: Mon, 23 Sep 2002
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: DOF tipping point 1:1 vs 1:2 etc. Re: Medium Format vs. 35mm
Robert Monaghan wrote:
> speaking of DOF, does anyone have a formula for when DOF goes from 1:1
> (as in 50% in front, 50% in back, as is usually the case in macro work)
> to 1:2, as in 1/3rd in front, 2/3rds in back of the plane of focus?
>
> thanks bobm
I'm not really sure what you are asking. But I will give it a try.
According to Jacobson, (ignoring some irrelevance about the exit pupil),
the formulas for DOF in front and back are respectively
CN(1 + M)/(M^2(1 + NC/(fM))) and CN(1 + M)/(M^2(1 - NC/(fM)))
where C, N, and f have the usual meanings and M is the magnification,
i.e., the ratio of image size to object size. If M is close to one,
then it is quite plausible that NC/(fM) would be small and could be
ignored, in which case these two quantities would be essentially the
same. (However, they are never exactly equal.) If you want the first to
be half the second, it follows from some algebra that you need
2/(1 + NC/(fM)) = 1/(1 - NC/(fM)) or
1 = 3NC/(fM) or
M = 3(NC)/f = 3f/H
where H = f^2/(NC) is the usual expression for the hyperfocal distance.
Since M = f/(D - f), some more algebra tells you this happens when
D = f + H/3.
(Jacobson mentions this specific case in his lens tutorial.
Note that often hyperfocal distances are relatively large, so one is
hardly doing closeup photography in this case. For example, under my
assumptions the hyperfocal distance for my 150 mm lens (4 x 5 format) at
f/8 is 75 feet, and close up would be less than 1.5 meters or about 5 feet.
If you want more generally r times as much depth of field in back as in
front, these formulas seem to say you should choose the magnification to be
(r+1)/(r-1) times NC/f or (r+1)/(r-1) times f/H.
Equivalently you should choose the distance so that
D = f + [(r -1)/(r+1)] H.
Let me know if that was what you were after.
P.S. Jacobson's formulas can't necessarily be used for distant objects.
For example, it is possible for the second denominator to become zero or
negative, in which case they no longer apply, and the rear depth of
field is infinite. Also, for distant objects, M is going to be quite
small, and because of the form of the expression, slight variations in
arithmetic, as those resulting from rounding, can make very big
differences in the results. (Of course they can be reformulated to
avoid such problems by putting in the formula for M and doing some
algebra.) Numerical issues like this often arise in applied
mathematics, and pure mathematicians are sometime unaware of them. It
is quite possible for an exact formula to be utterly useless because it
has been put in a form which makes accurate approximate calculations
impractical.
--
Leonard Evens [email protected]
Date: Mon, 23 Sep 2002
From: "Mxsmanic" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Almost nothing anyone says about DOF is always true.
"Q.G. de Bakker" [email protected] a �crit
> If 'calculating' DOF requires "very different"
> formulae when subject distance decreases appreciably,
> DOF itself must be a very different thing in
> those situations.
No. The latter does not follow from the former.
FWIW, aperture calculations become much more complex for large apertures
(beyond f/1.4 or so) as well, as the standard simplification of dividing
focal length by f-stop isn't exactly correct, and becomes very incorrect at
large apertures.
This is why it's impossible to go below f/0.5 for ordinary photographic
lenses, even though the simplified rule would imply that this is possible if
the aperture size is twice the focal length of the lens. In fact, the
aperture size has to be infinitely large at this f-stop. It is possible to
come close, however; at least one f/0.7 lens has been built. Even at f/1.0
the optical design and manufacturing issues become fairly overwhelming,
though.
From: "Q.G. de Bakker" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Almost nothing anyone says about DOF is always true.
Date: Mon, 23 Sep 2002
fbearl wrote:
> As you well understand, the formulae are not that complicated. But the
> choice of the value for circle of confusion leads to a circle of confusion.
> The values I have used are 1000 for 35mm, 1500 for medium format and 1720
> for 4x5. These are based on "true perspective", the distance at which the
> print should be viewed to give a versimilitude of the depth of field of the
> image as it was taken. The Swedish folks (the blads) seem to use a coc of
> 1500 for their lenses. However, on consumer lenses, esp. 35mm, the scale
> for coc seems to be about 100. Maybe OK for a 4x6 print viewed at 8", but
> not good for an 11x14 viewed at 4'.
Zeiss talks about the "international standard" defining the acceptable
maximum size of the CoC as 1/1000th of the image format's diagonal length,
yet in 35 mm photography only 1/1500th of the 43 mm diagonal is deemed
tolerable.
Date: Mon, 23 Sep 2002
From: "Mxsmanic" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Almost nothing anyone says about DOF is always true.
"Q.G. de Bakker" [email protected] a �crit
> Zeiss talks about the "international standard" defining
> the acceptable maximum size of the CoC as 1/1000th of
> the image format's diagonal length, yet in 35 mm photography
> only 1/1500th of the 43 mm diagonal is deemed tolerable.
Use 1 minute of arc, and much of this confusion will disappear.
Date: Mon, 23 Sep 2002
From: "Mxsmanic" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Almost nothing anyone says about DOF is always true.
"Q.G. de Bakker" [email protected] a �crit
> Using one set standard does not do justice to
> the variable thing DOF is, as you yourself
> have pointed out in "the other thread".
I myself have pointed out that all depth of field proceeds from the finite
acuity of human vision. Human eyes can see details no smaller than 30
seconds of arc in size as projected on the retina. "Depth of field" is the
illusion of sharpness that results when details projected onto the retina
are blurred to a degree that is below the threshold of visual acuity; in
other words, DOF is created by blurring of less than 30 seconds of arc on
the retina. Anything below that looks sharp. And any system of capture,
recording, printing, viewing, or display that produces such an image on the
retina will create the illusion of DOF as described. It is ultimately the
only thing that matters, and how you get to that point is irrelevant.
In practice, a value of 1-2 minutes of arc is used more often than 30
seconds, because 30 seconds is the absolute limit under perfect conditions
(imposed by the physical dimensions and spacing of cone cells in the
retina), and perfect viewing conditions almost never exist.
Date: Mon, 23 Sep 2002
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Almost nothing anyone says about DOF is always true.
...(quotes above posting)
Your basic point is well taken. It is the angular displacement that is
important and that eliminates having the worry about the viewing distance.
But I've always found the arithmetic confusing.
At 25 cm (or about 10 inches), 1 minute of arc is subtended by a chord
of approximately .073 mm. This is between one third and one fourth of
the usual (weak) standard of 0.25 mm for the diameter of the circle of
confusion in the print viewed at 25 cm. The way I understand it, from
experience with eyecharts, it is that the bars and spaces between them
in such a chart should be distinguishable if they subtend 1 minute of
arc. From that point of view, you need a pair, and the corresponding
resolution in terms of lp/mm would be the reciprocal of 2 x 0.073 =
0.146 mm or about 7 lp/mm. That is towards the low end of what is
considered tolerable. A circle of confusion of 0.25 mm would correspond
to 4 lp/mm.
The diameter of the circle of confusion apparently should be the
reciprocal of the resolution. I once convinced myself that was true,
but I've forgotten the explanation. I think it is because a series of
circles of confusion meeting at their edges has the same effect as a
sequence of line-space pairs. The light intensity in a circle of
confusion drops off smoothly from the center rather than suddenly going
from black to white.
--
Leonard Evens [email protected]
Date: Mon, 23 Sep 2002
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Almost nothing anyone says about DOF is always true.
fbearl wrote:
> "Leonard Evens" [email protected] wrote
> As you well understand, the formulae are not that complicated. But the
> choice of the value for circle of confusion leads to a circle of confusion.
Agreed.
> The values I have used are 1000 for 35mm, 1500 for medium format and 1720
> for 4x5.
I'm not sure what units you are using. Using millimeters, one might
take about
0.25 mm for a print viewed at about 10 inches
0.25/8 = .03125 mm for 35 mm negatives,
0.25/4 = .0625 mm for medium format negatives,
0.25/2 = 0.125 mm for 4 x 5 negatives
Some people with better eyes or higher standards might insist on using
half these values. Also, all I've done is to divided the print standard
by reasonable enlargement factors. Since there are many medium formats
and even different opinions of just how much to enlarge 35 mm or 4 x 5,
others might use slight variations of these number (8, 4, and 2). Also,
there may be other considerations except enlargement factor I don't know
about.
How are your numbers related to mm?
> These are based on "true perspective", the distance at which the
> print should be viewed to give a versimilitude of the depth of field of the
> image as it was taken. The Swedish folks (the blads) seem to use a coc of
> 1500 for their lenses. However, on consumer lenses, esp. 35mm, the scale
> for coc seems to be about 100. Maybe OK for a 4x6 print viewed at 8", but
> not good for an 11x14 viewed at 4'.
> Would you run those formulae past us one more time?
I'm not sure just what you would like. I've already summarized the
formulas, and there doesn't seem to be much point in repeating it again.
But if you elaborate, I will give it a shot.
> I would like to see
> the formulae and a coherent explanation that I could archive for future
> reference. I could use a math lesson.
> Thanks :>)
--
Leonard Evens [email protected]
From: [email protected] (DFleming)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: DOF tipping point 1:1 vs 1:2 etc. Re: Medium Format vs. 35mm
Date: 23 Sep 2002
[email protected] (Robert Monaghan) wrote
> speaking of DOF, does anyone have a formula for when DOF goes from 1:1
> (as in 50% in front, 50% in back, as is usually the case in macro work)
> to 1:2, as in 1/3rd in front, 2/3rds in back of the plane of focus?
>
> thanks bobm
When a lens is focused at 1/3rd the hyperfocal distance, 1/3rd of the
DOF is in front of the plane of focus, and 2/3rds is behind the plane
of focus. The derivation follows.
Using the simplified DOF equations (approximate equations when the
subject distance is much greater than the focal length, derivation at
http://www.dofmaster.com/equations2.html ):
Dn = (h*s)/(h+s)
Df = (h*s)/(h-s)
where:
Dn = near distance of acceptable sharpness
Df = far distance of acceptable sharpness
h = hyperfocal distance
s = focus distance
What we need to find is the distance s where:
(s - Dn) = (Df - s)/2
Solve for s:
s + s/2 = Df/2 + Dn
(3/2)*s = Df/2 + Dn
3*s = Df + 2*Dn
3*s = (h*s)/(h-s) + 2*(h*s)/(h+s)
3 = h/(h-s) + (2*h)/(h+s)
3 = [ h*(h+s) + (2*h*(h-s)) ] / [(h+s)*(h-s)]
3 = ( h*h + h*s + 2*h*h - 2*h*s ) / [(h+s)*(h-s)]
3 = ( 3*h*h - h*s ) / [(h+s)*(h-s)]
3 = 3 * [ (h*h) - (h*s)/3 ] / [(h+s)*(h-s)]
(h+s)*(h-s) = (h*h) - (h*s)/3
h*h - s*s = (h*h) - (h*s)/3
s*s = (h*s)/3
s = h/3
So, when the lens is focused at h/3, 1/3rd of the DOF is in front of
the subject and 2/3rds is behind the subject.
Don
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Depth of field Independent of Focal Length
Date: Thu, 26 Sep 2002
roland.rashleigh-berry wrote:
>
> But if you have seen hugh blowups from full fish-eye lenses within the same
> format it is obvious that the smaller focal length lens has a much greater
> depth of field even though the shots can be magnified up to full size.
>
> Could somebody pull out some equations to compare a 24mm lens with a 50mm
> lens but have the circles of confusion for the 24mm lens half the size that
> of the 50mm lens? We'll prove it with science. :-)
See the posting by D Fleming and the web site he refers to.
http://www.dofmaster.com/dof_imagesize.html
I can also show anyone who is interested just what is happening in terms
of the formulas given in Jacobson's lens tutorial, which is sometimes
quoted in support of the belief.
--
Leonard Evens [email protected]
From: [email protected] (DFleming)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Depth of field Independent of Focal Length
Date: 25 Sep 2002
[email protected] (RD) wrote
> Just whan you thought it was safe . . .
>
> "Contrary to popular belief, depth of field does not depend on the
> focal length of the lens. Depth of field is determined only by the
> magnification of the image, and although telephoto lenses magnify more
> than wide-angles (at the same distance), the depth of field in a
> picture taken by a 24mm lens will be the same as that in one taken by
> a 300mm lens if the subject is the same size in each."
>
> That's a direct quote from the "Basic Photography Guide" UK website of
> Robert Slade. It seems to agree with the earlier comments that DOF
> changes as a function of print size and viewing distance, both of
> which affect the perceived image magnification. Any comments regarding
> the simplicity of this statement?
This rule of thumb is approximate for relatively close focus
distances. By that I mean that it is approximate when the focus
distance for the shorter lens is less than about 1/4 the hyperfocal
distance for that lens.
The rule is fairly safe to use with longer lenses, say, 100mm and
300mm lenses on a 35mm camera, because the hyperfocal distance is
large for these lenses.
The following text shows some examples of when the rule of thumb works
and when it doesn't. A more technical discussion, with graphs
comparing several lenses, is here:
http://www.dofmaster.com/dof_imagesize.html
The rule can be shown to fail in many cases. For example, if the
shortest lens is focused at its hyperfocal distance (for a given
aperture), the depth of field extends from 1/2 the hyperfocal distance
to infinity. The depth of field for the longer lens won't extend to
infinity when you focus on the same object and keep the image size the
same.
As this is a forum for medium-format, let's use an example of
wide-angle (40mm) and telephoto (180mm) lenses on a 6x6 camera. Assume
these lenses are set to f/8. Further assume that the circle of
confusion used for the depth of field calculations is 0.045 mm. Here
are four examples:
1. Focus distance (3.7 feet) is about 1/4 the hyperfocal distance of
the 40mm lens.
2. Focus distance (7.4 feet) is about 1/2 the hyperfocal distance of
the 40mm lens.
3. Focus distance (11 feet) is about 3/4 the hyperfocal distance of
the 40mm lens.
4. Focus distance (14.7 feet) is the hyperfocal distance of the 40mm
lens.
We'll see that the rule of thumb works well for the first example. It
fails for the other examples.
Example 1, focus distance of shorter lens is about 1/4 the hyperfocal
distance of the shorter lens:
40mm lens at f/8, hyperfocal distance = 14.7 feet
Focus distance, 40mm lens: 3.7 feet
Focus distance, 180mm lens: 16.7 feet (= 180/40 * 3.7 feet, to keep
image size the same)
Depth of field, 40mm lens: 1.9 feet (.7 feet in front to 1.2 feet
behind subject)
Depth of field, 180mm lens: 1.8 feet (.8 feet in front to 1 foot
behind subject)
The depths of field for both lenses are about the same. There are
minor differences that have no real significance when you're shooting
in the field.
Example 2, when focus distance of shorter lens is about 1/2 the
hyperfocal distance for the lens:
40mm lens at f/8, hyperfocal distance = 14.7 feet
Focus distance, 40mm lens: 7.4 feet
Focus distance, 180mm lens: 33.3 feet (= 180/40 * 7.4 feet, to keep
image size the same)
Depth of field, 40mm lens: 10.1 feet (2.5 feet in front to 7.6 feet
behind subject)
Depth of field, 180mm lens: 7.4 feet (3.3 feet in front to 4.1 foot
behind subject)
For this example, the difference in total depths of field is a
little less than 3 feet. Note also that more of the depth of field is
behind the subject (7.6 feet) when the shorter lens is used, as
compared to the depth of field behind the subject (4.1 feet) when the
longer lens is used. The difference between the depths of field for
the lenses is significant, given the relatively short focus distance
for the 40mm lens.
Example 3, when focus distance of shorter lens is about 3/4 the
hyperfocal distance for the lens:
40mm lens at f/8, hyperfocal distance = 14.7 feet
Focus distance, 40mm lens: 11 feet
Focus distance, 180mm lens: 49.5 feet (= 180/40 * 11 feet, to keep
image size the same)
Depth of field, 40mm lens: 37.7 feet (4.7 feet in front to 33 feet
behind subject)
Depth of field, 180mm lens: 16.9 feet (7 feet in front to 9.9 foot
behind subject)
The depth of field for the 180mm lens is less than half the depth
of field of the 40mm lens.
Example 4, when focus distance of the shorter lens is the hyperfocal
distance for the lens:
40mm lens at f/8, hyperfocal distance = 14.7 feet
Focus distance, 40mm lens: 14.7 feet
Focus distance, 180mm lens: 66 feet (= 180/40 * 14.7 feet, to keep
image size the same)
Depth of field, 40mm lens: infinite (7.4 feet in front to infinity)
Depth of field, 180mm lens: 30.7 feet (12 feet in front to 18.7
foot behind subject)
The depth of field for the 40mm lens extends to infinity, while the
depth of field for the 180mm lens extends to only about 19 feet behind
the subject.
------------
Don
http://www.dofmaster.com - free depth of field and hyperfocal distance
calculators for Windows and Palm operating systems
From: Stefan Patric [email protected]
Subject: Re: Depth of field Independent of Focal Length
Newsgroups: rec.photo.equipment.medium-format
Date: Thu, 26 Sep 2002
RD wrote:
> Just whan you thought it was safe . . .
>
> "Contrary to popular belief, depth of field does not depend on the
> focal length of the lens. Depth of field is determined only by the
> magnification of the image, and although telephoto lenses magnify more
> than wide-angles (at the same distance), the depth of field in a
> picture taken by a 24mm lens will be the same as that in one taken by
> a 300mm lens if the subject is the same size in each."
>
> That's a direct quote from the "Basic Photography Guide" UK website of
> Robert Slade. It seems to agree with the earlier comments that DOF
> changes as a function of print size and viewing distance, both of
> which affect the perceived image magnification. Any comments regarding
> the simplicity of this statement?
That first sentence is patently false. Obviously, Robert Slade has
never study basic optical theory. Perhaps, he is referring to apparent
sharpness in a PRINT, which is different from, but related to, the
depth of field of a LENS, and DOES depend on the enlargement
magnification. Apples and oranges.
>From the M&M Photo Lab Index, Photographic Optics and Filter Data
section, Depth of Field of Camera Lenses, subsection: "Depth of field
varies according to the focal length and f-number of the lens, the
distance of the subject, and the acceptable degree of sharpness (circle
of confusion)." One only has to look at the formulas used to actually
calculate depth of field to see that Mr. Slade doesn't know what he's
talking about.
There is, however, a property of depth of field, and this is what the
latter part of the quoted paragraph is referring to, but which was
badly stated. For equal image reproduction ratios (on film, NOT in the
prints), f-stops, and circles of confusion, the depths of field of any
two lenses will be equal, regardless of the focal length of those
lenses. So, in the example used, the depth of field of the 300mm lens
will be the same as the 24, if the reproduction ratios ("subject
size"), f-stops and circles of confusion are equal. This seems to defy
reason until one realizes that the subject distances will be different.
So, in the example, the 300mm lens will be 12.5 times farther from the
subject as the 24mm lens. (This is necessary to get the same repro
ratio, since the 300 is 12.5 times longer than the 24.) And this
greater subject distance compensates for the 300mm lens' inherently
less depth of field (all things being equal).
Also, depth of field is independent of camera format. With all things
being equal, the depth of field, for example, of a 90mm lens is the
same whether it's a 4x5 lens or a 35mm one. So, a 90 designed for 4x5
has the same depth of field as a 35mm 90 or a 90 on medium format. Of
course, the image circles are different, but not the depths of field.
Depth Of Field Formulae
For subject distance u, circle of confusion diameter c', lens focal
length F, lens iris opening diameter A (note: NOT the f-stop number.
The actual diameter of the iris opening is used.)
(near dof point) D1 = c'u*2 / ( AF + uc' )
(far dof point) D2 = c'u*2 / (AF - uc' )
(near distance) L1 = u - D1
(far distance) L2 = u + D2
In normal usage, however, to compensate for the resolving power of the
eye, film format and enlargement ratio when making prints, the circle
of confusion is usually expressed as a ratio of the focal length of the
lens, instead of a particular dimension, i.e. 0.05 inch. So, by
accepted convention, c' = F x (1/1720). Substituting:
D1= u*2 / (1720A + u) and D2 = u*2 / (1720A - u)
--
Stefan Patric
[email protected]
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Depth of field is very simple for view cameras!
Date: Thu, 03 Oct 2002
I've been using my Toho 4 x 5 view camera for about three months now,
and I've been slowly unlearning my habits from many years of medium
format photography. I've had a Horseman Technical camera, which can
function as a 6 x 7 view camera, for over 30 years. In gauging depth of
field, I relied on the standard formulas based on distances on the
subject side of the lens For a 4 x 5 camera, I've realized---somewhat
belatedly---that it makes more sense to use formulas based on distances
on the image side of the lens. I'm sure this is well known to long time
large format photographers.
At f-number N, the range in focus is C*N*(M+1) on either side of the
focus position, where C is the diameter of the acceptable circle of
confusion in the negative, and M is the scale of reproduction. For
distant subjects, one may take M = 0, which simplifies the formula to
C*N. What had misled me was that this distance is quite small and hard
to estimate precisely, even for 4 x 5. For example, with C = .125 mm,
and N = 16, it is only 2 mm. But as I finally realized, all one has to
do is to put a scale on the focusing knob, which is necessarily geared
down. For my Toho, that multiplies distances by a factor of 4.8.
Subsequently to my self discovery, I found just exactly this method in a
large format web page, which elaborates on it considerably. But for me
at least, all that is needed is the equivalent of mm markings on the
focusing knob. The fomula is so simple that I can do the needed
arithmetic in my head, or at worst with a simple calculator. In use, one
just determines how far apart the positions for near and far focus are
on the focusing knob distance scale, and does some simple arithmetic to
determine the needed f-stop given one's choice of C. One then focuses
half way between the two. And, of course one does the best one can to
check everything at the taking aperture on the ground glass.
I've known all this theoretically for quite a long time, but for my
Horseman, I never saw it as a practical way to do anything. It has no
distance scale, and the focusing knob is not something one would think
of putting a scale on. Also because C is smaller and f-numbers often
smaller, the distances are much less. It does have a rangefinder system,
which I did use with the conventional formulas, but its range is too
narrow for fine measurements. So I mostly relied on visual estimates of
subject distances.
Sorry if this is old hat for everyone else, but it is new to me. And
I'm sure some view cameras manufacturers already build appropriate
scales into their cameras, but Toho doesn't.
--
Leonard Evens [email protected]
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
From: "Michael K. Davis" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: hyperfocal vs flatness Re: Medium Format vs. 35mm
Date: Fri, 4 Oct 2002
Hi Alan!
Alan [email protected] wrote:
: I have read an article, from the www, demonstrating that the sharpness
: of distant objects will be better if you focus more toward infinity
: and forego close focus if it is the distant objects that need to be
: sharp. Further, the loss in detail sharpness of closer objects is
: less visible as the objects are larger. If I can locate the link I'll
: post it. The pictorial example had a canon in the foreground (maybe
: someone has seen it) and I think windows in the distance. The author
: pointed out the resolution of the window frames when focused at
: infinity and hyperfocal.
You are referring to this link:
http://home.fox.nstn.ca/~hmmerk/DOFR.html
After reading it again, you might want to have a look at this very long
thread,
http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=002jtA
where Harold Merklinger's focus at infinity technique was exposed as being
convenient at the expense of image quality. I was the antagonist and the
argument I presented has not yet been successfully challenged.
(By the way, the loading of this page pauses midway for some reason, so
wait at least 30 seconds before trying to page down to the bottom...)
http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=002jtA
Mike Davis
--
Date: Sat, 05 Oct 2002
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: hyperfocal vs flatness Re: Medium Format vs. 35mm
...(quotes above post)
I found your comments very much to the point. I remember looking at
Merklinger's web page and being unable to figure out what he was saying.
There is some mathematical analysis there, but to me, a professional
mathematician, it was not clear what its point was. But maybe I'm just
getting old. (I've just tried to check Merlinger's web site again, but
it doesn't respond.)
I finally decided that the proponents of this approach are trying to
deal with situtions where for whatever reason there is insufficient
depth of field to encompass the whole scene. In such cases, they are
arguing for sacrificing objects in the foreground instead of objects in
the distance. The rationale for doing so might be that large image
elements without very fine detail may appear sharp even if slightly out
of focus. Since that is more likely to be true for objects in the
foreground than in the distance, it might be argued that one should
favor the distance. The easiest way to do that is to focus at infinity.
But as you point out, even if you decide not to focus at the hyperfocal
distance, you will do somewhat better, whatever your aims, by focusing
somewhat short of infinity.
It might also be added that for a 35 mm camera, often the distance on
the image side of the lens between the hyperfocal setting and the
infinity setting is very small. (It is essentiall independent of the
focal length and is about a quarter of a mm for a circle of confusion of
diameter .03 mm at f/8.) This is magnified by the gearing down in
rotational movement in focusing the lens. But often the difference
between setting at the hyperfocal distance and infinity is not large.
If there are any errors in film position or the distance scale, that
could affect how accurately the lens is focused when supposedly focused
at the hyperfocal distance.
Merklinger's comments about the Scheimpflug principle have also confused
me. He notes that the principle is a necessary but not sufficient
condition for determining the plane of focus. That is in fact true
since there are infinitely many planes which satisfy the condition. But
once you choose one point in that plane, that together with the
Scheimpflug principle, determine it uniquely. You can choose such a
point by focusing. The hinge rule is a practical aid in some
circumstances because it gives you an additional condition, which
together with the Scheimpflug principle determines the plane of best
focus. From a purely mathematical point of view, it is not necessary to
use it if one has focused precisely, but of course in the real world,
one often wants redundant information as a check.
> Mike Davis
--
Leonard Evens [email protected]
from leica topica mailing list:
Date: Wed, 04 Sep 2002
From: Jim Brick [email protected]
Subject: Re: Subjective test of four 50 mm lenses
Mike,
All 50mm lenses have the same depth of field, at the same f/stop, from the
same distance.
There are lens formulas that fudge this one way or another but it is only
detectable on an optical bench.
Go here to read my writing on DOF.
http://www.clearsightusa.com/depthofield.html
The sharper the lens, the less apparent DOF it will have. This is because
of the sharp-to-unsharp abruptness. Less sharp lenses tend to appear to
have more DOF because it is difficult to tell where the sharpness stops and
the fuzziness starts.
Jim
[Ed. note: re: MF vs. 35mm DOF stops loss]
From: "David J. Littleboy" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Compared to 35mm...
Date: Sat, 23 Nov 2002
"maf" [email protected] wrote:
> He is probably correct about not going back to the old thread, since in the
> current thread (different post) he answered the question correctly "but the
> 35 mm print will show more depth of field."
Actually, the answer in this thread got that much right, but it
overestimated the difference. It's only about 1/2 stop for 645 and 1 stop
for 6x7 for the same CoC on the same size print. However, MF is used for
larger, higher-quality prints than 35mm is (that's the whole point of MF,
IMHO), which means that the 2 to 3 stop less DOF estimate for MF is spot on.
David J. Littleboy
Tokyo, Japan
From: "David J. Littleboy" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Compared to 35mm...
Date: Sun, 24 Nov 2002
"David J. Littleboy" [email protected] wrote:
>
> Sorry. I wasn't clear. MF has a slightly narrower DOF for the same size
> prints, but you intend to make larger prints, so the DOF is even narrower.
Still not right. Someone pointed out (in private email) that I've been
misunderstanding the approximation I was using.
Let RATIO be the ratio of the size of the larger format to the smaller
format.
The number of stops you have to adjust to get the same DOF in the same size
print is:
delta-f = 2*log(RATIO)/log(2)
For 645, RATIO is 41/24, delta-f is 1.55 stops (not 1/2 stop)
For 6x7, RATIO is 56/24, delta-f is 2.44 stops (not 1 stop)
(Assuming 8x10 prints with no cropping in the short direction.)
> As someone else pointed out though, using a tripod and stopping down gets
> you almost to where you were with 35mm.
Hmm. I guess this is wrong too: the lenses at hand really want to be used at
f/8 or f/11, and are noticeably soft at f/22, so even for 645, stopping down
1.5 + may be difficult some of the time.
This also says that the DOF of lenses faster than 2.0 is going to be
radically shorter than for 35mm, a reason, I think, for not being interested
in such lenses.
David J. Littleboy
Tokyo, Japan
Date: Sun, 24 Nov 2002
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Compared to 35mm...
maf wrote:
>>maf wrote:
>>
>>
>>>I do a lot of landscape photography with 6x7 format. DOF and sharpness
>>
> is
>
>>>quite good at f/11 - f/16.
>>
>>Yes, but landscape tends to be at rather distant focal
>>points, even if using hyperfocal technique. The ongoing
>>discussion on 35mm vs MedFormat DOF becomes much more a
>>relevant thing to keep in mind when photographing at closer
>>distances.
>>Bill D. Casselberry
>
> No question about that, however I usually have some part of the image at a
> distance of about 15 feet transitioning up to infinity. Landscape
> photography typically uses wide angle lenses, which also helps with DOF.
>
> In a previous post, I mentioned that it is important to know what kind of
> photography is contemplated, the expected lighting conditions, and the speed
> of the film, to determine whether MF is feasible.
>
> But even with closer subjects, if one is using a tripod, stopping down to
> f/22 or more is an option. I suspect that the diffraction limitation issues
> are balanced by the lower magnification requirements of the larger MF
> negative. I am sure that the mathematicians can expound on this.
Well, for a change we seem to agree.
I also do a lot of 6x7 photography of landscapes and I have no problem
stopping down to f/16 in most situations. With my choice of coc in the
final print (.2 mm), I come up with the following hyperfocal distances
65 mm 15 ft
90 mm 30 ft
150 mm 83 ft
So with the 65 mm lens, if focusing at the hyperfocal distance, in
principle everything from 7 1/2 ft to infinity would be in focus, and
with the 90 mm lens everything from 15 ft to infinity would be in focus.
Of course, some people may be hypercritical and want to use a coc of
half the size I find acceptable. For them, the figures would have to be
doubled.
I can even often get by at f/8, particularly with the 65 mm lens.
As to diffraction, according to Jacobson, a good estimate for the
diameter of the Airy disc in mm is .00135 N where N is the f-number.
With N = 16, this comes out to 0.0216. To produce an 8 x 10 print, one
must enlarge 3.63 times, and that increases it to about 0.08, which is
a bit over one third of 0.2. On the other hand, perhaps the grain
sniffer referred to above, who insists on a coc of diameter 0.1, would
worry.
It was my impression that with 35 mm, it was usually at f/22 that
diffraction became a real problem. According to that standard, one
would have to stop down to f/45 for it to become a problem with 6 x 7.
It should be noted that David Littleboy does 6 x 4.5, and he is very
picky. Together those factors might explain the difference between him
and those of us using 6 x 7 and somewhat lower standards.
Finally, it should be pointed out that the problem with 6 x 7 (or other
medium or large format) vis-a-vis 35 mm is not really a depth of field
or diffraction. All that happens is the range of usable f-stops gets
shifted down toward smaller apertures by some number of f-stops. The
real problem is exposure time and hence subject movement. Of course,
the two are related since for the same film and lighting, the exposure
must be the same. But if you have a still subject or the movement is
sufficiently far away, there is no problem getting the depth of field
you want, whatever the format.
--
Leonard Evens [email protected]
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Compared to 35mm...
Date: Thu, 21 Nov 2002
Gregory L. Hansen wrote:
> I don't know much about cameras beyond the 35mm world, but I became
> curious about larger formats. How do the shooting parameters compare?
> Does anything special happen when you're focusing a much larger image onto
> your film?
The image is larger, so depending on the optics, it may be easier to
evaluate the scene.
>
> Say you have lenses that offer equal solid angle (or equal magnification,
> whatever, not equal focal lengths because I image effective focal lengths
> are different). I'm assuming the medium format world has their own set of
> lenses, but I don't really know. Would the 35mm and larger format have
> the same depth of field?
If the angle of view and point of view are the same, and you make the
same size final print viewed the same way, then 35 mm has roughly a two
stop advantage in depth of field over medium format. But this depends
on which format you use. It depends on the ratio of the linear
dimensions of the image. 35 mm has a short dimension of 24 mm. 645 has
a short dimension of about 45 mm and the ratio is 1.875. 6 x 6 has a
square format, but often people crop it to 4:5 aspect ratio, so it is
equivalent to 645. 6 x 7 has a short dimension of 56 mm and a ratio of
2.33. So the difference ranges from a little under two stops to about 2
1/2 stops.
> The same light intensity on the film at a given
> aperture? Same exposure times at a given aperture?
Yes.
> I'd imagine the
> medium format gives you better ability to crop and zoom in on small or
> distant features, sort of an ersatz telephoto if it came to that.
Up to a point. It depends on the quality of the lens how far you can do
that.
> Does
> medium format film come in the variety of film speeds that 35mm comes in,
> or is it all like pretty slow?
I haven't noticed any significant difference between the films that are
available. There are a few film types that are available only in 35 mm,
but not many. However, it may be more difficult to find 120 film. You
may have to order it from a large photo supplier if one isn't handy
where you live.
> I could imagine 6400 speed film that was
> amazingly sensitive but could still give non-grainy images if you don't
> try to enlarge it too much, simply because the negative is so big. But I
> suppose you couldn't just go to the local MotoPhoto to get it developed.
My local Motophoto develops 120 color film, and they do a good job.
They don't do b/w but they don't do that for 35 mm either. They will
also make 5 x 5 prints.
One thing you didn't mention is that you get fewer exposures on 120
film. The exact number depends on the format. 6 x 7 gives 10
exposures, 6 x 6 gives 12, and 645 gives 16. Compare that to 35 mm
which usually gives 24 or 36 or more. Medium format photographers tend
to be more deliberate with their shots because of this. On the other
hand, you don't have to wait as long to complete a roll of film and
develop it or have it developed. (There is 220 film which is roughly
twice as long, but it is not as readily available, and can't necessarily
be used in every camera.)
--
Leonard Evens [email protected]
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Compared to 35mm...
Date: Fri, 22 Nov 2002
roland.rashleigh-berry wrote:
> "Gregory L. Hansen" [email protected] wrote >
>>I don't know much about cameras beyond the 35mm world, but I became
>>curious about larger formats. How do the shooting parameters compare?
>>Does anything special happen when you're focusing a much larger image onto
>>your film?
>
>
> I'm sure somebody will correct me if I am wrong but here goes...
>
> First thing with these larger formats is that film flatness can be a
> problem. Certain;y 6x9 format can suffer from this. A smaller format like
> 6x4.5 less so. Also, since the focal lengths are longer then not only is
> depth of field less but also depth of focus as well to excaccerbate the film
> flatness problem.
Depth of focus, for distant objects, is N*c where N is the f-number and
c is the diameter of the acceptable circle of confusion in the film
plane. c is generally larger for medium format than for 35 mm because
for the same size final image viewed the same way, you enlarge less, so
you can tolerate a bigger circle of confusion in the film.
So tolerance for film plane errors is actually larger for medium format.
On the other hand, it is also harder to control film position.
> I tend to use my MF gear for landscape work and stop down
> to f16 to avoid these problems. This reduces contrast and resolution but the
> film area is larger so it compensates.
I haven't noticed any loss of quality with my medium format lenses if I
stop down to f/16. The need to enlarge roughly half as much decreases
the significance of diffraction.
--
Leonard Evens [email protected]
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: How to use your lens DOF scale creatively
Date: Wed, 20 Nov 2002
I've never had a depth of field scale on a medium format lens, and I
usually just ignored those on my 35 mm lenses. But since taking up
large format photography and thinking a bit about the matter, I realized
that a common large format technique can be used on medium and 35 mm
reflex cameras for which lenses have depth of field scales.
The method is as follows. Find the center of the dof scale. Looking on
the ground glass, find the points on the distance scale where that
center lies for the nearest part of the scene you want in focus and for
the most distant part of the scene you want in focus. Then place the
center of the scale half way in between. Since the f-stop markings on
the dof scale are symmetrically placed about the center, you can use the
scale markings to help you locate the midpoint on the lens barrel. Then
there will be a certain f-number on the scale sitting opposite the two
reference points. The next part is a little trickier. The lens
manufacturer has chosen a criterion for what is considered sufficiently
sharp. If you are willing to go with what the manufacturer chose, just
use the f-stop indication that the scale gives you. But you may wish to
be more demanding. Decide how much more demanding you want to be on the
basis of your experience with the camera and lens. This will give you a
factor. Say you decide you want to be 1.5 times as fussy. Just
multiply the indicated f-number by 1.5 and use that instead.
Note that the dof scale markings are really just a distance scale. They
work for distant objects and the formula is simple N*K where N is the
f-number and K is a scale factor determined by the manufacturer's
criterion for sharpness and the gearing ratio relating rotation of the
lens barrel to movement of the lens along its axis.
I apologize if the above is not as clear as it might be. While it is
very easy to do these things in practice, I find it hard to describe
them in words. And of course I don't suggest you drop what you are
doing now and use this instead if you are satisfied. Still a few people
might find it useful. I won't be one of them since my lenses don't have
dof scales.
--
Leonard Evens [email protected]
From: [email protected] (Bill Hilton)
Newsgroups: rec.photo.equipment.medium-format
Date: 20 Nov 2002
Subject: Re: How to use your lens DOF scale creatively
>From: Leonard Evens [email protected]
>I don't disagree with your main point. But I was trying to point out
>that the scale markings are proportional to the chosen coc. So all you
>need to know is the proper factor to multiply the numbers by.
This is true and I agree with what you wrote earlier, but the problem with
using the numbers on the barrel of MF lenses is precisely this, you "need to
know the proper factor to multiply the numbers by" since they are for 8x10"
prints (at least Mamiya agrees this is what they're doing). Different MF
makers seem to use different numbers too, so unless you know the fudge factor
they are worse than worthless since they are so misleading.
>One difficulty with using dof tables or calculators is that they give
>you distances from the lens to elements of the scene. Unless you have a
>laser rangefinder, it is very difficult to measure those distances.
For most landscape photography situations using wide to moderately wide lenses
the distances are so close that you can either estimate them accurately or
measure them quickly. For example, for a 35 mm lens (which I have for my 645)
the HF distance for the formula I use is either 6 or 7.4 ft @ f/22 (depending
on which coc I use), so I can set the focus point on the lens scale to one of
these distances and know I'm in focus from either 3 ft or 3.7 ft to infinity.
I can "guesstimate" 3 ft pretty well, I've found.
For 45 mm (I have this focal length in both 645 and 6x7 ... 43 mm anyway, close
enough) the close-focus distances are about 5 and 6 ft (actually 4.9 and 6.2
but you get the idea). Often I'll just point the lens just beyond the tripod
foot, set the focus point to 10 ft and know I'm sharp to infinity. So it's not
that big a problem with medium format landscape photography and shorter lenses.
If you had to accurately 'guess' the differences between say 50 ft and 40 ft it
would be difficult to be accurate, but when you're usually talking distances
inside 10 ft it's not that big a deal, in practice. Or so I've found. You can
always carry a 10 ft piece of string marked off in feet too so you can quickly
measure it. Once your lens is 100 mm or longer then you're right, you can't
accurately estimate the distance to HF (or at least I can't).
Bill
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.large-format,rec.photo.equipment.medium-format
Subject: How is coc related to lp/mm?
Date: Sun, 24 Nov 2002
One thing that has confused me for a while is the relationship between
the diameter c of the acceptable circle of confusion and an equivalent
figure in terms of the spatial frequency in lp/mm. Naively, one would
think that c should one half the wavelength or 1/(2s). Similarly, s
should be 1/2c. But it appears that most people use 1/c for the spatial
frequency. I think this depends on optical transfer function theory.
For example, if you look at www.photo.net/learn/optics/lensTutorial
you will see an OTF for a coc of diameter 0.03 mm. Taking 1/(2c) yields
a frequency of 16.7 lp/mm and taking 1/c yields a frequency of 33.3
lp/mm. At the former value, the OTC drops to about 75 percent, and at
the latter value it drops to about 20 percent. So I suppose one could
make arguments for either value, but perhaps 16.7 lp/mm = 1/(2c) would
be much too conservative.
Can anyone enlighten me further on these matters. Also a naive
explanation not based on OTFs would be appreciated, even if it isn't
accurate.
Another interesting insight from the graph in the Jacobson tutorial is
that drop off due to diffraction is very different. For example, at
f/22, which produces an Airy disc comparable in size to the .03 coc, you
have to go to about 55 lp/mm before you drop to 20 percent. The combined
effect is even more complicated. This suggests that simply comparing
circles of confusion and Airy discs or combining them by simple addition
or using root mean square may give overestimates of the combined effect.
--
Leonard Evens [email protected]
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 24 Nov 2002
Subject: Re: How is coc related to lp/mm?
> diameter c of the acceptable circle of confusion and an equivalent
>figure in terms of the spatial frequency in lp/mm.
C of C is always that value that will give 8 lp/mm at a view distance that
gives the correct perspective.
An example, if we use a 55mm lens on a 35mm and blow it up to 8x10 with a view
distance of 10 inches. The final print will have 8 lp/mm and will be viewed at
the correct perspective. Change any of the above values and you change the
circle of confussion.
So any C of C value takes into account view distance. Anyone who ever walked
up to a bill board (under 10 feet) would see that the images are very out of
focus. Walk back 100 feet and the image will sharpen up. From the highway at
over 100 ft the image looks very sharp.
That is how the word "confusion" comes into play. As we go through a view
distance, at some point in that distance, a fuzzy circle or fuzzy dot will no
longer be confusing and the image will look sharp.
Larry
From: [email protected] (Michael Gudzinowicz)
Newsgroups: rec.photo.equipment.large-format
Subject: Re: How is coc related to lp/mm?
Date: 25 Nov 2002
Leonard Evens [email protected] wrote:
> One thing that has confused me for a while is the relationship between
> the diameter c of the acceptable circle of confusion and an equivalent
> figure in terms of the spatial frequency in lp/mm. Naively, one would
> think that c should one half the wavelength or 1/(2s). Similarly, s
> should be 1/2c. But it appears that most people use 1/c for the spatial
> frequency. I think this depends on optical transfer function theory.
> For example, if you look at
> www.photo.net/learn/optics/lensTutorial
> you will see an OTF for a coc of diameter 0.03 mm. Taking 1/(2c) yields
> a frequency of 16.7 lp/mm and taking 1/c yields a frequency of 33.3
> lp/mm. At the former value, the OTC drops to about 75 percent, and at
> the latter value it drops to about 20 percent. So I suppose one could
> make arguments for either value, but perhaps 16.7 lp/mm = 1/(2c) would
> be much too conservative.
>
> Can anyone enlighten me further on these matters. Also a naive
> explanation not based on OTFs would be appreciated, even if it isn't
> accurate.
I don't have the references cited by David, however, he and Bob
Atkins have discussed this at length years ago. The discussions
should be in the archives.
The graph shows the effect of defocus alone (0.03mm coc), the
effect of diffraction alone at f/22, and the combined effects of
defocus sufficient to give a coc of 0.03 mm and diffraction at
f/22. I don't see the rationale of using this graph to
determine the formula for coc to lpmm conversion. However, it
does demonstrate phase reversals leading to spurious resolution
measurements.
A point source and lines are treated differently. If I remember
correctly, the line scenario does not use the "1.22" constant or
the inverse mentioned by David (0.823). If you consider the
Rayleigh resolution limit, rather than the approximation of
1600/Ne for the diffraction limit, one would use 2000/Ne for
lines. Larry, who responded in the medium format group thread,
has a reference and graph where the latter value is used.
Therefore the limit in lines per mm would be 1.22 / Airy disk
radius.
> Another interesting insight from the graph in the Jacobson tutorial is
> that drop off due to diffraction is very different. For example, at
> f/22, which produces an Airy disc comparable in size to the .03 coc, you
> have to go to about 55 lp/mm before you drop to 20 percent. The combined
> effect is even more complicated. This suggests that simply comparing
> circles of confusion and Airy discs or combining them by simple addition
> or using root mean square may give overestimates of the combined effect.
If you check Bob's posts, you'll find that one of point of
the discussions was that for a given value of defocus, there is
an optimum aperture at which sharpness is maximized. That
aperture is very close to that based on simple geometrical
considerations which were summarized in the spreadsheet I had
sent to you some time ago.
Over a couple of decades, I've done careful resolution lens tests
using "good" lenses which were limited primarily by diffraction
when stopped down. When the data was examined using nonlinear curve
fits for different models, the RMS approach frequently required or
led to values for maximum film resolution which were much less
than the actual test resolution on film. That absurdity did not
occur with the simple geometric model where sums of circles are
used. I'd suggest that you do you own tests using targets which
are more sensitive than the USAF target. Mine use gradations
of 1/2 ^ 1/12th power rather than the USAF 1/6th.
Also, you won't see an Airy disc using most camera lenses and
pictorial film. Many photographs of discs are made by focusing a
microscope on the image using a monochromatic source. David
mentions that one should integrate over the entire visible
spectrum. If you do that, a smooth curve results which resembles
the central peak of the 555 nm curve. However, there are no
minima/maxima - it tapers off smoothly. In other words, the
disc surrounding the peak doesn't exist. In other words, one just
gets a single spot. Although the model works well for astronomers
and Rayleigh, there are problems if one just assumes that visible
light is equivalent to a monochromatic source or filtered light.
That problem is further compounded by the nature of the emulsion.
The crystals in most emulsions are large, are randomly
distributed, and scatter light in the emulsion. The resultant
spot size may also vary with exposure. Generally, I place the
"white" portion of the target on zone 8 for both reflected and
transmission targets. Overall, the spot size is usually a bit
larger that predicted by the appropriate equations. So when using
film, I generally use 1/coc to get lpmm resolutions, and use the
Rayleigh formula for diffraction effects because it gives the
best fit to real world pictorial negatives. However, I'm
comparing spots made by the visible spectrum and not
monochromatic peaks and rings. Although the theoretical approach
is satisfying, film is a very poor detector which can't be
ignored.
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: How is coc related to lp/mm?
Date: Mon, 25 Nov 2002
...
> I don't have the references cited by David, however, he and Bob
> Atkins have discussed this at length years ago. The discussions
> should be in the archives.
Thank you. I will look in google groups, and I will try to check
Jacobson's references. I've looked at a couple of books on digital
image processing, and I now understand the mathematics in Jacobson's
tutorial.> The graph shows the effect of defocus alone (0.03mm coc), the
> effect of diffraction alone at f/22, and the combined effects of
> defocus sufficient to give a coc of 0.03 mm and diffraction at
> f/22. I don't see the rationale of using this graph to
> determine the formula for coc to lpmm conversion.
I've also looked in "Photographic Optics" by Arthur Cox. My copy is
over 30 years old, but I doubt if much has changed in this area since
then. He goes breifly into optical transfer functions by drawing graphs
without the accompanying analytic formulas. But he does give some sort
of rationale for converting to lp/mm. There is a certain cutoff
frequency where there is zero response, but it is a matter of convention
at which level above zero, you consider the response to be high enough
to be detectable. Apparently it is common to choose a response of about
20 percent. If you use the rule 1/coc for the spatial frequency, that
would be consistent with the graph in the Jacobson tutorial and also in
Cox at 20 percent response.
...
> Therefore the limit in lines per mm would be 1.22 / Airy disk
> radius.
I guess that the mathematics is pretty clear, but if you are going to
convert a function into a single number, there may be a lot of room for
interpretation just which mathematics you use and how you go about it.
Ultimately, if you really want to understand just what happens,
presumably you have to use the complete analysis using the OTF instead
of relying on one number and you have to be very specific about what you
are looking for.
> If you check Bob's posts, you'll find that one of point of
> the discussions was that for a given value of defocus, there is
> an optimum aperture at which sharpness is maximized. That
> aperture is very close to that based on simple geometrical
> considerations which were summarized in the spreadsheet I had
> sent to you some time ago.
Unfortunately, I work under Linux, and I have to make a special effort
to fire up a machine with Windows and Excel, so I haven't looked at it
yet. But I will give it a try.
...
> Rayleigh formula for diffraction effects because it gives the
> best fit to real world pictorial negatives. However, I'm
> comparing spots made by the visible spectrum and not
> monochromatic peaks and rings. Although the theoretical approach
> is satisfying, film is a very poor detector which can't be
> ignored.
You are clearly way beyond me in these matters, so I will just take your
word for it. Years ago, I did some primitive lens tests, but these days
I just assume that my lenses are more than good enough for my purposes.
So far, that has been the case. I also don't have much problem with
actually working with depth of field as a practical matter. I have
tentatively settled on a coc of diameter 0.1 mm for 4 x 5 film, and this
seems well within the range of what is considered acceptable. I may be
at the low end, but at present I am scanning my film, and my scanner is
probably the limiting factor in resolution. Diffraction is not really
an issue for me since if I were to stop down enough for it to be
significant, subject motion would be much more of a problem because of
the long exposures necessary, at least for the typical subjects I
> Rayleigh formula for diffraction effects because it gives the
> best fit to real world pictorial negatives. However, I'm
> comparing spots made by the visible spectrum and not
> monochromatic peaks and rings. Although the theoretical approach
> is satisfying, film is a very poor detector which can't be
> ignored.
You are clearly way beyond me in these matters, so I will just take your
word for it. Years ago, I did some primitive lens tests, but these days
I just assume that my lenses are more than good enough for my purposes.
So far, that has been the case. I also don't have much problem with
actually working with depth of field as a practical matter. I have
tentatively settled on a coc of diameter 0.1 mm for 4 x 5 film, and this
seems well within the range of what is considered acceptable. I may be
at the low end, but at present I am scanning my film, and my scanner is
probably the limiting factor in resolution. Diffraction is not really
an issue for me since if I were to stop down enough for it to be
significant, subject motion would be much more of a problem because of
the long exposures necessary, at least for the typical subjects I
photograph. I'm lucky if I can get by with f/22 in most cases.
The reason I asked the question was for communication. Often people say
things about resolution in terms of lp/mm, and it is not clear how to
convert from cocs to lp/mm. I think I have a much better hold on that
now thanks to my reading and your response.
--
Leonard Evens [email protected]
From leica user mailing list:
Date: Tue, 18 Mar 2003
From: Adam Bridge [email protected]
Subject: [Leica] Wavy Lenses and depth of field
There's an interesting article on Slash-Dot about wavy lenses which can increase
the depth of field by a factor of 10.
The site with the basic information has been "slash-dotted" (that is swamped by
requests in excess of its ability to respond) but it's an interesting article.
Of course we'll wonder what will happen to the bokeh...
http://slashdot.org/article.pl?sid=03/03/18/1547249&mode=thread&tid=126&tid=152
[Ed. note: - interesting review of lenslets for credit card thickness lenses and digicams;
also defocusing lenses which increase depth of field by tenfold etc. ;-) ]
From: George Kenney [[email protected]]
Sent: Sat 5/24/2003
To: [email protected]
Subject: Re: [HUG] A new Lens for the V-System
I found that article I mentioned. As Jim noted, this is not news for chip
designers and was, in fact, the subject under discussion here if only I'd
not been too dim to understand. Nevertheless, for people who prefer the
long explanation in English instead of the engineers' verbal shorthand, I
think this is a really interesting article. An additional benefit: It gives a few
comparative examples of photos with/without the induced blur which
leads, through the miracle of software, to sharper, greater depth of field
images. Btw, I think this process was just recently patented.
Thanks, Jim, for the note about the white paper. I guess this has
advanced a lot farther than I'd imagined.
http://www.sciencenews.org/20030329/bob9.asp
From: Leonard Evens [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: DOF Calculation Method for View Cameras
Date: Sun, 15 Jun 2003
Nicholas O. Lindan wrote:
> I am repeating this post under a new title. It was originally in the
> "photographing an Eclipse" thread - not the place you would expect to
> find depth of field info: Lets see, the sun is 92 million miles and the
> moon is .... gee, my table don't go that far...
>
> Herewith is an easy way to set DOF with a view camera:
>
> o focus for the farthest item to be in focus
> o note the position of the camera standard (light pencil mark, etc.)
> o focus for the closest item to be in focus and make another mark
> o set the camera to the spot between the two marks
There is an extensive discussion of this and related matters at
www.largeformatphotography.info
For the somewhat mathematically inclined, see also
math.northwestern.edu/~len/photos/pages/dof_essay.pdf
>
> Sinars (and possibly others) have little gadgets for doing this easily.
> The Sinar gadget also has a scale for reading out the required f-stop.
>
> Courtesy of Sinar, here's a table, for 4x5, that will determine
> the proper f-stop by measuring the distance between the two marks:
>
>
> f distance
> in cm.
The distances you give are in mm, not cm.
>
> 45 8.5
> 32 5.9
> 22 4.2
> 16 3.0
> 11 2.1
> 8 1.5
> 5.6 1.1
Sinar is apparently using a maximal acceptable circle of confusion of
diameter 0.1 mm. The actual formula they are using is quite simple.
Focus spread = 2 x f-number x diameter of coc
So for f-number 5.6, you get 2 x 5.6 x 0.1 = 1.12 mm
But the coc is really a matter of individual choice dependent on what
one expects to do with the resulting negative or transparency. A value
of 0.1 mm corresponds to a choice of 0.2 mm in an 8 x 10 print viewed at
about 10-12 inches (or a larger print viewed proportionately further
away). This also corresponds to 5 lp/mm at 10-12 inches. Some people
may want a smaller value, perhaps as low as 0.05 mm. Also, as one stops
down, diffraction becomes an issue. The large format photography web
page suggests a couple of methods of trying to balance defocus and
diffraction. The most popular one is due to Paul Hansma.
It should be noted that one can put a scale on the focusing knob of most
any view camera and accomplish the same thing that the Sinar has. A
disucssion of how I did that is in the document I wrote referred to above.
> The table is approximate as it was taken from measurements on my Sinar.
> It is also based on an 8x10 print (that's magazine size, this _is_ a
> commercial camera). I stop down an extra stop and don't have much
> problem with 20x24 prints. Open up one stop when shooting 8x10 film.
>
> Note this method is independent of focal length - neat, huh? Focal
> length got worked into the process by the movement it took to focus
> the lens near and far.
>
> For a mark making surface one can use a piece of tape with that
> "Post-It" removable adhesive. Or what the heck, glue on an unglazed
> ceramic tile - works great, and forever, with a pencil and eraser.
>
> A.I.: (Almost Useless) Don't use a ball-point pen, fountain pen etc.
> to write on ceramic: the ceramic will wear down and ruin the writing
> surface faster than you can figure the distance to the sun....
> --
> Nicholas O. Lindan, Cleveland, Ohio [email protected]
> Consulting Engineer: Electronics; Informatics; Photonics
--
Leonard Evens [email protected]
From: Paul van Walree [email protected]
To: [email protected]
Date: Sat, 12 Jul 2003
Subject: [Contax] VWDOF 2.0
A new version of the VWDOF calculator is available at
http://www.vanwalree.com/optics/vwdof.html
If you like it, thank muchan for it. There are many new features;
hopefully more than there are bugs :)
The concept of DOF is further illustrated at
http://www.vanwalree.com/optics/dof.html
Combined with the calculator user manual this is a lot of reading.
Not too much, I hope; some things just don't come easy.
From: "Nicholas O. Lindan" [email protected]
Newsgroups: rec.photo.digital,rec.photo.equipment.35mm,rec.photo.equipment.large-format,rec.photo.equipment.medium-format
Subject: Depth of field: From Zero to Infinite
Date: Tue, 18 Nov 2003
"David Littlewood" [email protected] wrote
> Nicholas O. Lindan [email protected] writes
> >"David Ruether" [email protected] wrote
> >
> > > DOF can be affected
> > > by lens resolution, with the sharper lens having slightly less DOF
> >
> >With pinhole photography having infinite depth of field and the
> >simplest lens of all.
>
> Or a zero DoF, depending on what size circle of confusion you define as
> "acceptable
A pinhole is uniformly, er, sharp at all distances. So depth of
field is infinite.
A lens only produces its smallest CoC at only one distance with things
getting worse at any other distance. If one takes the optimum CoC as the
limit of acceptability then the depth of field collapses to zero.
At the extreme, zero depth of field is only sharp at one distance
with no image production at any other distance. The picture would
be of an outline of objects that cut the forward focal plane (correct
term escapes me).
Confocal systems achieve infinite depth of field by using a
zero depth of field technique and build up the image as the
object is moved through the imaging plane. The major use
for this technique is microscopy of living cells and
ophthalmic retinal photography.
http://www.microscopyu.com/tutorials/java/virtual/confocal/index.html
There is someone who has taken marvelous insect photographs with
a confocal setup.
As an example of a do-it yourself confocal setup
* Make a vertical slit in a 35mm slide mount with two razor blades
* Project the slit-slide in a dark room
* Mount a camera on a tripod at right angles to the projector
and focus the camera on the light plane
* Open the shutter and move the object through the plane
of light. You will get a side-lit image -- extra projectors
(or a spinning light source) are needed for uniform
illumination.
Bingo - the whole object is imaged in focus and depth of
field is independent of lens aperture.
--
Nicholas O. Lindan, Cleveland, Ohio [email protected]
Consulting Engineer: Electronics; Informatics; Photonics.
".
> --
> David Littlewood
From: "Nicholas O. Lindan" [email protected]
Newsgroups: rec.photo.digital,rec.photo.equipment.35mm,rec.photo.equipment.large-format,rec.photo.equipment.medium-format
Subject: Re: Depth of field: From Zero to Infinite
Date: Tue, 18 Nov 2003
"Dennis O'Connor" [email protected] wrote
> Nicholas O. Lindan wrote:
> > Confocal.... http://www.thales-optem.com/pdf/ViewMedia.pdf
> Now, if I can just get my portrait subjects to hold REAL still, while I do
> this...
Feed me a line, please. It _has_ been done:
http://www.pbs.org/wgbh/nova/sheppard/mugshot.html
http://www.3dcrystalimages.com/Pages/3DPortraits.htm
This is a hot area in engineering/scientific imaging
as increased performance and decreased cost of lasers,
CCD imagers and computing power are moving the technology
out of the laboratory.
Confocal imaging: coming to a shopping mall near you!
--
Nicholas O. Lindan, Cleveland, Ohio [email protected]
Consulting Engineer: Electronics; Informatics; Photonics.
From: "Nicholas O. Lindan" [email protected]
Newsgroups: rec.photo.digital,rec.photo.equipment.35mm,rec.photo.equipment.large-format,rec.photo.equipment.medium-format
Subject: Re: Depth of field: From Zero to Infinite
Date: Tue, 18 Nov 2003
And, for the integrally minded:
http://www.thales-optem.com/pdf/ViewMedia.pdf
It does show a confocal comparison for a solid object, a pollen
grain.
--
Nicholas O. Lindan, Cleveland, Ohio [email protected]
From: "Nicholas O. Lindan" [email protected]
Newsgroups: rec.photo.digital,rec.photo.equipment.35mm,rec.photo.equipment.large-format,rec.photo.equipment.medium-format
Subject: Re: Depth of field: From Zero to Infinite
Date: Tue, 18 Nov 2003
A better explanation of confocal microscopy:
http://www.solarius-inc.com/html/confocal.html
Note that confocal microscopy is used for imaging transparent
(and murky) 3D objects such as living cells -- a bit
different from the usual photographic subject for most of
us.
--
Nicholas O. Lindan, Cleveland, Ohio [email protected]
Consulting Engineer: Electronics; Informatics; Photonics.
End of Page