Lens Resolution Testing
by Robert Monaghan
USAF 1951 Test Chart Resources:
USAF Lens Testing Archive
USAF Test Chart (4x5"; TIF)
USAF 1951 Chart (.PDF file)
USAF1951 Chart (autocad .dwg format,
thanks to Bertho Boman! (posting)
USAF1951 Chart Info (.png scan file,
thanks to Charles R. Batishko! (posting)
USAF Test Chart Resources (Bertho Boman)
Related Local Links:
Lens and Film Resolution Pages
MTF Charts
Pseudo MTF Tester Project Tips
Resolving Power Spreadsheet (Excel)
courtesy of Frank Loeffel (posting)
Related Links:
A Sharper Image
Digital Camera Resolution (using Nyquist Chart) [10/2000]
Digital vs Film (comparisons, etc.)
Image Detail from Scanners Study [1/2001]
In Search of Sharpness (QT Luong)
John Chapman's Lens Test Resource Site [8/2002]
Johnston's Lens Ramblings (sharpness)
Large Format Lens Spec Comparison Chart (+ lens tests) by Chris Perez
Leica M Lens Test Results [4/2002]
Lens Ratings - Medium Format Lenses
Lens Terminology contrast vs. resolution (photozone)
Lens Testers Anonymous
Lens Test Chart (Edmund Scientific Sales)
Lens Test Chart - USAF 1951 [Sinepatterns Sales]
Lens Testing Chart and How-to-use Info
Lens Testing How To Article (William E. Sadler)
Lens Testing Methodology (Robert Atkins)
Lens Tests from Popular Photography Magazine (Sigma)
Lens Tests on Medium Format Lenses (and lrg fmt - C.Perez..)
Lens Tutorial by David Jacobson
Lens Tutorial for Large Format
Norman Koren's Lens Testing Charts and Resources (recommended!)
Visual (+Math)
Review of MTF... (Norman Koren) [2/2001]
Selecting the Sharpest Aperture (Ken Rockwell) [9/2002]
User Lens Test Reports:
Lens Testing (35mm vs. 645/6x6) by Fernando Carello
Lens Testing in the Backyard by Jim Bullock
We have archived the USAF 1951 lens testing materials and some related
links and posts here. Mr. Charles Sleicher (Email [email protected]) has produced a very worthwhile lens
testing chart, recently reviewed in Popular Photography's December
1999 issue by Mr. Herbert Keppler in his SLR Column. This chart is
especially valuable for color as well as black and white chart
elements. Moreover, Mr. Sleicher provides a sample slide for comparison
and evaluation. This slide is critical in helping new lens testers have a
standard by which to compare their lens tests results. The results should
be much more comparable lens test results from using this resource!
Notes:
From Modern Photography, June 1965, Bennett Sherman, Techniques Tomorrow, p.31:
What about the difference between the popularly priced lens and the very
expensive one? First of all, there is not a very great difference between
the optical performance. Most lenses are very nearly the same optical
designs, such as the familiar Biotar types. In the expensive lens, an
extra effort is made to keep the focal length of the manufactured lens
very close to the design value. In the less expensive types, the focal
length may vary a bit more. There can also be a small variation in the
correction qualities for close ups, and the less expensive lens might show
a bit more variation of sharpness at various apertures. You'll probably
never notice it in everyday shooting, but careful testing including
resolution charts, can show up these slight differences. Because of close
tolerances in manufacturing and testing, the more expensive optics show
a greater uniformity of performance, lens to lens. [italics in
original] In any case, careful testing can tell you what to expect from
your lens, and quickly identify a clunker.
Good Old MF and LF Lenses are Still Good Lenses... |
---|
But for monochrome, at f/8 and below (for MF) or at f/16 and below (for LF), there is likely to be no discernible difference in performance between a good old lens, and a good, modern lens in any reasonable focal length. |
Large Format is limited by film flatness... |
---|
35mm is limited by film quality, while LF is limited by film flatness and the need to use small apertures |
At the beginning of Hicks and Schultz's Medium and Large Format Handbook, they
make the point that medium format lenses are currently well matched to film resolution.
As the quote above suggests, 35mm users are limited by film resolutions
and film flatness and location issues. Large format users are limited
by film flatness too, but also by the diffraction effects of LF
lenses used at small apertures (f/45..). In medium format, film flatness is still an issue, but
the MF lenses resolutions (at 70-80 lpmm) and color film resolutions are a better match than
either the 35mm or LF cases. If you want an excellent LF lens, consider the 110mm f/5.6 super
symmar XL aspheric and similar lenses of high resolution and quality.
Zeiss Lenses Deliver up to 200 lpmm on fine grain films... |
---|
Fleischer/Mueller mentioned how manufacturers of film say (as one reason why they don't make more high resolution films) they don't believe that lenses are good enough to show what a high resolution film is capable of. Fleischer/Mueller, being a Zeiss man, of course rebukes by pointing to the 1996 Photokina Zeiss exhibition, in which they displayed photos, made from Ektar 25 (capable of 200 lp/mm) negatives, and made using Zeiss lenses, that show lenses in fact can use the film's high resolution. Every last bit of it, in fact. (See posting) |
The worst film in their test resolved 90 lp/mm (Kodak Portra 800). The best,
but no longer available, film they tested resolved 250 lp/mm (Agfaortho 25).
The best film they tested that is still available resolved 180 lp/mm (Tmax
100) (Velvia in second place with 160 lp/mm and Agfa Portrait XPS 160, Kodak
Portra 400 BW and Kodak Portra 160 VC a shared third place, with 150 lp/mm).
(See posting) (Ed. note - see source: March 2003 issue of Zeiss Camera Lens News, nr. 19.) |
In 35mm, the prime normal lenses are capable of delivering 100 lpmm (see
tables). In medium format, the lenses can deliver 70-80 lpmm.
Large format lenses can reach 80 lpmm centrally, though 60 lpmm is more typical, and even
40 lpmm or less is "not unknown". By comparison, most 35mm zooms are lucky to reach 50 lpmm.
Unfortunately, if you start out with a lens capable of delivering 100 lpmm in 35mm, you will
be lucky to realize 50-60 lpmm in the end due to various degrading factors.
For medium format, this falloff goes from about 80 lpmm to 40-50 lpmm. Large format users will
similarly often achieve ~40 lpmm results. The larger area of LF and MF formats provide an offset
for the higher resolutions achieved in 35mm. Some of these limiting factors are highlighted on
our Beating 50 lpmm pages.
The key point here is that medium format is again the happy middle format. The MF lenses
are a good match for available color film resolutions. While 35mm lenses are higher resolution
than most MF or LF lenses, they are limited by the resolution of color films and format size or
enlargeability issues.
Conversely, LF users have even worse problems with film flatness than
MF users, along with the disadvantage of running into diffraction limits at typical slower f/stops
in use. So MF gets the most out of color films, without losing too much resolution to diffraction
(as in LF) or lacking enough film area (as with 35mm) to produce high quality color enlargements
for portraiture and similar work...
From Zeiss Camera Lens News #4 Spring 1998 (see
archives)
Refer to the article for the full explanation of the above ten tips to higher quality...
Before You "Upgrade" Your Old Lenses.... |
---|
Any professional quality 35mm camera made since the 1960s, any professional quality roll-film camera made since the 1950s, and any professional large format camera, ever, should deliver sharpness which cannot be improved upon. Likewise with lenses: most lenses made since the 1970s, and many medium format and large format lenses made since the 1950s, will deliver the quality you need. Unless you want extreme wide angle, ultra fast lenses or zooms, where progress has been significant, there is no need to buy the latest and best. Forego your next equipment "upgrade" which will probably be illusory anyway, and spend the money on materials. |
From Lenses for 35mm (Kodak Workshop Series KW-18 1998, p. 27, Artur Landt):
One caveat: Particularly with short lenses (ed. note, wide angles), use of apertures smaller
than about f/8 reduces image quality, due to refraction* caused by the edges of the diaphragm
blades.
(*Ed. note: I think he means diffraction effects become an issue with smaller physical openings
such as f/11 and f/16 with wide angle lenses, as the absolute size of the aperture hole is
quite small and diffraction is more of a problem. [note: translated from 1993 German original]
This caveat is important, because many moderate priced wide angle lenses of marginal
performance continue to improve as they are stopped down, even to f/11 and f/16 or f/22.
Past f/8 or so, you begin to be strongly limited in resolution by diffraction limits. In
other words, setting your wide angle to f/16 for the most depth of field may not
provide you with the best resolution. This is something you should check out with each
wide angle lens. With better quality wide angles, resolution should be highest at mid-f/stops.)
From Lenses for 35mm (Kodak Workshop Series KW-18 1998, p. 29, Artur Landt):
In actual applications, several factors will reduce theoretical resolving power considerably.
These include low subject contrast, camera shake, subject movement, inaccurate focusing, and
imperfect film flatness. Photographers may encounter mulitiple factors in one image, such as
diffused lighting, slight overexposure, and diffraction from a small lens opening.
From Lenses for 35mm (Kodak Workshop Series KW-18 1998, p. 33, Artur Landt):
Photographers who use color negative films to make prints up to 8x10 inches can safely do without
tests, since the differences in image quality will not be noticeable at these small enlargements.
Those shooting color slide films or producing larger prints (black and white or color) should
definitely evaluate their lenses' image quality. This is even more important for professional
photographers whose pictures must meet the highest requirements.
f/stop | center | radial 10 deg. | radial 20 deg. | tangent 10 deg | tangent 20 deg |
2 | 713 | 702 | 668 | 680 | 590 |
2.8 | 510 | 502 | 480 | 488 | 424 |
4 | 356 | 351 | 334 | 340 | 295 |
5.6 | 255 | 251 | 240 | 244 | 212 |
8 | 178 | 175 | 167 | 170 | 148 |
11 | 130 | 128 | 122 | 124 | 108 |
16 | 89 | 88 | 84 | 85 | 74 |
22 | 65 | 64 | 61 | 62 | 54 |
The above table of ideal (maximum obtainable aerial) resolution versus f/stop is interesting
chiefly for its extension to tangential and radial angles of 10 degrees and 20 degrees
from the center values. At f/2, our ideal lens may be capable of (aerial) center resolution
of 713 lines per mm. But tangential resolution at 20 degree angles falls to 590 lpmm, or about
82% of the center value. Similar losses can be seen for other f/stops in the corners versus
the center resolution values. Now you have an insight into another reason why edge
resolution is less than center values for even ideal lenses.
A few years ago I used a resolution chart to test resolving power of some
lenses that included two in the range of your interest. One lens that
stood out for sharpness and contrast was the Sigma 70-210 f2.8. Optically
it was if anything slightly better that the equivalent Nikon; both were
excellent. The Nikon, however, is much more ruggedly made. I am still
using the Sigma and have made excellent photographs with it. I also have
several excellent Nikon lenses. Because I once purchased two lenses that I
was not satisfied with, I now usually test a lens before buying it. It has
saved me much grief. An aricle on resolution charts will appear in the
December issue of Popular Photography or email me for info.
Dear Robert,
I have discovered your useful and informative web site belatedly, but I
have a number of things to say about it. First of all, let me tell you
that I am the designer of the Sleicher chart that David Jacobsen and Bob
Atkins have refered to. By trade I am a chemical engineer, and I am
retired from the University of Washington, where I taught for over 30
years. Since retirement I have been able to devote time to photography. I
do mostly nature photography - wildlife and scenics. I have had modest
commercial success as a stock photographer and seller of prints. My
credits include the National Geographic, the Sierra Club Wildlife
Calendar, Sunset Magazine, an Audubon calendar, winner of the 1998 Annual
Photography contest in the wildlife division of Nature's Best magazine,
local pulications, and several exhibits.
About the table you attributed to David Swager - if I click on the link
to Swager's resolution page, I get a page from a pamphlet by Brian Geyer
of Really Right Stuff. The numbers in the table come from me. They appear
in the instructions that I send out with my chart, and were used by Brian
with my knowledge and support. I think it would be a good idea if you
noted that this table is strictly empirical and subject to change if more
or better data become available. Even better, you may wish to present my
original table, which has an additional column for Technical Pan.
... I can send you one of my charts. The chart comes with (1)
detailed instructions, (2) a set of targets designed for wide angle
lenses, and (3) a slide of the chart that has targets that are resolved at
100 l/mm, which helps users to evaluate their microscope or other
measuring equipment. I have received many complements on the clarity and
completeness of the instructions.
---------
[Ed. note: the following is quoted from Terry's posting:
From: terry roth [email protected]
Subject: lens test chart
Date: 1998-05-18
I was entering a posting on the Sleicher test chart, but my carrier was
dropped, so am reposting. The title should actually convey the idea that
the chart measures resolution to 160 lines/mm, but it is the resolution of
the entire camera system, including camera, camera support, lens, film,
technique, etc.
There are 84 targets with 14 pairs of grids in 4 colors (including black)
on this 2X3 foot chart. I have attained resolution of 120 l/mm with my
Mamiya 7, on Kodak tech pan developed in TD-3. There is a hint of
resolution at 140 l/mm, but there is some residual astigmatism and I am
not sure the vertical group is resolved (at 100X in a Nikon lab
microscope.) Needless to say, this is very nearly at the limits of even
Tech pan--the grain, nearly imperceptible in normal size enlargements, in
beginning to overwhelm the grid at this magnification. The grid lines
(there are 10 horizontal and 10 vertical lines in each of the 14 groups
(20 to 160 lines/mm), and the lines of the 160 l/mm grid are less than
0.00015 inches apart on the film.
I got the target at a local pro shop, it is also available for $28 from
Mr. Sleicher , a 10-page instruction manual with helpful info on achieving
highest resolution is included.
His address is
Charles Sleicher
5002 Harold PL NE
Seattle Wa 98105.
I am not affiliated in any way, other than being a very satisfied user. He
suggests using a high quality microscope to view the negatives, but I have
found that the enlarger using a good quality enlarging lens, (Nikon,
Componon, or Rodagon) and a good focusing magnifer with the enlarger all
the way at the top of the column gives comparable results, but a little
more difficult to center the various targets at the edges.
[end quote]
Dry Plates vs. Digital Surprises... |
---|
You may think that dry plates also are history, but in my day job (holography) we use dry plates every day. We would _love_ to go digital, but we need about 5000 line-pairs/mm of resolution to match the performance we get from dry plates. Digital detectors are still about two orders of magnitude away from that requirement. ... from posting by Helge Nareid |
From: ted andresen [email protected]
Newsgroups:
rec.photo.equipment.large-format,rec.photo.equipment.medium-format,rec.photo.equ
ipment.35mm
Subject: Here's an easy way to measure resolution.
Date: Mon, 29 Nov 1999
If you'd like an easy way to measure resolution you should check out the
bottom of the website below. It contains a resolution chart and a guide
for doing the calculation. Look for the link "Resolution" at the bottom
of the page.
http://members.aol.com/Tjacmc/
Ted Andresen
St. Petersburg, FL 33703-1721,
From: [email protected] (H.Gunnarsson)
Newsgroups: rec.photo.equipment.misc
Subject: Re: Could you tell me some homepages about lens test ?
Date: Thu, 8 Oct 1998
Check out http://www.photodo.com/lens , choose "products & tests"
--
H�kan Gunnarsson
G�teborg/Gothenburg, Sweden
From: [email protected] (HRphoto)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Lens Sharpness!!! [was Re: Nikkor AF 24-120/3.5 not
professional?]
Date: 23 Apr 1999
>the best lens will never do itself justice if there is an >inferior film in the camera.
But the best lenses will still outperform inferior lenses, even with low
resolution films. See:
http://www.f32.com
under Articles: Manufacture and Performance of Photographic Lenses.
Heinz
[Ed. note: Michael's note is interesting in showing 4750 lpmm resolution
possible with film and microscope objectives!...]
Date: Mon, 10 Jan 2000
From: Michael Abbey [email protected]
To: Robert Monaghan [email protected]
Subject: Re: This 35mm vs 4x5 myth
Hello Robert:
Yes. I was wondering if I would get a rise from someone. I am a
professional
photomicrographer, so I am referring to microscope objectives. With a
Nikon
100X microscope objective of numerical aperture of 1.35 and 405nm light I
can resolve 4750 L/mm of subject onto film. The relationship of numerical
aperture to F stop is based on the assumption of the same refractive index
the microscope objective is immersed in oil with a 1.515 RI.) so true
comparison with an air lens is not really accurate. I find dogmatic
statements of superiority for lenses made in these groups so absurd that I
just have to post. I actually made money on a bet at a local camera store
with someone who was sure that Canon lenses were sharper than Nikon, and
of
course Canon doesn't make microscope lenses. I have Zeiss, Nikon, and
Leitz
Plan Apo objectives that all perform the same, and the format is
irrelevant.
I can shoot on to 35mm or 4X5 with the same resolution. One of my
favorites is a Zeiss 63X na1.4 which has a 2.57mm focal length, a 0.09mm
working distance,and an F-stop (in oil) of 0.357, this will push almost
5000 L/mm.
note my depth of field is 1/10 of a micron.)
For your information the
relationship of F stop to na is : The reciprocal of twice the F stop
equals na, and the reciprocal of twice the na equals the F stop, this
assumes the same RI on media. I also shoot 4X5, 6X7, and 35mm in the
regular terrestrial world with Nikon, Schneider ( I love my 6X Aspheric
loupe), Pentax, Leica, and Rodenstock lenses and I think choosing the best
tool for the job is what counts. Never let equipment limit your
imagination, that is what clients are for!
Regards
Michael
[Ed. note: Mr. Puts is a noted lens testing expert...]
From Leica Mailing List:
Date: Fri, 28 Apr 2000
From: Erwin Puts [email protected]
Subject: [Leica] sharpness & optical quality
> In other words: sharpness is NOT everything > in photography. What about "residual" tonal gradation, > shadow detail, highlights' subtlety, color accuracy when its acutance / > edge resolution goes down somewhat at f/2.8 or f/2 ? > >Right on. When will people realise this? Most modern lenses are 'sharp' >enough. The occasional massive step forward in resolution (35 lux -> 35 lux >asph) excepted. The character resides in the rest of their qualities. Add >bokeh to the above list, for a start. Judging a lens by its sharpness is >like judging a car by the size of its engine.
These comments can be read and heard often. They are not
representative of current thinking about image quality and
demonstrate an embarrassing lack of insight in the true nature of the
topics discussed.
I cannot think of any person or book discussing optical quality in a
serious way who will uphold the notion that "sharpness" has relevance
to image quality. Sharpness is often equated with resolution, which
is not the case. No one has ever been able to define 'sharpness' in a
consistent or measurable way. Fact is that "sharpness" does not
exist. The notion of sharpness impression does exist, but that is a
psychological phenomenon, loosely related to acutance. Resolution
only refers to the ability to distinguish between two adjacent
objects, the smaller the distance between two objects, the higher the
resolution. It has no direct relation to image quality, but PopPhoto
still uses it as a criterion for optical performance. The measurement
of resolution is so dependent on so many uncontrollable parameters,
that no one would propose resolution as a discriminating
characteristic for optical systems. The suggestion that the Summilux
asph has a much higher resolution value than the previous version is
not true. Contrast is higher, but not resolution, which is only
marginally improved due to the higher contrast.
It is remarkable that the sharpness topic is discussed often by
persons who wish to denounce the value of the concept, while most
persons who discuss optical quality do not even think of using this
concept.
Now bokeh, which is just a new word for the older concept of
'rendition of unsharpness areas' is a very imprecise notion and is
based on perception and personal judgment and appreciation. The
sharpness impression is also imprecise and based on perception etc.
So if one argues in favour of bokeh or highlights' subtlety or
residual tonal gradation (whatever that may be) or any other
impressionistic criterion, we are not proposing anything new. I do
not see the added value in replacing the vague and obsolete notion of
sharpness with another concept just as vague and unreliably related
to the real optical quality of a lens.
A lens cannot be characterized by one simple criterion, measurable or
not. And the definition and assessment of the optical quality of a
lens is a major undertaking that defies any attempt to simplification.
Indeed a lens has a character, just like an individual, but we are
all aware that the description of a person's character is a highly
subjective and dangerous activity. So is the description of the
character of a lens.
The recent discussion about the perceived differences between the
Nokton and Summilux 50 clearly demonstrates the pitfalls. The Nokton
lens was not allowed to show its qualities in that approach.
When proposing characteristics for evaluation of a lens, we should be
aware that this area of discourse can be studied from several
perspectives and levels. Any photographer can choose whatever lens
(s)he wants, based on whatever arguments. A discussion of these
choices may be enlightening as we get a glimpse into a
photographers' personal decision chain. This is however no substitute
for a comprehensive assessment of the optical quality of a lens,
based on current thinking and theory of image quality.
Erwin
Date: Tue, 22 Aug 2000
From: "Christopher M. Perez" [email protected]
Newsgroups: rec.photo.equipment.medium-format,rec.photo.equipment.large-format
Subject: Re: Final Resolution - the real answer?
pico wrote:
> Perhaps a well informed expert can help with this question. It springs from > a post I saw elsewhere. > ... > The situation: The lens can resolve max 100 lpmm (measured how, as > an aerial image?) and the film can resolve the same max 100 lpmm. > > - what is the resolution potential of the resulting image on that film? > > lens == 100 lpmm > film == 100 lpmm > result == ?
100 l/mm. Why? Film doesn't behave like some of us have been lead to
believe. The formula that's been floating around (something like
1/lens_resolution x 1/film_resolution = 1/'real'_resolution) does not
accurately describe the situation.
Hopefully there's a physicist reading these threads who can properly
define/describe reality.
-x
Date: Tue, 22 Aug 2000
From: [email protected] (Richard Knoppow)
Newsgroups: rec.photo.equipment.medium-format,rec.photo.equipment.large-format
Subject: Re: Final Resolution - the real answer?
"pico" [email protected] wrote:
>Resolution - the last gasp? > >Perhaps a well informed expert can help with this question. It springs from >a post I saw elsewhere. > >This exercise has nothing to do with esthetics, personal preferences, >real-world consumer lenses and so forth - it is a question of the >scientific test of "resolution" under the following idealized conditions. > >The film is monochromatic, the light source is matched to the film's >singular >color sensitivity and the photographed object is an idealized >black-and-white >resolution test target and we assume it is focused properly to the film. > >The situation: The lens can resolve max 100 lpmm (measured how, as >an aerial image?) and the film can resolve the same max 100 lpmm. > >- what is the resolution potential of the resulting image on that film? > >lens == 100 lpmm >film == 100 lpmm >result == ? > >My intuition tell me that the max resolution possible is significantly >less than 100 lpmm. > >=== >and extending the question, each factor contributing to the final outcome >must (I assume) obey the same mathematics. If a print is made, then the >resolution of the enlarging lens (and paper) is a factor. If, for example, >the enlarging lens is capable of only as much as the outcome of the >elements above, then the math is carried on serially - correct? What is the >end-result of this 100 lpmm lens when the best component after the lens >is no better than 100 lpmm? > >lens == 100 lpmm >film == 100 lpmm >enlarging lens == 100 lpmm >result == ?
If I understand what you are asking correctly it is essentially how
to calculate the final resolution of a system when the resolution of
the parts are known.
The problem is that there is no simple solution because the
resolution limit as ususally stated it only a point on a curve who's
shape can vary tremendously. The combined resolution must be
calculated as a convolution of the curves of the various components.
There are some rules of thumb but they are based on assumptions
which may not be true in practical cases.
For lens + film the rule of thumb is 1/T = 1/L + 1/F Where T =
Total resolution, L = Lens resolution, F = Film resolution.
Another rule of thumb is to use the square root of the sum of the
squares. Neither of these is really accurate, or even very meaningful.
If you are familiar with sound equipment it is very much like asking
for a single number to characterize the frequency responce of a chain
of components. Even defining the cut off point at some fixed
attenuation can't be done in a general sense without a knowledge of
the slopes of the curves.
Using monochromatic light is not really a simplifying factor here.
This is probably not a very satisfactory answer, especially since
resolution numbers get bandied about all the time by photographers.
Have a look at the Schneider of America web page for some MTF curves
on their lenses. http://www.schneideroptics.com
The curve for film is generally simpler but is affected by edge
effects and scattering, see Kodak's film data sheets for examples. You
could probably combine curves like this graphically. It would be an
interesting excercize.
As far as enlarging resolution, it depends on the contrast of the
film at the stated resolution limit. If you put a resolution chart in
the film gate the resolution of the image will be whatever the lens
will do. If the film has the same contrast at its limit as the chart
(it won't) we will get the same value.
Generally, paper has high enough resolution not to be a factor but,
again, the combination of paper and lens will be similar to lens and
paper.
It should also be noted that the "resolution limit" is often defined
as the point where the image contrast falls below a defined criterion,
it is not an absolute limit. Just as the frequency response of an
audio component may continue out beyond its "cut-off" the resolution
of a lens may continue beyond its cut-off, but at a very low level.
In the case of lenses this can be an important factor. A lens can be
designed for maximum resolution regardless of contrast, or for best
contrast at mid resolution frequencies (using the language of the MTF
graph) but poorer performance at the high resolution end. i.e., the
contrast is high throughout a large portion of the MTF curve but falls
off very fast at the high end. The high resolution lens may have much
lower mid range contrast but maintains some contrast to higher values.
A somewhat similar effect can be gotten with film. Edge effects,
i.e., variations in development at the the border between a high and
low density area, are affected strongly by developers. By exagerating
this effect the "acutance" of the film is increased by increasing the
contrast locally. The same effect, however, tends to suppress fine
detail, i.e., lower resolution. In other words it changes the shape of
the MTF curve.
No rule of thumb can account for combining such values.
Add to this the fact that for visual presentation, the eye
interprets edge contrast as sharpness. So a relatively low resolution
but high contrast image will look sharper than one of low contrast
that actually has more detail in it.
This is _not_ a trivial problem.
---
Richard Knoppow
Los Angeles, Ca.
[email protected]
Date: Sat, 9 Sep 2000
From: A_M [email protected]
Newsgroups: rec.photo.technique.nature,rec.photo.digital
Subject: Re: lens, film resolution, MTF limits
Roger N Clark
[email protected] says...
> Discussion in the photo groups over the last few months > raised the question of what MTF level gives what cycles/mm > For example, the MTF limit at f/11 is 182 line pairs/mm.
Quoting from "Photographic Lenses Tutorial" at
http://www.graflex.org/lenses/photographic-lenses-tutorial.html
"...If we again assume 555 nm, this comes out to 1482/N lpmm, which is in
close agreement with the widely used rule of thumb that the resolution is
diffraction limited to 1500/N lpmm..."
In other words at F11 the resolution is diffraction limited to 1500/11
lp/mm which is 136 lp/mm, not 182 lp/mm. At F11, 555 nm and 136 lp/mm
the MTF (i.e. the contrast) is zero. The usable range stops somewhere
before that. A few percent MTF aren't enough.
However this is just the diffraction limit. No lens is perfect and the
lens MTF might reach zero at a spatial frequency lower than the
diffraction limit.
> For an > unphased system of image detail (what we have with photos), > 2x Nyquist sampling is necessary resulting in the need for > 7020 DPI sampling for 50% MTF response and over 18,000 dpi at > the MTF limit (but no film has this resolution).
Once again, the usable range ends before the diffraction limit and in any
case well before the MTF reaches 0%. I have a source quoting an 80 lp/mm
limit, another quoting 30 and 50 lp/mm limits (for APS and 35 mm
respectively).
--
Alfred Molon
Email address is alfred_molon at csi.com
Date: Fri, 08 Sep 2000
From: Chris Kelly [email protected]
Newsgroups: rec.photo.technique.nature,rec.photo.digital
Subject: Re: lens, film resolution, MTF limits
A_M,
There seems to be a lot of information out there. There are some
people out there, however, who take "actually looking themselves"
very seriously. Bryan over at RRS is one. He used to have be in
the habit of shooting from heavy tripod, with mirror locked up,
with cable release, on Velvia, and looking at the actual results with
a microscope. The Nikkor 85mm 1.4, for instance, comes in at
around 100 lpm. Some are even sharper. I have shot some VERY
sharp lenses with all the correct technique on several different film,
ProviaF, Astia, and Velvia and there is a definite difference.
Intererestingly enough, I have shot on some lenses which do not
have excellent MTF ratings, with proper technique, and the film
still makes a difference. The resolution of Velvia still effects the
outcome.
Oh, on your last statment. Practical experience of MANY
professionals in the field, as well as testers, show the 50lp/mm
to be totally false. Someone is making a mistake in the math.
I have a feeling that your sources are in the digital community and
have an interest in making the film look worse much worse than
it is.
Chris
...
Date: 28 Aug 2000
From: Jay Wenner [email protected]
Newsgroups: rec.photo.equipment.medium-format,rec.photo.equipment.large-format
Subject: Re: Final Resolution - the real answer?
In rec.photo.equipment.large-format Nicholas O. Lindan
[email protected]
wrote:
: This isn't my area, but I'll take a qualitative stab at it. : In general these types of functions add as a root mean square: : MTF system = 1 / sqrt (1/MTF^2 + 1/mtf^2) : So if you have 2 100 lpm mtf's then the system has an mtf of 70.7 : If you have 3 then the result is 57.7.
I've enjoyed reading this thread because it's like having a bunch of
experimentalists argue with the theory guys. The original question was
in simplified so a theory guy could answer it, but all the
experimentalists
got in and said it couldn't be done. Perfect.
It seems to me the above answer is right. If the noises introduced by
each component are independent, then the final noise should be an RMS.
Jay Wenner
Date: 05 Sep 2000
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Final Resolution - the real answer?
Hi
The performance of a lens will vary with distance. You will find that a
lens that has a range of 5 ft to infinity will max out at about 20 ft at
about 500 l/mm center resolution in daylight and be down to about 350 l/mm
at the 5ft marker and the infinity marker. You can actually make a graph
of this and most Professional Quality optic programs will give this graph
in the outcomes section of the printout.
Larry
Date: 05 Sep 2000
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Final Resolution - the real answer?
Hi
I have a resolution chart done by Kodak the might help you with the
answer.
Just about all name brand lenses will do 500 l/mm at about f4. A perfect
lens in daylight will do the following
f1=2000 l/mm f2=1000 l/mm f4=500 l/mm f8=250 l/mm f16=150 l/mm
Now most name brand lenses are not perfect. So a typical name brand lens
would have these values.
f1=100 l/mm f2=300 l/mm f4=500 l/mm f8=250 l/mm f16=125 l/mm
Email me and I will send you a chart/graph done by Kodak on typical lens
resolutions.
Larry
Date: 05 Sep 2000
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Final Resolution - the real answer?
Hi
Yes it is areial resolution in daylight. Resolution will vary with the
wavelength of light so pictures taken at noon at wide settings (bigger
then f4) have generally higher resolution then pictures taken at dusk at
the same fstops.
Again I have a graph made by Kodak that shows how this works. Email me
and I will send it to you.
Larry
Date: Wed, 6 Sep 2000
From: [email protected]
To: [email protected]
Subject: Re: Final Resolution - the real answer?
Hi
You will see that the graph already has a doted line through it with one
end connected to 550 wavelength (eye wavelength) and the other end
connected to f-2. The read out is 750l/mm. If you use the 500 wavelenght
number (summer noon daylight wavelength) then it reads out 1000l/mm at f2.
Larry
Date: Thu, 07 Sep 2000
From: Roger N Clark [email protected]
Newsgroups: rec.photo.technique.nature,rec.photo.digital
Subject: lens, film resolution, MTF limits
I've made a major update to my image/scan detail page at:
http://www.users.uswest.net/~rnclark/scandetail.htm
In particular, I've found the modulation transfer function
for a diffraction limited optical system and added discussion
concerning image detail. The results show that 35mm film
is the limit, not lens resolution.
Discussion in the photo groups over the last few months
raised the question of what MTF level gives what cycles/mm
For example, the MTF limit at f/11 is 182 line pairs/mm. For an
unphased system of image detail (what we have with photos),
2x Nyquist sampling is necessary resulting in the need for
7020 DPI sampling for 50% MTF response and over 18,000 dpi at
the MTF limit (but no film has this resolution). Other
MTF levels and f/ratios are on my web page.
The MTF model results are in complete agreement with observed
scan detail comparisons between 4x5 and 35mm film formats and
explain why the larger film format shows more detail.
I've also reorganized the digital camera detail discussion,
and made direct comparison of 35mm film and digital camera images
with the same test target (no change in results from previous
page, but I believe a better presentation). The results show
digital image sensors have a long way to go to match film image detail.
If reports of new finer grained films become reality, film will
pull further ahead, with an f/11 35mm system capable of delivering
more than 600 megapixels if film grain/resolution were not the limit!
Comments welcome.
Roger Clark
http://www.users.uswest.net/~rnclark (Home Page Photography)
Date: Wed, 13 Sep 2000
From: Paul Butzi [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: New Chip "better than film"?
(excellent reasoning snipped)
>So my conclusion is that it is possible to make images with >large-format cameras that no digital camera can imitate, unless the >digital camera has a sensor (be it CCD or CMOS) that's as large as the >film in the large-format camera. That may happen eventually, but I >don't think it will be any time real soon.
So far sensors are small because smaller sensors mean cheaper
sensors, with higher yield in the manufacturing process.
But now, the 1.2cm sensors have pretty much reached the limit of
how many pixels can be profitably crammed into the sensor. That
limit is imposed by the limit of resolution of the lens. At a certain
point it becomes cheaper to make the sensor larger than to make
the lens better.
Consider, for instance, the 'last years model' Olympus C-2000
I own. The sensor size is approximately 1cmx1.3cm. The
maximum image resolution I can get is 1600x1200, which means
that on the sensor, the pixel packing density is about 120
pixels/mm. If we use the Nyquist criterion, that means we can
resolve something like 60 line pairs/mm. Let's be charitable
and assume that the zoom lens on this camera can actually
deliver this to the sensor.
Now we wish to build NEXT years camera - it will have,
instead of a 2.1 megapixel sensor, a 8 megapixel sensor.
Now, instead of 1600x1200, we get an image of 3200x2400.
We need to increase the pixel packing density by a factor
of two. If we keep the 1cmx1.3cm format, the linear pixel
density will be 240 pixels/mm, and all of a sudden we're
asking the lens to deliver 120 line pairs/mm. Uh, oh.
The alternative is to increase the size of the sensor.
My conclusion is that sensors will get larger until resolution
demands are met. For very large resolution demands,
the image plane will probably always be scanned with
a smaller sensor (the way flatbed scanner cameras
currently do).
-Paul
Newly updated and moved web site at:
http://www.butzi.net
Date: Tue, 19 Sep 2000
From: "Ray Smith" [email protected]
Newsgroups: uk.rec.photo.misc
Subject: Re: Digital Equivalent resolution of film
Hi Folks,
I've spent some time delving around on the net, and found a very useful
guy called Harry Adam who is a research scientist at Kodak. Here are some
figures he's given:
Kodak have a range of defined targets for defining image quality. The
Kodak 6 megapixel cameras only *just* produce a 6 x 4 print image that
meets their defined standard for quality. A quality exposed slide (he
quoted for ISO 25, so we're talking about the good end here!) is
considered to still meet the quality target when producing a 30 inch by 20
inch print, i.e some 25 times greater! So Kodak consider that a 150 mega
pixel camera would be needed to reproduce an equivalent quality image to
that of a slow slide film.
Now here comes the good part! A slide film will resolve up to 10000 levels
of luminance! So in round terms, we would need a fourteen bits per channel
scan to define the slide. To put that another way, we need effectively six
bytes per pixel.
The answer to my question is therefore around 900Mbytes of data on a frame
of ISO 25 slide film.
For a typical 100ASA negative film, the figure drops to around 150Mbytes,
due to the combined loss of resolution and colour depth.
Cheers
Ray
[Ed. note: Special thanks to Dr. John Owlett for sharing these
points!...]
Date: Fri, 13 Oct 2000
From: [email protected]
To: Robert Monaghan [email protected]
Subject: Digital Photographs, Megapixels, Infinitives, and Other Rambling
Hi Bob,
Rick Housh and I continued the discussion about the megapixel equivalent
of 35mm, for which you provided me with some pointers, on NikonDigital
and offline. Here are a couple of extracts from my latest note to him.
[----- snip -----]
you wrote
> On that subject, the D1 or one of the Kodak or Fuji Nikon-bodied SLR > models was very tempting but, aside from cost and mostly because of > your well-taken comments about the sharpness of the current "low" > megapixel crop, I decided not to expend a shameful amount of money > on a top-of-the line model yet. That said, I also decided that if > lens mounting compatibility was not an issue I shouldn't restrict > myself to Nikon. After lots of research I bought a Kodak DC290, > which is a glorified (although quite so) 2.2 megapixel point and shoot.
My initial estimate of 6 megapixels was based upon an information content
of 40 lp/mm. This came from a Leitz white paper I read once, which said
that perceived sharpness was primarily influenced by resolution up to
this limit. This limit, together with the eye's resolution of 5 lp/mm at
a distance of one foot, coincides with the 8x rule of thumb I learnt in
the early 70s:
If you use prime lenses, a slow film, and a tripod, you can enlarge your
negatives up to 8x and get a sharp print.
In the years since, this rule has remained remarkably stable. In an
informal test in "Practical Photography" last year, the technical editor
concluded that up to 8x10, he could hardly tell 35mm from medium format;
from there up to 12x16 he could tell 35mm from medium format, but could
hardly tell 6x6 from 6x7; at 16x20 he could tell 6x6 from 6x7.
Now, this rule argues against the softlensophobia of the members
(including me!) of the NikonMF List. It doesn't much matter whose prime
lenses, whose slow film, or whose tripod you use. And if you want to
enlarge by more than 8x,
o the best initial improvement comes from a better tripod, o the greatest improvement comes from a better film, and o better lenses only help when you've got the other factors right.
In his medium format megasite, Bob Monaghan has an article on "Beating
the 50 lines per mm Resolution Limit" and how difficult it is to do
... on anything except photographs of test charts. 50 lp/mm would allow
for 10x enlargement. The first promise of 10x enlargement I know of came
in a Meyer Optik advertisement of the late 20s -- 70 years later it is
still state of the art!
Well almost. There are some films that allow higher resolutions on
ordinary subjects ... provided you have the tripod, the technique, and
the lenses to take advantage of them. They are the Kodachromes (which
were way ahead of their time) at 60 lp/mm, Velvia at 80 lp/mm, and
Technical Pan at 100 lp/mm. You may not be able to see the extra
resolution on an 8x10 held 12 inches from your eyes, but it's there
and you can see it in larger enlargements and at closer distances.
All this means that 6 is not the only perfect number:
o to match normal amateur use of 35mm, you need 6 megapixels o to match critical use of Kodachrome, you need 12 megapixels o to match critical use of Velvia, you need 22 megapixels o to match critical use of Technical Pan, you need 35 megapixels
None of this shakes my belief that, once a 6 megapixel SLR becomes
available for a thousand dollars, the quality and price will be good
enough for advanced amateurs to change to digital in droves.
Translating this into 35mm film scanner terms:
o to match normal amateur use of 35mm, you need 2000 DPI o to match critical use of Kodachrome, you need 3000 DPI o to match critical use of Velvia, you need 4000 DPI o to match critical use of Technical Pan, you need 5000 DPI
This implies that scanning 35mm film can, at present, give better
results than digital cameras, and that the Polaroid scanner that
Mackie uses and the rumoured new Nikon LS3000 are round about all
you need for colour work which is the equal of a wet darkroom.
All this so far addresses "sharpness" by which I mean "accurate
representation of fine detail" and which is affected enormously by
both resolution and contrast. It doesn't address the question of
accurate colour.
Some people have argued that when you take colour into account you need
four times as many megapixels, to take account of the RGBG filtering on
one-shot digital backs. I think that's unlikely. There's no theoretical
reason that I know of why sharpness and accurate colour should be linked.
And none of the practical results I've been able to track down show
anywhere near as many megapixels being needed.
Depth is also a different matter. Computer geeks like me are so wedded
to the 8-bit byte that is natural for us to use three bytes (one red, one
green, one blue) for 16 million "true colours". But it seems that the
eye has a finer perception than that, and that 12-bit depth is needed to
capture all nuances.
There. I have finally written all these thoughts down. I find it
fascinating and I hope you won't mind my using you as a sounding board.
I'm not sure whether it's worth posting on NikonDigital though. Is this
stuff just too specialized?
[----- snip -----]
> One thing I have noticed is that the terms "sharpness" and "resolution" > are not synonymous in the digital photography world. In fact, they are > somewhat contradictory, as "digital sharpness" specifically refers to > the contrast between adjacent edges in the image scene, and other > factors affecting the somewhat subjective ability to render objects in > the scene more distinct from each other. In some cases this sense of > acceptable sharpness causes artifacts in the image which are considered > acceptable in the goal of achieving maximum "sharpness". So if you're > talking to a dedicated digital photographer about resolution he will > understand it in the traditional sense, but sharpness to him is a > different animal.
Yes, I've noticed that too. To a still photographer, it's always
possible to get more detail. When technique, tripods, films, and lenses
run out of steam, you can always move to a larger format. So to us,
picture A is sharper than picture B if the fine detail in B is clearer
in A, and there is fine detail in A which was not visible at all in B.
There is no notion of absolute sharpness in a landscape.
A cartoon, however, can be absolutely sharp. You can have a picture
which contains ALL the information in the cartoon. Changing to a 4x5
camera cannot increase the detail you gather about the cartoon. It can
only gather more detail about the piece of paper on which the cartoon is
drawn.
The first to make this distinction, I think, were the early television
engineers, who had serious limits on the amount of detail they could
convey, but knew that their viewers wanted a sharp picture. For good
or ill they developed the MTF.
In my megapixel essay above, I therefore defined sharpness as "accurate
representation of fine detail", which seems closer to what still
photographers mean by the word.
A phrase that should be consigned to the dustbin is "tack sharp". Its
meaning varies from person to person, and it is hopelessly over-used.
Tacks should be restricted to use as toothpicks by grasshoppers sitting
on railroad tracks.
[----- snip -----]
Later,
John
Dr John Owlett
Senior Internet Security Consultant, IBM Global Services
[email protected] http://www.ibm.com/security/services/
Manis K. Banerjee wrote:
> What exactly is "contrast" when referred to in respect of a lens? In > case of a film the word "contrast" perhaps relates to its ability to > render different colour shades accurately without any merging of the > colours. The same may be true in case of B&W where the shade and the > highlight are clearly distinuished at the borderline. Or am I wrong in > this? But a lens, after all, is a piece of glass. It passes through the > light at different intensities. depending on the subject, to form the > right image at the film plane. So if each of these light elements is > accurately transferred to form the right image would it not mean that > the lens is "sharp"? What exactly is the difference between sharpness > and contrast?
It's a good question, Manis. By which I mean, of course, that the answer
is a long one. :-)
In fact it's part of Owl's Very Long Answer to the question of how many
megapixels there are in a frame of 35mm: when you compare silver-halide
photography with digital photography, you have to think again about what
sharpness is.
But let's start with contrast. You've put it rather well in saying that
it's "where the shade and the highlight are clearly distinguished at the
borderline." A clearly defined line is sharp; a fuzzy line is not sharp.
As you say, contrast can be reduced if the film cannot render the line
clearly. It can also be reduced if the lens flares.
A lens does more than pass light through to the film. It absorbs a tiny
amount; and it reflects quite a bit. Consider a single-element lens with
two glass-to-air surfaces, the front and the back:
o Some of the light striking the front is reflected, and disappears
in the surrounding air.
o Most of the light striking the front is transmitted through both
front and back to reach the film in the right place.
o Some of the light transmitted through the front is reflected by the
back, goes out through the front again, and disappears in the
surrounding air.
o Some, only a very little, of the light transmitted through the
front is reflected by the back, then reflected again by the front,
and then transmitted through the back on a second attempt, to reach
the film ... in quite the wrong place.
(Would one of you at the back please give Piglet a poke? He's beginning
to snore.)
Flare can give rise to UFOs -- Unwanted Flaring Objects, which appear as
brightly coloured polygons on the picture -- but most flare is scattered.
It shows up as a reduction in contrast.
With a single element lens, things aren't too bad. But if you have a two
element lens, things are six times as bad:
o light can be reflected at the back of the front element and then
the front of the front element;
o and at the front of the back element, and the back of the front
element;
o and at the front of the back element, and at the front of the front
element;
o and at the back of the back element, then at any one of the three
surfaces it meets on its way out.
("Ouch," said Pooh. And even Kanga looked glassy-eyed.)
The cure for this lies in the multiple coatings that are now normal on
all lenses. They control the reflections, improve the contrast, and make
multi-element lenses possible.
"Has he finished now?" Eeyore asked Rabbit, almost hopefully, thinking
that, as Owl's lectures went, that one went quite quickly.
"No, that was only contrast," replied Rabbit. "He promised -- or
threatened, I'm not sure which -- to talk about sharpness as well."
Owl cleared his throat and continued:
To a silver-halide photographer, contrast is an important part of what
makes a picture sharp; and so is resolution. Picture A is seen as being
sharper than picture B if
(a) the detail in picture B is present in picture A, and the contrast
of that detail is higher, and
(b) there is detail in picture A that is not visible in picture B.
In this view, sharpness is the ability to show fine detail clearly, and
is a combination of resolution and sharpness.
When the TV engineers came on the scene, they realized that there was not
much they could do about resolution, but that it was possible to give
their customers a sharper picture by managing the contrast. For good or
ill, they gave us the modulation transfer function -- the MTF -- which
gives a measure of overall sharpness.
To many TV engineers, and to many digital photography enthusiasts, if the
detail in picture B is present in picture C, and the contrast of that
detail is higher -- but there is no new detail -- then picture C is not
only sharper than picture B, but it is sharper than picture A as well.
Aaargh!
To me, as a completely impartial dinosaur (I agree with those
palaeontologists who believe that the dinosaurs did not die out, but
evolved into owls), that seems like cheating. You can have a digital
photograph which appears supremely sharp because all the detail that is
there is high contrast, and any detail which cannot be high contrast is
not there.
Don't get me wrong. If you are taking pictures for a client, and
pictures like that are what meets the client's need, then take pictures
like that. Equally, digital image enhancement techniques -- which can
never add information missing from the original picture -- can make an
indecipherable picture clear, and that can be very useful.
Though I haven't yet worked out why this process is known as unsharp
masking.
"Has he finished now?" asked Eeyore again.
"I think so," said Kanga.
"Then we must be thankful for small mercies," said Eeyore. "We must be
thankful that he didn't give us this part of his Very Long Answer last
time. And that he hasn't given us more of it to-day."
"Well just one final point on sharpness," said Owl, "and that is that we
should outlaw the phrase 'tack sharp'. It means something different to
everyone who uses it, and is in danger of becoming a clich�. Tacks
should go back to their niche of being used by grasshoppers to pick their
teeth while sitting on railroad tracks."
-------------------------------------------------------------------------
"... and so we see that all nouns beginning "GE" or "GI" which came into
English from French -- such as Gillette, George, and Gitzo -- should be
pronounced with a "J" sound," concluded Owl.
It was the Feast of Stephen (Boxing Day, if you prefer) and so the
animals were not outside in the Hundred Acre Wood. The snow was not
deep and crisp and even -- even in stories for children it is raining
at Christmastime in the South of England -- and Christopher Robin had
brought the animals indoors into his playroom where he could chair
their discussion group in the dry.
"Mum, is Owl all-knowing?" asked Roo quietly.
"Of course not, Dear," Kanga replied. "Not even he thinks that ... well,
probably not. But he never allows a lack of knowledge to get in the way
of holding an opinion."
"Christopher Robin," said Owl. "I never did get a chance to finish my
lecture on megapixels. Do you think everyone would like to hear it now?"
"Silly old Owl," said Christopher Robin. "I don't know whether everyone
wants to hear it, but I'm sure they will let you give it. What exactly
do you want to talk about?"
"Well, you see, Kanga and Rabbit explained how many pixels there are in
in a frame of 35mm film: as Vlad the Anti-Virus said, if your photography
is recording N line-pairs per millimetre, then the whole frame contains
3456 times N squared pixels...."
"I knew it!" snorted Eeyore. "Owl is descending into algebra. We'll
have inverse trigonometric functions before long, you mark my words."
"... What I want to look at is how many megapixels you need for a sharp
print."
"Not so fast, Owl," said Piglet. "Vlad said more than that. He said
that you had to triple or quadripple ... quadropull ... quad-something
that number to take colour into account."
"Oh, good grief!" said Eeyore. "Now we're for it. Asking Owl to slow
down is like ... is like ... well, all I can say is that a journey of
a thousand miles begins with an hour's lecture from Owl on map-reading."
"That's a good point, Piglet," said Owl:
************************************************************************* * * EXTREME ESTIMATES * * As Vlad pointed out, pixels in a digital camera are grouped in fours: * one for red, two for green, one for blue. And he suggested that you'd * therefore need four times as many pixels to capture colour as well as * sharpness. But I'm not convinced by that argument. * * I don't think the two issues are independent. If you have four times * as many pixels, you are not just capturing colour, you are increasing * the sharpness at the same time. And it's not at all clear that you * need as many pixels for accurate colour as you do for sharpness. * * Vlad's suggestion of 100 lp/mm is high, but should be achievable, even * on ordinary contrast subjects, if you use perfectionist technique and * Technical Pan film. The resulting 35 megapixels won't be achieved by * digital cameras for a few years. The best sensors on sale are still * a little less than half that. * * * William Oliver, a pathologist quoted by Rabbit earlier, demands 125 * lp/mm, and thus 54 megapixels. I can see why: in a pathology lab he * can control the lighting and increase the contrast; moreover, he needs * to be right first time -- after he's taken his pictures and the subject * has been cremated, he can't take any more. * * Even higher resolutions are possible. Stewart Bell, who tests lenses * for "Amateur Photographer" says that, on Technical Pan, "both Canon * and Nikon 50mm f/1.4 objectives turn in resolution figures in the * neighbourhood of 400 lines per millimetre" (25 Dec 1999 issue, page * 65). That's about half a gigapixel! But probably only at f/4 on * well-lit high-contrast test charts. * * Conversely, Rob the Remarkable quoted an estimate that 1200 pixels per * inch -- 24 lp/mm or 1.9 megapixels in all -- was enough to match 35mm * photography. That's a fairly gloomy view of amateur technique ... or * perhaps it's a realistic view of autofocus zoom compact capability. * * It seems to be seriously hard work to exceed 50 lp/mm -- * * http://www.smu.edu/~rmonagha/mf/limits.html * * -- and I suspect that ordinary careful amateur photography runs at * about 40 lp/mm: 6 megapixels in all. * *************************************************************************
"Being gloomy about amateur technique seems like a good idea to me," said
Eeyore. He sounded almost cheerful at the prospect. "No tripod and
supermarket film should keep the resolution down, even if they choose
a processor carefully. Which they won't."
"Come on; come on," bounced Tigger, "we've heard some of that before.
What's this about the megapixels you need for a sharp print?"
"Well the thing is," said Owl, "the number of pixels you can see in a
print depends on how far it is normal for the print to be from your eyes.
So, before we can count the pixels, we have to think about what 'normal'
is."
"We also have to think about what the meaning of 'is' is," he added.
"No, we don't," said Kanga. "We don't have to do that at all. You're
teasing the American gray squirrels again, and it Does Not Help."
"OK," said Owl, "but we do need to think about normality, and about the
difference between perspective and angle of view."
************************************************************************* * * PERSPECTIVE AND ANGLE OF VIEW * * This Autumn, I visited a castle in Scotland which has a ruined arch at * the gate. As I approached, the castle itself was picturesquely framed * in the ruined arch, so I decided to take a photograph. I knew that I * was probably the hundredth person that day who had decided to take that * picture, but I consoled myself that I was probably the first that day * to use a tripod. * * My F3/T had a 'normal' -- 50mm -- lens on it, but this had too * narrow an angle of view: I couldn't see the arch in the viewfinder. * "Shoeleather zoom" I thought (I'm not sure why: shoes won't fit on * my claws) and I fluttered backwards a few metres. * * That didn't work either. The arch was nicely around the edge of the * picture, but it no longer framed the castle: the perspective was * different, and the top of the castle was blocked off by the top of the * arch. It wasn't perspective I needed to change, but angle of view. * * So I moved that tripod back to its first position and changed the lens * for one set at 35mm focal length; and got the picture I wanted. * * Perspective is determined ONLY by how far you are from the subject. * Angle of View is determined by the focal length of the lens. Often, * the difference doesn't matter much, but in this case it did. * *************************************************************************
-------------------------------------------------------------------------
Eeyore was having what was, for him, a wonderful time barracking Owl;
so he muttered sotto voce (which means it's supposed to sound quiet and
everyone is supposed to hear it), "And now Owl -- Owl of all creatures --
is going to tell us what normal is."
So he did:
************************************************************************* * * WHAT IS NORMAL? * * Last century, in the 1800s, photographers began to think about what was * a normal angle of view. After all, if you stand outside and look at * the landscape, you can see things far to the left, to the right, and to * the sky without moving your eyeball. But if you take a picture with an * ultra-wide lens that captures all you can see ... the picture looks * flat, and not at all how you remember it. * * Unless, that is, you make a huge print and stand very close to it. * * So what they did was to make landscape prints of different sizes * to see how far people held them away from their eyes. And, grossly * oversimplifying what they found, the distance from eye to print was * about the same as the length of the diagonal of the print. * * So, using similar triangles and a little high-school optics (rays * through the optical centre of a simple lens are not diverted), they * inferred that a 'normal' lens should have a focal length equal to the * length of the diagonal of the negative. * * In the case of 35mm film, that's a normal lens of 43mm focal length. * * In the 1800s, as now, you have only to say something clearly for people * to know why they disagree with you. So there were some who said this * distance -- 320mm (12.8") for quarto prints of 250mm x 200mm (10" x 8") * -- was too great, and some who said it was too little. * * Those who said it was too little preferred the angle of view of a * quarto print at a normal READING distance of 400mm (16"). For them, * a 'normal' lens would have a focal length of 54mm. * * Those who said it was too great preferred the angle of view of a quarto * print at a normal SCRUTINIZING distance of 250mm (10"). For them, a * 'normal' lens would have a focal length of 34mm. * * And, of course, the debate over whether a normal lens should be 50mm or * 35mm has not been resolved yet, 150 years later. * * That is normal angle of view, not normal perspective. There isn't * really a normal perspective of a mountain: when you're up close, the * foothills seem much bigger ... but that's not seen as being wrong. * * The human head, on the other hand, DOES have a normal perspective. * It's about 250mm (10") high so, if you want to fill a 35mm frame with * it using a 43mm lens, you have to be 300mm (12") away. That's awfully * close, and the nose will seem much too large. The angle of view may be * 'normal', but the perspective is quite wrong. * * When two human animals stand and talk to each other, they stand with * their heads about 900mm (36") apart. That is 'normal' perspective, and * is the equivalent of using a 135mm lens for a head shot. * * In practice, many human photographers prefer 105mm lenses or 85mm * lenses as portrait lenses. That's the same perspective, but different * angle of view: a head-and-shoulders shot or a head-and-chest shot * respectively. * *************************************************************************
"I told you!" said Eeyore. "I told you! I warned you that there would
be some inverse trigonometric functions, and there were some hidden away
in there."
Pooh shuffled quietly across to Christopher Robin. "I'm not sure I've
got all this," he said. "Is Owl saying that fallen arches make it
difficult to climb a Perspex mountain?"
"Silly old Bear," said Christopher Robin gently. "You don't need to
worry. All that Owl needs from that last section is that different
people hold pictures at different distances from their eyes."
"Oh," said Pooh. "Why didn't he say so then?"
Owl cleared his throat and continued:
************************************************************************* * * HOW MANY MEGAPIXELS CAN YOU SEE IN A PRINT? * * If you hold a 250mm x 200 mm (10" x 8") print at a 'normal' distance of * 320mm (12.8") from your eyes -- and you have average vision -- then you * can resolve 5 lp/mm. * * But, because people do not all agree that 320mm is normal, they do not * all quote the same number of line pairs per millimetre: * * o at 400mm (16") you can resolve 4 lp/mm * o at 320mm (12.8") you can resolve 5 lp/mm * o at 250mm (10") you can resolve 6 lp/mm * o at 200mm (8") you can resolve 8 lp/mm * * Usually it is those who are most rigorous about print quality who use * the higher requirements. For example, in the Really Right Stuff * catalogue, Bryan Geyer uses 6 lp/mm. And Leitz uses 8 lp/mm. * * Now that we have these numbers we can convert them to megapixels quite * quickly. At 4 lp/mm, each millimetre needs 8 pixels; 250mm needs 2000 * pixels; 200mm needs 1600 pixels; so a 250mm x 200mm (10" x 8") print * contains 3.2 megapixels. * * So, to the nearest megapixel, in a 250mm x 200mm (10" x 8") print: * * o at 400mm (16") you can see 3 megapixels * o at 320mm (12.8") you can see 5 megapixels * o at 250mm (10") you can see 7 megapixels * o at 200mm (8") you can see 13 megapixels * * It's particularly interesting that ordinary careful amateur photography * -- at 6 megapixels -- should give just a little more than enough for * a sharp 250x200 (10x8) print viewed at normal distance. It confirms a * rule of thumb I first heard when I was an owlet: that if you use * * o slow film, * o a prime lens, and * o a tripod * * you can safely enlarge your negatives by a factor of 8. With modern * films, Nikon's best lenses, a substantial tripod, and punctilious * technique, you can break that rule now, but for most of us it's as true * as it was in the 1920s. * *************************************************************************
"Tell me, Christopher Robin," said Pooh. "Owl speaks so much in these
discussions ... are the stories about me and my friends, or about Owl and
his friends?"
"It isn't really like that," answered Christopher Robin. "None of us is
more important than any other. The stories are about you and your
friends, Owl and his friends, Rabbit and his friends and relations, and
all of you ... equally. Maybe Owl is more prominent in photographic
discussions, but all of us -- confused, timid, confident, bouncy, gloomy,
and calm -- can find someone like us in the Hundred Acre Wood."
"I think I understand," said Pooh. "So it's right that Eeyore should
have been more prominent in to-day's discussion."
"Absolutely," said Christopher Robin. "It's particularly important at
this time of year that we should stop and think about a donkey. Merry
Christmas, Pooh. Merry Christmas, everybody."
[Ed. note: the following post highlights the issue of focus
shifting; lenses need to be focused at their working aperture for best
results...]
From Contax Mailing List:
Date: Sun, 15 Oct 2000
From: Bob Shell [email protected]
Subject: Re: [CONTAX] AF accuracy
Focus shift has always been a problem with manual focus cameras. Norm
Goldberg wrote a series of articles on this topic some years ago. I think
they ran in Popular Photography magazine. Norm even invented and patented
a system which would move the focusing screen as the aperture ring was
changed to compensate for the problem. With this system lenses would have
had a cam connected to the aperture ring which worked with a system of
levers in the camera to move the screen. It was completely mechanical,
simple, and would have been reliable. Each lens's cam would have been
manufactured to match the focus characteristics of the lens. I believe
some info on this was also in Norm's book.
I have not seen Norm in years and don't even know if he is still with us,
but it only took a few minutes listening to him to learn that this was a
real problem for which he had devised a real solution. He never could get
any of the camera companies to adopt the system, though. I always thought
that was a shame.
Bob
[Ed. note: included message:]
From: adam forrester [email protected]
Date: Sat, 14 Oct 2000
To: [email protected]
Subject: [CONTAX] AF accuracy
2)focus shift on stopping down is not taken into account by af software
otherwise how can the canon results be explained or for that matter it
would
imply manual focus cameras cannot achieve as accurate focus as af cameras
due to focus shift not being taken into account.The increase in depth of
field covers focus shift.
[Ed. note: Mr. Erwin Puts is a noted lens testing expert, author of many
articles on Leica and related topics (as well as a CDROM) worth checking
out!]
There are a few strands of arguments mixed up after my original post. Let
me try to entangle and clarify them. And I apologize for being partly
resonsible as I was not clear enough, I am afraid.
First the basic point: I stated that in silverbased 35mm photography a
resolution of 40 lines was enough for god quality/handheld imagery. I
still stand firmly behind this statement.
40 lines times 24mm times 40 lines times 36mm equals 1.3 million image
points. For those with a historical archive, we can look at Barnack's
article about why he selected the 24x36mm format. And you will find the
identical reasoning. Barnack arrived at a maximum count of 1 million image
points as the norm for leica photography.
Incidentally 40 lines per mm give you image points 1/40mm small, that is
image points of of 0.025mm.
As we all know the image points on the negative are assumed to be points
with a diameter of 0.03mm (the circle of confusion). And Barnack used the
diameter of the CoC in his calculation. So we may safely state that the
1.3 million image points in a 35mm negative are assumed by the inventor of
the format to be good enough for normal situations.
OF COURSE; current emulsions have much higher resolutions and when we use
200 lines/mm as the optimum (corresponding to the 100 lp/mm of the best
leica lenses), we get a much higher count: 24 x 200 x 36 x 200 = 34.5
million image points.
Back to the proposition. The 1.3 milion image points are 960 x 1440 image
points. By calling them picture elements, some were quick to assume that
the 960x1440 grid or matrix of image points is identical to a 960 x 1440
sensor array (CCD).
This however is NOT the case. And by identifying my grid of image points
with a small scale CCD of indeed very modest quality, the matter gets very
confusing.
When you digitize a picture (by scanning) or capture (CCD), we use a
sensor array of CCD elements of a certain dimension. Every individual
sampled (digitized) element is called a pixel. BUT: pixels are inherently
without physical dimensions! Here is the missing link. Image points as I
used them in the resolution calculation have a finite, but well defined
diameter of 0.03mm and these points are physically located in the
emulsion. Not so in digital capture. A pixel is identified in an image
file by its spatial coordinates (rownumber/columnnumber) and the
attributes of colour and brightness. Jim Brick will surely confirm this.
As it is so important I will repeat: pixels have NO physical diamensions
and have only a reference in spatial coordinates. In other words: the size
of the original sensor that is used for sampling is not imcrporated into
the data of the image file.
The file tells you that at columnlocation 15 and rowlocation 5, there is
associated a color and a brightness. This info is the same irrespective of
the size of the individual sensor (large or small). The digitized picture
is represented in the image file as a matrix of N by M dimensions (rows,
columns). Of course the finer the sensor grid (size of sensor elements)
the larger the matrix. But large or small, this matrix is without any
physical dimension.
We simply have a list like this one:
This list would be the same if our individual sensor elements are 3 micron
or 30 micron in diameter. Of coure a 3 micron element can capture more
detail than can a 30 micron sensor.
The simple notation: this CCD has 960 x 1440 picture elements (sensor
elements), gives you no clues about the size of individual sensors. It
only defines the size of the N x M matrix in the file.
We all know that the pixel array in de digial camera, needs more than one
sensor element to capture the necessary information about location,
brightness and colour. Generally we need 4 sensor elements to define one
cell in the matrix above. The CCD of 960 x 1440 sensor elements can
capture only information for a 480 x 720 matrix in the image file.
When do dimensionless pixels get their physical size or dimension?
The physical dimension is attached to the matrix when we print or display
on a screen. Easiest to understand is the monitor sceen. This has a fixed
size of screen dots of 0.26mm (some more some less!!). Normally we say 96
dots per inch for this output medium. And now we add dimensions to our
dimensionless pixes in the matrix.The dot size is by the way almost
identical to the diameter of the CoC.
Because we now make a one to one correspondence between the N x M matrix
and the screen dots. If we have a 100 by 100 matrix (again for the sake of
argument), we get on the screen a 100 x 0.26 by 100 x 0.26mm image,
composed of 10000 pixels. The size is 26 by 26mm and the resolution is
1/0.26 = 3.8 lines per mm. It is of no importance if the screen itself is
a 14 inch or a 21 inch monitor.
If we need to fill a whole screen with a picture of 20x25cm, with this
resolution and dot size, we need a matrix of 200mm/0.26mm by 250mm/0.26mm
that is 769 x 961 or rougly 800 x 1000.
Now we can combine matrix size, print size and resolution. If we want a
print (in this case screen) of dimension 20x25cm, that can show image
details of size 0.26mm (the value of the CoC or threshold of sharpness
impression) at a normal viewing distance, we need an image file of 800 x
1000 rows and columns (or dimensionless pixels).
Now we have to add the size of the sensor elements that we used to
digitize the image in the first place. If these sensors had a diameter of
0.26mm too, we may assume that the sensors could only registrate quite
course detail, like outlines of buildings or cars or people at a distance.
If the sensors had a diameter of 0.026mm, we can registrate much finer
detail. but to cover the same image area, we need ten times as much pixels
or matrix elements. Now we need a matrix of 8000 by 10000 rows and
columns!
With film and a resolution of 40 lines/mm on the negative (that is 1.3
million image pounts) I can easily enlarge 8 times to get a 20x25cm print
and then hold 5 lines per mm in the enlargement.
To get that level of detail in a digital print I need a matrix of 200mm
times 5 lines by 250mm times 5 lines equals 1000 by 1250 true matrix
cells.
With current CCD chip technology that is a CCD count of 2000 by 2500
pixels.
This number is close to the current 3 Megabyte standard of digital
cameras.
If I would improve my image quality on the negative I can with the same
film and lens get a resolution of 10 lines in the print over a 30 by 40cm
print.
That would amount to a matrix of 3000 by 4000 or 12 million times two or a
CCD with 24 Mb of pixels.
I hope this clarifies.
My original intent was to show that the normal quality of handheld 35mm
photography can be quantified as having a resolution of 40 lines/mm or 1.3
million image points.
To transfer this requirement to the digital world needs a CCD with at
least a 3 MB pixel count, when you put the dimensions in the equation and
not only the number of sensors in rows and columns. The whole argument
based on the equivalence of my 1.3 image points to a ccd with 960 by 1440
sensors is invalid because of its lack of accounting for the physical
size/dimensions.
My original intent has not been to say that digital technology will
surpass in the near future the detail defintion and clarity of a good
silverbased negative. On the conrary. Above you can see that a 24 Mb chip
would be needed to come close to film performance.
Erwin
From Leica Mailing List:
In a wide ranging BW comparison I selected films from 25 to 125 ISO and a
score of developers from D76 at one end (fine grain solvent) and Rodinal
at the other end (acutance, grain preserving). I used (in this case) test
charts, as I wanted to compare definition, grain pattern and cut-off
frequency, that is the level of detail that cannot be resolved anymore.
The first surprise: under the microscope I could see almost the same level
of detail, when enlarging 10 or 400 (!!) times. Even at 400 times (which
would amount to a print of more than 12 meter wide, surpassing clearly
Newtons Big Nudes), quite fine detail was preserved in the grain and could
be clearly seen. To use some (relative) figures. The test pattern has a
number of 25 and half of that resolution would be 12.5. Intermediate steps
are 20 and 16, meaning a reduction of 25% per step. At enlargement 10 the
25-pattern could be seen very clearly and at 400 times the grain of course
would kill this level of resolution, but still 16 could be just detected.
Now imagine a digital camera that without alteration can produce prints of
A4 format and 12 meters wide?
The influence of the developer is relatively modest. With the same film
and same enlargement, the best could do '25'and the worst could do '16.
For critical work this may be important, but for general work it is of
lesser importance.
Film choice gave remarkable results. The 25 and 50 ISO films (exception TP
and the mystery film) could hardly resolve more details than the 100 ISO
class, '34' could be seen, but '40' not. Here the limit is the lens, not
the film! So while the '25' could resolve 30% more than the 100 ISO, we
are
talking about a level that is hardly attainable in practice.
The two best ISO100 films are D100 and TMX in that order. D100 resolved
25% more than TMX (same developer of course), and had more pronounced
grain, but also more edge contrast. The worst D100 combination was below
the level of the best TMX combination of course. This shows how close the
performance is nowadays.
'The most important factor that lowered resolution is camera shake, which
can influence the resolution by more than 100%, so in fact much more than
the influence of film and/or developer.
So the trend in the industry to neglect the 25ISO film for 35mm work is
understandable, as the very high performance of the best ISO100 film would
exceed the performance of most lenses and even at enlargements of 100
times (print size almost 4 meter wide) everything the lens has been able
to record on the film can be seen and printed. This is a remarkable
achievement of current silverbased emulsions and given this level of
quality the optical industry has to come up with some tricks to push film
to its limits. Only a lens like the Apo-Telyt 4/250 would be able to
exploit all what silver has to offer.
The best all round combination? D100 and FX39 or Xtol, but good old D76 is
very close. Even Rodinal gives amazing performance with this film.
At a level down in the definition of very fine detail, the relative number
'12,5' or '10', I have to say that almost every film in the ISO100/125
class can handle this, APX100, and PlusX too. With good clarity of details
and very fine performance. More to follow
Erwin
From Rollei Mailing List;
you wrote:
Guitars or fish?
The large format folks argue Schneider vs: Rodenstock with an occasional
rooter for Nikon or Fuji. If Zeiss still made LF lenses what a donnybrook
we would have. I have one LF lens that is less than fifty years old. Some
of the old-timers were pretty astonishing.
Measuring the actual performance of actual lenses is not easy. I have
recently been reading some papers from a symposium on the subject held by
the National Bureau of Standards some forty years ago. It seems to be very
easy to be mislead by simple chart testing. I had a long argument with a
fellow on the large-format group about this some time ago. He claimed that
any name-brand lens would resolve 400lp/mm at f/4 and had observed aerial
images demonstrating this. He also claimed he could see the resolution
drop when viewing in late afternoon redish light. I suspect he was seeing
what is called false resolution. It turns out this can lead to
observations of double the actual resolution, even more under some
circumstances.
Lenses seldom approach the Rayleigh limit of resolution due to
diffraction at larger stops, even exceedingly well corrected ones.
Generally loss of resolution due to residual aberration will far exceed
the diffraction loss so getting anything near 400 lp/mm with any lens is
rare.
----
[Ed. note: Mr. Erwin Puts is a noted lens tester and author on Leica
related topics and lens testing in general...]
It has been proposed on this list that 'tonality' is more important than
'sharpness'. But do we really know what we are talking about. These are
fuzzy terms of which we assume the user knows what he is talking about. On
the assumption that these concepts are related to the finished print (one
of the endresults of the photographic imaging chain), let us delve
somewhat deeper into these terms. Are they meaningful and are they indeed
at different ends of a scale.
'Sharpness' as most will know does not exist as such. We have a sharpness
impression as our visual system scans a pattern and the sharpness
impression is greatest if there are steep bounderies between adjacent
areas of equal brightness. We can measure these edges with a
microdensitometer and it is commonly called acutance, correctly describing
the phenomenon.Websters Dictionary defines acutance as a measure of the
steepness or abruptness of an edge in a photographic image. At least
Webster knows what photographers are talking about.
Now 'tonality'. Any print has a maximum white and maximum black, that can
be measured: whiter than the physical base of the whitening agent in the
emulsion is impossible and blacker than the full expose of all silver is
impossible too. So the maximum tonal range lays always between pure white
and black. To stay within the realm of measurements first, it is clear
than the progression from white to black happens in steps of more or less
grey.
Fechner's Law states that the eye can only discriminate between two values
of grey if the relative difference is 2%. Now evading the equations as
they involve logarithms, it can be calculated that between the pure white
and black there are 243 different grey tones (values) that an observer can
distinguish in a print. We can define tonality as the maximum range of
grey tones that can be distinguished and a good print should be such that
all these shades of grey are present. If a print has a lesser range of
grey shades, we may say that such a print has a limited tonality.
Comparing both concepts acutance or edge steepness (sharpness) and maximum
range of grey values (tonality), is there anything that suggests that both
are at odds with each other. Not at all! Both are present in a print and
it may even be argued that both do enhance each other.
The implication behind the statement might be that some lenses are better
in the rendering of tonality than in definition of sharpness. This would
be wrong! A lens than can record fine details with high clarity will also
record any subtle differences in grey value, where a lesser lens would
have flare and low microcontrast and other defects than would wash out
fine differences in tonal values.
So a lens good for the recording of fine details (acutance) is also good
for tonality.
Back to the original proposition: sharpness and tonality can not be used
as two concepts that are mutually exclusive and not even as a slider on a
scale: more of A means less of B.
It is the fuzziness in the original statement that may produce confusion
in sketching a dichotomy that does exist.
Erwin
From Leica Mailing List:
One of the interesting aspects of this film is ots high resolution: after
several trials I got 150 to 175 linepairs/mm. I used the Apo-Telyt 3.4/135
and a testpattern which among others had small objects (lines and circles
and more complicated figures). The chart was photographed at 4 meters
distance and this film/lens combination gave details that on the negative
were a mere 0.003 mm. That is 3 micron. Can you image how close two
details are that are separated by a three thousands of a mm? At this level
it is most difficult to make observations that are absolutely sure, as the
margins are so small. To be on the safe side. let us go for a resolution
of 120 line pairs/mm, that is details of 0.004mm. Not a big improvement,
you might think, but in this type of analysis, it gives a safety margin.
We know that the eye can resolve at best, under ideal circumstances, 6
linepairs/mm. Given the large differences between individuals, most
handbooks go for 3 linepairs/mm as an average. How much do we have to
enlarge to see these minute details in the negative. For the high figure
it would mean 20 times and for the lower figure 40 times.
Most darkroom workers will have noticed that the jump from 10 times
enlargement to 15 times or higher, induces a severe loss of contrast (the
lens cannot handle it, the enlarger light is spread out over a larger
area, giving flare and reduction of micro-contrast). The definition of
very fine detail automatically suffers.
In this case we are glad to have a high contrast Leica lens, as this
degradation can be limited. The Focotar however has to be changed as the
Focotar cannot handle these bigger enlargements.
Consider this: at 10 times enlargement the eye can detect details that on
film are between 30 and 60 linepairs/mm. Anything smaller than this limit
cannot be seen. A solid black or grey dot may be indeed a dot or it may
contain micro structures. At 20 times (or better at 40 times) the solid
dot shows even smaller detail and we see details in the details. With a
lesser camera lens the enlargement would indeed show no further
information, the black dot stays black and will be somewhat larger. This
state of affairs does also indicate that testing lenses when done at
moderate enlargements (10 to 15 times) may not be representative of the
true image performance.
Some pople will argue that they never go beyond 10 times, so any
additional information capacity is useless. This at first sight makes
sense. But if we use an analogy from music, we know that music sounds
better if the low and high tones that we may not be able to hear are
present, adding to the volume and richness of the sound. The same works
for film. You may not see the details, but you may sense that there is a
hidden dimension beneath the surface and indeed, when enlarging more, the
information pops up. Optical performance along the whole line of spatial
frequencies is more or less linear: if a lens is good at recording the
fine details, it is automatically less good at recording even finer
structures. But the reverse os true too: if a lens is excellent at
recording very fine structures, it is automatically outstanding at the
definition of fine details. So the better lens may have an information
capacity that exceeds normal expectations, but you get, as a bonus so to
speak, excellent image quality at the level of perception we habitually
use. AND you have the ability to go successfully beyond the self imposed
limits.
There is a persuasive argument, that goes along the line that what was
good enough for Cartier-Bresson should be good enough for us (technically
speaking). Here I wish to disagree. HCB and other masters of 35mm
photography brought us a new vision and greater perceptual awareness of
the reality around us.
Our technical capabilities have been improved significantly and now we can
define our own view of the world ith our enhanced recording capabilities
and I think we should try to do this. Explore new worlds as Trekkies would
have it. We can get stuck in just proving that it can be done or we may
integrate high resolution photography into a new type of imagery, where
all aspects of contrast, fine grain, high definition and full tonality
will be supportive of our visual language statements.
Erwin
From: "Q.G. de Bakker" [email protected]
Tom Thackrey wrote:
Amazing how people tend to forget they are addressing the
rec.photo.EQUIPMENT.35mm newsgroup ;-)
If you don't want to read peoples opinions on the latest gear, you've
strayed into the wrong kind of newsgroup.
At the moment, my newsreader has 24,564 messages in this newsgroup, and
only 20 in the rec.photo.technique.art group. And you're message is one of
the 24,564...
From Leica Topica Mailing List:
[email protected] wrote:
Different lenses for digital cameras. It takes a 6 megapixel sensor to
come close to the 24x36mm size of a 35mm film frame. These are expensive
and require an inordinate amount of electronics to handle. I believe one
or more of the $15,000 Kodak/Canon/Nikon cameras uses this sensor. Ever
pick up one of those behemoths?
The problem of using an M lens on a digital camera is that you have to
limit the bandwidth of the lens to be no more than 70% of 100lp/mm (5
micron pixels.) This is why many manufacturers (Schneider, Rodenstock,
etc) make special lenses for use on digital cameras. It is agreed that
with non periodic subjects such as people, animals, landscapes, this
problem does not become blatant. But if you are recording a subject using
the best f/stop of the M ASPH/APO lens and there is fine detail that you
want to record, you won't like what you see.
They also make sensors with the pixel active areas (the photo transistors)
butted together with no space in between. This is extremely expensive and
used in military and astronomical applications where data falling between
pixels cannot be missed.
Film in interesting and unique as an image recording medium because the
"pixels" are randomly distributed and shaped grains of photosensitive
crystals. Because the grains are randomly shaped and distributed the
regularity which causes the moire/aliasing in a CCD detector is not a
problem with film. Therefore the lens can have much better resolution
than the film without causing these effects.
Making smaller pixels is not possible. Three microns is about the limit
because there is not enough room for an electron collector (capacitor) for
each pixel of sufficient size to give meaningful results. The signal to
noise ratio goes way down. And the dynamic range is non existent. Anything
under one micron is beginning to interfere with the wavelength of light
and then you have a whole new set of problems.
As I've said many times before, the electron microscope was invented
because the optical microscope could not increase in power as it and the
wave length of light were interfering with each other. Same problem we
have in recording light values in small pixels. They should invent a new
technology that has randomly distributed and shaped pixels that can be
disbursed at any frequency. Oh, I forgot. They already have.
It is called film.
Jim
From Leica Topica Mailing List:
Schneider has a white paper on optics for digital imaging at:
http://www.schneideroptics.com/pressreleases/kina.htm
It covers sharpness and resolution on film as well.
Jim
From: [email protected] (Hemi4268)
Hi
This is a very difficult question to answer unless you know how to measure
resolution. Just like focal lengths have a value, so does resolution.
For photography, resolution usually ranges between 10 lines per millimeter
(l/mm) to about 200 l/mm with the average in todays photography being
about 70l/mm
Wth this info you can calculate the detail by taking the focal length of
the lens and multiply it by the resolution and then dividing it by the
distance.
So a image taken with a 50mm lens that has 100 l/mm at 10 ft has exactly
the same detail as a image taken with a 100mm lens that has 50 l/mm
resolution taken again at 10 ft. So generally, the 35mm negative is going
to have to have twice the resolution as a 6x7, in order for it to be the
same quality as the 6x7.
Larry
Date: Wed, 13 Dec 2000
PBurian wrote:
Peter, I don't fully agree with your opinion about Side-by-Side tests.
Sure, without an optical bench and loads of equipment you cannot get
scientifically valid results. But I fell, such results are not really
needed and not even always appropriate when comparing lenses.
I know that e.g. the german magazine "Color Foto" has the equipment and
does scientific testing. I don't doubt their results, I'm quite sure they
don't cheat. But then, do their results really give a means to say: "Lens
A is better than lens B"? They meter MTF. Nothing wrong with that. But,
they do it at 10,20 and 40 lp/mm. What about 5, 15, 25, 35 lp/mm? What
about 60 lp/mm? They do their measurements at a certain focusing distance,
but what about other distances? In the few cases they DID meter at several
focusing distances, the results varied greatly. These were always macro
lenses, one measurement close up (1:1) and one at their "normal" distance,
so these variations are probably at the extreme end. But still ... Same
goes for their measurements of distortion, vignetting, flare ... (you name
it). The data are scientific, but FAR from complete.
Then they boil everything down to one or two numbers. If you want detail,
you still have to look at the MTF diagrams which are only a tiny, small
fraction of the "real" MTF function. Their numbers simply cannot tell the
whole story.
OTOH, a side-by-side test done by a seasoned pro in the field sure cannot
be scientific. But his conclusions are A LOT more useful to me than the
scientific numbers, if I use the lenses for the same kind of shots.
Magazine tests rarely tell how good a lens works with a certain TC, with
extension tubes or whatever. What about flare under various conditions?
What about handling, feel, AF speed ...? What about moisture, rain, high
or low temperatures ...? All these little things that are infinitely more
relevant for real picture taking (as opposed to shooting test charts on an
optical bench) than the 2% higher MTF value or 0.1% lower distortion of
lens A compared to lens B. The result of such a non scientific test
doesn't even have to be: "Lens A is better than lens B". Just talk about
your findings and leave the rest to the reader. Stating a personal opinion
marked as such would of course be ok as well.
Thomas Bantel
Date: Mon, 25 Dec 2000
Please go to href="http://wdn.com/~johnchap/lenstest/testlens.htm" for
some description and links to the instructions�which you can download
and print out.� Also Included in the material are the resolution charts
which are� needed.
�
The material will instruct you how to set up the charts, shoot the
tests, and interpret the results.� I have� always found it very
enlightening how certain lenses do.� The results you will get may very
well surprise you. I have found several�inexpensive lenses with
excellent optical quality.� On the other hand, I have found some very
expensive lenses that are not nearly as� good as their much less
expensive brothers.
�
If you encounter any problems, have any questions, or have any advice
how I can improve access and�usability of the site, please email me.
I have been testing the new Gigabit high resolution film since several
months and what began as a simple filmtest, has now been expanded into a
fresh exploration of film and lens capabilities. My results surprise me as
much as I hope it will surprise the readers of this report. An open mind
is necessary, however, as many established ideas have to be discarded!
The Gigabit film has been promoted as a new dimenson in Hi-res BW
photography and a useable resolution of 600 to 900 lp/mm (every line pair
a black and a white line or space) has been quoted. In several usergroups
this claim has been discussed to the extreme: could it be possible
theoretically and if so could there be lenses that can use this
capability.
First the basic fact: the resolution of the eye: under ideal circumstances
the eye can resolve at a distance of 25cm at most 6 to 10 lp/mm and here
we take the absolute limit, which is reached when the eye is using its
vernier acuity resolving power. Six lp/mm would be according to all
ophtalmic handbooks a good average. This means that any detail which is
smaller than 6 lp/mm cannot be detected as separate lines and will be seen
as a grey patch. A simple calculation shows that to be able to see details
that on a film are recorded with 600 lp/mm, we need a magnification factor
of 100 times.
I started using a high quality microscope with a magnification of 40, 100
and 400 times. Then I used a testchart with lines and circles (as the
pattern itself will influence the resolution limit). I set up the Leica in
front of the testchart at a distance that gives a negative magnification
of 100 times. The idea was that when using the microscope at M=+100, I
would be able to see the resolution pattern that is closest to the 600
lp/mm. I used the Apo 135 as this one is capable of resolving at least 300
lp/mm at a acceptable contrast and even 450 lp/mm at a very low contrast
(less than 10%).
To ensure optimum results I used of course the center portion of the
negative and to make sure film flatness and focussing errors are not a
problem Iused the following setup. A Siemens Star was used to check
accurate focusing (a phase shift in the pattern will indicate a focus
error) and I did extensive focussing bracketing by marking the distance on
the lens and taping scaled paper on the mount to accurately make my
bracketing. Result one is that the lens and the camera focussed extremely
accurately even at 13,5 meters with a 135mm lens, which is reassuring in
itself.
I shot three films and checked every negative under the microscope to find
the best results. As resolution tests always involve errors in viewing, I
used the best 10 results and averaged the numbers to get a result that is
at least in principle reproduceable by anyone.
To my utter surprise the microscope revealed at most 60 lp/mm!! (IN one
instance I read off 90 lp/mm, but that is the extreme case, which I could
not reproduce later on). So 60 lp/mm is the practical maximum resolution?
But: what is the limiting factor here: lens or film and where are those
mythical 600 lp/mm?
Back again to the testchart, now with a Summicron 50 of latest design.
Repeat the whole procedure and I got somewhat lower figures, around 50
lp/mm.
Now some people assert (wrongly, but it is common view) that leica lenses
are optimized for contrast and not resolution. To counter this (you see
what you have to do to get a reliable test with meaningful figures) I used
a Canon F1 with a 1.8/50mm lens, reputedly a high res lens. Results did
not differ: no significant and statistically relevent difference in
resolution between the Leica and Canon lens. I used as a third test a
TriELmar at the 50 position and got again results around 50 lp/mm. So
obviously the lens is not the limiting factor and so the results found (60
lp/mm) must be film based. Could the spectral composition of the light be
a factor.
I used all light sources, flash, daylight sun and tungsten halogen light
to look for differences: none found.
Then we need comparison films to investigate further. APX 25, APX 100,
Delta 100, Pan F, TMax100 and Technical Pan were selected.
As Gigabit film is standard Agfa Copex Microfiche film (no new emulsion
this Gigabit! Only new is the developer, but wait and see later) and TP is
also a micro-film, which are "forced" to go for continious tone negatives,
I wanted to use films designed form the start as continious tone film as a
comparison.
Surprise two: using the same procedures and testing developer-film combos
with D76, FX39, Rodinal, TMax, the best film combo for resolution is
TMax100 in FX39, which even surpasses the APX25!
To be continued!
Erwin
The first check
From Leica Mailing List;
This study is not designed to be a lens test. The use of several lenses
was necessary to ensure that the employment of a specific lens would not
unduly impair the validity of the results. It may come as a surprise that
the humble Canon lens, which is a cheap noname in the market, is as good
as a current Summicron (when resolution on axis or close to the axis is
concerned). Optical (MTF) tests did show the Canon to be very well made
(no decentring etc) and indeed had a slightly higher aberration content
than the Summicron, specifically spherical aberration softness of smaller
details.
For a resolution test this is not such a big problem. It does indicate the
the market value of a lens does not have to be related to its true optical
capabilities. Market value is related to scarcity and that elusive value
of What-Peope-Want-To-Believe.
Back to the real topic.
What I try to establish is system performance and not the absolute
performance of a lens or a film. Generally any photographer uses a film
and a lens for pictorial or continious tone pictures and so the absolute
resolution figure of a microfilm (like Gigabit=Agfa Copex and TechPan) of
600 or 400 linepairs/mm is a figure to be interpreted. This figure is
based on two assumptions: a film gradient of 3 and the possible resolution
of black-white patterns, like letters. To make these films useful for
pictorial photography, the gradient (characteristic curve or CI-value) has
to drop to a more normal value of CI=0.6. The Gigabit film just does this
and uses a specific developer that gets a CI of 0.5 or even lower. The
true speed point of this film is ISO20 and when using it as EI=40 (as
recommended) and developing it to a CI of 0.5, we get the classical
pattern of underdevelopment and underexposure, good for the dynamic range
(overexposure latitude) and resolution, as overexposure will kill any
attempt to record very fine details.
I used TechPan as a companion film and when developing both films (Gig and
TP) in Technidol for the same development time, I got identical results.
The much discussed dynamic range of the Gigabit film can be had with the
TP too if developed to this effect.
Studying resolution patterns at several densities, I noticed that all
these hi-res films are very sensitive to small variations in exposure
(plus and minus). Best resolution you get at densities around D=0.4 to
D=0.6, which generally is one stop below the grey card value. That may be
the reason that hi-res films often are underexposed.
You can only compare film capabilities when films are developed to the
same CI-value, so my first attempt was to find the correct exposure and
development time for all films to be comparable. Having established that I
selected the lens for a new series of tests. (My third attempt and many
rolls of film used).
Officially best results with a lens are at the infinity position (which is
defined optically as the location of the object where the rays are
perpendicular to each other and reach the lens at right angles). Normally
this at 50 to 100 times the focal length. For all lenses I made tests at a
distance of between 5 meter (50mm) and 13.5 meter (135m).
Now is 13.5 meter quite inconvenient, so I checked with Leica what would
be the closest range without dropping significantly in optical quality.
They proposed that 5 meter would be appropriate for the 135mm.
The apo-90mm was also used, at 5 meter and 4 meter. You should realise
that films in the speed range of ISO 16, 25 and 40 are difficult to test.
At longer distances the flash or the tungsten lamps will be at low power
and using shutter speeds of 1/4 sec and an aperture of 3.4 or 4 is not
quite convenenient.
So I did another thing and ordered new resolution charts (the same Leica
are using btw), which make for easier and more accurate testing at the
distances to be used.
With these new goodies (test patterns and film/exposure/development resuts
for optimal result and comparison), I made extensive series of tests with
both lenses at several distances and the following combinations of
film-developer.
Gigabit in Gigabit developer, in Rodinal, in Technidol and in FX39
TechPan in Rodinal, in Technidol and in FX39.
Seven combinations in all, with two lenses at three distances and optimum
apertures, including exposure and focus bracketing. That kept me "happy"
for a while. Here I have to include some observations about resolution
figures in general. The difference between 80 lp/mm and 100lp/mm looks big
when you interpret in numerically. If you look at the resolution patterns
themselves on which the calculation is based, you see a much smaller
difference. In other words, there is a law of diminishing returns when you
progress in the higher resolution numbers.
My test was setup as follows: the original test patterns have real
patterns of 1 and 2 linepairs /mm on the print. They go in steps of 1.1,
1.25, 1.4, 1.6 and 1.8 lp/mm. So if I use a reduction of 40 times and can
clearly detect in the subsequent enlargement a pattern of 1 lp/mm and the
next pattern (1.1) is blurred, then my film/lens-resolution is 40 lp/mm.
I used reductions of 30 and 40 and 100 times, with the main attention to
the 30 and 40 times enlargements: that would still imply a print of 1meter
wide from a 35mm negative and who does this on a daily basis.
The results? Next
Erwin
------------------------------
The results?
The best resolution I could get (averaged over some observations to
account for eye fatigue and some unavoidable subjectivity when deciding
if a certain pattern is just resolvable or not):
Every film/developer combination could easily produce 60 lp/mm IF and
WHEN
developed and exposed and focused accurately. If any of the three
variables
is off the optimum value, results drop to 40 lp/mm or even worse.
Of course you need to address the issue of grain and tonal range and
dynamic range etc and here we see that best are TechPan and Gigabit (but
only whenexposed/developed correctly!), with APX and TMax100 (ex equo) and
just a small step behind the hi-res films.
If you do realise that Tmax100 has a speed of ISO100 and the others barely
get ISO25, that the tonal rendition of Tmax is excellent and that it has a
dynamic range of 7 to 8 stops, when developed appropriately, the overall
winner should be Tmax100!.
The startling conclusion for me is that there seems to be a threshold of
useable resolution of about 60 lp/mm. This value is attainable when using
outstanding equipment in the right combination of all elements of the
imaging chain.
The use of hi-res films will give you an advantage in grain when enlarging
beyond 15 times, but you will hardly get better useable resolution. To
move beyond this threshold to values of 80 or 90 lp/mm, can be done, but
the additional care, accuracy and control over all parameters is extremely
demanding and you may question if this additional amount of control and
energy is worth the effort.
To be realistic: a true 60 lp/mm with high contrast (high MTF values) on
the 35mm negative will deliver a 15 times print with a print resolution of
4 lp/mm, which is exremely high and needs a very attentive eye to detect.
.
I have to withdraw my previous conclusion that the film is the limiting
value in the imaging chain. More study and calculations and these results
with several lenses at several distances do indicate that the optical
quality of the lens is the limiting factor and even more so the user
capability and accuracy of his control over the imaging chain.
The Kodak people do agree that there is not that much practical difference
in 35mm photography between TP and TM100, when used to its full
capabilities.
There is an empirical equation to find the system resolution of lens and
film resolution. This calculation shows that a film resolution of 200
lp/mm and a lens resolution of 100 lp/mm (like apo 90 or apo 135)gives a
system resolution of 90 lp/mm!! Precisely the result I get experimentally.
That shows that my rsults are close to what can be done, using the
equations that Kodak and others use, based on their experiments and
testing.
Feed into the equation a film resolution of 400 lp/mm and you get a system
resolution of 100 lp/mm, again within the boundries of my results. Use a
lens with a useable resolution of 50 lp/mm (like the Canon or Summicron
50) and you get a system resolution of 50 lp/mm again a confirmation of my
previous testing results and why I changed from Summicron 50 to apo90 for
this type of testing.
The equation tells you that for a lens with a resolution of about 50 lp/mm
the resolution of the film is irrelevant! Film resolution becomes an
important topic when using high quality lenses and being able to use the
optical capabilities of the system!
This conclusion will have strong imapct on the domain of handheld
photography and it also explains why Tri-X and handheld shooting is such a
happy and impressive combo..
The next stage will be the test of the apo 280 and the R8. Here I have the
absolute pinnacle of lens quality and resolution.
Next part will discuss the implications: so do not rush to hasty
conclusions and let the results as they stand, sink in and reflect a bit
on what is being unveiled here.
Erwin
From Leica Mailing List:
I started this investigation to find the truth about the claim of the
Gigabit people that by providing a film with a resolution of 600 to 900
lp/mm would improve the quality of current BW photography significantly if
not dramatically.
The use of document film in a POTA type developer for pictorial results is
not new and indeed with TechPan we have a solution that is claimed to
provide the 35mm user with medium format quality.
So I compared the Gigabit with the Techpan as a natural comparison. My
selection of Tm100, instead of D100, was motivated by the fact that TM100
has slightly finer grain than D100 and a straighter curve, making it more
in line with the GB and TP film curves, which are quite straight too. As
GB stresses its capabilities to not diminish in quality when overexposing,
the TM100 was the better comparison.
A second argument was the role the TM100 (and Tgrains) has played in
emulsion technology and its boost in quality. I also wanted to make
readers aware of its excellent quality and induce them to use this film.
There is no need to cultivate a mono-culture here in film choices.
Resolution does not tell all of a film/lens capability. As I have often
noted, no single number or one-dimensional approach can capture the full
spectrum of a lens/film combo. It is well known that the lens MTF is a
excellent representation of the residual aberrations of a lens. Film MTF
is based on acutance measures. The combination of film and lens MTF (the
so-called cascading function) gives you a number, which is the result of
combining the MTF value of the lens (at a certain resolution) and the MTF
value of the film (at a certain resolution). MTF values and resolution
values are (as I noted in my Viewfinder articles) proportionally related.
A high MTF value is also a high resolution value. But if I would say that
the film/lens combo would provide at 50 lp/mm a contrast transfer value of
20% and at 100 lp/mm a value of 3%, most readers would have trouble
interpreting these values.
So for practical evaluations of film/lens systems, the resolution figure
is a fine approximation of useable image quality when doing high
resolution and high magnification photography, providing high quality
optics are used (those with high MTF values over the relevant spatial
frequencies).
It does not make sense to do this exercise for Tri-X films and Leica
lenses as the resolution of the TX will be reduced significantly by grain
patterns to a low level. Generally my preliminary conclusions are these:
(subject of course to further knowledge I may get during the rest of the
study).
If you use films with a relative low resolution figure of 100lp/m and
less, the quality of the optics is significant for image quality and with
a moderate control of all components, you may get 10 to 30 lp/mm on the
negative.
When using films that have a resolution figure of 200 lp/mm, the quality
of the optics is decisive and here with a tight control of all components,
you get around 60 to 70 lp/mm on the negative.
Films with even higher resolution (which automatically implies a very high
MTF value at the lower spatial frequencies) can deliver up to 100lp/mm,
but with topclass lenses and an extremely high level of control of all
components.
I also noted that the grain pattern is more likely to influence the result
than does resolution. So TP may be as good as TM in practical resolution,
but will win in the fine grain area when enlarging above 15 times. SO the
end result may be more pleasing with TP, but you do get the same sharpness
impression and detail definition.
Quite surprising is the fact that slight errors in focussing accuracy and
exposure have more impact than film flatness. At least so it seems in the
middle of the negative. I did not check the corners of the negative area,
because I have not yet found a reliable way to align the camera back and
the test charts absolutely parallel. Some people have reported that their
Leica negs are always less sharp in the left side of the negative, where
the film chamber is located. This could indicate problems with the film
feed. I have not been able to check this in a reliable way, so do not see
this as fact. If you note these problems, let me know.
Some other results. The BJP tested recently the TP, PanF, TM100, D100,
APX25 and old Pan-X and these are the results at 20 times enlargement:
finest grain TP, and with only marginal differences in grain size the
rest. In sharpness it is again TP and then the rest within very close
distance. The BJP summed it up as follows: T-grains have given the medium
speed films like TM100 and D100 properties very close to those of a
previous generation (APX, PanF and Pan-X). With TP a fraction ahead but at
ISO16 to 25 this is not honest when compared to ISO100!
A German magazine (no longer in existence: it was too good and informative
for the market) tested these films also 10 years earlier and used an
elaborate direct contact masking to test resolution. Their results: TP and
APX25 and Agfa Ortho and TM100 could all resolve patterns around 100
lp/mm. But the quality of the pattern was different: clearly
differentiated with TP and just visible with TM100 and not visible with
PanF.
Here lies the danger of using resolution patterns without some additional
qualifications. TP and TM100 both resolve 100 lp/mm, but where the TM100
is here at its limit, the TP could go to 250lp/mm. but this value is not
useable in pictorial 35mm photography with the current best lenses! Just
resolvable also implies that defocusing, camera shake, overexposure etc
have a much greater impact on the result than when using TP.
It is not as easy as you think and to close now with a paraphrase of a
great optical/emulsion scientist: it is self defeating to try to capture a
film or a lens in one number!
Erwin
Date: Mon, 01 Jan 2001
"Lawrence Thompson" [email protected] wrote:
Your question contains an assumption that something is wrong and a new
lens will fix it. Hidden therein is the possibility that what is wrong
is not completely diagnosed.
If you check the archives, you will find a lot of discussion of this
topic. One thread goes approximately, as follows: The ultimate image
depends on technique (of the photographer and the lab), a correctly
aligned camera, no motion during the exposure, the lens, the film, the
paper, the viewer and the subject. Lens performance (you mention lpm),
depends on the subject contrast, the aperture and the threshold of
acceptable contrast. Film performance depends on the contrast of the
image presented by the lens, the resolving capability of the film and
the threshold for an acceptable result. For the lens-film combination,
a typical synthesis is (1/J) = ((1/L) + (1/F)). In this, J is the
resolution of the combination of film and lens, L is the lens
resolution and F is the film resolution (in l/mm). For this to make
any sense, there are a number of other assumptions that could also be
stated.
As a worked example, consider the 60 l/mm lens and a 125 l/mm film.
The result is then (1/J) = (1/60) + (1/125), or J=40.5; if instead you
obtain a modern wonder that produces 80 l/mm, then J=49. Naturally, if
the Optar could produce better than 60, you would not want to change.
It might produce this at about f:22. The new wonder would then not
produce much better at f:22 (diffraction limit). Consequently, to
achieve better, you need f:16. Is that enough depth of field?
More important, what about the rest of the system that produces your
pictures? You need to consider how you proved that it is the lens that
limits your results. Is the rangefinder is off? Is the groundglass
back registered with all of your film holders? Is it parallel to the
plane of the subject and the plane of the lens? Is there any
possibility of motion during the exposure? Is the enlarging lens much
better than the taking lens (otherwise, there is another similar
reduction in performance when used jointly with the film, as when the
taking lens is used with the film). What about the capability of the
paper? Is the enlarger perfectly aligned with a glass carrier and
correctly focused? If, for example, one piece of glass or the other is
hazy, it will reduce the contrast and thus the apparent resolution.
(note: this is not guaranteed to be a complete analysis, merely a
starting point.)
I guess my point is that I would want to try a lot of tests before
concluding that an expensive lens is the solution. I would not
necessarily believe what I was told (in general) about Optars (or
Nikons, or others), but would, instead, see what I could achieve,
myself. In considering resolution measurements, if you have not seen
it, the web site: http://www.hevanet.com/cperez/testing.html is a very
useful resource. A related point in improving the resulting image is
whether 135 mm is the best focal length. Perhaps having a longer or
shorter one would contribute more to the quality of the results than a
second (different) lens of the same (or nearly so) focal length.
From Leica Mailing List:
I came across a small handbook from Fuji that had a formula for
calculating overall resolution and here it is if somebody is
interested.
1/R = 1/r + 1/r + 1/r etc
"R" is the resolution of the total system and the "r"s are the
resolution of individual factors. In the formula the "r"s are
indexed 1 and 2 and 3 etc., but I did not think I would manage to
write this into the e-mail, so I did not try. Instead, look at a
practical example.
If the lens has a resolution of 200 (Lpm) and the film a resolution
of 50, then the overall resolution is 40 calculated with the above
formula as follows.
1/R = 1/200 + 1/50
1/R = 1/200 + 4/200
1/R = 5/200
1/R = 1/40 and R is thus 40 (Lpm)
To me this means that you will get the best payout in form of
improved technical picture quality if you get working on the weakest
point in the chain. This probably seldom is your lens, if you have a
Leica, even if it is non-asph. I wonder how many lines a shaky hand
can resolve, and how many a steady hand.
Sorry if you knew this already. Now go back to shooting film.
Chris
From Leica Mailing List:
MTF Measurememts on lenses with and without filters show: a drop in
contrast of 1 to 2 % when the filter is of excellent quality. Such a drop
is not visible in practice
A drop in contrast around 10%, when the filter is of bad quality. Such a
drop is clearly visible in practice. A drop of 10% is equal to the image
degradation that you get when a lens element is decentred.
Hopefully the filter discussion can be continued on a factual basis now.
Erwin
Date: Tue, 27 Feb 2001
Robert--
I just came across your lens resolution page, and plan to go
through it later (after work); maybe add a link from my new page. I've
just written a page that explains MTF resolution measurements and
simulates the sharpness of a photographic imaging system (film, lens,
scanner, sharpening algorithm) based on published data and accurate
computer models. It's a unique visual approach-- I've never seen anything
like it, but who knows what I'll find on your page? The URL is
http://www.normankoren.com/Tutorials/MTF.html
If you like it, I'd appreciate if you include it as a link on your page.
I'm still revising it, so I'm open to suggestions.
Thanks,
Norman Koren
From: [email protected] (Richard Knoppow)
[email protected] (ZR) wrote:
There is a good primer on photographic optics and lens faq on David
Jacobson's home page at: http://www.photo.net/photo/jacobson/
I think it has a little on resolution testing in it.
Resolution is one way to measure lens performance, but by itself has
limited utility.
The simple measure using a bar chart (there are several types) can
be done photographically, or visually.
The chart is set up at a prescribed distance from the lens so that
the image size at the focal plane is constant, i.e., independant of
focal length. For lenses optimized for subjects at infinity the error
will be negligible as long as the target is more than about 20 times
the focal length distant from the lens.
For photographic examination the target is photographed with due
care to insure perfect focus. The resulting resolution will be the
combined resolution of film and lens. There is no exact formula for
calculating this but it is approximately 1/T = 1/F + 1/L where T +
Total resolution, F = Film resolution, and L = Lens resolution.
Obviously if the two are equal the resulting resolution will half of
either.
Visual examination is done with a very high quality magnifier,
preferably one which does not have such a small aperture that it acts
as a stop for the lens. If it does the result will not be accurate
since the magnifier will make the lens appear better than it is.
The magnifier is used to look at the "aerial image", that is, the
image which would normaly fall on the film, but exists in space at the
point where the film would be. This eliminates the reduction of
resolution due to film loss, but brings a bunch of potential errors of
its own. One of the greatest errors with either system is false
resolution. That is an optical illusion that makes it appear that the
bars are sharp where, in fact, they are not and you are seeing an
overlapping of blured images. Various patterns of bars have been
devised to reduce or eliminate this problem, but great care on the
part of the observer is still necessary.
Lens performance needs to be checked across the field. There are
many lenses which are sharp at, or near, the center of the field, but
which have very poor performance away from it. Generally, all lenses
have poorer performance as you move away from the center, its due to
optical principles. However, there is a reasonable level of perforance
which can be expected. Stopping down reduces some of the aberrations,
especially away from the center, but some lenses deliver good
correction at larger stops than others.
Most of the aberrations show up as lowering the resolution, but some
don't, such as geometrical distortion. Also, the contrast with which
the lens is able to render edges of objects varies with residual
aberrations. The eye percieves edge contrast as sharpness. A
high-resolution but low contrast lens is percieved to be less sharp
than one with poorer resolution but better contrast. To some extent
the curve of the variation of resolution indicates this effect. This
kind of curve is often called a Modulation Transfer Function, or MTF
curve.
Resolution, and even the more complete MTF, is only one criteria of
lens performance. While it is very valuable it is not enough by itself
to completely characterise a lens.
---
Richard Knoppow
From Rollei Mailing List;
No matter how "good" or "special" is a lens design a lot about
the quality it delivers is based on the film being held in the
proper place. We don't generally have "pressure plates". We
have "film Channels". Another consideration is the one pointed
out by Master Lens Expert Norman Goldberg, in Madison Wisc.
He points out that as you change your lens aperture you also
change your lens focus position. I believe he pointed this
out about 20 years ago in Popular Photography magazine.
So, you reach a point in lens design where "getting better"
might not count for anything.
Ed
From Rollei Mailing List;
you wrote:
If he is talking about overall resolution of lens plus film I doubt very
much it has doubled. Film resolution probably has doubled since the
1930's.
Typical pictorial films of that time had perhaps 50 to 80 lp/mm, several
modern films have over 100 lp/mm and a couple nearly 200. This must be
qualified by saying that this is from high contrast (1000:1) targets.
Lens resolution will not have changed by that much but, again, there are
so many qualifications that its hard to be definite.
The resolution of a lens varies with the stop and with angular distance
from the center of the image, generally going down away from the center.
It also is different for radial and tangential lines.
There is a balance somewhere between the fact that several aberrations
are reduced by stopping down and the loss of resolution by diffraction
which increases as a lens is stopped down. At some point there is an
optimum stop. This stop varies with the angular distance since some of the
aberrations do.
That said there are better lenses than those of the 1930's. In fact a
1941 vintage Kodak Ektar is better than a contemporary Zeiss Tessar but
the difference is visible only when wide open.
In short, I doubt if the typical resolution on film is double that of
sixty years ago, although the very best lenses on the very best film
probably make it. Certainly its not tripple.
If you can get an actual 60 lp/mm on film you are doing very well.
----
From Contax Mailing List;
[email protected] wrote:
The resolution of modern colour film is high. Fujifilm claims that Provia
100F (RDP III) has a (high contrast) resolution of 140 lp/mm. It has been
claimed that the 100MP and the 120 645 lens have resolutions in the
160-180 lp/mm range (see Blake Ziegler's recent post). These numbers would
appear to support your argument, but we need to put them into context.
The first thing to note is that such a high lens' resolution is
(probably?) measured for the central part of the image and certainly only
applies to an image (or a part of it) that is perfectly focussed. In
practice there will be focussing errors from many causes. More
significantly, the accepted definition of depth of field allows for a
departure from exact focus which would reduce the achieved resolution
(just inside the limits of DOF) to 33 lp/mm (.03mm CoC) for 35 mm and to
12.5 lp/mm for 645 . So over a large part of our images we are not going
to get anything like those theroetical definitions on the film.
Secondly, even if we somehow achieved a resolution limited only by the
film (140 lp/mm), then differences between formats would only start to
become evident at enlargements bigger than about 12 times magnification
(ie a print in which the smaller dimension is 300mm - 11 inches - for 35mm
film), because of the limited resolving power of the eye. Conventional
wisdom has it that you can detect the difference between formats at much
smaller enlargements than that.
I think we have to look for other reasons why Medium Format might, in
practice, give better results than 35mm. In my view, we need to look at
the resolution of the devices used for viewing the film - projector
lenses, enlarger lenses, digital scanners etc. If one of these becomes
the resolution-limiting element, then the resulting definition will depend
proportionally on the film size: medium format will be almost 1.8 times
better than 35mm, but only if the resolution of the 'viewing device' is
equal in each case. I very much doubt that your enlarger lens has a
resolution approaching that of the film - and don't forget there is a
limit to how flat you can keep the film in the enlarger/projector and how
well you can focus it.
My 'digital' friends tell me that I must scan at a ppi at least twice the
spacial frequency of the detail I want to resolve. So to make the film
resolution the determining element I need to scan at 280 pixels per mm or
7000 ppi! Maybe a good drum scanner can do that, but not a scanner that I
could afford! (and I don't much relish the thought of handling the 220
Mbyte (RGB) or 300 Mbyte (CMYK) files that would result from scanning my
35mm slides either :-(). If I scan both formats at the same, but smaller
resolution than this, then 645 will obviously win. Affordable scanners for
35 mm go to 4000 ppi, but the MF capable ones are more limited. So even
scanning is going to reduce the MF advantage somewhat.
Alan
[Ed. note: an important point, that some reviewers use aerial lens
resolution...]
[email protected] (VT) wrote:
I went to the Amateur Photographer board to ask if the AP lens testing
was done with film or filmless - this was a reply I got:
QUOTE:
...... replied to your post 'AP Lens Test Resolution' at the site:
Amateur Photographer Boards. The Url of this forum is
http://www.amateurphotographer.com/cgi-bin/wwwthreads/wwwthreads.pl.
The reply was:
If my memory is correct, about 6 months ago there was an article in
the AP that showed the gear that Dr. Bell uses. I think that it was
filmless. I must confess to not completely understanding the method of
testing at low contrast, but if you can find the article it may keep
you occupied fo some time.
UNQUOTE
If that's the case, then the calculation I had posted is probably
correct - so the resultant high contrast resolution figures from AP
and Pop Photo do manage to concur pretty closely:
(another good example of "techobabble")
From ROllei Mailing List;
....
The fluffiness of the visible grain is really a matter of the stochastic
distribution of grains rather than the shape of individual grains, which
are submicroscopic. High sulfite developers generally also do not have
much grain clumping. Clumping takes place in developers which cause
substantial emulsion swelling. The developing silver crystals are attacted
to each other and migrate a short distance through the softened gelatin.
The effect varies with temperature, the exact nature of the emulsion, etc.
Developers with high salt content tend to reduce the amount of swelling
and consequent grain clumping. Also remember than the grain in the print
is from light going through the clear spaces between the grain.
A sharp grain pattern does not necessarily indicate a sharp image. There
is an optical illusion caused by superimposing a sharp texture pattern
over an image. The eye tends to interpret the combination as sharp. you
can try this by making a combination print from a low grain but not very
sharp image (or a sharp image which is slightly out of focus, wich will
also blur the grain) with a sharp grain pattern. The combination print
will look sharper than a print without the grain. I think the popularity
of Rodinal for 35mm use stems from this. Rodinal will also generate a lot
of edge effects when diluted, which also give the illusion of sharpness
(acutance).
Because sulfite tends to recombine the reaction products of the developing
agent high sulfite developer also do not have much acutance effects.
Whether these are desirable or not depends on what you are doing. For
instance, for photogrpahic sound recording they are death but can make
small format negatives much sharper looking.
----
From Nikon MF Mailing List;
Scott Perkins wrote:
They're very good questions.
Actually, the correct term for what you describe is "resolution".
Contrast is the difference, in terms of lightness and darkness, between
two areas of an image. Many people incorrectly use "sharpness", which is
a little subjective, as a synonym for resolution.
However, sharpness is a more complex attribute than resolution, and is
used to describe a phenomenon which involves both resolution and changes
in contrast between the boundaries or edges of adjacent light and dark
objects in an image. For example, you can increase the sharpness of an
image with a digital image editor using a function which does exactly
that, without affecting resolution. It finds edges of objects, and
emphasizes the contrast between them, without increasing the resolution of
the image. This makes the edges look "sharper" and may give the
*appearance* of higher resolution. However, true resolution is fixed, and
can never be improved from the original captured image. If a detail is
not captured on the original image it is lost. However, it is possible to
create the appearance of higher resolution by "interpolating" the original
image statistically. This involves creating a new picture element or
"pixel" between two others which is a statistical average of the values of
the two it comes between. This will provide a higher apparent resolution,
but ironically, often at the cost of sharpness, and may or may not be true
to the original scene.
A lens can have a very high resolution, but still have low contrast.
Many would subjectively describe such lens as less sharp than another lens
with somewhat lower resolution, but higher contrast.
As the distinction between film and digital photography narrows, it
becomes more necessary not to confuse sharpness and resolution.
- Rick Housh -
From Nikon MF Mailing List;
Mark Stephan wrote:
I do it with both. Print film first, because it's fast and convenient, and
will give you a quick check to see if the lens is working properly. You
can (and should) look at the negatives to check exposure consistency and
sharpness (using a loupe). If you have a really good lab which you trust,
you can do a rough check of sharpness by looking at the prints, but don't
rely too strongly on prints other than the quick check.
Slides will give you a much better idea of the color rendition of the
lens, exposure accuracy, sharpness and so on. I'd shoot a few frames with
a known lens that you have, so you'll be able to have side-by-side
comparisons with a known constant, that will give you a useful idea of the
new lens' contrast and color. Use a good loupe to check for sharpness at
both centers and corners. You don't necessarily need to use Velvia, but I
would use a good ISO 100 film such as Provia, or any of the Ektachromes. I
used to use Kodachrome for tests, but the processing turnaround is now too
slow compared to the same-day service I can get with E-6 films.
--
Date: Mon, 23 Oct 2000
I have finally gotten my lens testing site, which I had to move last
December, back up and, I believe, operational. Please go to
href="http://wdn.com/~johnchap/lenstest/testlens.htm" for some
description and links to the instructions which you can download and
print out. Also Included in the material are the resolution charts
which are needed.
The material will instruct you how to set up the charts, shoot the
tests, and interpret the results. I have always found it very
enlightening how certain lenses do. The results you will get may very
well surprise you. I have found several inexpensive lenses with
excellent optical quality. On the other hand, I have found some very
expensive lenses that are not nearly as good as their much less
expensive brothers.
If you encounter any problems, have any questions, or have any advice
how I can improve access and usability of the site, please email me.
[Ed. note: somehow, I didn't get a chart attachment - filtered by
overzealous virus scanner?]
Hi
As you can see from the chart, final resolution of a lens will vary with
F-Stop and wavelength. In general, a normal name brand lens will do about
500 l/mm at about f-4 in noon summer sun. This will translate into about
80 l/mm using Tmax400 and about 125 l/mm using Tmax100. Films generally
follow a regression curve of ASA Vs resolution. All this can be imputed
into a computer to give various resolution values with various inputs of
film ASA, f-stop and time of day.
Larry
From: "Miro" [email protected]
The universal standard is the MTF diagram. And it is done at each f-stop.
Factors in a good lens.
================
1/ Resolution (sharpness)
If you consider that a $3,000 lens will last you a lifetime then it doesnt
seem to be such a bad investment if you want most of the things in my
list. The average camera will eat $5000 in film over 10 years (if you
include the prints) and that means you may as well go for good shots.
Then you come down to knowing which f-stop is the best balance. For many
Zeiss and Leica lenses they publish the data on how to use the lens. So
you always know that at 35mm and f5.6 you get max contrast and max sharp
for a specific lens.
The worst thing that happens to photos is in fact shutter shake, the
mirror hitting the chassis pre-exposure and bumping the whole camera. A
simple monopod may look nerdy but it will increase your lens sharpness
greatly.
I'm probably biased since I have played with many peoples toys and seen
the difference.
"woodsie" [email protected] wrote
Date: Wed, 11 Jul 2001
....
Simplest way is to photograph a newspaper tacked to the wall. Make
sure your camera is aimed directly, i.e., film plane is parallel to the
wall, and that the newspaper fills the frame as much as possible. Use a
tripod and a faster shutter speed to keep camera shake out of the
question. Flash is fine. Shoot at different distances, with your lens set
at widest aperture, then a couple stopped down a bit. Adjusting focus
slightly over and under can help too. Write down everything you did for
later comparisons.
The best thing is to use slide film for this, and examine it later
under at least a 6x loupe, 10x or more will help determine detail even
better. In a pinch, a 50mm lens held backwards works as a good 6x loupe.
Examine the center and corners of the frame - it's not uncommon for the
corners of any lens to be a little less sharp.
I say use slide film because you're not introducing the results from
your lab's enlarger. If their printing process isn't sharp, you'll never
get good results from print film. You can also try another lab with print
film shot under the same conditions, just to see.
You can also make a direct viewer with a piece of clear acetate, or a
frame of blank (clear) slide film, that's been sanded with very fine
sandpaper to make it milky. Open the camera back and place this directly
over the film opening, being careful of the delicate shutter, and lock the
shutter open on bulb. The sanding will have created a focusing screen, and
you can see exactly how sharp the image is when it hits the film. Again,
use a loupe. This will let you see whether there's a difference between
your viewfinder and your film. Make sure the viewer lays absolutely flat
on the film rails, and the sanding is towards the lens. Any lab that
processes slide film should be able to provide a clear slide film leader
trimoff.
Hope this helps. Good luck!
- Al.
From Nikon MF Mailing List:
[email protected] writes
Hi Scott,
I am by no means an expert in optics so I will try to keep things
simple.
The "finest possible detail" with "a minimum of fuzzyness" is provided
by several factors, including the lens, film, camera and their user.
The RESOLUTION of the lens, or of the film, or of the paper it's printed
on defines how small a detail can be satisfactorily represented.
Resolution is affected by CONTRAST; for example it's a lot easier to
resolve fine detail if the contrast between highlights, mid-tones and
shadows is high. If the contrast is low, with all tones in a narrow
band, resolving fine detail becomes difficult. A good lens will
transmit the image to the film with minimal loss of contrast, and will
make a much better job of representing low contrast detail. A poor lens
may not show low contrast detail at all.
Because resolution and contrast are so inextricably linked, lens
designers use a measurement called MTF (mass transfer function) which
indicates how much image information a lens can pass.
Other factors also influence the representation of small details, such
as ABERRATION. The term "aberration" includes a number of factors which
can degrade the ability of the lens to transmit information to the film.
For example, "chromatic aberration" can introduce colour fringing and
others such as spherical aberration and coma can reduce overall
sharpness. These aberrations can be partially, sometimes almost wholly
eliminated by good optical design, but too much correction of spherical
aberration may lead to a lens that provides good sharpness and contrast
but has a poor rendition of out of focus parts of the shot.
It's the subjective quality of the lens's rendition of out-of-focus
elements of the picture, especially out-of-focus highlights.
I could say a lot more but I was limited to "just a sentence"!!
Best regards,
--
Date: Wed, 30 May 2001
I couldn't find a resource on the web that describes a viable method
for rigorously testing barrel and pincushion distortion. I'll just
test at different distances. There is a lot of sample variation in
lenses -- Bob Monaghan is fond of saying that one lens model varies
more than two different models possibly do -- so test before buying.
For example, the lens I'm interested in -- Minolta 24-105/3.5-4.5 --
was tested as having 1.2% barrel to .90% pincushion by one magazine,
but 2.7% barrel to 1.3% pincushion by another magazine.
Joseph Meehan [email protected] wrote:
Trouble is, the USAF lens test chart has its most detailed resolution
lines in the center. However using a copy table, it should be easy
enough to move the USAF chart between the center and four corners.
Here is the USAF chart in Postscript format:
http://www.photo.net/photo/optics/USAF1951.ps
Date: Fri, 15 Feb 2002
From: Jan [email protected]
To: [email protected]
Subject: Re: [Rollei] Define "Contrast" and "Sharpness"
yes, a tautology is something tautological!
Id rather say:
contrast means measuring how dark(-grey) black subjects/areas) will be reproduced
and how light(grey) white areas/subjects will be reproduced, (good contrast lens will
reproduce black as black, not dark-/medium-grey, and white as white not light-/medium grey),
and sharpness/resolution refers to the finest pattern reproduceable disregarding the fact,
that black has turned into 49.5%reflection and white has turned into 50.5% reflection,
the finer the pattern got.
The rest in fact is best seen in examples as in Zeiss's propaganda pamphlets.
Jan
Dirk-Roger Schmitt wrote:
> o.k.
>
> very easy.
>
> You just need a graph of contrast vs spatial frequency or vs. spatial
> wavelength for each lens to discuss differences.
>
> That is the only way.
> you wrote:
> >Okay all you optics nerds. Here yer chance to show some of us lesser
> >scientific underlings a thing or two. My question is this-
> > What exactly is contrast, and sharpness. Explain how they relate to
> > each other in the context of a discussion of a lens - (example - the
> > differenc ebetween a id 50's Tessar, and a F series Type 4 Planar) -
> >Roberto
Date: Fri, 15 Feb 2002
From: Richard Knoppow [email protected]
To: [email protected]
Subject: Re: [Rollei] Define "Contrast" and "Sharpness"
... (quoting above..)
The spacial frequency or modulation transfer curves are important in
showing the difference. A high contrast lens has good edge contrast, a very
similar effect to what is called acutance in film. A high resolution lens
may have lower edge contrast. Visually, the high contrast lens will look
sharper even though it will have lower resolution.
The high contrast lens will typically have an MTF curve which is fairly
flat up to some value of resolution and falls off quickly above that. The
high resolution lens may start falling off sooner but will fall off at a
lower rate so that, at some high value of resolution, it has greater
contrast than the high contrast lens.
The energy distribution of the lens in the form of a strehl ratio, or
what in German is Light Mountain, shows the effect also. The high
resolution lens will have a narrow main peak but many side peaks. The high
contrast lens is apodized so that the main peak is somewhat wider but there
are no side peaks. This is probably a better measure of the effect than the
MTF curve.
The physical cause is the balance of diffraction and aberrations in the
lens. The spherical aberration curve with respect to image height can be
varied by the designer. It affects the maximum blur spot size. The
adjustment for minimum blur spot may result in zonal spherical.
Supposedly Zeiss and Leitz had different opinions as to what was the best
characteristic for pictorial photographic lenses. Also supposedly, Japanese
designers tend to compromise toward higher edge contrast and German
designers go for higher resolution. One would have to measure a number of
lenses to tell if any of this is true.
This is not new stuff. Kingslake illustrates it in a book on lenses he
wrote in 1948. _Lenses in Photography_ either edition, Rudolf Kingslake.
Generally, the better corrected a lens the better its contrast whatever
the compromise between contrast and resolution.
----
Richard Knoppow
Los Angeles, CA, USA
[email protected]
Date: Sat, 16 Feb 2002
From: caveteursus [email protected]
To: [email protected]
Subject: [medium-format] Re: Contax: 35mm Equiv. Focal Lengths for 645AF lenses?
Years ago, I puchased a USAF resolution target from Edmund
Scientific -- still have it. I find that most of my MF lenses (Zeiss
and Schneider) are quite comparable to the Leitz lenses -- so you
could mount a 150mm TeleXenar on a Leica or Contax and get a very
nice negative, although the expense of an adapater wouldn't make it
worth it.
BTW, those lenses best able to "resolve" are the 50mm DR Summicron,
50mm Summicron R, 55mmm 2.8 Micro Nikkor, 35mm Summaron (surprised
me!), 150mm APO Symmar and the 150mm TeleXenar. The ancient micro
Nikkor resolves close to 100 LPM at 0.5 meters, but mid 40's when
used in the true macro mode.
I know it's not Modulation Transfer (see the excellent articles on
Same in Photo Techniques), but I don't use Nikon's any more, save for
the F3 which I use for duping (75mm APO Rodagon) and copy work.
--- In medium-format@y..., "loslosbaby"
From: David Littlewood [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Diffraction and B&W
Date: Wed, 20 Feb 2002
Ken Durling
From: "Nick Thomas" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: The Sharpest Lens
Date: Thu, 21 Feb 2002
Michael Reichmann wrote
> Wondering what is the sharpest lens around?
Undoubtedly the lenses used in photolithographic steppers - how about
0.13 micron resolution? (something like 3800 lpm)
Absolutely no good for anything else however.
Nick Thomas
From: Tony Polson [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Lens testing realities - a bit shrill, but a good read.
Date: Tue, 19 Mar 2002
"Mxsmanic" [email protected] wrote:
>"Paul Rubin" [email protected] wrote ...
>
>> I'm skeptical that's true even for Zeiss
>> and Leica, but I don't use those lenses
>> and don't know.
>
>The last Zeiss lens I had contained a little sheet of paper in the box
>with the signature of the person who tested it.
Unless you are talking about your Dad's Hasselblad, you have been
conned. Royally.
The Carl Zeiss lenses for 35mm are all supplied with a signed test
certificate, but it is EMPHATICALLY NOT for the particular lens that
you bought. The only manufacturer of lenses for 35mm who has ever
done that was the original Angenieux.
From: Gary Frost [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Lens testing realities - a bit shrill, but a good read.
Date: Tue, 19 Mar 2002
In my experience there IS NOT a direct correlation between lens
price and lens performance. The most obvious example is a 50mm
prime for under $100 compared to a $1000+, 16+ element zoom.
There are many times a $100 lens can outperform a $1000 lens.
I have many other examples.....
The 'cheap' AF mount has much more alignment slop compared to
a good quality metal helical but the $1000 zoom with all those
elements in various zooming spacing is more likely to suffer from
sample variation and wear than the $100 prime.
Each lens really needs to be considered individually, regardless of
price.
For lens testing in general, the most relevent are the tests you
do yourself. I tend to agree though that much of the 'smoke and mirrors'
on the subject is manufacturers dodging the crap they put out.
In general I would like to see more hard data on lenses: distortion,
vignette, color abbr, MTF, etc. These things can be objectively
measured and published. There is still sample variation, but the
data is extremely relevent to lens performance. This data is provided
for most large format lenses.
Gary Frost
Mxsmanic wrote:
>
> "Meryl Arbing" [email protected] wrote ...
>
> > There are a lot of people who think they can
> > get $1000 worth of value out of a $100 lens
> > and they are the same who tend to discount
> > lens tests.
>
> True, but there are also people who like to believe that there are
> glaring differences between a $1000 lens from their preferred
> manufacturer and a $1000 lens from any other manufacturer, and they tend
> to _overemphasize_ lens tests in their desperation to find support for
> their preconceived notions.
>
> In reality, of course, you generally get what you pay for. A $1000 lens
> is going to be a lot better than a $100 lens, as a general rule, and
> this will be obvious even without a lens test. Conversely, two $1000
> lenses will tend to be very similar in quality, and it will be very
> difficult to find differences even with careful lens testing, and often
> the winner will be pretty much random, with little consistent
> correlation with any particular manufacturer (i.e., for a given pair of
> lenses, sometimes Leica might win, and sometimes Zeiss, and sometimes
> Nikon).
>
> Ultimately, the only _important_ differences are the ones you can see
> without a fancy lens test. So $1000 lenses are different from $100
> lenses, no matter what anyone might like to believe to the contrary,
> because it's easy to see the differences in ordinary photos. But two
> $1000 lenses are unlikely to be significantly different, because the
> only distinction between them is likely to be visible exclusively in
> precise lens tests, and not in real-world photos. The higher the price,
> the less variation there will be between brands of lenses. Cheap lenses
> can vary much more dramatically between brands, and even among different
> samples of the very same model.
>
> > In fact, there is a simple way out of the lens
> > testing dilemma. I just look for the name Zeiss
> > and don't worry about whether I', getting a "lemon"
> > or not...others will look for Leica...or Canon L
> > series...etc. and forget the 3rd party compromise
> > lenses where you have no idea what you are going to
> > get.
>
> This is an excellent technique. All of the top lenses from the top
> manufacturers are consistently excellent, so any of them will usually be
> fine for real-world work. Of course, they all cost a fortune, too. But
> you get what you pay for.
From: [email protected] (brian)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: cost of lpmm URL etc. Re: Lens testing realities
Date: 22 Mar 2002
Deep-UV stepper lens, 5000 line pairs/mm for only $1,000,000 ($200/linepair/mm)
High-power microscope objective: $10 (used) / 1000lp/mm = $0.01/lp/mm
Brian
[email protected] (Robert Monaghan) wrote...
> Chris Perez's cost of quality for lpmm is at:
> http://www.hevanet.com/cperez/proposal.html
>
> Kodak 620 Special w/100mm = $15(used) / 63 l/mm = $0.23 / line / mm
> Mamiya C220Pro w/80mm - $225(used) / 67 l/mm = $3.35 / line / mm
> Bronica SQA w/ 80mm - $800(used) / 67 l/mm = $11.90 / line / mm
> etc.
>
> while there are other things than lpmm by which I evaluate lenses, it is
> worth noting that there are some stunning buys out there in surprisingly
> decent buy undervalued cameras and lenses such as the 100mm f/4.5 kodak
> anastigmat cited above...
>
> fyi
>
> bobm
From: [email protected] (brian)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: cost of lpmm URL etc. Re: Lens testing realities
Date: 24 Mar 2002
David Littlewood [email protected] wrote
> brian [email protected] writes
> >Deep-UV stepper lens, 5000 line pairs/mm for only $1,000,000 ($200/linepair/mm)
> >
> >High-power microscope objective: $10 (used) / 1000lp/mm = $0.01/lp/mm
> >
> LOL.
>
> BTW, I wish I could get (good) high powered microscope objectives for
> $10. I am just hoping to buy one (a Zeiss 40/0.65, BTW) for �100 - about
> $145.
Hi David:
Actually I was thinking of an old Bausch and Lomb or perhaps Edmund
Scientific objective. Some of them are actually pretty good, and
certainly without the expensive cachet of Zeiss.
By the way, almost forgot another example: CD player objective, 1000
linepairs/mm (yes, really!), free if you want to dig up an old
computer or Sony Discman from a landfill. Actually, I think that
every participant in this NG already owns one. I've got at least half
a dozen that are laying around getting little or no use anymore. The
only catch is that it is monochromatic and covers a very small field
of view. Coma is corrected, but the field of view is limited by field
curvature and astigmatism to only a degree or two. It never ceases to
amaze me how many people actually own and use a diffraction-limited
f/1 lens without even knowing about it!! Infinitely better correction
than any 35mm lens ever has had or ever will have.
Brian
From contax mailing list:
Date: Wed, 23 Jan 2002
From: muchan [email protected]
Subject: Re: [Contax] On lens performance
Jasper tiong wrote:
> Can I get a "diploma" in optics if i can half
> understand 20% of the messages in "on lens performance
> " thread ? ;))
I confess, I'm less than 10%, so will fail the diploma.
I need a special "complimental" lesson, to translate these numbers
to the words I think I know.
In Shutterbug, once I read one of 10 advices from Roger Hicks,
"Learn the difference between resolution, contrast and acutance".
And I think I learned it, (but not to the level of diploma.)
And this thread, it seems me, is talking about these things in
mathematical numbers?
Please someone
(1) correct me if I'm wrong and
(2) probide me the "theoretical" version of following.
A. "Acutance", is how each "grain" is seen distinct from other neighbor
"grain", or from the "non-grain" area of the film/paper.
And it is basically attribute of the film, developping process (and printing
device and technique). Better the "accutance", the viewer feel that the
overall image as sharp, and contrasty.
-- always, the word "sharp" is unclear for me...
B. Resolution, is how much information a image has. Limited with the grain size,
but till the resolution reaches to grain size, lens's resolution decide the
image resolution. More resolution has an image, more details in small area you
can see. But high resolution with low contrast make the image looks "flat"...
C. Contrast, seems me very contradictive... it is the differnce between
bright part and dark part. When the dynamic range of the lens is narrow,
that is, if the contrast of the lens is low, the image would separate more
dominant white and black part, and (when accutance is strong) the viewer
may feel the image more "contrasty". (somewhere I read it is "Macro-contrast"
of the image). When the dynamic range of the lens is wide, that is, if the
contrast of the lens is high, the image will show more gradual tones between
the white and black, making overall image less "contrasty", but seeing the
small details, the small differences of darker and brighter part is also seen,
(micro-contrast) so improves the overall "resolution" of the image...
hmmmm, I think I'm very deeply confused.
D. From this (confused) point of view, if the lens has "high resolution", "low dynamic
range", and "high accutance" film/process is used, the scene looks very "contrasty",
and "sharp" for Nikon users... it is rather "edgy" than being "sharp"... for Contaxians?
Zeiss's designers are more conscious about "high dynamic range" of optics, that improves
the "micro-contrast" and so also the "resolution", which is "sharp" for different viewer,
but less "sharp" for the other viewers...
hmmmm, I think I'm also having deep prejudices...
Again, please someone correct the wrong part of above A, B, C, D, and tell me which
values, (like "10 lp/mm", "100 lp/mm", 0.005, "Rf=0.40", "0.64", "MTF=0.97") or terms
(like "white level", "black level") related to the A, B, C, D or to their "corrected
versions"!
muchan (confused and prejudiced)
From Rollei Mailing List:
Date: Mon, 25 Mar 2002
From: Richard Knoppow [email protected]
Subject: RE: [Rollei] Schneider - Zeiss debate
you wrote:
>> with all respect, these comparisons ain't worth nuthin, since
>> they are only
>> practical and without any figures and numbers or even scientific
>> test-patterns ;-)
>
>Jan, I fully agree. That's the reason why I wrote that the site owner was
>suggesting(!) that his(!) Schneider is contrastier than his(!) Zeiss ;-)
>
>I actually have tried a more scientific approach of my "bookshelf" test by
>including two USAF1951 test target and shoot from open aperture to f16. I
>was doing this with the Rolleikin (why wasting 120 when I can only evaluate
>35mm ;-). But during my attempts, I broke a spring in the Rolleikin counter
>when advancing the film. This left me really frustrated and in my
>frustration, I did not completely rewinded the film before opening the
>camera and my tests with the 2.8F Planar was wasted :(
>After that, I started "repairing" the counter but could not get this counter
>to count again :(
>I probably still have the results of the 3.5F Planar on film but I need to
>finish this roll first with the 3.5E3 and Vb(which is now taken apart for
>cleaning).
>
>> But to become a little more serious, I think you can't tell much
>> by judging scanned
>> prints of unknown proveniance. I'd need some Velvia (or Provia
>> 100F) slides and my
>> magnifier (or even better Technical pan negatives).
>> Jan
>
>What kind of magnifier do you suggest using?
>Siu Fai
Good Scotch (Oh, wait, that's pacifier not magnifier).
Tests like this tell you something about the system performance,
including film and any focus errors introduced by the finder and film
flatness, but not a whole lot about lenses, unless they are grossly
different.
Real lens testing should be done by examining the aerial image,
preferably using an electronic sensor of some sort although visual
examination using a very high quality low power microscope was the
traditional method for decades.
It takes very little defocusing to affect a 100 l/mm resolution chart.
The shape of MTF curves can be useful in analysing the type of aberration
a lens has and will show defocusing effects. But to be really meaningful
many MTF curves must be taken at various points of the image, from center
to edge, and at a variety of stops.
To really answer a question like Xenar vs: Tessar or Xenotar vs: Planar
would require plotting the spherical aberration, astigmatism, chromatic
aberration, etc., using an optical bench. MTF will also vary with color,
depending on the degree of chromatic correction of the lens.
Lenses also have an effect similar to acutance in film. If you look at an
MTF curve as a frequency response curve, as for a loudspeaker, its shape
will affect the appearance of the image. Some lenses start to fall off in
the midrange but fall off slowly, so that they still have output at high
resolution numbers. Some stay fairly flat in the mid range but fall off
rapidly after some point. The second type is apt to appear sharper to the
eye than the first kind, although its _limiting_ resolution may be much
lower.
The shape and cut off of the resolution curve is often tailored for
lenses used on electronic sensors to avoid aliasing.
The total resolution of a combination of lens and film can not be
calculated in any simple manner. The curves for the two must be combined by
convolving. The old rule of 1/t = 1/L + 1/F or using the square root of the
sum of the squares is _very_ approximate as applied to the limiting
resoluton but not really useful in giving any idea of the apparent sharpness.
On film resolution can be measured but the possible errors of the method
must be understood and, preferably, eliminated as much as possible. The use
of glass plates or a vacuum back is mandatory of film curvature is to be
eliminated. The glass reseau available on some models of Rolleiflex
certainly eliminates film curvature and wandering of the film plane, but
introduces a flat plate, which brings some errors along with it. This is
not a trivial problem.
On a related subject, many years ago Kodak built a special camera for
measuring film resolution (its described in the Journal of the Optical
Society of America somewhere). In about 1948 they discovered the special
lens used in this camera had substantial and unpredicted aberrations. A new
lens was designed with limiting resolution of 500 lp/mm, and all Kodak
films were re-measured. The result was a near doubling of the resolution
numbers for the films.
The lens is described by Rudolf Kingslake in on if his lens design books
and I think in another JOSA paper.
Also: A single number for resolution is quite misleading. For what limit
is it stated? For electronic transmission and filters the bandwidth is
often stated as the half amplitude value but for sharp cut off filters
limites for other amplitude values must be stated to characterise the
system. Optical systems seem to use something like 10% response rather than
50% response. It probably depends on the application. For visual
determination, using the usual bar chart, the limit is one of visual
contrast. A caveat, bar charts can display false resolution. One fellow who
posted to the large-format group, insists that any decent lens is capable
of 400 l/mm resolution in the center of its field. This is simply hogwash.
What he was seeing was some false resolution due, probably, to the
aberrations in the microscope system he was using to examine the aerial
image. Even calculated values for ordinary lenses are never even close to
this value.
----
Richard Knoppow
Los Angeles, CA, USA
[email protected]
From: "Q.G. de Bakker" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Questions about medium-format vs. 35mm
Date: Sat, 16 Mar 2002
Mxsmanic wrote:
> > ... and decreasing maximum attainable resolution
> > considerably (roughly halved every two stops you
> > close your lens down)!
>
> And how many MF lenses are truly diffraction-limited, even at small
> apertures? (The only case in which this would be an issue.)
Practically all!
Diffraction increases at quite a pace when stopping down. Any lens that will
be able of doing, say, a modest 80 lp/mm at f/8 (theoretical limit at this
f-stop is 200 lp/mm) will be reduced to a maximum of 28 lpmm at f/22.
Staying with this example, and in view of the fact that DOF standards do
assume 30 lp/mm as the limit for acceptable unsharpness, it is clear that
though at f/22 everything, near and distant, will be equally "sharp",
maximum (!) resolution will be below what was deemed acceptable...
Date: Tue, 09 Apr 2002
From: Robert Feinman [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Really nice explanation of sharpness and resolution
Take a look at
http://www.luminous-landscape.com/sharpness.htm
For a much better description (than mine) of what an inkjet needs to
reproduce to be sharp.
--
Robert Feinman, Ph.D
[email protected]
Panoramic Photographs: http://robertdfeinman.com
From nikon Mailing List:
From: "Richard Simmonds" [email protected]
Subject: RE: [Nikon] re:digital v film
Date: Wed, 27 Feb 2002
Suzy
Billboards are usually printed at 75dpi and some at 150dpi
Richard Simmonds
Gratitude Media
[email protected]
From: Ron Todd [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: How do I test lens resolution?
Date: Thu, 21 Mar 2002
...
This is the site:
http://takinami.com/yoshihiko/photo/lens_test/procedure.html
It is the property and work of Mr. Yoshihiko Takinami.
From: "roland.rashleigh-berry" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Understanding Image Sharpness
Date: Tue, 26 Mar 2002
I saw a very good article about this on the Internet
http://www.normankoren.com/Tutorials/MTF.html
From minolta mailing list:
Date: Tue, 30 Apr 2002
From: David Kilpatrick [email protected]
Subject: Re: Re: Minolta lens test?- Question once more
latruz at [email protected] wrote:
> So, after two days of discussion we are in the same point :(.
> Nobody knows nothig, except our own experiences, not very precize
> because of different conditions of taking pictures, films, labs, ect.
>
> My question is:
> 1. Does anyone in our group have equipment and metodology to perform
> tests (ONLY MINOLTA LENS!)? Maybe it shouldn't be very sofisticated -
> but with the same metodology and conditions?
> 2. If not - do we have any source to compare MINOLTA lens according
> to SCIENTIFIC (NOT from popular magazines) tests?
All of this is not strictly true anyway. In Europe the Technical Image Press
Association (of which I am no longer a member, since they hold their annual
meeting on my wife's birthday every year and that rather prevented me going)
have a deal with the Zeiss/Hasselblad optical facilities.
Zeiss test a batch of samples (I believe 3 or 5 examples) of each lens and
provide a fully detailed report on MTF, vignetting, geometry, colour
transmission etc.
Any magazine which is a member can buy a set of tests from them - for
example, all current 28mm f2 lenses from all makers - and then put these
into their own format.
I believe that Practical Photography in the UK uses the TIPA arrangement,
and that FOTO in Sweden does the same. I find FOTO's use of the lens tests
much better but occasionally a Swedish term confuses me. Swedish is like
Scottish anyway so 'bra' with a dot on the a is just 'braw' which is a
sensible word to use for a good lens.
You can tell the Zeiss/TIPA tests quite easily but the publishers will
reformat the graphs and give them their own style. I tend to trust these
tests. On the whole, Minolta come out well but Zeiss and Leica are often
ahead. Sometimes a Canon or Nikon lens will be better than the equivalent
Minolta, sometimes the other way round. Generally the Cosina, Soligor,
Tamron and so on show for what they are. Sigma can do well with specific
designs, Tokina seem pretty good generally. Recently the Voigtlander (Cosina
made) rangefinder lenses have been getting some incredible results.
Whether or not FOTO is still using TIPA data I do not know; if they are not,
then the tests look equally good and my guess would be that they are using
Hasselblad's labs. Certainly the Swedish FOTO group tests, which appear
almost every month and always cover a large batch of lenses, are superior to
anything I've seen in English language mags. If I had to buy the rights to
translate and reprint lens tests, this would where I would go.
David Kilpatrick
Minolta Club
http://www.freelancephotographer.co.uk/
From Minolta Mailing List:
Date: Tue, 30 Apr 2002
From: "ctgardener" [email protected]
Subject: Re: Minolta lens test?- Question once more
--- In Minolta@y..., "latruz" pawel.walczak@c... wrote:
> So, after two days of discussion we are in the same point :(.
> Nobody knows nothig, except our own experiences, not very precize
> because of different conditions of taking pictures, films, labs,
ect.
I wouldn't say that at all ... between photodo.com, photozone.de,
various tests at http://www.par.univie.ac.at/~bob/photo/lenses/ and
other sources you can occasionally find by searching (photo.net, etc)
as well as informed opinions here on this list, if you're reasonably
skilled at reading between the lines, it's really not hard to decide
between different lenses. If you have specific requirements (focal
length, speed, etc.) you're usually not going to find many lenses
that are so close in "deciding factors" (size, weight, price, minimum
focus distance, number of aperture blades, AF speed, lens speed,
sharpness - center sharpness, corner sharpness, sharpness wide open
across various focal lengths - distortion, flare-resistance, bokeh,
color reproduction, and whatever else I'm missing) that you really
need controlled scientific comparisons to see which one is marginally
better than the others. It takes a lot of effort to come up with
ratings like you see on photodo (not to mention the fact that one
person or group needs EVERY lens for a consistent comparison - and
then you're still subject to product variations), and when you're
done, all you have is an analysis of resolving capacity, which is
still only a starting point ... if I went by mtf charts, I would have
bought the Sigma 24mm over the Minolta 24mm, but other comments gave
me the impression I'd get better results consistently (in varying
lighting conditions) with the Minolta ... can't tell you why,
exactly, other than I remember some discussion about flare with the
Sigma ... can't even tell you I wouldn't be happier with the
Sigma ... but I did some research, made a decision, and it's done &
over with. I've gone through this exercise enough times to wish
there were mtf charts available for every lens, but they're not
necessary to making informed decisions; they're of limited use, and
asking anyone to do the tests is just asking far too much.
- Dennis
From minolta mailing list:
Date: Tue, 30 Apr 2002
From: "dseelyjr" [email protected]
Subject: Re: Minolta lens test?- Question once more
It is certainly true that we all only know our own experiences - even if our
lens was tested "objectively" (e.g., for MTF), that test really only holds for
that one lens. Extending that one test result to every lens of that model
manufactured and assembled at different times in different places is difficult,
even with statistical controls in place. I'd love to see a test of twelve
lenses of the same model, bought randomly at different stores / different times
to get some idea of the sample variance. Maybe some magazine could do this,
then sell the "slightly used" (but tested!) lenses with their corresponding MTF
curves, etc. And you're right that our experiences are imperfect; when I "test"
a lens I try to compare it to a known quantity, usually a lens I consider to be
"excellent", because otherwise I wouldn't know hwo to assign a particular lens
five stars, or four stars, or ... I think the real conclusion is that test
results can only be used as a guide. This isn't particularly satisfying, since
that means I may never really know that my particular lens is "excellent" or
just "very good", but if I'm satisfied with, does it really matter?
Dan
-- In Minolta@y..., "latruz" pawel.walczak@c... wrote:
> So, after two days of discussion we are in the same point :(.
> Nobody knows nothig, except our own experiences, not very precize
> because of different conditions of taking pictures, films, labs, ect.
>
> My question is:
> 1. Does anyone in our group have equipment and metodology to perform
> tests (ONLY MINOLTA LENS!)? Maybe it shouldn't be very sofisticated -
> but with the same metodology and conditions?
> 2. If not - do we have any source to compare MINOLTA lens according
> to SCIENTIFIC (NOT from popular magazines) tests?
>
> Pawel
From minolta mailing list:
Date: Wed, 1 May 2002
From: [email protected]
Subject: Re: Minolta lens test?- Question once more
Well, I did comment on exactly this yesterday (under the headline "Wanted
..."). Since then I have found a test chart website, which among other
charts contains one that should be useful for lens tests, although it is
said to be aimed at digital still cameras.
http://www.esser-test-charts.com/photo/index.htm
The one I am referring to is TE170. Somebody has experiences with such
charts?
regards,
ake
From: [email protected] (brian)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Resolution threshold
Date: 19 May 2002
Robert:
System performance at f/2 may well be film-limited on-axis, or in a
central image circle about 10mm in diameter, but certainly not in the
edges and corners of the frame. What is needed is an area-weighted
measure. I did check through my collection of lens prescriptions and
discovered that the 55mm/2.8 micro Nikkor will just barely resolve
650lp/mm near the optical axis if you toss out the violet and deep red
wavelengths in the MTF calculation. Off-axis is a completely
different story, of course.
Also, I wasn't aware that the Foveon 16MP chip has actually made it to
market yet. The physically larger 16MP Kodak chip is widely
available.
Brian
...
From: "Q.G. de Bakker" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: "35mm frame sized sensor"
Date: Sat, 25 May 2002
Mxsmanic wrote:
> That's why only amateurs buy Technical Pan, Kodachrome, and Velvia.
By the way: i've just unearthed an old post from my files in which Zeiss'
Kornelius Fleischer mentioned that testing at Carl Zeiss, in search for a
replacement for the Kodak Ektar 25 film they used at Zeiss as test film (200
lpmm), showed that Velvia (exposed at ISO 40) reaches 160 lpmm. Very good.
However, Kodak Portra VC, rated at ISO 160, is showing the same 160 lpmm!
I know, i know, one is a reversal film, the other isn't. But it illustrates
that the "slow = high resolution" dictum is rapidly losing validity.
From: [email protected] (TDuffy8486)
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Resolution vs Fstop
Date: 22 May 2002
Just an observation...I believe it was Joe Englander who, years ago, published
a test in Camera and Darkroom magazine of different format lenses (35, 2 1/4,
4x5) for apparent sharpness of enlarged prints.
An interesting by-product of the test was that the optimum fstop for resolution
for a 210mm apo-symar (probably still one of the sharpest lenses available for
large format) was f11. Resolution suffered when stopping down further. This is
quite a bit larger an fstop then people have been quoting in this thread.
No question that out of focus parts of negatives are much less sharp than
diffracted f90 shots, but, conditions permitting, we ought to be trying to
shoot toward the f11 end of the scale then the f45 side.
I find that for average outdoor shots, my 5x7 format negatives are signifcantly
sharper than my 8x10 shots because my lens is stopped down about 2 stops less
and therefore closer to the lens' optimum aperture.
Take care,
Tom Duffy
From: [email protected] (Richard Knoppow)
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Resolution vs Fstop
Date: Fri, 24 May 2002
...
This is an interesting table, thanks.
Just to stir things up further please note that the diffraction
limit varies with image angle, the resolution dropping off as you move
away from the center. The diffraction loss varies differently for
radial and for tangential lines. The reason is obviousl from
observation of the stop at angles from the optical axis.
For practical lenses the actual drop off does not quite follow the
formula for pure diffraction partly because of aberrations and partly
because of varying amounts of pupil magnification.
---
Richard Knoppow
Los Angeles, CA, USA.
[email protected]
From: [email protected] (Richard Knoppow)
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Resolution vs Fstop
Date: Sat, 25 May 2002
"Leonard Evens" [email protected] wrote:
>"Richard Knoppow" [email protected] wrote:
>
>> This is an interesting table, thanks. Just to stir things up further
>> please note that the diffraction
>> limit varies with image angle, the resolution dropping off as you move
>> away from the center. The diffraction loss varies differently for radial
>> and for tangential lines. The reason is obviousl from observation of the
>> stop at angles from the optical axis.
>> For practical lenses the actual drop off does not quite follow the
>> formula for pure diffraction partly because of aberrations and partly
>> because of varying amounts of pupil magnification. ---
>> Richard Knoppow
>> Los Angeles, CA, USA.
>> [email protected]
>
>Could you give a reference which gives estimates of the drop off from the
>center of the image?
>--
>Leonard Evens [email protected] 847-491-5537
>Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
For instance _Photographic Optics_ Allen R. Greenleaf 1950, New
York, The Macmillan Company
A chart appears on p.30 giving theoretical radial and tangential
resolving power for stops from f/1 to f/64 and for image half angles
from 0deg to 45deg.
Greenleaf gives approximate formula of:
Optical Axis (10^6)/(1.22 lambda F) in lines per millimeter
For extra-axial points
Tangential (10^6 cos^3 theta)/(1.22 lambda F)
Radial (10^6 cos theta)/(1.22 lambda F)
Where lambda is wavelength in mu
Theta is the half angle
F is the f/stop
Greenleaf does not give derivations for the formulae.
For example (from the chart, calculated for 589.3 mu)
f/8
0deg = 174 l/mm
10deg R=171 l/mm T=166 l/mm
20deg R=163 l/mm T=145 l/mm
30deg R=143 l/mm T=113 l/mm
40deg R=133 l/mm T=78 l/mm
Undoubtedly distorted pupils will affect this but I don't have a
specific reference for that.
---
Richard Knoppow
Los Angeles, CA, USA.
[email protected]
Date: Fri, 21 Jun 2002
From: Christopher Perez [email protected]
Subject: Re: Just a Cherry Picking Minute
Newsgroups: rec.photo.equipment.medium-format
What you say might be true of MF glass. And based upon what I've
v.recently seen, you're probably right. The exception being the newer
glass for Mamiya's 6 and 7-series rangefinders. Fuji's 690 lens also
appears to be a match for the Germans. That's recent history.
OTOH, I've inspected literally at 1000's of negatives testing lenses
against USAF resolution charts and can say, with firm conviction, the
following:
For Large Format work, modern (post 1970's) optics show no visible or
consistant difference in contrast or resolution between German and Japanese
optics. In fact, the most consistantly fine LF optics have come from Fuji
for the lenses I've seen.
Prior to that, Kodak appears to have made the finest commercially
available lenses in the world (during the 1950's).
Yes, I'm anal about optics. And I'm pleasantly surprised by Zeiss lenses.
Though I don't yet know now different it'll be to print from them compared
to Mamiya's 7-series, and rather doubt I'll be able to tell the difference
between them in practice... which is where this whole conversation began,
right? :-)
- Chris
Mike wrote:
> In general German optics have always had more contrast then Japanese
> optics
> due to the material used in making the glass. Higher contrast will also
> have the effect of giving the appearance of increased resolution.
>
...
From: John Garand [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: kiev lenses.. Re: marketing
Date: Tue, 25 Jun 2002
"Q.G. de Bakker" [email protected] WROTE:
>So why should third party testers (even though they have checked their
>procedures and equipment with the bureau of standards) be exempt from this?
They aren't exempt. And most will note that their results apply to
the particular item being tested and other examples may differ.
But then our options seem to be limited to (though I'm sure I've
missed something really obvious):
1. Believe the manufacturer's ad copy; or,
2. Believe that nothing is true; or,
3. Buy one of everything and test for ourselves; or,
4. Take the reported results of 3d party testers as the most
objective data readily available (though not relying on only one or
two reports if more are available).
Under the circumstances, it seems the rational thing to do is #4. Of
course we humans will frequently take those reports which agree with
our cherished notions as validating our position and ignore any
reports which disagree with our pre-determined decisions. :-))
Date: Sat, 15 Jun 2002
To: [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: re: magazine tests
In theory, Modern Photogr.'s lens tests have been done using the same
lens test charts and techniques during many years of testing (they were
merged with Pop Photo some years ago), by professional lens testers with
lots of experience (e.g., Mr. Art Kramer..), and so represent one of the
few data bases of lens test data which can be compared over the years.
They had a lab with lots of high $$ instrumentation, and did a variety
of tests in house such as contrast and precise focal length testing etc.
They also did strip down reports, sound level testing, and so on.
Nonetheless, I have the same suspicions at times when I see a sudden
out-lying data point in the middle of a test run, or one that is amazingly
high (like the rating of 91 lpmm for the Schneider 140-280mm zoom for
hasselblad at f/11 and 280mm (perhaps a typo, it was 81 lpmm at f/8, and
even that is amazingly higher than their tests of other Zeiss lenses).
However, they (Modern Photography) did address this issue in Oct. 1965,
when they tested some Konica lenses from same production runs and also
had them tested by the U.S. National Bureau of Standards. They viewed
the results as validating their conclusions and test methodology since
the same basic pattern of lens performance data in their view was reported
by both the outside US national lab and their USAF test chart procedures.
The other response might be that their results gained popularity as they
often matched what folks found when testing various lenses or using them.
Since I am arguing that lenses vary a lot, the validity of these single
or few test points is an obvious issue. But having just posted a series
of over a dozen comparisons of different tests of the same lens designs
by brand (Contax, Canon, Leica, Nikon, Minolta, Pentax, Chinon, Fujinon..)
I am impressed by the amount of variation suggested by all these tests out
there. Conversely, you can also see examples where the lenses are very
similar in performance and optimal settings (e.g., Konica) and tradeoffs.
In short, I gotta use the data I can get ;-)
grins bobm
From: "Kinon O'Cann" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Lens Quality--How Do You Rate It?
Date: Thu, 11 Jul 2002
The differences that are most visible are sharpness, particularly in the
corners, contrast, and color rendition. I've seen photos taken with Leica
lenses that show more shadow detail than some poor lenses. Yes, the
differences are slight, and in some cases impossible to see, but there are
visible differences.
"Jeremy 1952" [email protected] wrote...
> > There are real differences, but they are also slight. That slight
> > difference however can be important to those who see the difference.
> >
>
> You say that there are differences, but you do not state what those
> differences are. Not meaning to single out your particular response, but it
> is typical of what people say when asked why high-priced lenses are better.
>
> If you compared photos shot with a good quality OEM lens from Canon, Nikon,
> Pentax, Olympus versus the Carl Zeiss or Leitz product, WHOULD YOU ACTUALLY
> BE ABLE TO SEE A DIFFERENCE IN THE PHOTOGRAPH?
>
> What would that difference look like?
>
> Would the image be sharper? Would the colors "jump off the print," because
> they were so much more saturated?
>
> Using the example of a f/1.4 50mm normal lens, if I were to chuck my SMC
> Takumar ($75.00) for the $1,995.00 Leitz for the Leicaflex, what specific
> improvements would I get in my photographs (assuming that all other factors
> such as technique, film type, use of tripod and lens hood were all the
> same)??
>
> I would really like to know if there are any significant differences. I
> suspect that there are not. So, my question remains: Can one see a
> difference in the photo? What does one look for?
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Lens Quality--How Do You Rate It?
Date: Thu, 11 Jul 2002
"Jeremy 1952" [email protected] wrote:
>> There are real differences, but they are also slight. That slight
>> difference however can be important to those who see the difference.
>>
>
>You say that there are differences, but you do not state what those
>differences are. Not meaning to single out your particular response, but it
>is typical of what people say when asked why high-priced lenses are better.
>
>If you compared photos shot with a good quality OEM lens from Canon, Nikon,
>Pentax, Olympus versus the Carl Zeiss or Leitz product, WHOULD YOU ACTUALLY
>BE ABLE TO SEE A DIFFERENCE IN THE PHOTOGRAPH?
>
>What would that difference look like?
>
>Would the image be sharper? Would the colors "jump off the print," because
>they were so much more saturated?
>
>Using the example of a f/1.4 50mm normal lens, if I were to chuck my SMC
>Takumar ($75.00) for the $1,995.00 Leitz for the Leicaflex, what specific
>improvements would I get in my photographs (assuming that all other factors
>such as technique, film type, use of tripod and lens hood were all the
>same)??
>
>I would really like to know if there are any significant differences. I
>suspect that there are not. So, my question remains: Can one see a
>difference in the photo? What does one look for?
You ask interesting questions, and it is surprising that
this thread has not yet devolved into a "my brand is better
than yours" series of posts...;-) As an inveterate "tester"
of lenses (see: www.ferrario.com/ruether/slemn.html for
part of it - short the Canons, Leitz, Minolta, etc. and
the many "el-cheapo" lenses that were left out, for various
reasons...;-), I think that there are subtle differences
among brands' best lenses, as there are somewhat less subtle
differences among lenses within most brands' offerings...
As for the $2000 Leitz 50mm f1.4 vs. the Zeiss, Nikkor,
older Pentax, etc., I suspect that by f4 or so, you could
not tell the difference in prints or slides shot with any of
them, unless one brand's version had a noticeable color cast
or peculiar contrast characteristic. At f2, you may be able
to spot differences, but they may not always favor the more
expensive...;-) Much of pricing is marketing, and a high
price does not guarantee high image quality in a particular
type of lens (and the reverse...), but it can reflect
manufacturing consistency, or manufacturing difficulty (if
unusual, as for the f1 50mms, which are both very expensive
and not very good optically - a bad combination...;-).
Personally, I chose the Nikkor line many years ago since it
was (generally) affordable, offered a very wide range of
lenses (including many unique to the line), most lenses in
the line (even the "cheap" ones) were at least very good
optically (with few "klunkers", unlike with most other
lines in which some important focal-lengths were not well
represented), manufacturing consistency was good, and Nikon
tried to make most of the lenses compatible with all bodies,
old and new (including AF bodies). Other manufacturers'
individual lenses could sometimes be better than a
particular Nikkor, but in general, Nikon offered a wider
range of lenses of a higher average quality, with less
"obsolescence" than other lines did. Times change, and I
now have more respect for the Canon line than I did, and
less for some of Nikon's recent cheap offerings, alas...
(but I still wonder why people liked the pre-aspheric
Leitz rfdr offerings, though - I was not impressed with
some of these...;-). Brand choice is probably at
least 9/10 the result of marketing for most people, and
once made, the choice is often defended quite irrationally
(I write on the video NGs a lot, and compare Mini-DV
camcorders, and often experience first-hand the ill effects
of "brand religion" when people disagree with my findings,
however well supported they are...;-), so I expect this
thread may soon be filled with "Zeiss is better" or similar
comments...;-)
David Ruether
[email protected]
http://www.ferrario.com/ruether
Hey, check out www.visitithaca.com too...!
From: "Jeremy 1952" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Lens Quality--How Do You Rate It?
Date: Thu, 11 Jul 2002
...(above post quoted)
I became intrigued about this topic when I read an article posted on Robert
Monaghan's web site about a test run by one of the photography magazines
comparing a Leitz normal lens with a 1974 SMA Takumar and also with a
Minolta normal lens from that same period. Both the Minolta and the Pentax
lenses had significantly improved LPM over the Leitz offering.
So, I began to wonder, what was all the fuss about with regard to the Leitz
and Zeiss lenses? Without wanting to ignite a war over whose brand is
"best," I really would like to know if there is any way to objectively
quantify the alleged superiority of the German lenses. As you can see, most
of the responses were vague.
As for build quality, I can see where certain styles of photography (such as
war correspondent) might require extra heavy duty lenses, but I can buy a
lot of replacement lenses for the price of that Leitz f/1.4!!!
It appears to me that the margin of superiority may be slight, and comes at
a tremendous cost. Not that I have anything against Leitz and Zeiss, but I
can't seem to identify the quality (qualities) of their lenses that cause
some photographers to be so passionate about them.
If we can't define what it is that makes a high priced lens that much better
than a reasonably-priced OEM lens from one of the major manufacturers, how
can we justify spending all that money on them?
I suppose that I, too, am trying to justify my decision not to upgrade from
my 1970-era SMC Takumars. I don't require auto focus, and the screw mount
causes me no problem at all. My results seem fine, at least to me. I just
wonder what improvement, if any, there would be if I took the same shot,
say, on a Leicaflex with that $2000 normal lens?
I suspect that the photos might be streaked--from my tears shed after having
spent all that money for virtually nothing in the way of improved results .
. . :-)
Newsgroups: rec.photo.equipment.35mm
From: Steffen Kluge [email protected]
Subject: Re: Lens Quality--How Do You Rate It?
Date: Thu, 11 Jul 2002
According to Jeremy 1952 [email protected]:
>So, I began to wonder, what was all the fuss about with regard to the Leitz
>and Zeiss lenses? Without wanting to ignite a war over whose brand is
>"best," I really would like to know if there is any way to objectively
>quantify the alleged superiority of the German lenses. As you can see, most
>of the responses were vague.
Among the few photographers I know who switched to Leica from
other brands (mostly Nikon), the only commonly cited advantage of
Leitz lenses (besides lots of various little bit and pieces of
like and dislike here and there) was the out-of-focus rendering.
Bokeh or whatever they call it.
In terms of LPM there may be quite a few lenses by other makers
out there that beat Leitz lenses in center sharpness by some
margin, and there are probably a few that are just as good across
the field - for a lot less money. But most cheaper brands seem to
have a hard time eliminating those nasty out-of-focus artifacts,
the price of super sharpness, that show structure and pattern
where there shouldn't be any. I could imagine that a perfectly
smooth rendition of OOF areas greatly enhances the subjective
impression of depth in a photograph, and that seems to be what
Leica users are after.
For the record, I'm not partial to Leitz optics myself, I've
never owned anything Leica. I'm just guessing here.
Cheers
Steffen.
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Lens Quality--How Do You Rate It?
Date: Thu, 11 Jul 2002
...(quotes above post)
Yes, this aspect does vary, even within one lens
line - though I actually prefer "bad bokeh" for
the better DOF when I'm trying for universal focus,
and for shooting some kinds of images that depend on
"bad bokeh" (see www.ferrario.com/ruether/sunplant1.html).
So, one must not assume that "good bokeh" is always
good, and that "bad bokeh" is always bad...;-) I often
do not like the look of images made with "good bokeh"
lenses - things can look a bit diffuse and soft for
me...
David Ruether
[email protected]
http://www.ferrario.com/ruether
Hey, check out www.visitithaca.com too...!
Newsgroups: rec.photo.equipment.35mm
From: [email protected] (Stephen M. Dunn)
Subject: Re: How To Test My Lenses
Date: Fri, 19 Jul 2002
Bill Tuthill [email protected] writes:
$I've tried all but the Edmunds chart, and think that next time I'll use
$a newspaper taped to the wall. All you're really trying to decide is
$if a new lens is better than your old lens.
For an example of what the traditional test (newspaper taped to the
wall) can tell you, and some comments on things it can't, see my
newspaper-taped-to-the-wall comparison of three of my lenses at
http://www.stevedunn.ca/photos/writings/eflenses.html
--
Stephen M. Dunn [email protected]
From: [email protected] (ArtKramr)
Newsgroups: rec.photo.equipment.35mm
Date: 22 Jul 2002
Subject: Re: How To Test My Lenses
>Subject: Re: How To Test My Lenses
>From: "Alan Chan" [email protected]
>Date: 7/21/02
>
>I have not followed this closely so I might miss something. What I have
>noticed is that most people who so concerned about sharpness or resolution
>of their lenses, often missed other characteristics of their lenses which
>might affect the final output in practice. A sharper lens is not necessarily
>a better lens imo.
>
>> Yes - that is why I test. The results tell me if the lens
>> is substandard or not (a simple comparison of the two short
>> edges of subsequent frames, shot at the same wide stop and
>> focus, of the same distant detailed subject, tells me if
>> the lens is optically well-aligned, and if so, it is likely
>> also to be about as sharp as another good sample of the
>> same lens), and how it compares with other similar-FL
>> lenses... In the process of simple alignment, wide vs.
>> middle aperture sharpness, and corner vs. center sharpness
>> testing on film, you can also get info on distortion, flare,
>
Your post brings to mind a series of landmark optical tests that Kodak ran some
years back. They were known as the JND tests. JND stands for "Just Noticeable
Difference". Kodak was out to determine how small a difference in results the
human eye could detect. They made hundreds of prints and showed them to a wide
variety of viewers from professional photographers to amateurs to casual
snapshooters. The got the information they wanted, but something else emerged
that was even more important than the original test goals. They found out that
prints that were of higher contrast were perceived as being sharper even though
they were actually less sharp. Of course, in lens design we balance contrast
and resolution. If the greatest sharpness is wanted, it is at the price of
contrast and conversely. As a result Kodak designed the Commercial Ektars which
optimized contrast. Nobody noticed that they weren't all that sharp.
Arthur Kramer
Visit my WW II B-26 website at:
http://www.coastcomp.com/artkramer
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: How To Test My Lenses
Date: Mon, 22 Jul 2002
...(quoting above post)
I suspect that commonly, "sharpness" is taken as the
subjective combination of resolution and contrast...
In video this is a common problem - people have
trouble seeing that certain camcorders have (rather
obviously) lower resolution capability than some others
that have lower (and more suitable for general use)
image contrast. With lenses for stills, though, it is
preferable to maximize both contrast and resolution
(through good design), though ultimately there is a
need to trade off one for the other, and arrive at the
most ideal balance.
One example supporting what you report: I bought a
Rollei 35 with a 4-element Tessar-type lens that tests
not very high in resolution toward the image corners.
I sold it in favor of the model with a 5-element lens
that had obviously higher resolution in the corners.
My slides never had the "snap" of those shot with the
earlier camera, *even in the corners*. I sold the
"better" version and returned to using the "worse"...;-)
David Ruether
[email protected]
http://www.ferrario.com/ruether
From: [email protected] (john chapman)
Newsgroups: rec.photo.equipment,rec.photo.equipment.35mm
Subject: Lens Testing Instructions
Date: 28 Jul 2002
After a gap of several years I have reposted instructions for testing
lenses, as well as a resolution chart that can be downloaded and
printed for use. To access the kit go to:
http://members.cox.net/lenstestr1/lenstest.htm
Comments, suggestions, criticisms can be sent to [email protected]
From: [email protected] (john chapman)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: How To Test My Lenses
Date: 25 Jul 2002
I buy many, if not most, of my lenses used, generally off ebay. I
test every one of them. In every test I include my benchmark lens
(Nikon 50/1.4) to ensure consistency between lens tests. If the newly
purchased lens fails to meet what I consider minimum standards, I will
turn around and resell it, generally at what it cost me. I only buy
new lenses if they come with a generous return policy.
Besides general photographs, my tests include resolution targets shot
at three different distances. My procedures generally follow those
specified in the lens testing kits that were once sold by Modern
Photography before its demise. The use of these targets, rather than
newsprint, allows me to objectively measure differences between
lenses, rather than only being able to conclude that A is sharper than
B.
I frequently find, particularly with Tokina lenses, that my results
show worse results than the published tests by photo mags such as POP
Photo. While, as succintly articulated elsewhere, there are a number
of factors that make one lens better than another, my own view is that
if the damn lens cannot resolve (so I can read them with a 30x loupe)
those little lines at relatively high LPM both in the center and the
frame edge, then I do not want the lens. While I have not kept real
track of it, my impression is that those lenses that have not tested
well perform noticeably poorly and this poor performance is most
noticeable at the edges.
In some cases I have tested multiple samples of the same lens, or
variations of the same lens (for example, I think I have tested maybe
4 variations of the Sigma 28-70/2.8/2.6-2.800). There have been
neglible variations between these samples. One of the worst lenses I
have ever tested was a brand new Nikon 80-400VR. I returned that lens
to the dealer. The results were so poor that I can only hope my
sample was defected, rather than representative of that make and
model. Interestingly, POP gave it an OK optically.
"Meryl Arbing" [email protected] wrote
> This has always been a strange one for me. There are some people who have
> made an entire Internet career out of "poo-pooing" lens tests by not only
> photographic magazines...but manufacturers as well...and yet they seem to
> think that unstructured random tests by individuals somehow have some
> validity.
> I would ask you what you are going to do with the results of your "test"? Do
> you have other examples of the same lens to compare? (Beware the bugaboo of
> "sample variation"!!!) SO...do you buy 10 lenses at a time..test them all
> and return 9 once you have picked the 'cherry' or do you buy/test/return one
> lens after another until you have tested every identical lens in the
> store...or in the city?? What if you find the first lens you tested is the
> best...do you go back and demand THAT specific serial number!!?
> If a $300,000 piece of lens test equipment isn't accurate enough to produce
> meaningful results then what do we expect from a photo of a page from your
> local newspaper taped to the wall or a laser printed (8.5x11") "test
> chart"?? If controlled laboratory test are not consistent enough..why do
> people suggest that completely uncontrolled, "real-world" shap shots prove
> anything?
From Rollei Mailing List:
Date: Fri, 5 Jul 2002
From: [email protected]
Subject: [Rollei] FTM curves for the 2,8FX planar exist !!!
I recently picked the advertising brochure of the 2,8FX TLR. Now the
schematic cut-through diagram of the 2,8, 5-element 80mm planar is
correctly shown ; and the vintage 2.8 cut-through with a 2.8 Xenotar
also but things are now clear.
A good surprise : FTM curves, illumination and distorsion performance
are shown, Zeiss-style !!!!
BUT... Too bad, the diagrams are **not legended** !!! so nobody knows
for which aperture which results are obtained.
Well RUGers now that we almost have those data, this will help
re-starting one of our favourite controversies : is the 5-element 2.8
TLR planar better or not that his 7-element SLR brother ;-);-);-)
--
Emmanuel BIGLER
[email protected]
From: Bill Tuthill [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: How To Test My Lenses
Date: Thu, 18 Jul 2002
Jeremy 1952 [email protected] wrote:
> I don't believe that I've ever seen lens testing charts available for
> purchase (maybe I just never looked hard enough before).
>
> What type of test would be best, and where can the test charts be purchased?
Edmunds Scientific sells a good one.
If you have an inkjet printer, the Koren test pattern works nicely:
http://www.normankoren.com/Tutorials/MTF5.html
If you have a laser printer, photo.net has a PostScript USAF1951 chart.
I've tried all but the Edmunds chart, and think that next time I'll use
a newspaper taped to the wall. All you're really trying to decide is
if a new lens is better than your old lens.
From: "Meryl Arbing" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: How To Test My Lenses
Date: Thu, 18 Jul 2002
This has always been a strange one for me. There are some people who have
made an entire Internet career out of "poo-pooing" lens tests by not only
photographic magazines...but manufacturers as well...and yet they seem to
think that unstructured random tests by individuals somehow have some
validity.
I would ask you what you are going to do with the results of your "test"? Do
you have other examples of the same lens to compare? (Beware the bugaboo of
"sample variation"!!!) SO...do you buy 10 lenses at a time..test them all
and return 9 once you have picked the 'cherry' or do you buy/test/return one
lens after another until you have tested every identical lens in the
store...or in the city?? What if you find the first lens you tested is the
best...do you go back and demand THAT specific serial number!!?
If a $300,000 piece of lens test equipment isn't accurate enough to produce
meaningful results then what do we expect from a photo of a page from your
local newspaper taped to the wall or a laser printed (8.5x11") "test
chart"?? If controlled laboratory test are not consistent enough..why do
people suggest that completely uncontrolled, "real-world" shap shots prove
anything?
All that testing does is to measure how closely a particular item conforms
to the design specifications. It is cheaper to produce a product with loose
specifications because more of them will pass the loose requirements
expected of them...but you get a wide variation. Products with tight
specifications and rigorous testing have one major
advantage...consistency...you always know what you are going to get and it
becomes less and less necessary for individuals to try to do their own
testing.
"Jeremy 1952" [email protected] wrote...
> Many posters have commented that one should test one's own lenses, rather
> than rely upon reviews. Makes sense to me.
>
> I don't believe that I've ever seen lens testing charts available for
> purchase (maybe I just never looked hard enough before).
>
> What type of test would be best, and where can the test charts be purchased?
From: [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: How To Test My Lenses
Date: Tue, 23 Jul 2002
David and I have debated this issue several times years ago. I agree with
Art that most people tend to prefer high contrast to high resolution. I
think the reason is that high contrast simplifies the image slightly while
high resolution makes it slightly more complex. I tend to like simplicity
in graphic images. The differences can be subtle, of course. I'm one of
the people who values high contrast while David seems to be on the
resolution side. Yet we both use a lot of Nikkors which typically favor
contrast over resolution in the design tradeoff. Glad to see you've dusted
off the old 35mm cameras, David. I thought you had gone over completely to
video. Just kidding.
Fred
Photo Forums
http://www.photoforums.net
"Neuman - Ruether" [email protected] wrote
> [email protected] (ArtKramr)
> wrote:
>
> >>Subject: Re: How To Test My Lenses
> >>From: "Alan Chan" [email protected]
> >>Date: 7/21/02
> >>
> >>I have not followed this closely so I might miss something. What I have
> >>noticed is that most people who so concerned about sharpness or resolution
> >>of their lenses, often missed other characteristics of their lenses which
> >>might affect the final output in practice. A sharper lens is not necessarily
> >>a better lens imo.
>
> >>> Yes - that is why I test. The results tell me if the lens
> >>> is substandard or not (a simple comparison of the two short
> >>> edges of subsequent frames, shot at the same wide stop and
> >>> focus, of the same distant detailed subject, tells me if
> >>> the lens is optically well-aligned, and if so, it is likely
> >>> also to be about as sharp as another good sample of the
> >>> same lens), and how it compares with other similar-FL
> >>> lenses... In the process of simple alignment, wide vs.
> >>> middle aperture sharpness, and corner vs. center sharpness
> >>> testing on film, you can also get info on distortion, flare,
> >>> etc. by using simple VF tests.
> >>> (See: www.ferrario.com/ruether/slemn.html.)
> >>> DR
>
> >Your post brings to mind a series of landmark optical tests that Kodak ran some
> >years back. They were known as the JND tests. JND stands for "Just Noticeable
> >Difference". Kodak was out to determine how small a difference in results the
> >human eye could detect. They made hundreds of prints and showed them to a wide
> >variety of viewers from professional photographers to amateurs to casual
> >snapshooters. The got the information they wanted, but something else emerged
> >that was even more important than the original test goals. They found out that
> >prints that were of higher contrast were perceived as being sharper even though
> >they were actually less sharp. Of course, in lens design we balance contrast
> >and resolution. If the greatest sharpness [resolution] is wanted, it is at the price of
> >contrast and conversely. As a result Kodak designed the Commercial Ektars which
> >optimized contrast. Nobody noticed that they weren't all that sharp.
> >Arthur Kramer
> >Visit my WW II B-26 website at:
> >http://www.coastcomp.com/artkramer
>
> I suspect that commonly, "sharpness" is taken as the
> subjective combination of resolution and contrast...
> In video this is a common problem - people have
> trouble seeing that certain camcorders have (rather
> obviously) lower resolution capability than some others
> that have lower (and more suitable for general use)
> image contrast. With lenses for stills, though, it is
> preferable to maximize both contrast and resolution
> (through good design), though ultimately there is a
> need to trade off one for the other, and arrive at the
> most ideal balance.
> One example supporting what you report: I bought a
> Rollei 35 with a 4-element Tessar-type lens that tests
> not very high in resolution toward the image corners.
> I sold it in favor of the model with a 5-element lens
> that had obviously higher resolution in the corners.
> My slides never had the "snap" of those shot with the
> earlier camera, *even in the corners*. I sold the
> "better" version and returned to using the "worse"...;-)
>
> David Ruether
> [email protected]
> http://www.ferrario.com/ruether
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: How To Test My Lenses
Date: Tue, 23 Jul 2002
...(quotes above)
It is good to hear from you! (We used to correspond
rather often, back a-ways...) I'm not sure I quite
agree with some of the above, though...;-) As I pointed
out in my response to AK (with the Rollei 35 examples,
from about 20-25 years ago...), I prefer(ed) sharpness,
which is the combination of both resolution and contrast,
and when there is a distinct contest/tradeoff between
these (as with the two best Rollei 35 lenses, the
Tessar-type and the Sonnar-type [the Triotar was
despicable! ;-]), I also favor(ed) good contrast - but I
think a lens would have to show VERY, VERY high contrast
combined with VERY, VERY low resolution to actually
contribute much to the simplification of an image (and I
do like nicely-rendered textures...;-). Looking at your
fine, graphically-simplified images, I think it is choice
of lighting and framing that accomplishes the formal
simplification you seek, not the lack of lens
resolution...;-) As for the rest, I shoot video for fun
now, (and profit), with stills reserved for profit only,
but I keep up with matters of stills, have still a lot
to blab about in still-photography, and add to the reviews
of lenses (see the 17-35/18-35, and the revised 24-120
reviews, under "I babble" on my web page - and I update
the "SUBJECTIVE Lens Evaluations [Mostly Nikkors]"
article often), and I'm still changing my working methods
for shooting stills (an eye problem forced some of this,
darn!) and learning new things (always the hard way,
alas...;-), and writing again here (in very long, run-on
sentences) after being off for a while with video
interests...;-)
David Ruether
[email protected]
http://www.ferrario.com/ruether
From: "William E. Graham" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: How To Test My Lenses
Date: Tue, 23 Jul 2002
[email protected] wrote...
> David and I have debated this issue several times years ago. I agree with
> Art that most people tend to prefer high contrast to high resolution.
Yes....I have a cheap program called Adobe
Photodelux....About the only thing it's good for is changing
the brightness and contrast of my pictures, and eliminating
"red eye". I notice that I bump up the contrast of almost
everything I run through it by about 10%. You might say it
turns my $200 lenses into $2000 lenses.....
From: "Jeremy 1952" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Lens Quality--How Do You Rate It?
Date: Fri, 12 Jul 2002
> Does the lens give you the image quality that YOU want? If so, it is a good
> lens!!!
That certainly is everyone's bottom line, but I was attempting to get a list
of features or empirical measurements that could be used in comparing one
brand of lens against another.
It is very difficult to determine what level of image quality is "good
enough" without comparison with other lenses.
As you can see from most of the responses, the typical respindent says that
expensive lenses are "better," but is apparently unable to specify exactly
WHAT is better.
Assuming that build quality is not an issue (i.e. the lens is not going to
be used in a stressful environment), we are left with optical quality. It
appears that the end products of these lenses--the photographs--are
virtually indistinguishable from each other, at least at "normal"
enlargement sizes.
Using my previous examples of a SMC Takumar 50mm f/1.4 ( $75.00 on eBay)
versus its Leitz equivalent ($1,995.00) it does not appear that the majority
of photographers would be getting significantly improved results with the
Leitz offering--even though it is priced 30 times more than the Takumar.
(I really wish that a Leitz or Carl Zeiss lens designer could give us
his/her point of view on this. Maybe we're missing something.)
Even comparing a current new Pentax, Minolta, Canon or Nikon lens against
the hyper-priced German lenses will probably reveal that the end product is
much the same quality--for a lot less money (I'm speculating about this, but
I do believe it to be true.)
I wish that someone could just point to two photos, side-by-side, and say,
"HERE is why the German lens is better than the
Pentax/Canon/Nikon/Minolta/etc. equivalent."
At this point, it sure seems that the small margin of improvement--if indeed
one exists--costs an awful lot more. If money is no object, then that's
fine. But for most of us who have mundane things to pay like mortgages and
auto expenses, it doesn't seem as though the German lenses give us much bang
for the buck.
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Lens Quality--How Do You Rate It?
Date: Fri, 12 Jul 2002
"Jeremy 1952"
From: "Jeremy 1952" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Lens Quality--How Do You Rate It?
Date: Fri, 12 Jul 2002
...(quotes above posting)
OK, I accept your explanation that it would be difficult to compare two
lenses unless the subjects were exactly identical. BUT, that is what I
meant when I said that I wished someone could point out the differences
between lenses in two side-by-side photos. I was envisioning having two
photos, shot of the same subject, shot at the same time, with camera mounted
side-by-side, using the same type film.
My point was that I expected to notice little, if any, difference between
the two photos.
Contrast that with another scenario--the expensive German lens versus a
really cheap, no name lens (dare I suggest Quanteray, as a possible
example?). Take photos of the same subject side-by-side and produce 11x14
enlargements of each. Do you think that there would be a discernable
difference between the two photos? I suspect that one would, indeed, notice
some difference.
Whether one would elect to pay a lot more for the German lens, given the
amount of difference in the quality of photo pruduced, is a matter for the
user to decide. But, I suspect that once a photographer uses a
Pentax/Nikon/Canon/Minolta/etc OEM lens, he or she will not be seeing much
of a difference between that and a Leitz or a Zeiss equivalent.
I admit that there are differences in quality, even between the run of
lenses by the same manufacturer. But, one would hope that, for the money
charged, there would be some discernable consistent quality difference (one
that is MEASURABLE) between a $2000 lens and a $300 one.
Now I'm beginning to wonder whether there is all that much difference
between the Nikon/Canon/Pentax/Minolta lenses and the
Quanteray/Tamron/Sigma/etc. But that's another subject altogether :-)
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Lens Quality--How Do You Rate It?
Date: Fri, 12 Jul 2002
...(quotes of above thread..)
>OK, I accept your explanation that it would be difficult to compare two
>lenses unless the subjects were exactly identical. BUT, that is what I
>meant when I said that I wished someone could point out the differences
>between lenses in two side-by-side photos. I was envisioning having two
>photos, shot of the same subject, shot at the same time, with camera mounted
>side-by-side, using the same type film.
This would be nice, but to be interesting, it should
include, say, all the major mfgrs' 50mm lenses of all
speeds, and would be difficult to put this whole
shebang together - and it would still not necessarily
show optimum, or even average, samples of each lens
(though at f4 and smaller, this is unlikely to cause
problems...). I can't do the above, and I think few
others have the resources to do it, and I suspect that
with 50mm lenses for 35mm, you are right in that you
would not see any important differences. This would
likely change with some particular FLs, like 14-15mm
lenses, or lens types, like 28-85s, though...
>My point was that I expected to notice little, if any, difference between
>the two photos.
Yes, I have agreed with this...
>Contrast that with another scenario--the expensive German lens versus a
>really cheap, no name lens (dare I suggest Quanteray, as a possible
>example?). Take photos of the same subject side-by-side and produce 11x14
>enlargements of each. Do you think that there would be a discernable
>difference between the two photos? I suspect that one would, indeed, notice
>some difference.
Possibly, particularly at the edges and corners of
wide-angles and short-to-medium-FL zooms. With these,
even stopped down to f11, you may well see differences
(but those differences may not always be what you expect).
Even with the easier-to-design slower tele zooms, the
"cheap-brand" ones rarely have the wirey, "snappy" image
of the best Nikkors.
BTW, once again, "German" lenses are often "Japanese"
in reality...;-) "German" does not necessarily indicate
"better", though the German brands are often considerably
more expensive. Maybe "expensive" vs. "inexpensive"...?
>Whether one would elect to pay a lot more for the German lens, given the
>amount of difference in the quality of photo pruduced, is a matter for the
>user to decide. But, I suspect that once a photographer uses a
>Pentax/Nikon/Canon/Minolta/etc OEM lens, he or she will not be seeing much
>of a difference between that and a Leitz or a Zeiss equivalent.
I think this depends more on FL and lens types. For
instance, a couple of years ago I bought a used
Minolta Maxxum system, with 28mm, 50 macro, 35-105,
Even with the easier-to-design slower tele zooms, the
"cheap-brand" ones rarely have the wirey, "snappy" image
of the best Nikkors.
BTW, once again, "German" lenses are often "Japanese"
in reality...;-) "German" does not necessarily indicate
"better", though the German brands are often considerably
more expensive. Maybe "expensive" vs. "inexpensive"...?
>Whether one would elect to pay a lot more for the German lens, given the
>amount of difference in the quality of photo pruduced, is a matter for the
>user to decide. But, I suspect that once a photographer uses a
>Pentax/Nikon/Canon/Minolta/etc OEM lens, he or she will not be seeing much
>of a difference between that and a Leitz or a Zeiss equivalent.
I think this depends more on FL and lens types. For
instance, a couple of years ago I bought a used
Minolta Maxxum system, with 28mm, 50 macro, 35-105,
and 100-300 Minolta lenses. When tested, the macro
was fully the equal of the Nikkor equivalent, but
the other three lenses were niticeably inferior in
both resolution and contrast compared with the
similar-type Nikkors. All looked good in prints, but
the negatives, examined under a 10X magnifier, showed
clear differences in quality between most of the
lenses compared of the two brands... In other words,
a Minolta owner could be happy with all of the lenses
checked, but if one were trying to own the "best",
most clearly fell short of the Nikkors...
>I admit that there are differences in quality, even between the run of
>lenses by the same manufacturer. But, one would hope that, for the money
>charged, there would be some discernable consistent quality difference (one
>that is MEASURABLE) between a $2000 lens and a $300 one.
Marketing is everything, and it separates people
from their money. No other assumption about
pricing and quality differences should be inferred...;-)
>Now I'm beginning to wonder whether there is all that much difference
>between the Nikon/Canon/Pentax/Minolta lenses and the
>Quanteray/Tamron/Sigma/etc. But that's another subject altogether :-)
See: www.ferrario.com/ruether/slemn.html ...;-)
More: as I pointed out earlier, optical quality
across the line and breadth of offerings was
probably greatest in the Nikon line, though other
main-line brands produced lens lines that were
certainly good enough for most purposes (but
sometimes were without good lenses in important
particular FLs and speeds), and sometimes had
particular outstanding lenses. The "off-brand"
manufacturers often produced particular lenses
that approached the performance of a few of the
best of the main-line offerings, but the line
consistency was not good enough to depend on
them for all lenses of interest. Currently, the
pressures of the need to sell quantities has led
Nikon to offer lower-than-normal-for-Nikon lenses,
while Canon and Leitz are finally living up to
their marketing-hyped reputations - and some
few "off-brand" products are finally about as
good as the best by others... The above is a
bunch of generalities, but it may give you
some bit of an answer to your question. With
particular types of lenses, especially ones that
are unusual or difficult to design/build, you
may see obvious optical quality differences
(a complex issue in itself...), but with simpler
lenses, the differences may be subtle or even
invisible except under very close examination,
and their importance depends on the particular
tastes and preferences of the examiner...
In other words, compare 10 different 50mm f1.4s,
85-90mms, or 100-105s at f4 with essentially flat
subjects, with all other conditions the same,
and, regardless of price, you probably will not
see any obvious differences; compare 14-15mms,
full-frame fisheyes, wide-range wide-to-tele
zooms, etc., and at f4 or even f8 you will
probably see obvious differences, but higher
price may not buy higher image quality (it may
bring the reverse...).
David Ruether
[email protected]
http://www.ferrario.com/ruether
From: [email protected] (ArtKramr)
Newsgroups: rec.photo.equipment.medium-format
Date: 13 Jul 2002
Subject: Re: hassy gear going, going.. Re: Why has no one improved on the
>Subject: Re: hassy gear going, going.. Re: Why has no one improved on the
>From: John Stafford [email protected]
>Date: 7/13/02
>Some of the rarified group who by nature are interested only in sharpness
>seem to have an extra sensory nerve extending from the eye, to brain to
>anus.
Kodak did an extensive set of tests using thousands of viewers regarding
sharpness. They concluded that contrast was more important to a sense of
sharpness than sharpness itself. Thus Kodak Commercial Ektar lenses were
designed with an emphasis on contrast at the cost of sharpness. But you are
right. With modern lenses and films. sharpness has ceased to be a problem.
Arthur Kramer
Visit my WW II B-26 website at:
http://www.coastcomp.com/artkramer
From: fotocord [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: at least i'm happy
Date: Mon, 08 Jul 2002
Godfrey DiGiorgi wrote:
> Seems like an awful lot of boring work to little purpose. If you just went
> out and shot a few test rolls of real photos, you'd find out the same
> stuff.
But you've thrown other variables into the mix. And it's easy to see more
DOF as "better" even though sharpness may have been lost.
>I often get a lot of very nice photographs doing test rolls,
> learning a lens. Why waste the film on resolution charts when you *might*
> get something far better?
Seeing as how we are talking about ~ a half a roll for each lens, that
isn't much of a waste IMHO. Since you process your own B&W like I do, what
are we talking about $1.50 a lens? Cost is no object on buying a camera but
now you're worried about $1.50?
from hasselblad mailing list:
Date: Thu, 25 Jul 2002
From: Henry Posner/B&H Photo-Video [email protected]
Subject: [HUG] Re: Auto Focus or Focus Confirmation
you wrote:
> My experience is would tend to validate your thoughts in that some lens
>do better at small F/stops than others. Over the years I have taken photos
>with my Canon macro lens at F 32. I have a series butterfly photographs
>that are enlarged to 11"x14". They were taken at F 32 with my Canon 100MM
>micro.
Unless you're examining them side-by-side with similar images
taken at other aperture/shutter speed combos which resulted in the same
exposure, you cannot really say that you've achieved optimum image quality.
And, there's other things besides aperture which determine overall image
quality and sharpness. A 1 stop underexposure with the best lens will
compare unfavorably with the same hardware and the proper exposure even at
the smallest aperture
> People who view them almost always remark how sharp they are.
People who eat my mother's cooking always say how yummy it is, but
frankly she's a mediocre cook who sets a very convivial table. Either
they're being polite or they're enjoying the evening overall. Who else puts
sweetened cranberry relish in meat balls?
--
regards,
Henry Posner
Director of Sales and Training
B&H Photo-Video, and Pro-Audio Inc.
http://www.bhphotovideo.com
from hasselblad mailing list:
Date: Sat, 29 Jun 2002
From: David Gerhardt [email protected]
Subject: Re: lenses to lust after test Re: [HUG] Pictures with 250/350
"Ing. Ragnar Hansen AS" [email protected] wrote:
> I am sorry to not be able to put some pictures from this lens on the net, but
I do not have a scanner, and the quality of the pictures from this lens is best
seen on slides or copies. As Ernst Wildi said to me some months ago; "I think
there is something missing when I look at digital photos."
Especially if you're looking at web-based image files that are typically on
the order of 1-2MB. I doubt if one could tell the difference between an
average 35mm image and one (similarly composed) from a 250SA. In general,
images on a computer screen can look very good; and then when you print them
you see why 1MB isn't enough data for "photo-quality". ...And most web
sites shy away from image files of 40+ MB in size.
Sorry for the rambling; just don't expect to see why "SA" lenses cost so
much on files formatted for web browsers ;-)
--
David Gerhardt
[email protected]
From zeiss interest mailing list:
Date: Fri, 21 Jun 2002
From: [email protected]
Subject: Re: Re: Lens testing
Just a very brief reposte to Bill's advice.
The interest of doing some testing of the classic lenses resides, I think, in
being able to make comparisons outside one's own sphere of reference. Thus
uniform standards and methods must be involved at some point, and therein
resides the interest of test targets, as otherwise useless and uninteresting as
they may be. Thus, saying it differently, a single person's criteria may be met
by subjective testing, satisfying specific needs and based on individual
criteria. But the daunting obstacle to Bill's (and my) quest for group-wide
lens testing, where several of us could share the load and expense and time, is
that there is little way to make meaningful across-the-board comparisons from
one person to the next. There has to be an element of unity somewhere.
Give me a method to overcome this obstacle and let's take it from there. The
topic is of immense interest to me, andjudging by past comments, to many of us.
Yours in Zeiss
charlie
"William B. Lurie" [email protected] writes:
>Sean, John's points are all well taken. In truth, testing
>your lenses in a system where you test the lens
>exclusive of everything else that is involved with
>making a finished picture, is extremely difficult, and
>involves techniques and equipment far beyond
>accessibility for any one photographer. In addition
>to the lens, you have its focussing arrangement and
>the film, as a minimum.
>
>But isn't it really the result that you'd like to evaluate and
>compare? The finished product? It certainly isn't a
>mathematical or scientific evaluation, but if you plan and
>take a bunch of pictures, using your best methods, and see
>how you like them, that should tell you whatyou want to know.
>
>As for test charts, they exist in black and white, and with
>grey scales also, and these can be included in your test
>pictures.** �Some people use newspapers as test objects,
>because the letters come in a variety of sizes. As for film,
>I personally favor slide film if you are looking for detail
>and sharpness. In the past, I used to use a low-power
>microscope to examine the image on film. Even a hand-held
>jeweler's loupe will tell you an awful lot.
>
>In conclusion, I apologize. I've ranted and raved on this
>very subject before, and tried to get this group interested
>in doing something constructive, as a group, and gave up.
>Maybe somebody younger will pick up the baton, some time.
>
> � � � � � � � �Bill Lurie
>
>Reference: U. S. Department of Commerce
> � � � � � � � � � �National Bureau of Standards
> � � � � � � � � � �NBS Special Publication 374
>
> � � � � � � � � � �"Method for Determining the Resolving
> � � � � � � � � � � �Power of Photographic Lenses"
>
>Granted, there is more to it than resolving power.
>MTF testing adds something, but it is beyond the
>scope of any amateur. And, according to authorities
>on this subject, resolving power represents 98 to 99%
>of what you really care about, and is well within your
>scope to test.
From: [email protected] (ArtKramr)
Newsgroups: rec.photo.equipment.35mm
Date: 06 Sep 2002
Subject: Re: Popular Photography
>Subject: Re: Popular Photography
>From: "William E. Graham" [email protected]
>Date: 9/5/02
>
>One of the big problems with Consumer Reports is that they are
>generalists, and specialize in nothing. But, if you realize this, and
>carefully read the criteria that they are using for their evaluations of
>any given product line, you can frequently get some useful information
>out of them. I have used them to good end in the past for things like
>choosing the paint to use when recovering my deck (over $200 worth of
>paint), and the built-in dishwasher to replace my old one....In other
>purchases, I might read their evaluations, but only with a huge grain of
>salt.....
>"John" [email protected] wrote
>> NickC [email protected] wrote:
>> >
>> >I wrote to Consumer Reports and told them of my findings and never got
>> >an answer. I called the company that made the U210 and was informed
>> >that they too saw the article and had written a protest. Consumer
>> >Reports did nothing. I canceled my subscription and never bought
>> >another mag from them.
>>
>> They're the ones that called Kodak's Ektar lenses "mediocre
>> to poor" . Some of the best lenses ever made.
>>
>> Regards,
>>
>> John S. Douglas - Photographer, Webmaster & Computer Tech
>> Website --- http://www.darkroompro.net
Theres' a reason why they did that. Not an excuse a reason, Y'see they dindn't
know any better,. The Commercial Ektar designs were the results of Kodak's JND
research which showed that lenses that exhibited high contrast (at the cost of
sharpness) were percieved as being sharper than lenses that actually were
sharper but had lower contrast. Using this research, Kodak designed the
Commercial Ektars and the rest is history. But consumer Reports tested only
for sharpness not taking contrast into consideration. Thus the conclusion
they reached. But photographers all knew better.
Arthur Kramer
Visit my WW II B-26 website at:
http://www.coastcomp.com/artkramer
From: [email protected] (Ted Harris)
Newsgroups: rec.photo.equipment.large-format
Date: 11 Sep 2002
Subject: Re: Rodenstock APO SIRONAR S 150mm F5.6
To further add to what Steve said. Quoting from the book "Image Clarity: High
Resolution Photography" .....
"Sharpness is an aspect of image quality quite different from resolution.
Whereas resolution is determined by how small image featheres are, sharpness is
determined by how distinct their outlines are. ..... The percerption of
sharpness is influenced mainly by edge contrast sometimes called edge
definition .....
For more detailed discussions look at the highly technical discussions and
mathematics of accutance.
Cheers,
Ted
Ted Harris
Resource Strategy
Henniker, New Hampshire
From: James Horn [email protected]
Newsgroups: sci.optics
Subject: Re: USAF Test Bar Target
Date: 17 Sep 2002
http://www.efg2.com/Lab/ImageProcessing/TestTargets/
We use 'em here at Wescam all the time. Delighted to help!
Jim Horn, WB9SYN/6
From: "Paul Meier-Wang" [email protected]
Newsgroups: sci.optics
Subject: Re: USAF Test Bar Target
Date: Wed, 18 Sep 2002
Try Sine Patterns in Pittsford NY
http://www.sinepatterns.com
...
From: Bertho Boman [email protected]
Newsgroups: sci.optics
Subject: Re: USAF Test Bar Target
Date: Fri, 20 Sep 2002
[email protected] (pooja) wrote:
>Hi Bertho,....
>
>Kindly mail me the Target to the following email id.
>
>[email protected]
>
>or [email protected]
Will do!
I added the USAF 1951 Resolution target in AutoCAD format on my
website. There is also a link to a site that has a ton of
information on the subject. (And everything else about photography
too)
See:
http://www.vinland.com/USAF-1951.html
Bertho
From: Bertho Boman [email protected]
Newsgroups: sci.optics
Subject: Re: USAF Test Bar Target
Date: Wed, 18 Sep 2002
Jim Klein [email protected] wrote:
>Hi,
>
>I need the specifications forn the USAF Test Target (or a big version
>of one). This is the target with the bars and numbers on it.
>
>I'm going to add it so it can be pixelated at arbitrary resolution and
>used as an object which can be imaged in Roadrunner. Who do I call to
>get this?
>
>Jim klein
I made an accurate USAF-1951 target in AutoCad that I can email you if
interested.
The bars and spacing are correct, If you want an oversized one, you
can scale it to any size since it is in a CAD format.
If anyone has a copy of the actual specifications, I would like to
obtain a copy. I am curious about the specification for the text
part.
Bertho Boman
From: "Optiker" [email protected]
Newsgroups: sci.optics
Subject: Re: USAF Test Bar Target
Date: Thu, 19 Sep 2002
If you have an older copy of the Photonics Spectra Photonics (Design and
Applications) Handbook, page H-28 in the 1998 edition is the article on the
USAF 1951 target. It gives a half-page of textual info, then tables showing
info on the contrast and the resolution of the sets of bars with a
black-on-white image of the target for reference.
The article references Gurley Precision Instruments, but I don't find this
info on their web site.
I've scanned the page from the Photonics Spectra Handbook and am forwarding
it to you at [email protected]. I've converted to a binary image and
saved as a PNG to reduce the full-page scan to a relatively small (161 KB)
file, so it's a little fuzzy, but readable.
Optiker
"Jim Klein" [email protected] wrote...
> Hi,
>
> I need the specifications forn the USAF Test Target (or a big version
> of one). This is the target with the bars and numbers on it.
>
> I'm going to add it so it can be pixelated at arbitrary resolution and
> used as an object which can be imaged in Roadrunner. Who do I call to
> get this?
>
> Jim klein
Date: Thu, 26 Sep 2002
From: Kevin Kalsbeek [email protected]
To: Russian camera list [email protected]
Subject: [Russiancamera] Erwin Puts on lens contrast
Hi,
Heard from E.P. today, and now things are a lot more clear. Here is his
reply:
I cannot blame you for not understanding the topic of 'lens contrast' as
used by most people in newsgroups and elsewhere. My simple rule is: If you
cannot explain it, you do not understand it and if you do not understand it,
you should not talk about it. Free after Wittenstein!.
Lens contrast is the result of the optical correction of the lens, including
the coating of the lens surfaces. Generally yu may say that a lens with high
contrast is well corrected and will let all light rays from the object pass
through the lens without any stary light and unwanted deviations of the
paths of the rays. That is al light from a very white part of the object
will only be focused on the negative in the correponding area and nowhere
else. And on a black part of the object there will be no light. In this case
you have a lens with a 100% contrast, as the black and white are indeed
balcka nd white. If the lens has a lower contrast, than some of the light
rays intended for the white part will be deflected and will fall on the
black part, which then is no longer pure black but a bit grey. Such a lens
has a lower contrast.
That is all there is. But in practical photography the quality of the image
is degraded by many factors: exposure, focusing, position of the light
source, use of lens hood etc. Really informative statements about the
inherent contrast of a lens (the contrast that can be attributed to the lens
itself) can be made only if you test a lens in a situation where all
parameters can be controlled.
Erwin
Erwin's answer brings to mind a possibly simple and controllable way of
testing contrast. Will have to do some thinking, and perhaps some
testing as time allows.
Regards,
Kevin K
From leica mailing list:
Date: Thu, 17 Oct 2002
From: Stephen Gandy [email protected]
Subject: [Leica] Re: [CVUG] Lens comparison
the lens test results often have as much to do with the tester as the lens,
something which has been discussed many times on practically every photo
mailing list.
however, the highest performing lens ever tested by Pop Photo was the Cosina
Voigtlander 50/3.5, I think in the April 2002 issue In the same issue the
new Nikkor 45/2.8 introduced with the Nikon FM3A performed almost as well.
I believe it is also manufactured by Cosina.
when all is said and done though, it's the photog that makes the quality
shot, not the lens. in other words, there is more benefit in learning to
be a better photog, than trying to figure out which is the best lens.
Stephen
...
Date: Sun, 27 Oct 2002
From: muchan [email protected]
To: [email protected]
Subject: Re: [Contax] Re: Bishi or Honwaka?
...
> I was careful not to mention the trade names of the lenses H. M.
> tested, but one of them was a Leica, which by his defintion (even
> light distribution across the CoC) he called neutral. That is not say
> his result would be typical of all Leica lenses.
Leica lenses in general surely have beautiful bokeh, but I think Zeiss
in general has "softer" look. Partly because, (just it seems to me),
that Zeiss designer tend to be modest, to try "plateau" peak of MTF graph
over wider range of aperture, but Leica designer tends to aim "one best
aperture" of the given lens which surpass all the other aperture setting,
or "point peak" in the MTF measurering, at that point, it may over the
performance of Zeiss rival, but rest of apertures, "plateau" peak of
Zeiss lens may win... Quality of Bokeh is somehow summary of optics when
not everythig is going right, (at least focus is not at right position),
modest design may help keeping good compromis when things are not optimal...
...
From: [email protected] (brian)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Is Zeiss batch testing lenses?
Date: 1 Nov 2002
...(query about batch lens testing..)
Hi Bob:
I have no idea what Zeiss does with their photographic lenses, but I
do challenge the notion that 100% testing is rare and expensive. I
once toured a factory in China in which small scanner lenses were 100%
MTF tested using some fairly clever and economical equipment. I think
that for this application is was critical that the lenses actually
come very close to the design performance. It took only about 10
seconds to completely evaluate each lens. Mind you, these lenses
undoubtedly cost less than $10!
Another tidbit; on page 213 of "Eyes of Nikon" (a Nikon publication
from 1985), there is the following quote:
"As an additional benefit, MTF testing is incredibly fast (individual
testing takes only 6 seconds!), so every single Nikkor or Nikon Series
E lens coming off the production line can be tested."
Brian
www.caldwellphotographic.com
From: "David J. Littleboy" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: why MF won't get better ;-) Re: What is depth of focus?
Date: Mon, 11 Nov 2002
"Q.G. de Bakker" [email protected] wrote:
> Where the "sweet spot" of a lens is depends on the particular design. Some
> lenses perform best wide open, others need moderate stopping down. There is
> no hard and vast rule.
It's worse than that: resolution changes _differently_ with aperture at the
center, edge, and corner for every different lens. (In some lenses it seems
the corners are still getting better while the center is starting to get
worse as you go from f/8 to f/16.) I presume you know this site, but for
people who don't:
http://www.hevanet.com/cperez/MF_testing.html
David J. Littleboy
Tokyo, Japan
From: [email protected] (FLEXARET2)
Newsgroups: rec.photo.equipment.medium-format
Date: 12 Nov 2002
Subject: Re: Tweaking Focus on Med Format
It is funny how through the years that Nikkor lenses for Bronica focal plane
shutter models, Carl Zeiss Jena lenses for Pentacon 6 and other
fine optics have been given less than sensational sharpness ratings in reviews,
when the cameras were at fault.
In my opinion, all medium format users should check their finder
focus readings along with what the film is getting at the film plane.
This done by using a piece of groundglass (on the rails at the camera film
aperture) with a loupe epoxyed to it with the shutter at B - camera on a tripod
etc.
There may be a need to tweak the position of the finder groundglass.
As for interchangeable backs - one can only replace them or replace the film
insert - unless someone figures out other adjustments.
Through the years many medium format users have been unaware of this
problem which was covered up by stopping down the lens and getting greater
depth of focus to make for a sharper image.
- Sam Sherman
From: [email protected] (ArtKramr)
Newsgroups: rec.photo.equipment.medium-format
Date: 28 Nov 2002
Subject: Re: How is coc related to lp/mm?
...
If you take a series of prints of equal lines/mm but some are of greater
contrast than others, the higher contrast prints will be perceived to be
sharper as proven by Kodak's JND tests.
Arthur Kramer
Visit my WW II B-26 website at:
http://www.coastcomp.com/artkramer
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 05 Dec 2002
Subject: Re: Kodak Supermatic Flash Shutters
>Don't you guys measure aerial resolution rather than use on-print metrics?
Yes, aerial resolution can be measured by using an optical bench. Targets can
be of various contrasts but the 1000:1 is mostly used.
Another way is to run an actual thru focus test or use of a focus wedge with
high reso film. Then, the results (MTF) are read out on a micro-D.
Or did you actually put the Supermatic against the Zeiss (Hasselblad) lens
>in the same manner?
The Haselblad lens is more or less designed for high MTF readings at somewhat
low resolution values(40 l/mm and lower). This gives very nice results with
fine lace on a wedding dress at 4 to 5 times blowup or 8x8 or 10x10 print size.
Most true recon lenses give high MTF values at very high reresolutions (100
l/mm and up). This is great for 50 to 100 times blowups. Something the
Hasselblad doesn't do so well but believe it or not, that $20 Tourist 101 lens
did great at.
Larry
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 06 Dec 2002
Subject: Re: Kodak Supermatic Flash Shutters
>That's very helpful information, Larry. I might just consider making a
>camera for that recon lens.
Some recon lenses are like the Hasselblad, not very high resolution but have
high MTF values (higher then 100%) at lower resolution values under 40 l/mm.
Again, these types of lenses making 5x5 original negatives, give great results
contacted or at lower magnifications under 10X.
Why so? Well, many of these rather short focal length lenses are highly
correct to measure stuff on the ground They are not intended to be used to
tell what grade bolts were used to put the stuff on the ground together.
Larry
From: "David J. Littleboy" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Closing down aperture reduces resolution of lens?
Date: Wed, 25 Dec 2002
"Tony Spadaro" [email protected] wrote:
> And if you zre anal retentive enough to believe the validity of your own
> assinine "test" why the hell are you wasting time on 35mm when obviously
> only 8x10 or larger will give you the results you feel you need?
Despite the rudeness, there is a hint of a real issue lurking here; namely
you are effectively claiming that the loss of quality due to diffraction is
of the same order of magnitude as the other issues (grain, film resolution,
degradation in scanning/printing) that limit the quality of larger prints. I
don't have links to equivalent tests for 35mm lenses, but if you look at the
lp/mm values in:
http://www.hevanet.com/cperez/MF_testing.html
You'll notice that a lot of these lenses are pretty funky at f/22, but that
f/16 varies with the lens. (Here "pretty funky" means "20 to 30% lower
resolution than f/5.6 or f/8".) A lot of lenses hold up quite nicely at
f/16.
Is this enough to make a difference in a 9x or 10x (8x10, maybe with a touch
of cropping) prints? I suspect that these numbers aren't enough to tell,
since they are (presumably) 10% MTF for high contrast targets, and you
really need 50% MTF for low contrast targets* to be meaningful on a print of
real-world subject matter...
*: And I suspect such numbers for most lenses would be incredibly
depressing.
David J. Littleboy
Tokyo, Japan
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Closing down aperture reduces resolution of lens?
Date: Wed, 25 Dec 2002
Don Stauffer [email protected] wrote:
>I suspect that motion blur accounts for more degradation than
>diffraction for all hand-held shots. I remember a test done somewhere
>on handheld versus tripod even for 60th and 125th exposures. There WAS
>considerable blur. The tests as I remember used a pretty good film and
>was in either a photo mag or book.
Yes. Even tripod-mounting does not guarantee optimum
results, though... When shooting handheld for best
sharpness, I would take about five frames of the
same thing, and select for sharpness - and even under
difficult conditions, there would usually be one good
frame (and even under ideal conditions, there would
usually be one bad frame...). Some of us *do* care
about image sharpness, and take measures to optimize
it, like avoiding too-small/wide stops, camera-shake,
poor lens samples/types, etc....;-) And, it is possible
to learn how to hold a camera steady during an exposure.
The first step is buying a good 10X magnifyer, and
figuring out which frames are really sharp, and which
ones aren't - and why. Eventually, the percentage
of sharp frames goes up...
David Ruether
[email protected]
http://www.ferrario.com/ruether
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Closing down aperture reduces resolution of lens?
Date: Wed, 25 Dec 2002
"Tony Spadaro" [email protected] wrote:
> Because I'm obviously a much ruder person than you - and one who assumes,
>perhaps wrongly, that any sane person looks at his pictures in a fairly
>normal manner instead of studying each and eery one with a 10x loupe looking
>for the smallest imperfection. Dust on the subject is a bigger problem than
>diffraction on a 50mm macro lens at 1:1 with an aperture of f32.
> Diffraction is simply not an important factor at any aperture the lenses
>are designed to use.
I look at film, not prints at 10X (10X on a print
reveals nothing of interest...). If you think f32
on a Micro-Nikkor is sharp, I've got a bridge in
Brooklyn for sale, at a VERY good price....;-)
Or, run an aperture sequence from f8 to f32, look
at the results on film with a good 10X (hardly a
taxing magnification), and you will see a BIG
difference in the image going from f8 to f32 (or
even from f16 to f32). If you don't care that
the image is noticeably sharper at f16 or wider,
that's OK for you, but not for some of us who are
looking for the best sharpness (as one of many
image aspects to be optimized...). I'm assuming
that you mean the "marked" f32 on the lens (which
will give you an effective stop of f64 at 1:1 with
normal extension). If you mean the lens-marked f16
(which becomes effectively f32 at 1:1), I would
agree that this is an acceptable stop to use for
macro work with some gear... BTW, a big air
hand-syringe works wonders for cleaning macro
subject areas... Also BTW, I have some macro
images you may find interesting, at:
http://www.ferrario.com/ruether/phun.html
("Bugs"), all "hand-held", at magnifications
up to 3X, including some insects caught in the
air...
David Ruether
[email protected]
http://www.ferrario.com/ruether
From: David Littlewood [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Closing down aperture reduces resolution of lens?
Date: Mon, 23 Dec 2002
Robert Monaghan
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Closing down aperture reduces resolution of lens?
Date: Mon, 23 Dec 2002
"David J. Littleboy"
[email protected]> wrote:
>"Tony Spadaro" [email protected] wrote...
>> Diffraction causes small apertures to lose sharpness, but in general not
>> enough to worry about. Most lenses will not close down to the point where
>> the image is noticably damaged by diffraction.
>If you are scanning Provia 100F, Velvia, Tech Pan at 4000 dpi, you should be
>able to notice the difference between f/8 and f/16 or f/22...
>David J. Littleboy
>Tokyo, Japan
Even in low-resolution systems, like video cameras,
the effects of diffraction can easily be seen. At
this URL are samples of stills shot with a video
camera (Sony VX2000), where stopping down below
about f4 gradually softens image-center detail.
The surprise for me, after putting up this web page
(using the still-photo function of the camera), was
finding out that the motion-video images (which are
even lower in resolution) also clearly showed the
effects of diffraction...
See: www.ferrario.com/ruether/diffraction.htm.
David Ruether
[email protected]
http://www.ferrario.com/ruether
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Closing down aperture reduces resolution of lens?
Date: Mon, 23 Dec 2002
"Tony Spadaro" [email protected] wrote:
>I use to have a test of this up on my old website. The difference between f8
>(or possibly 11) and f22 was barely noticable using a 2700 dpi scanner. I'm
>planning to reshoot for the test and scanning at 4000 dpi before putting it
>on the new website but need a warm windless day for the shooting. I used
>Provia F100.
With B&W film-tests, using a good 10X magnifier, it is
easy to see the decline in resolution stop-by-stop from
f8 to smaller stops, with f16 generally being the last
"good" stop for optimum results at the image center with
35mm... (not that you cannot "stretch" to f22, but the
results are noticeably less sharp than at f8...). The
differences may be less obvious with a reduction in
resolution with copying, as with scanning...
David Ruether
[email protected]
http://www.ferrario.com/ruether
From: David Kilpatrick [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: National Geographic Rates Film 400% better than latest Nikon D1X Digital
Date: Mon, 6 Jan 2003
Bill Tuthill wrote:
> Per Inge Oestmoen [email protected] wrote:
>
>>VT wrote:
>>
>>>But one runs into aliasing and other loss problems there too
>>>(note the Nikon CoolScan 4000D tested at only about 60lp/mm when
>>>theoretically at 4000dpi it should have done closer to 79lp/mm......)
>>
>>Nikon LS-4000 only manages 60lp/mm? That is very bad. Do you have any
>>source for this claim? Not that I doubt it, but I would like to
>>document the shortcomings of the LS-4000.
It should only achieve 60 lppm (line PAIRS per millimetre for those who
don't grasp that this definition changed around 1966 when Geoffrey
Crawley was publishing his work on MTF - before that people used to
quote LINES PER MILLIMETRE and hence the figures given for early Leica
lenses are unfeasibly high, and should be halved to match modern lppm
readings).
It has 160 LINES (not LINE PAIRS) per millimetre scan density, deriving
this from 4000 lpi (lines per inch - lines, not line pairs). This equals
80 line pairs, which would be a 100 per cent contrast theoretical
resolution. 60 lppm at a normal film contrast level sounds like a
reasonable factor for light scatter, physical depth of emulsion,
diffraction limiting etc.
So that's not really a surprisingly low figure, just to be expected.
David
From russian camera mailing list:
Date: Fri, 20 Dec 2002
From: [email protected]
Subject: RE: Russian lens tests
"njp66 [email protected] wrote:
>The Soviets did give figures for the resolution of their lenses but
>the figures seem absurdly low.
>Was this based on what they could see on a print, rather than
>examining the negative through a loupe?
The Soviet tests are not lines/mm they are line pairs/mm. which means that the
approximate corelation is double the number.
--
Nathan Dayton
www.commiecameras.com
From russian camera mailing list:
Date: Fri, 20 Dec 2002
From: Bob Shell [email protected]
Subject: Re: Russian lens tests
[email protected] wrote:
> The Soviet tests are not lines/mm they are line pairs/mm. which means
> that the approximate corelation is double the number.
> --
> Nathan Dayton
> www.commiecameras.com
>
Line pairs per millimeter is the standard measurement used in optical
testing. Often you will see it called lines per millimeter, even though
it means line pairs. I don't think this is why the Russian test numbers
are so low, even though we know the lens performance is much better than
the numbers would indicate.
In their latest issue Popular Photography put some vintage Nikon lenses
through the same tests they use on modern lenses, and the old stuff did
really well in most cases. It would be great if we could talk them into
testing some Russian glass.
Bob
From russian camera mailing list:
Date: Fri, 20 Dec 2002
From: "Jay Y Javier" [email protected]
Subject: Re: RE: Russian lens tests
Nathan
I've found some soviet references so far which do say "line pairs per
millimeter" - a TENTO catalogue, a Zorki-1 passport, and several lens passports
for J-9 and I-61 lenses. Did something get lost in the translation?
Jay
From russian camera mailing list:
Date: Fri, 20 Dec 2002
From: "tigerarm2000 [email protected]
Subject: Re: Russian lens tests
...
Chinese national standard of 35mm lens resolution is also very low.
37 l/mm at center and 22 l/mm at edges is good enough to be rated a
first grade lens(J1).
Use the method to test a Nikon normal lens would give a resolution
figure of about 50 L/mm at center.
I think the resolution of film plays a very important role in the
tests.
Zhang
From russian camera mailing list:
Date: Fri, 20 Dec 2002
From: [email protected]
Subject: Re: Russian lens tests
The Soviets only gave a single center/edge or center/corner measurement,
probably with the lens wide open. Resolution should be better with each
lens stopped down two or three stops.
For practical purposes with 35mm what counts most is not ultimate
resolution but contrast (MTF) at about 20 lpm, which means how cleanly
light and dark points that end up spaced about 1/1000" apart on a
negative are tonally separated. Not coincidentally, this is the circule
of confusion used to compute depth of field. In a 10x enlargement those
points will be about 1/100" apart. What happens out beyond 20 lpm is of
diminishing importance in influencing our subjective sense of sharpness.
Of course you should adjust this critical point as film size and
enlargement ratio changes. For instance if you routinely enlarge 35mm to
16x20 the critical resolution becomes about 32 lpm.
From russian camera mailing list:
Date: Fri, 20 Dec 2002
From: Paul Shinkawa [email protected]
Subject: Re: Re: Russian lens tests
Dave:
Thank you! That is the clearest explanation I've read
on resolution and testing with respect to 35mm.
-Paul
...
From russian camera mailing list:
Date: Fri, 20 Dec 2002
From: Kevin Kalsbeek [email protected]
Subject: Re: Russian lens tests
Neil,
They used a very different system! One source, Vade Mecum, says it represents
microns, rather that line, or line paires per millimeter.
They work nicely though, don't they! B^)
Kevin
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 04 Feb 2003
Subject: Re: Q: Mamiya RZ lens resolution?
>Anyone have a suggestion on where I can find Mamiya RZ lens USAF resolution
>test chart results?
You will not get this info because it's not really worth anything. First, reso
targets comes in many contrasts. Second, resolution values can be for the lens
itself or the result of any lens film combination.
Then theres center vs edge readings. Finally, MTF values are usually not taken
right off a USAF target as they are rather hard to scan.
I will give you some idea of what your lens is doing. Center resolution on
must name brand lenses will run about 500 lines per millmeter at f-4 with 1000
to 1 contrast target using noon summer sunlight.
This would result in about 100 lines per millmeter using Tmax 100 film with a
focus position within 16 microns or about 1/1500 of an inch of true focus.
Larry
From: [email protected] (Bill Hilton)
Newsgroups: rec.photo.equipment.medium-format
Date: 05 Feb 2003
Subject: Re: Q: Mamiya RZ lens resolution?
>From: "Christoper M Perez" [email protected]
>What verifiable approach would you suggest a person take for evaluating the
>performance differences between camera systems? Something that, as a
>consumer, can be reviewed in the cold light of day? Where can such
>information be found?
You can find sites with MF lens data on Reverend Bob's site. Here are a couple
of links ...
http://www.photodo.com/prod/lens/mediumlenses.shtml Photodo site, which tries
to rank lenses by a point system. Many people whine about the single-point
sample size (probably guys whose lenses don't score well?) but at least you
have the same guy testing every lens with the same method. 913 MF lenses in
the data base, the ones I own in the Pentax and Mamiya systems seem to
generally agree with his results.
http://www.geocities.com/Yosemite/8917/lensrat3.html ... a smaller number of
lenses but worth comparing to Photodo. I got a fatal javascript error on this
page so didn't poke around so don't know the testing methodology.
http://www.hevanet.com/cperez/MF_testing.html Resolution numbers, provided by
someone named "Christopher Perez", which, I can't help but notice, is your name
as well. Surely you're not the same guy?
And here's the source of all these links, Bob Monaghan's great medium format
site ...
http://medfmt.8k.com/mf/links.html
http://medfmt.8k.com/mf/zmisc.html
Bill
Date: Wed, 05 Feb 2003
From: "Christoper M Perez" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Q: Mamiya RZ lens resolution?
Hi Bill,
Thanx for the reference information.
On Photodo, are they using published data? Or running their own MTF tests?
The reason I ask is that one of their articles says:
"Please have in mind when you compare the Zeiss and Canon articles about MTF
that they have a different approach to the subject. Zeiss exhibits
achievments from the real world lenses, while Canon exhibits design ideals
for each lens..."
If a person is comparing 'real world' data with 'design ideals', it seems to
me that there is room for error in interpretation. When I read Zeiss MTF,
I'm fairly certain that I'm looking at manufactured test samples. But when
I read Mamiya's MTF charts, I'm not sure if this is design or live test
examples.
The geocities site says:
"Note that these ratings differs slightly from those at Photodo. The old
ratings are 0.1 - 0.3 higher.
These tests are from the Swedish magazine "Aktuell Fotografi". They started
MTF lens tests in 1990 at the optical lab at Hasselblad.
The tests are done at 10 cycles/mm ("contrast"), 20 cycles/mm and 40
cycles/mm ("sharpness"). Both tangential and radial MTF values (0 - 1) are
read at 0, 3, 6, 9, 12, 15, 18 and 21 mm off the optical axis. A MTF value
of 1 equals a 100% transfered contrast which is impossible. Very good lenses
have MTF values over 0.80. For some lenses the values drop to almost zero in
corners wide open. Each lens is measured at f/4 and f/8..."
So it appears that these folks have taken actual examples into Hasselblad's
lab to test. While it may be a collection of "one" that gets tested, it's a
better place to be than with no data. At least that's what I think.
Yes, I am the person in your third comment below. I was rather hoping that
someone had taken the time to do the kind of system test on a Mamiya RB/RZ
lens collection that Kerry (with other kind souls) and I had done on similar
items. Without spending $2k for a new RZ with a single lens, I was hoping
to find relevant test data that I could work from. I suppose I could go out
and rent a camera to test. But the philosophy to date has been to test only
those systems that my friends and I have actually owned. The whole effort
has gotten quite expensive. Time to acquire, test, then sell as well as
managing any profit/losss gets sometimes 'out of hand'.
Regards - Chris
...
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 05 Feb 2003
Subject: Re: Q: Mamiya RZ lens resolution?
>Try to find out the criteria used by NASA when they were shopping for
>camera systems to use in space - should give you a starting point.
>
I can't say anything about NASA but since I was chief of the CIA spy camera
program from 1985 to 1996 now retired I might have a few insights.
First, just about all name brand lenses have about the same resolution although
for special applications, suppliers would allow us to source select. This
means we would take delivery of say ten 180mm lenses and select 5 sending the
rest back.
Biggest issue was with the camera bodies. Only about 2 out of 10 camera bodies
focused well enough to meet specs. This means the other 8 camera bodies had
mirrors that were not in sync with true focus. In other words, when you
focused on something at 20 ft you were actually focused at 16 or maybe 25 ft.
This is a full 100 micron focus error that won't really clear up until stopped
down to at least f11.
Just about all image quality problems can be traced to the camera bodies. It's
very rare indeed to trace image quality problems to a lens from a major
manufacture.
Larry
Date: Wed, 05 Feb 2003
From: "Christoper M Perez" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Q: Mamiya RZ lens resolution?
These are interesting observations. Thank you.
For the 4x5 lenses Kerry and I have tested we have used camera bodies with
properly aligned ground glass. The gg on my Linhof Super TechIII is quite
nearly 'spot on' with respect to industry specifications and my film
holders. The test results I post are fairly representative of what that
'system' will return under specific conditions, leaving the lens as the
biggest test variable.
For 120 format testing the variability you mention comes squarely into play.
You may have noted that one of my gripes with my current Hasselblad body is
that it's mirror required alignment before I could get a good 'system' test
result. Now I'm confident that when I focus on an object that I'm within an
inch of the actual plane of focus. For serious hobby work this is
sufficient.
Granted, USAF charts won't give you is a clear indication of a lenses
ability to return good contrast. However, the USAF chart method provides
one measure of performance that is verifiable. Of course there are other
dimensions to be taken into account. Contrast is one dimension that has
already been mentioned. Camera system film plane alignment is another.
Scene contrast ratio is yet another (Kerry and I have used the same lighting
setup since 'day one' and feel we have minimized this variable about as far
as we can without getting absolutely ridiculous and is set at a typical
daylight 6:1 contrast ratio).
It appears that there are very few source of independent tests of Mamiya
optics. Oh well. It was worth the asking.
Regards - Chris
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 06 Feb 2003
Subject: Re: Q: Mamiya RZ lens resolution?
>"out of the box, brand new,
>you should put a lens and body on an optical bench and optimise the
>assembly"
Lenses no, bodies yes.
Larry
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 12 Feb 2003
Subject: Re: Enlarging limitations
>I've never met a 1000 to 1 contrast subject that I wanted to photograph,
>though. (And no one else has, either.)
In sunlight you have lots of highlights against deep shadows. Some of these
highlights [are] coming right off glass and chrome on buildings and cars. If you
measure them, they are even higher than 1000 to 1.
Larry
From: "David J. Littleboy" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Enlarging limitations
Date: Wed, 12 Feb 2003
"Hemi4268" [email protected] wrote
> >I've never met a 1000 to 1 contrast subject that I wanted to photograph,
> >though. (And no one else has, either.)
>
> In sunlight you have lots fo highlights against deep shadowns. Some of these
> highlights coming right off glass and chrome on buildings and cars. If you
> measure them, they are even higher then 1000 to 1.
That's a _scene_ with a total overall contrast range of 1000:1. That's not a
detailed pattern with 1000:1 at 100 lp/mm at the film plane.
As an aside, my 55 lp/mm estimate may be way too high for what we actually
get on film. The 1Ds is a lousy 40 lp/mm (for 1000:1 targets
From: [email protected] (brian)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Enlarging limitations
Date: 14 Feb 2003
[email protected] (Hemi4268) wrote
> > (Scanner and
> >enlarger lens limitations.)
>
> No such limitations.
>
> Larry
Enlargers and scanners use imaging lenses. Imaging lenses are subject
to limitations caused by aberrations, diffraction, and mundane things
like lack of perfect film flatness.
You also mentioned earlier that ordinary 35mm lenses can deliver
400lp/mm at f/4. This is true only for certain extraordinary lenses
used over a restricted waveband and only for a very narrow region near
the optical axis. Even then the image contrast at 400 lp/mm will just
about be zero because you're very close to the Rayleigh limit for
green light.
It sounds like you just aren't very familiar with photography. Or
perhaps you're just an argumentative jerk.
Brian
www.caldwellphotographic.com
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 14 Feb 2003
Subject: Re: Enlarging limitations
>You also mentioned earlier that ordinary 35mm lenses can deliver
>400lp/mm at f/4. This is true only for certain extraordinary lenses
>used over a restricted waveband and only for a very narrow region near
>the optical axis.
Center region maybe but 400 l/mm is available at f-4 in full noon summer
sunlight with most name brand lenses.
Actally make that 500 l/mm at 20% MTF. All you need is an optical bench and a
sunlight source. Works everytime.
Although it good that you say what you say since most if not all the other
people in this forum say it's not possible at all. Extraordinary lenses or
not.
Larry
From: [email protected] (brian)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Enlarging limitations
Date: 14 Feb 2003
[email protected] (Hemi4268) wrote
> >You also mentioned earlier that ordinary 35mm lenses can deliver
> >400lp/mm at f/4. This is true only for certain extraordinary lenses
> >used over a restricted waveband and only for a very narrow region near
> >the optical axis.
>
> Center region maybe but 400 l/mm is available at f-4 in full noon summer
> sunlight with most name brand lenses.
>
> Actally make that 500 l/mm at 20% MTF. All you need is an optical bench and a
> sunlight source. Works everytime.
A fully diffraction-limited f/4 lens cannot have a contrast of 20% at
500 cycles unless you use fairly deep UV light of about 340nm or
shorter. Not even remotely close to visible. Is that what all that
nonsense about "full noon summer sunlight" is about? By the way, most
ordinary photographic lenses are poorly corrected for UV that short
because of spherochromatism. That is if they can transmit the light
at all.
Brian
www.caldwellphotographic.com
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 15 Feb 2003
Subject: Re: Enlarging limitations
>A fully diffraction-limited f/4 lens cannot have a contrast of 20% at
>500 cycles unless you use fairly deep UV light of about 340nm or
>shorter. Not even remotely close to visible.
If you look at the tables in the SPSE standards handbook you will see the
wavelength distribution for noon summer sunlight. Lots and lots of UV. The
stuff you get burned with.
Also on page 976 you have a lens chart that clearly shows 500 l/mm f-4
performance with 425 nm of point or 525 nm line light source. Something noon
summer sun has lot of. If you don't believe me just leave your sun block home
next time you go to the beach.
Larry
From: [email protected] (brian)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Enlarging limitations
Date: 15 Feb 2003
...
> Larry
Assuming your "500 l/mm" is the same thing as 500 cycles per
millimeter, then at 425nm the diffraction limited modulation for f/4
is only 0.056, which is alot less than the 0.20 you quoted earlier.
To get the modulation higher you must use a shorter wavelength. At
525nm the Dawes limit is about 475 cycles/mm, so there can be no
response at all at 500 cycles.
By the way, what sort of detector are you using on your handy dandy
optical bench to detect UV light?
Brian
www.caldwellphotographic.com
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 16 Feb 2003
Subject: Re: Enlarging limitations
> At
>525nm the Dawes limit is about 475 cycles/mm, so there can be no
>response at all at 500 cycles.
So here we are discussing differences between 475 lp/mm and 500 lp/mm at 20%
10% and 0% when just about everyone else thinks 55 lp/mm is about the limit. I
think I will quit while I am ahead. You win, 475 lp/mm or l/mm which ever
blows your skirt up it is.
By the way, what sort of detector are you using on your handy dandy
>optical bench to detect UV light?
That handy dandy detector is a 100x optical bench microscope and the eye. It
works in the visable short wavelengths which you know, sunlight has lots of in
the summer time.
Larry
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 16 Feb 2003
Subject: Re: Enlarging limitations
>You still haven't told us how to get more than 55 lp/mm in an actual
>photograph...
Actually you are right. Porta max resolution is around 55 l/mm. I got it to
60 but again testing in with noon summer sun.
Porta along with original VPS is not a high resolution film. It's strengths
though are it's it ability to make everyone look like they just came home from
vacation.
It also has very good tone reproduction holding detail in white wedding dress
even if overexposed.
Last, it has a big MTF boost 110 %+ at rather low resolution levels giving a
high quality print look at rather low magnification levels (about 4x). Making
10x10 Hasselblad prints looking super in a wedding album.
Larry
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 15 Feb 2003
Subject: Re: Enlarging limitations
>Ain't too many rez charts hanging on gallery walls - yet!
>
>Let's do a photo calendar of lens rez test charts. Judging by the yield of
>usenet responses to mild rez-bait, I'll bet it would sell.
>
Shows how much you know. Reso charts in a photograph would really be too small
to see. Even at 20 ft with a 50mm lens a 1 inch by 1inch chart would be about
1/100 of an inch on the negative and about 1/20 of an inch on a 8x10 print.
Small enought to spot out or hide in leaves if one wanted to.
All my students were required to place these 1 by 1 inch targets in the
photographs submitted for crit.
Average resolutions for even student work ranged from 175 l/mm for Ektar 25 to
about 20 l/mm for tri-x pushed 2 stops. Then of course if the cameras were
screwed up somehow, resolutions were usually below 10 l/mm. I could count on a
class of 25 srudents to have at least 5 cameras screwed up enough that the
cameras themselves could be declared legally blind.
Larry
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 15 Feb 2003
Subject: Re: Enlarging limitations
>Can you folks decide which unit to use, lp/mm or l/mm?
Depends on when you went to photo or optics school. Before 1975 or so it was
l/mm and after 1977 it's lp/mm. I think the change came when digital came in
with their dots per inch.
The dots are up against each other and lines are not. The p was added to
express that each set of line has a same size space between them.
I like the old way. It's obvious my friend went to U of R after 1975.
Larry
Date: Thu, 23 Jan 2003
From: Leonard Evens [email protected]
Newsgroups:rec.photo.equipment.large-format,rec.photo.equipment.medium-format,rec.photo.darkroom
Subject: Re: Venting some disappointment
Gregory Blank wrote:
> You don't state which lenses but I know my Apo Symar is sharp at f/32 and 45.
> Whats more I would state I have never seen a truely sharp print using an
> f/ stop
> below f/32 with any of my humble lenses. Maybe your lenses were dropped?
> Or maybe your enlarger is out.
> Or maybe you just have superhuman eye sight which enables you to focus at
> and extra
> wide apeture.....and get better than normal results. :-)
The trouble is that no one ever objectively qualified what he means by
statements about sharpness. A perfect lens at f/45 would deliver
approximately 1500/45 = 33 lp/mm. If you looked at the aerial image so
as to avoid worrying about the resolution of the film, you would begin
to see the image degrade as something over 6 X magnification. If you
made a 24 x 30 print from an 4 x 5 negative and looked at it very
closely, you might just begin to see some degradation of the image.
On the other hand if you viewed it as normal viewing distance (the
diagonal of the print), it would look very sharp.
I think almost any statement about the sharpness of a lens at a given
f/stop is probably true under some conditions.
>
> Michael Briggs [email protected] wrote:
>
>>Gregory Blank wrote:
>>
>>>f/11 is nothing on a 4x5 use f/32 or f/45
>>>I never with exception for portraiture & film exposure tests go wider than
>>>f/22
>>
>>I disagree with this advice. I have tested two modern, high quality LF
>>f5.6 lenses for 4x5 and found them to have seriously degraded resolution
>>at f45. If you need the depth-of-field, by all means use f45. The
>>best apertures for both lenses were f16 and f22, with f32 being slight
>>worse. If you don't need the dof of f45, I would recommend a wider
>>aperture.
>>
>>--Michael
--
Leonard Evens [email protected]
Date: Thu, 23 Jan 2003
From: Robert Feinman [email protected]
Newsgroups:
rec.photo.equipment.large-format,rec.photo.equipment.medium-format,rec.photo.darkroom
Subject: Re: Venting some disappointment
Lots of discussions of this topic. My 2 cents:
1. Camera shake or misfocus of even a very slight amount can affect the quality
of an image. Even on a tripod a 4x5 may not be completely stable; and how well
do middle-aged eyes focus closeup on the groundglass?
2. I doubt that a modern lens would be noticeably better than an older symmar.
I've never seen a really bad LF lens made since the 1950's especially at f11 or
smaller. Modern lenses may have higher contrast, less flare or bigger image
circle, but sharpness hasn't changed much.
3. I think your experiment was reasonable. Shoot the way you usually do and the
same for your son. You're testing your working conditions, not Kodak's film.
4. In my experience the difference between 6x7 and 4x5 is no longer significant
at enlargements up to 16x20 or so. This is one of the reasons I've almost
completely stopped using 4x5. The added difficulties are not worth the effort
(especially in the winter!)
5. Digital adjustments can fix perspective problems caused by a tilting back. I
have a couple of Photoshop tips illustrating this on my web site.
6. LF still has the advantage if you need to adjust the plane of focus and you
don't have a tilt/shift lens for your MF camera.
7. The Rolleis have some of the sharpest lenses around. The sharpest picture I
ever shot was with a Rollei (3 feet high - you can see grain, but it looks
"sharp").
Collin Brendemuehl wrote:
> All that work with the big stuff for really no better picture.
> (a little tighter grain structure but not as contrasty)
--
Robert D Feinman
[email protected]
Landscapes, Cityscapes, Panoramic Photographs: http://robertdfeinman.com
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 24 Jan 2003
Subject: Re: Question about enlargements.
>and it is tack sharp
Actually tack sharp is not sharp at all being about 4 lines per millmeter. The
eye sees at least 8 lp/mm at 10 inches. If you look at a tack from 10 inches,
you will see a rounded point.
The textbook answer to this question starts with the eye. Since the eye sees 8
lines, an 8 line negative contacted printed will look sharp. This same
negative will look somewhat fuzzy at 2 times enlargment resulting in 4 lines of
resolution at the print. At 4 times enlargment, the print is very fuzzy.
Most standard brand lenses will produce 500 lines of center resolution at f4 in
noon summer sun.
A standard 100 ASA film will usually capture 50 to 100 of these lines or 10 to
20 percent of what the lens sees.
A typical 35mm 64 line negative will do about a 8 times enlargment or somethin
slightly bigger then a 8x10 sharp print.
Larry
From camera makers mailing list:
From: "Brian Swale" [email protected]
To: [email protected]
Date: Mon, 18 Nov 2002
Subject: [Cameramakers] Re: Nikkor process lenses
Hello folks,
I am also on the Olympus camera list, and the nikkor process lenses got a
mention there.
I just had a response from a fellow list-member which throws a little
monochromatic light on the topic, so I share it.
Brian wrote in connection with a Nikkor Process lens:
>I visited that, and was suitably awed, then took the link to the base URL..
There I selected one link to a site that looked interesting, and it is of a
micro-nikkor with a resolving power of from 700 - 800 lines per mm. Amazing.
Here's that URL
http://homepage2.nifty.com/akiyanroom/redbook-e/collection/oguri1.html
He wrote
"The trick here is these lenses only produce that resolution with
monochromatic light ("e-line" 530nM) and probably only at full aperture.
That is, in designing a photo-lithographic lens (process lens) they don't have
to worry about chromatic aberration at all, so diffraction is the limiting factor.
More modern process lenses are optimized for different (shorter) wavelengths
than e-line and are faster since these lenses are generally diffraction limited,
meaning if you stop them down resolution drops substantially.
If you tested the OM's (Zuikos) with monochromatic light they would also
have better resolution since chromatic aberration would be removed, although
they probably have compromised designs, since they must work with full
spectrum light.
In a normal full spectrum lens, the optimum resolution aperture is where the
diffraction fuzziness matches the chromatic fuzziness!
If you stop down more diffraction limits while if you open up, chromatic
aberration limits."
Regards,
Tim Hughes
Hope that helps people understand these lenses some more.
Brian
From ZICG mailing list:
Date: Fri, 25 Oct 2002
From: Edward Meyers [email protected]
Subject: Re: Re: Zeiss Triotar 100mm/
Good, better best. Describing useful lens quality
brings in to play (as you know) the camera. If film is not
positioned properly a great lens may perfom badly.
The Zeiss Contarex (I'm told) has a very good film channel.
This takes advantage of the wonderful lenses made for it.
Experts in this area questioned the design of smaller
SLR cameras, such as the Olympis OMs, which had short
film channels. Film needs to be stabilized before reaching
the exposure position, I am told.
Testing for resolution doesn't tell you about contrast.
And testing for contrast alone doesn't tell the whole story
on lens performance. You gotta take pictures. Then the
enlarging lens and printing or projection comes into
play.
Funny story. When I was the lens tester (resolution, only)
at Modern Photography magazine in the early 1960s, I tested
the curent enlarging lenses in actual use. Rodenstock did
not test out as best, at that time. The importer of the
Rodenstock lenses threatened to pull their advertising unless
I reported a "correction". I agreed to retest the rodenstock
enlarging lenses in the importer's darkroom along with their
expert, Rudy Simmons, one of the famous Omega enlarger
brothers. I showed up at his darkroom in Long Island City,
New York. We went into the darkroom and immediately the
room began to shake. What is that? Rudy explained that there
was a printing company next door and when the presses
were running (most of the day) the darkroom vibrated.
How were we supposed to test the enlarging lenses then?
Rudy looked at his watch and said, "let's go to lunch".
We did, and the case was closed. Ed
From: "Richard Knoppow" [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: testing a new (to me) lens leads to questions
Date: Tue, 19 Nov 2002
...
In general the resolution limit with 35mm film will be the
film, with LF, the lens. There are several problems which
can happen with LF to spoil the resolution. Vibration of the
camera and film holders with wrong film position are
probably the most common. Another possiblilty is movement of
the back when the film holder is put in. Any of these can
destroy the inherent sharpness of the lens.
If we compare the actual resolution of lenses probably
35mm lenses have higher resolution than LF lenses,
especially since they are mostly used at larger stops.
However, when magnification of the image is taken into
account the resolution on the print is usually higher for
the larger negative.
My personal experience is that any size larger than 35mm
will usually show better image detail and, for B&W at least,
smoother tonal rendition.
I don't have a Fujinon lens such as you have so I can't
comment on it directly, however Fuji lenses have an
excellent reputation. I think many WA lenses do not have
quite the performance of longer FL, and smaller coverage
types, but I don't think this is what you are observing.
The film holders can be checked with a depth micrometer.
The distance from the edge of the holder to the backplate
_with film in the holder_ should be the same as the distance
from the reference surface to the ground glass of the back.
Note that many spring backs make contact with the camera
back (the reference surface) only at the corners.
Vibration can be difficult to stop and is worse for larger
cameras. Its helpful to damp the camera with your hand just
before making the exposure.
As far as exposure its the 35mm camera is more likely to
be accurate than a LF lens with an old shutter. Its helpful
to have some means of testing the shutter for actual speed
and for consistency. Many shutters, especially Ilex and
similar, tend to have mechanical hysteresis causing the
speed to vary from one time to the next. Ilex actually
recommended firing their shutters a couple of times before
making the actual exposure.
The tester I suggest is one sold by Calumet. Its the size
of an exposure meter. The last I looked it was about $80 US.
It doesn't matter too much if the shutter speed is not what
is marked as long as its consistent an you know what it
actually is. This tester can also be used for 35mm cameras
and for checking strobe flash duration.
You can also check lens quality by examining the aerial
image. This is simply using a good quality loup to look at
the image projected in air from the lens without the
interspersed ground glass. You will usually be able to see
the effects of stopping down, especially at the corners,
this way. Ten power is about the lower limit for
magnification for LF lenses. The ideal magnifier is a low
power microscope, but these are not exactly household items.
I've found a triplet type magnifier is good enough.
I hope some of this rambling is helpful.
---
Richard Knoppow
Los Angeles, CA, USA
[email protected]
From: Jim Klein [email protected]
Newsgroups: sci.optics
Subject: Re: Optic satellites' resolution
Date: Wed, 01 Jan 2003
Hi,
Angular resolution is given by 2.44 times the wavelength and divided
by the aperture diameter. Then multiply that by the distance to the
target to get the resolution in target units. Simple optics from
Jenkins and White.
Jim Klein
...
From: Joe Gwinn [email protected]
Newsgroups: sci.optics
Subject: Re: Optic satellites' resolution
Date: Wed, 01 Jan 2003
Luca Morandini
From: "David J. Littleboy" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Enlarging limitations
Date: Tue, 18 Feb 2003
"Robert Monaghan"
From: Lassi [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Enlarging limitations
Date: Tue, 18 Feb 2003
"David J. Littleboy" wrote:
...
> (Actually, I suspect that your MTF combination charts may be understating
> total system performance. I've also seen the rule (1/MTFs)^2 = (1/MTFa)^2 +
> (1/MTFb)^2, which predicts larger numbers than you are showing, e.g. 100
> lp/mm lens + 100 lp/mm film would be 70, not 50, lp/mm.)
Agreed. I've always considered the circle of confusion as a
probabilistic deviation, because quantum mechanics says so. Essentially
equivalent to a noise limit. When computing the average deviation, you
add squares and take square root, if the distributions are about
Gaussian. When there are several steps in the process (camera lens,
film, enlarger lens, paper), each step contributes a square to the sum.
-- Lassi
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 18 Feb 2003
Subject: Re: Enlarging limitations
> (based on diffraction limits, use 1600/f# from
>Dawes by some authors..)
Remember that this limit of 1600/f# is for bees wax candle light, the standard
when these figures were derived.
Now add noon sumers sun to the calculation and you will find 2000/f# is not
unreansonable.
Although I do agree with you completely if your bases is the the Dawes limit.
One of the nice things about the SPIE theoretical resolution limit graph (page
976), is that it allows you to correct for wavelength were as the Dawes limit
does not.
Larry
From: [email protected] (Robert Monaghan)
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Enlarging limitations
Date: 20 Feb 2003
um, Hicks was basing his work on his empirical studies of lenses as a
lens reviewer, author of books on lenses and so on, work done recently
with current lenses from the 1970s onward (some were older Leica etc.
lenses, but still, this isn't 18th or 19th c. optics work ;-)
the key point IMHO is that he specifies a rather stellar contrast point
(50%) for this formula; many lens tests and formulas (e.g., 1600/F#..)
either fail to specify the allowed contrast level, or use levels more
like 10% or so for the cutoff. If you drop the allowable %contrast enough,
you can get values up to 2000/f# and beyond, but the photo contrast
qualities will be marginal or poor. Hicks formula provides excellent
contrast along with high resolution potential values which are more useful
to practicing photographers than the more optimistic values based on Dawes
(1600/f#) and other lower contrast% methodologies. On my 25cm telescope, I
use Dawes limit, but with 35mm lenses, I use Hicks formula ;-) ;-)
I agree with the point that a more blue light source is likely to produce
higher resolution values, both in theory, and in practice (as I noted by
citing Skudrzyk's experience with higher resolution values using flash
than electric lighting, and my suspicion that part of this is due to
vibration improvements, as he suggested, but also part may be the bluer
light source. The amount of resolution increase was significant - about
35-50%, but this was only visible with the best films etc.)
bobm
From camera makers mailing list:
From: "Uptown Gallery" [email protected]
To: [email protected]
Date: Mon, 13 Jan 2003
Subject: [Cameramakers] Opti-Copy bad news received
The QC guy went to lunch with the inventor (small world!) whose son had
received my 'hunch' letter. I say hunch because I looked up the name of the
inventor in an online phone directory in the city where the inventor lived.
His son was the recipient of the letter, passed it to Dad, who passed it to
the QC guy over lunch.
The QC guy had accidentally and embarrassedly deleted my email request but
gladly replied.
It is a monochromatic lens...
Here's his cut & pasted reply.
It was used by Opti-Copy on the
"Imposer" camera which was designed to take color seperation negatives or
reflective black and white copy and project them onto a film mounted on an
x-y stepping table in order to make one piece imposed film flats ready to be
contacted onto plates for printing. The lens was very flat field and of the
highest resolution - we were able to reproduce a 150 line screen within +/-
1% dot value over a 15" x 21" image. The model that you have is an 8 element
lens, I have no idea where any drawings of the lens design went to after
Opti-Copy closed, so I can't help you on that. The lens was highly corrected
for 1:1 reproduction, but did a good job from 25% to 150% ratios if stopped
down to F:13.5 ( at 1:1 the "sweet spot" of the lens was F:9.5). This may
seem a little odd that the lens would be used at 1:1 to copy when other
"Step and Repeat" competitors (such as Meisomex or Dai Nippon Screen) would
contact print the negative to the plate, but projection was much faster as
time to load and expose each film was much shorter. Originally we used a 19"
Goertz or Schneider Apo Artar on the Imposer but in the early 80's had this
lens designed. It was first a 23" focal length but later we redesigned at
27" focal length. These lenses cost us about $13,000 per copy when ordered
in groups of 10 and we probably made over 500 imposers to use this lens.
Now for the bad news - Although the lens is very good at what it was
designed for, it is not a color corrected lens. In other words it was
designed for very high resolution using monochromatic light. I forget the
exact wavelengths it was designed to transmit, but it was in the green to
yellow wavelength as we were using film that was not sensitive to red
safelights. I did a test with one of these lenses and focused an image using
green light, then blue light and finally red light to project and found that
the focus distance would shift by almost a half inch at 1:1. I wanted to see
how the lens would preform at higher magnifications so I tested one on a
different camera we made that was designed for enlarging (this camera had an
overhead track 36' long and could project a 12 x 18 film using condensor and
point source light onto a 70" x 14' wide copyboard). At 8x magnifacation the
image was best described as "Mush" even using green light. These lenses
were like this, very good at what they were designed to do, but outside of
that were pretty useless. Sad to say, but the lens will probably not work
for your use.
I tested quite a few lenses while working for Opti-Copy and from the
information you supplied about making experimental large format cameras
would probably suggest using older Apochromatic process lenses of 450mm
focal length and longer for your use. These are names like Apo-Artar,
Apo-Ronar and the like. Most will be f:9 or f:11 four element symetrical
lenses and all will be optimised at 1:1 but do a fairly good job when
stopped down to f:22 when used at high magnifications (Stop down further and
diffraction will cause the image to soften). Remember that these lenses will
cover a much larger area at 1:1 than say at 10x or infinity focus.
Occasionally there will be a 47.5 inch "Red-Dot" Artar come up on e-bay
which would easily cover 18 x 20 at infinity. There are a few "wide angle" 6
element lenses of this type but 360mm seems to be the high focal length for
this style. Enlarger lenses such as Schneider Componon-S or Rodenstock
Rodagon-G were also made only to 360mm lengths and would only cover 8x10 or
streching it a little would cover 11 x 14. Of the Apo style symetrical
lenses, one of the better performers that I tested was the Apo-Germinar
which was a 6 element lens made by Jena - the East German factory of Zeiss.
These seemed to preform better at magnification than the others and are
often seen on e-bay also.
I hope it's good for something...
You guys told me so....
Murray
From: "John Cremati" [email protected]
To: [email protected]
Subject: Re: [Cameramakers] Opti-Copy bad news received
Date: Mon, 13 Jan 2003
I once had a very rare 55 mm lens called a " Ultra Micro Nikor"
Supposedly it had a extremely high resolution at extreme magnifications...
Unfortunately I found out that it was same as your lens, optimized for just
one color with little use for conventional photography........ It was used
in the circuit board industry.. I had immediately listed it on E-bay and
sold it but now I wish I had tried a few things first...
Some alternative photography or silver images may in fact be quite
attractive with the effect this lens will create...Why don't you try a
little experimenting before throwing in the towel...
From: "Christopher M. Perez" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: RZ 110mm lens performance - comments (long)
Date: Mon, 03 Mar 2003
I have been thinking about the quality of Mamiya SLR lenses. [whine
mode on] This after my Hasselblad 500C/M has been in the shop 4 times
over the past 6 months. The front and rear of the 'blad body was 3 thou
sandths of an inch out of alignment. The mirror gave a focusing error
of 1 foot at 5 feet. A barndoor spring bent after the camera, wrapped
in foam in a plastic case, slid off another case and dropped 6 inches
(yes SIX!) to the floor. Then one of the film backs started to overlap
the first and second frame (that 'ol clutch problem). What's next? A
lens spring failure?[whine mode off]
At a photoswap this past weekend I carped about my 500C/M fate with
Hasselblad users. Some do weddings. Others do street photography. And
others do portraits. ALL said 'it's just the beginning... keep a pair
and a spare on hand... they'll be passing each other on their way to the
repair shop...' Ouch! I'm just a hobbiest who's neurotic about
resolution. [yep, whine mode is still off]
I came across a decent Mamiya RZ with 110mm Z f/2.8. In search of
better MF camera system reliability, I bought it, took it home and
tested the lens. Here's what I found. Using a USAF resolution chart
from Edmound Scientific to read lines per mm (l/mm), TMax100 film,and
souped in D-76:
Center/Middle/Edge/F-Stop
-------------------------
68 60 60 f/2.8
60 68 68 f/4
76 85 68 f/5.6
76 76 68 f/8
76 68 76 f/11
68 68 68 f/16
54 54 54 f/22
Since I'm completely neurotic about such things, here are a few
apertures comparing various lenses I've tested over the years (
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 03 Mar 2003
Subject: Re: RZ 110mm lens performance - comments (long)
> Using a USAF resolution chart
>from Edmound Scientific to read lines per mm (l/mm), TMax100 film,and
>souped in D-76:
>
>Center/Middle/Edge/F-Stop
>-------------------------
>68 60 60 f/2.8
>60 68 68 f/4
>76 85 68 f/5.6
>76 76 68 f/8
>76 68 76 f/11
>68 68 68 f/16
>54 54 54 f/22
>
In order to do this test right you need many more targets then just one. They
should be placed at various distances to cover any possible mirror error. Just
about all cameras have some mirror error.
What I do is place targets at 16 ft, 18 ft, 20 ft, 23 ft and 26 ft. Sometimes
even 29 ft. I then focus at 20 ft. This is called a thru focus test.
Each distance from 20 ft shows about a 50 micron focus error. If you find that
the 20 ft target is about 60 l/mm and the 23 ft target is 80, then you have
about a 50 micron error. This is typical of most cameras.
Just the fact that your higher resolutions are down around f-5.6 and f-8 tells
me that you have this error. You best reso should be at f-4
You can figure the most error allowed is to square the f-stop.
So f-1 would be 1 micron and f-2 at 4 microns and f-4 at 16 and so on. You
see, at f-8 or 64 microns of allowable focus error your 50 micron mirror error
is wiped out.
Larry
From: "Dan Beaty" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: RZ 110mm lens performance - comments (long)
Date: Tue, 4 Mar 2003
Christopher M. Perez [email protected] wrote
> However, after roasting the language in the Mamiya brochure, take a look
> at the Mamiya 110 Z edge performance. It is consistantly higher than
> anything I've seen thus far. So I'm left wondering; has Mamiya traded
> center performance better edge resolution?
>
> Thanx for listening - Chris
Chris,
Some tradeoffs are necessary when a lens needs to cover a larger area of
film. When considering that the 6x7 format is considerably larger than the
6x6, I feel that RZ lenses compare very favourably to Hassleblad on your
tests.
Of course portrait photographers would need less edge sharpness than
architectural and landscape photographers. I have enlarged nature photos
from the RZ 110 to 30"x40" from Kodak Ektar 25 film, and the detail at the
edge of the frame is much appreciated.
When compared to the cost of the Hassleblad system, I find that the RZ is
extremely versatile and provides both value and extremely high quality. But
the main trade off that most point out is the weight and size of the RZ. If
you do not need 6x7 for the additional enlarging potential and you are
shooting for fun, the Hassleblad still might be best for you.
I agree that many other factors come into play besides resolution numbers,
but they do give some indication. Why not take some photos with the normal
lenses of the same subjects, using the same film at various apertures? Find
which images give the same depth of field in a given subject, and enlarged
to the same print size.
That would be a practical way to see the real differences between these
cameras and lenses.
Dan Beaty
Columbus, Ohio USA
www.livingtruth.com
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Sharp at f8????
Date: Mon, 17 Mar 2003
"Dane Brickman" [email protected] wrote:
>It's generally accepted that the best results are obtained around 5.6 and
>f8.
>
>The way I read the material, that would apply regardless of the aperture
>"range" of the lens... like with the Sony, you'd shoot stopped down for max
>effect.
>
>Anyone have any experience with the Sony in this
>regard? Or know of any material that may take a contrary view with lenses in
>general?
????????????
As with most "rules of thumb", the numerous exceptions
make them nearly useless... As for the above, you do
not say what "Sony"... (camcorder, camera, CCD size,
etc.), and the rule of thumb varies with format, and
lens type. For 35mm, f8 is more often than not the
optimum aperture for center to corner optimization,
though there are many exceptions (WAs and zooms often
require more stopping down for corners; some lenses
perform optimally at f5.6 or even at f4). For other
formats and good gear, the optimum stop may often be
much smaller (large format), or wider (good video
lenses).
David Ruether
[email protected]
http://www.ferrario.com/ruether
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Target for testing lenses
Date: Sun, 16 Mar 2003
[email protected] (Bob S) wrote:
>"Yi-Zen Chu; Yiren Qu" [email protected]
>> Hello everyone,
>>
>> I am wondering if there's any test patterns I can find on the internet I
>> could print out with my printer (HP 710C) so that I can paste it on the
>> wall and test my about-to-arrive Nikkor?
>>
>> On a similar note, how else should I test my lens? What kind of subjects
>> do people use? What about color rendition?
>>
>> Thanks!
>>
>> Yi-Zen
>Since I don't take pictures of test patterns I take shots of my normal
>subjects, people, cars, etc and loupe the slides. Tells me more than
>a lens test pattern.
>
>Bob
I agree. I use the same distant (solves focus shift with
angle problems) detailed scene, shot twice without focus
change - once with the shutter release up (vertical frame),
once with the camera inverted (same framing). This allows
you to compare a horizon line imaged the same way across
the frame line. I also shoot a detailed horizon line
placed diagonally across the frame from corner to corner,
then reversed without changing focus. These, at f2/2.8
F5.6, and f11 for a quick check. I also shoot straight
down at an even texture like concrete, usually at f5.6,
to check closer-focus performance (often different from
infinity-focus performance, which is one reason I dislike
chart-testing [also, chart-testing is too critical of
*exact* focus for accuracy]). (See:
www.ferrario.com/ruether/slemn.html.)
David Ruether
[email protected]
http://www.ferrario.com/ruether
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Sharp at f8????
Date: Mon, 17 Mar 2003
"Bhup" [email protected] wrote:
>Only if you mount the camera on a good tripod. have a lens hood
>then its generally 2 stops from your max. depends on lens
>I only know of one lens which breaks this rule the Nikon 180mm AFn ED this
>one seems sharp wide open
The "two stops from maximum" is so rarely true, it should
have disappeared as a valid "rule of thumb" LONG ago...
For 35mm lenses, "f5.6-8" is more commonly true, but there
are a FEW exceptions on the wider side, and many on the
smaller (especially with super-wides and zooms, for the
image corners). BTW, many good lenses are sharp wide-open,
but their sharpest stop center-to-corner still is around
f5.6-8...
David Ruether
[email protected]
http://www.ferrario.com/ruether
From: "David J. Littleboy" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Sharp at f8????
Date: Tue, 18 Mar 2003
Dane Brickman" [email protected] wrote
>
> Yeah, that's what I always thought too... but looking over a Zeiss
> publication, it seems that sharpness and range of f-stops have nothing to do
> with each other. I'm wondering if I'm misreading this
> http://www.dantestella.com/zeiss/resolution.html
That's Zeiss hagiography, and doesn't mention imperfections. Lens defects
mean that lens performance is usually pretty poor at wide open. As you stop
down, you use less of the lens, and the imperfections cause less
degradation. At which point diffraction comes into play.
Look at the wide open vs. f/8 MTF charts at
http://www.photodo.com/nav/prodindex.html for just about any lens. Even the
sainted Zeiss Contax G Planar 45/2.0, one of the best lenses ever made, is
pretty poor at f/2.0.
Here are some numbers to look at:
http://www.hevanet.com/cperez/MF_testing.html
David J. Littleboy
Tokyo, Japan
Date: Sat, 5 Apr 2003
To: Russiancamera-user [email protected]
Subject: Re: [Russiancamera] Re: Some resolution numbers
From: Zhang XK [email protected]
Hi Jay and Kevin,
there is a Chinese national standard for testing photo lens film
resolutions (GB9917-88) that specified resolution figures for various lenses
for various grades. For example:
For fixed normal lenses for 135mm cameras with focul lengthes = 38mm-61mm,
J0 40 lp/mm center 25 lp/mm edges
J1 36 lp/mm 22 lp/mm
J2 32 lp/mm 16 lp/mm
J3 28 lp/mm 12 lp/mm
And standard film of GB100/21 with resolution of 95-120 lp/mm should be
used.
I assume the figure the author has quoted used this method.
Many modern lens tests used high resolution films of 150-200 lp/mm.
I have seen some test reports by using high resolution films to test some
Russian Normal lens such as Helios-44 against some of the top brand Japanese
Normal lenses and the results are very comparable.
If I can find these reports, I will post them here for reference.
Regards
> Hi Jay,
> Very true, but from what little I have been able to find out,
> the GOST system was different, just as the old ASA system was
> different from the DIN systems for film. Likely it was a
> combination of factors that resulted in the lower numbers.
> I rather expect that the GOST test methods specified a
> standardized film for this. Whether the older film was of
> lower resolution is problematic. I sort of think that the
> older thick emulsion films might have actually been of higher
> resolution than the average film these days, but then, I have
> no
> proof.
> Take care!
> Kevin
Date: Sat, 5 Apr 2003
To: Russiancamera-user [email protected]
Subject: Re: [Russiancamera] Re: Some resolution numbers
From: "Rob K." [email protected]
Hi Kevin,
I've seen a testreport giving 50/30 for the 44M7.
other data :
- Volna 9 : 47/30
- Summicron 2/50 : 50/27
- Helios 103 : 55 !/28
1) I don't know if this were serious tests or just a collection of data.
Therefore I'm very interested in Zhangs info.
2) Russian tests are always with fully open aperture, so at f8 the results
are higher.
Regards Rob K.
From: [email protected] (Neuman - Ruether)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Digital SLR,s and quality lense
Date: Thu, 17 Apr 2003
Jim Davis [email protected] wrote:
>"P�l Jensen" [email protected] wrote
>>> If the digital SLR,s like my S2 only use the 2/3 center portion of the
>>frame
>>> then lense choice should be slightly "less critical" than film SLR,s
>>> because corner softness is a given on less expensive lenses?
>>> This woulds see another small benefit of digital SLR,s.
>>>
>>> My less expensive lense are now just a little better on the SLR?
>>No. It is usually the other way around. All things equal, the smaller the
>>format the bigger part the lens resolution provide of the total resolution
>>(lens + medium resolution). This is why lens quality is seldom an issue for
>>MF and LF photography as the lens resolution is a smaller contributor to the
>>overall quality compared to smaller formats. This means that if you have a
>>small digital sensor and a large digital sensor, lens quality differences
>>will be more apparent on the smaller sensor.
>Consider the image circle any lens projects onto the image plane.
>Where exactly is it always softest? Why, it's at the edges. These
>edges are cropped from the smaller digital sensor. Need I say more?
>Jim Davis
>Nature Photography
>http://www.kjsl.com/~jbdavis/
Uh, you're both "right"...;-)
While inferior lenses do tend to be worse at the edges
than at the center, they are often also worse in the center
at wider stops than better lenses. All lenses perform about
the same once they hit diffraction limiting, but for a
good lens, this may be around f5.6 in the center and f8-11
in the "corners" (at the edge of the image circle, appearing
in the full-frame rectangular image), while for the poorer
lens, diffraction-limiting may not appear until around
f11-16 in the center, and may never be reached in the
corners with the available stops... Also, for a given imager
resolution, doubling the lens resolution does not double the
final image resolution - the resolution of the lens and
sensor are interactive, with the resultant resolution likely
less (and never greater) than the lesser of the two (which
is why fairly poor lenses and really good lenses do not look
"wildly" different in the image - though they do look
different...;-).
David Ruether
[email protected]
http://www.ferrario.com/ruether
From: Mxsmanic [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Goodby Nikon F Mount :-(
Date: Fri, 13 Jun 2003
McLeod writes:
> There's a lot more to a good digital image than resolution.
The importance of images characteristics varies with their magnitudes.
In other words, poor resolution is important, because it is highly
visible and may ruin an image. But very high resolution is practically
insignificant, because anything above a certain threshold is more than
enough, and at that point other image characteristics are more visible
and important.
So you need to avoid low resolution, but you don't necessarily need very
high resolution. Overall, you just need a good balance of all image
characteristics, based on the general image quality that you want and
expect.
--
Transpose hotmail and mxsmanic in my e-mail address to reach me directly.
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 23 Jun 2003
Subject: Re: polaroid & Max HQ Re: Lenses - yesterday, today and tomorrow
> it amazes me
>how much emphasis is placed upon resolution.
Resolution is alot like Horsepower. One could say the same thing for cars.
"It amazes me how much emphasis is placed upon horse power when color, price,
looks and how the car handles are important".
Well ususally high horsepower comes right along with color, price, looks and so
on. Same with resolution. Where you have high resolution, good contrast,
color and so on are not far behind.
Larry
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 28 Jun 2003
Subject: Re: Lenses - yesterday, today and tomorrow
>Again?????
>And please, could you explain why I'm getting 30 to 76 lp/mm *with the
>same film*, depending on the aperture and the lens?
Simple. A typical name brand $200 lens will do about the following in noon
summer sun.
f-2 350 l/mm
f-4 500 l/mm
f-8 250 l/mm
f-16 125 l/mm
f-32 64 l/mm
Now go through all the f stops with the calculation (FRxLR)/(FR+LR) using 100
l/mm for the film.
Your answers might range between 30 to 70 l/mm.
Also, you must allow for possible focus errors in your camera system. To do
this, set the camera at 20 ft but photograph targets at ranges from 16 ft to 26
ft which would be about a 400 micron focus spread.
Most store purchased cameras have at least 50 microns of focus error. Its not
uncommon to have as much as 150 microns of error. So what you see in focus on
the mirror and ground glass is not exactly the focus on the film.
Larry
From: Christoph Breitkopf [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Good 35mm and 6x7 film vs. digital comparison
Date: 03 Jun 2003
Gordon Moat [email protected] writes:
> It also seems that changes have occurred, even at Leica. Around the time of the M6
> introduction, that family of lenses started to take on a higher contrast
> appearance in the final images. I am not sure why that direction was taken, but
> many examples that I have seen from newer Leica lenses show less of the smooth
> tonality of the older lenses. The few designs that have not changed give largely
> the same results as older gear. I prefer the look created with the older lenses,
> though some consider the results to be softer.
Some information about the changes in the M-lenses are in
this very interesting (even for non-Leica people such as me)
article by Erwin Puts:
http://www.leica-camera.com/imperia/md/content/pdf/msystem/49.pdf
He says among other things, that the bokeh has somewhat suffered
for better sharpness.
(Also available in German:
http://www.leica-camera.com/imperia/md/content/pdf/msystem/48.pdf)
Regards,
Chris
From: "William D. Tallman" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Lens tests
Date: Sun, 06 Jul 2003
Don Stauffer wrote:
> I've used the old NBS (now NIST) targets, no longer available, and the
> Edmund Scientific poster chart. The later is about the only reasonable
> price one available other than by downloading. It is easy to use, the
> instructions are printed right on chart.
Okay, I'll check that out. Thanks.
> What I worry about with downloaded charts is, is my printer going to be
> the limiting factor?
Thought about that. Takinami's site addresses this by pointing out the
limiting test patterns. In the USAF chart, IIRC, pattern 1,4 is said to be
150 lpm and 1,1 to be 100 lpm. I printed the chart on an HP Office Jet
135, using glossy paper and 1200dpi. There was somewhat better clarity
than the same pattern printed on plain paper at 300dpi, but both trigram
pairs were clearly visible through a loupe. There was, however, some
amount of fuzziness.
Then I thought about simply increasing the distance between lens and
targets, thereby providing the lens with more sharply defined patterns.
That seems doable.
WRT Koren's patterns, he has factored in these parameters. His tests use
patterns printed out on an Epson 2200, which I have. I figure there should
be some significance to the data acquired, but to determine exactly what
assumes more knowledge than I have (an analysis of his work, which is
apparently unavailable).
Thanks!
Bill Tallman
From: "Malcolm Stewart" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Lens tests
Date: Sun, 6 Jul 2003
William D. Tallman [email protected] wrote
> Malcolm Stewart wrote:
>
> > If you're testing the lens, go for the finest grain B&W film, exposed
> > optimally
> > for this type of test and processed correctly.
> What's "S.D"? Never mind, I'll google it...
From: "Malcolm Stewart" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Lens tests
Date: Sun, 6 Jul 2003
If you're testing the lens, go for the finest grain B&W film, exposed optimally
for this type of test and processed correctly. Tests will probably be necessary
to establish what these are. You'll also need to eliminate camera shake because
the best B&W films (as above) & some lenses can resolve hundreds of lpm. You'll
also have to decide how to get different levels of contrast (and glare perhaps)
into your target. It won't be easy, and must be reproducible time after time!
You'll need at least a reasonable microscope for viewing the targets (at
extinction) on your negatives. Will your tests be confined to "the focused
plane" or will you examine curvature of field, coma and astigmatism, colour
aberrations, linear distortion etc. Are you familiar with working out the S.D.
of your observations?
Good luck, I've been there, and it's not easy to do time after time with real
certainty.
If you're testing a system of lenses and film together, just use your standard
film etc. ...and enjoy!
(My sample of the 50 f1.4M is one of the best 50s in my collection (when tested
with slide film) with a very clean image.)
--
M Stewart
Milton Keynes, UK
www.megalith.freeserve.co.uk/oddimage.htm
From: "William D. Tallman" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Lens tests
Date: Sat, 05 Jul 2003
Decided to test my Pentax lenses; want to see just how bad those Pentax-M
lenses really are... {grin}
There's a variety of resources available on the web, including targets for
downloading and printing on a high resolution printer. USAF-1951 targets
are available at http://www.takinami.com. Norman Koren's site
(www.normankoren.com) offers much information and some lens test targets
apparently intended to provide enough data to calculate MTF figures. And
I'm sure there's more, but that's what I discovered in a cursory search.
Anyone had any experience with these downloadable targets, and/or with
Norman Koren's testing system? Any other systems/products recommended?
Haven't decided what film to use. Recommended are slide and fine grained
B&W film. In both cases, I gather that having the film processed and left
uncut (and unmounted) is preferred. Dunno whether a choice made on
processing convenience is valid; who does slide film that will leave it
unmounted, and/or do I want to fool around with doing my own B&W processing
when I don't have a darkroom setup. Is there a more valid consideration
for choosing?
Thanks all....
Bill Tallman
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 29 Jun 2003
Subject: Re: Lenses - yesterday, today and tomorrow
>My texts all say, with caveats that don't matter in this context, that
>the best possible in air is f/1600 where f is the f stop number, e.g.,
>5.6.
If you read the books closely you will see that 1600 changes with the
wavelength.
It's like the sunny 16 rule. It's really f22 in the summer with summer sun
being about 10,000 foot candles. Then f11 in the winter when the winter sun is
about 2500 foot candles. Over all average is 5000.
So noon summer sun, the figure is more like 2000 and winter sun, around
Christmas time, it's down around 1800 or so. The 1600 figure is from a
standard of a bees wax candle which is around 3600K and the method of general
lighting when the figure was first published.
Larry
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 18 Aug 2003
Subject: Re: zeiss ikonta for shooting?
>I've had excellent results shooting landscapes with the Tessar at f/11 or
>smaller. The Novar at f/11 was soft in the corners.
After running resolution targets on at least 100 folders of all types, I have
found out about 80% of the lenses are soft in the corners regardless of lens
manufacture.
Best seem to be the Kodak Tourist Rapid 400 shuttered 101mm lens plusI have
found several lowly Novar f4.5's not far behind.
Larry
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 21 Aug 2003
Subject: Re: zeiss ikonta for shooting?
>I guess if you're main interest is shooting small objects at medium
>distances in the center of the frame then your test proves cheap lenses are
>as good as the more expencive versions.
It seems to me, I never said anything about edge sharpness. I just said I
found several Novars that did a good job.
I do test for both center and edge sharpness. All you need is more then one
target and I do have more then one target.
Out of the 100 or so folders I have tested,
80% really have no edge sharpness at all regardless of lens design. Just good
for 1:1 prints you see in most old albums. About another 15% have at least
some edge sharpness for an 3x blowup to 8x12.
Last, about 5% have enough edge sharpness to go 16x24 which includes several
Novars I have tested plus just about every Kodak Tourist camera with the 101
lens.
Larry
From: [email protected] (Hemi4268)
Newsgroups: rec.photo.equipment.medium-format
Date: 20 Aug 2003
Subject: Re: zeiss ikonta for shooting?
>So explain a tessar? It's not symetrical but still has 4 elements..
This is easy. A copy lens for 1:1 work is symetrical. A lens intended for
distance work will have smaller back elements to correct for distortion at a
distance. A slight non symetrical design is for simi-close work such as table
top setups. The greater the non symetrical design, the more distance the lens
is designed to work in. So you will find for landscape work, the front set of
elements of a lens will be rather large to the back set of elements.
Larry
From Leica Mailing List:
Date: Thu, 02 Nov 2000
From: imx
location colour brightness
15,1 133 23
15,2 4 19
15,3 778 111
Date: Sun, 5 Nov 2000
From: "Erwin Puts" [email protected]
Subject: [Leica] more about film
Date: Sun, 05 Nov 2000
From: Richard Knoppow [email protected]
Subject: Re: [Rollei] zeiss vs schneider
>At 03:38 PM 11/4/00 -0800, you wrote:
>> we will include a plain brown dog
>
> I want my Bass-O-Matic!
>
>John Hicks
>
>[email protected]
Richard Knoppow
Los Angeles,Ca.
[email protected]
Date: Thu, 9 Nov 2000
From: "Erwin Puts" [email protected]
Subject: [Leica] Sharpness and tonality
Date: Sun, 12 Nov 2000
From: "Erwin Puts" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: who cares what the pros use?
> [...]
> You're right on one level. We spend too much time arguing about the latest
> gear when the most important factor is the brain behind the lens.
Date: Mon, 04 Dec 2000
From: Jim Brick [email protected]
>Jim,
>
>What about a Digital camera accepting M lenses ?
>
>Lucien
Date: Mon, 04 Dec 2000
From: Jim Brick [email protected]
Subject: Optics for digital imaging white paper
Newsgroups: rec.photo.equipment.medium-format
Date: 11 Dec 2000
Subject: Re: Leica vs Medium Format??
From: Thomas Bantel [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Sigma vs. Canon 180mm Macro Lens?
> >>to let you do a side-by-side extensive field test between these two lens?
>
> As I said in another thread, I feel that Side-by-Side tests should only be done
> by publishers that have a full optical bench.
>
> Otherwise, the results may not be scientifically valid.
>
> Some European magazines have the equipment to do that, and perhaps one of them
> will publish such a Review. People here often mention the outcome of such
> reports.
>
> Peter Burian
From: johnchap [email protected]
Newsgroups: rec.photo.equipment.35mm,rec.photo.equipment
Subject: Lens Testing Kit
[Ed.note: Mr. Erwin Puts is a noted lens tester and Leica expert who has
published numerous articles and even a CD-ROM on Leica lenses etc...]
From Leica Mailing List:
Date: Sat, 30 Dec 2000
From: imx [email protected]
Subject: [Leica] Exploring the limits (part 1 of several)
Date: Mon, 1 Jan 2001
From: "Erwin Puts" [email protected]
Subject: [Leica] exploring the limits (part 2)
From Leica Mailing List;
Date: Mon, 1 Jan 2001
From: "Erwin Puts" [email protected]
Subject: [Leica] exploring the limits (part 3)
90 lp/mm in Gigabit with Rodinal (both Agfa products).
90 lp/mm with TechPan in FX39
80 lp/mm with TechPan in Technidol LC (both Kodak products)
80 lp/mm in APX25 and TX39
70 to 80 lp/mm in Tmax100 in FX39.
Date: Wed, 3 Jan 2001
From: "Erwin Puts" [email protected]
Subject: [Leica] limits part 4
From: [email protected]
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Resolving power of a 135mm Optar
> Folks,
> I've got a 135mm Optar on my Crown Graflex. I'm trying to determine if it
> makes sense to spend another $300-$400 on a newer 135mm-150mm lens.
> Nikkor-W, Schneiders and Rodenstock are all recommended but the question is
> what do I get for the additional dollars.
>
> I've been told the resolving power of the Optar is only 60 lpm and my film
> choice, TMAX 100 and 400 should resolve to 125 lpm and 200 lpm respectively.
> If this is true then I'd make the investment in the newer lens otherwise
> I'll keep the Optar.
>
> The newer lens will also provide better contrast but I think I can manage
> that with development, paper grade and Photoshop.
>
> Can anyone verify the resolving power of the Optar ? I did not find the
> info on www.graflex.org, many thanks
Date: Fri, 19 Jan 2001
From: Christer Almqvist [email protected]
Subject: [Leica] resolution limits
- --
Christer Almqvist
Date: Tue, 27 Feb 2001
From: imx [email protected]
Subject: [Leica] Filter measurements with MTF
From: "Koren, Norman L" [email protected]
To: "'[email protected]'" [email protected]
Subject: Lens, film resolution, etc.
Newsgroups: rec.photo.equipment.large-format
Date: Mon, 26 Feb 2001
Subject: Re: lpmm ... tutor ??
>Could someone explain to me the basic principal of the lpmm measure ?
>(a short "101" of this term).
>
>How does it relate to the Airforce bar target ?
>
>Thanks.
>
>ZR
Los Angeles, Ca.
[email protected]
Date: Tue, 17 Apr 2001
From: Edward Meyers [email protected]
Subject: RE: [Rollei] Slightly OT: Quality of Zeiss lens design versus new
CAD designs
Date: Thu, 26 Apr 2001
From: Richard Knoppow [email protected]
Subject: Re: [Rollei] Lens Resolution, Old and New Lenses
>Dear All,
>Someone over on the Cosina/Voigtlander list wrote in to say that the 1960s
>Tessar on a Rollei T roughly resolves twice as much as a Tessar of the
>1930s. Similarly, todays lens designs might resolve three times as much as
>a 1930s lens. One reason for this is the ever increasing resolution of
>films over time to which lens designers have responded. Is this broadly
>correct?
>
>sincerely,
>
>David Morris
>David Morris ([email protected])
Richard Knoppow
Los Angeles,Ca.
[email protected]
Date: Thu, 5 Apr 2001
From: "Alan NAYLOR" [email protected]
Subject: Re: [CONTAX] Med Formt
> "MF lenses don't have near as high MTF as 35mm lenses (simply
> because of optical properties that go hand in hand with having to cover a
> larger image plane), so though you have a larger negative, you actually have
> to use "more" of it to equal 35mm 'quality'."
>
> True...most of the time. But if one compares exceptional lenses, such as the
> 100 Makro for 35mm, or 120 for 645, they both will outperform the film
> itself. In that case, then film size alone is indeed the deciding factor. On
> lenses the quality of Zeiss (and I'm talking about both 35mm and MF lenses),
> the MTF factor will still be considerably offset by the larger film size. MF
> always wins! (Unless one has a Holga, I guess.)
Date: Thu, 28 Jun 2001
From: [email protected] (VT)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: SQF Testing - was Re: Pop Photo tests Tamron 24-135
>Just thought of this
>
>Resolution testing can be done using
>film -- which Pop Photo does
>or
>filmless -- where one literally views the target grid through the lens
>- ie: aerial resolution - the latter usually results in much higher
>figures, since there is no degradation due to film.
>There are formulas for "system" resolution which varies depending on
>the source - but usually it is:
>
>1/(Rs**2) = 1/(Rl**2) +1/(Rf**2)
>
>where
>Rs = system resolution
>Rl = lens res
>Rf = film res
>
>If we used the same film as Pop Photo of TMX at 200lp/mm and the
>highest res figure from Amateur Photographer of 122 lp/mm
>
>Then using the formula
>1/(Rs**2)= 1/(200**2) + 1/(122**2)
>results in
>Rs = 79 lp/mm
>which is pretty close to the Pop Photo high figure.
--
Vincent
[email protected]
Date: Thu, 12 Jul 2001
From: Richard Knoppow [email protected]
Subject: Re: [Rollei] APX 100 in D 23 1:1
Richard KnoppowB
Los Angeles,Ca.
[email protected]
Date: Fri, 20 Jul 2001
From: Rick Housh [email protected]
Subject: Re: sharpness and bokeh
>Please excuse my rookie questions.
>as far as I know I just want the very clearest and finest possible small
>detail in my photos with a minimum of fuzzyness.
>( like when you get a magnifying glass to see more )
>
>I assume that is "sharpness". I'm not sure how that differs
>from contrast ? ?
Date: Thu, 24 May 2001
From: John Albino [email protected]
Subject: Re: Re:Lens testing (was Nikkor 85/2 AIS)
>Should lens testing be done with print or slide film?
John Albino
mailto:[email protected]
From: johnchap [email protected]
Newsgroups: rec.photo.equipment.35mm,rec.photo.equipment
Subject: Lens Testing Instructions Repost
Date: Tue, 12 Sep 2000
From: [email protected]
To: [email protected]
Subject: Re: Final Resolution - the real answer?
Date: Wed, 11 Jul 2001
Newsgroups: aus.photo
Subject: Re: good vs bad lens?
2/ Contrast (solid colour definition)
3/ Flare (multi-coat reflections - Pentax is king)
4/ Even lighting across whole frame (vignetting) - hard to do on very fast
lenses
5/ vignetting - blur caused by 'rainbow' on edges - ED or APO glass is the
cure
6/ fast lenses have more uses than slow lenses f2.8 is better than f5.6 on
an SLR
7/ Focus speed - if AF
8/ Noise during focus
9/ Weight
10/ Filter size (odd sizes are very expensive)
11/ Prime or zoom - very very very few zooms actually do the job.
12/ Normal or macro prime focal - macro lenses for close up shooting but
make perfect hi-res lenses.
> sorry for the generalised question but i hear alot of people in here talk
> about 'good quality' lens...what exactly are the sort of things that makes
> one lens better than another??
From: "Al Denelsbeck" [email protected]
Newsgroups: rec.photo.misc
Subject: Re: Minolta... sharpness test.
Date: Sat, 21 Jul 2001
From: Tony P [email protected]
Subject: Re: sharpness and bokeh
>Date: Fri, 20 Jul 2001
> From: Scott Perkins [email protected]
>Subject: sharpness and bokeh
>
>Please excuse my rookie questions.
>
>as far as I know I just want the very clearest and finest possible small
>detail in my photos with a minimum of fuzzyness.
>( like when you get a magnifying glass to see more )
>
>I assume that is "sharpness". I'm not sure how that differs
>from contrast ? ?
>just a sentence please. what is bokeh ?
>I'd like to know if I like it : )
Tony P
From: Bill Tuthill [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Testing barrel/pincushion, distance?
> With a nice flat lens test chart. Well that does not really quite
> do it unless you intentionally change the distance over several
> exposures. BTW I would not be surprised if that also was somewhat
> dependent on the distance.
From: [email protected] (Robert Monaghan)
Newsgroups: rec.photo.equipment.medium-format
Subject: sample variation, beating 100 lpm, Re: Photodo or PhotoDon't ???
Date: 1 Oct 2001
sample variation is probably one order of magnitude more important than
any modest variation in setting focal plane positions during MTF testing
;-) I've been collecting more data from lens resolution and contrast
testing, and variations of 12% in resolution are "nominal" (based on USAF
1951 charts). As near as I can extrapolate, it looks like prosumer lenses
vary circa +/- 1/2 grade out of 5 (cf. photodo) during reported tests.
Moreover focusing errors of as little as 2mm rotation (1mm on many AF
lenses) (see mf/critical.html) can reduce lens resolution by up to 50%, or
more (at f/5.6 or faster). I suspect this explains many of the sometimes
anomalously high or low lens resolution values reported during a series of
such tests on many lenses (e.g., Pop Photo..).
another interesting study http://people.smu.edu/rmonagha/mf/limits.html#100
shows how many cheapy 50mm f/1.8 optics can deliver over 100 lpmm (!!) on
HCF copy film and microektachrome color slide films. Many such lenses are
evidently running into film resolution limitations rather than inherent
lens resolution limits (see mf/lenslpm.html for film resolution values).
One tester consistently gets 1/3rd higher resolution values by using
strobe lighting of targets (Dr. Eugen J. Skudrzyk) which helps eliminate
the effects of small camera vibration at these high resolution testing
limits.
My impression is that we are really testing our procedures for testing the
lenses, rather than really testing the lenses ;-) And even if we had a
very careful lens test report, it would only apply to that particular
lens, and probably not to the one we end up buying and using ;-)
grins bobm
--
From: [email protected] (Stuart Bobb)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Any tips on testing lens
Date: Mon, 01 Oct 2001
>
>Check that the diaphragm stops down instantly when the relevant linkage
>is operated. Excess oil can affect the operation of the iris diaphragm
>blades even before it becomes visible.
Good stuff! A few more in this space are:
a) Check that the aperture springs back open fully. I've seen lenses where
the blades looked fine but after stopping down to f16 or f22 they did not
spring 100% back open.
b) Make the the blades come in cleanly to form a hexagon (octagon, etc). If
some blades don't go in as far as others, the resulting polygon will be oddly
shaped. How odd does it need to be before you get problems? I don't know.
Two other simple tests.
Check the close focus. Does it focus as closely as the specifications for
the lens state? Check infinity -- find something that is quite a distance
off and make sure that your split image (or whatever focusing aid you have)
comes into alignment _before_ you run out of ability to turn the focus
barrel.
In other words, confirm that infinity is a bit further away than 90 or 100
feet. :-)
Look through the lens at a brightly lit white wall. Is the glass giving
everything a hue or tint that maybe isn't expected?
Stuart
From: "Webmarketing" [email protected]>
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Photodo or PhotoDon't ???
Date: Mon, 1 Oct 2001
The only really useful way to test lenses meaningfully is to make photos of
known repeatable subjects and then view the negatives or chromes with a high
power loupe. It's the way I test my lenses. Only in that way can you see
the effects of lens design on corner sharpness, contrast, distortion and
aberrations. There are always tradeoffs. In order to get a high MTF
(corner sharpness) one has to give up some contrast and vice versa. In
order to get rid of coma you need to give up some MTF etc. etc. These are
just a couple of examples. No magazine does this kind of testing and
Photodo certainly doesn't either. I agree with the individual who said that
a personal test is the only really meaningful one except that I would
include the need to shoot known and repeatable subjects and view the film
directly. Good shooting.
Fred
Maplewood Photography
"T P" please.reply@newsgroup> wrote...
> "Jay Washington" [email protected]> wrote:
>
> > Okay, I'm not trying to start a flame war, but I would like someone to
name
> > me another site that is something other than just a list of user
opinions.
> > I'm not asking as a challenge, I would like as many sites as possible
that
> > at least gives me some objective data and is not sponsored by the
> > manufacturers. Really, I would.
>
> Nikon is well served by sites that give more useful information than
> Photodo. I don't know about other lens manufacturers.
>
> > I don't understand why photodo get such a
> > bad rap. No one is constantly slamming Popular Photography's
reviews--maybe
> > we all know those are skewed, but I really am confused as to why
Photodo, a
> > free source of information, always gets rained on.
>
> Because its subjective approach to testing lenses, and their
> presentation of test data as an overall rating, are profoundly
> misleading. "Bait and Switch Monthly" is already to be subjective,
> biased and grossly misleading. Most people already know that, and
> distrust the magazine, so there's no need to repeat it on here.
>
> > For me at least, they've
> > been accurate--even after the fact (I bought a Canon 20-35mm non-L and
was
> > surprised by the lack of sharpness--when I found out about photodo, it's
> > rating matched my experiences). Every lens I've purchased since then
with a
> > 4.0 or greater photodo rating has been exceptionally sharp (Canon 85
1.8,
> > Canon 50 1.8, Sigma 105 2.8). My Canon 24mm TS is not so sharp, but
it's
> > okay and is amazing for what it does (perspective correction). This
okay
> > sharpness matches with the photodo rating. I guess I've just been lucky
> > that photodo has been accurate for my use.
>
> There is much, much more to learn about a lens than its MTF curves. If
> you really think MTF is the be-all and end-all of lens testing, then
> maybe it's time you learned a little more.
>
> > Anyway, good luck with whatever sources you use before you purchase.
>
> The best and most reliable test of all is to test it *yourself*, after
> you buy it, and return it if it isn't up to scratch.
>
> Always buy from a store with a good returns policy, and test with a good
> slide film such as Fuji Provia 100F. Use a lightbox and a good loupe.
>
> --
> Best regards,
> TP
From: "Meryl Arbing" [email protected]>
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Photodo or PhotoDon't ???
Date: Tue, 2 Oct 2001
I have found that there are two extreme factions on the subject of lens
testing. On one side are those who absolutely refuse to believe any test
results at all and claim that it is, in fact, impossible to obtain any
meaningful test results because of "sample variation". They are the ones
that claim that the only thing that matters is their subjective
determination of the quality of a lens based upon their actual results.
While this approach may seem reasonable on the surface, it is doomed to
failure. If it were true that lab testing of a specific lens is unreliable
because of sample variation then would it not also be true that subjective
"field" testing of a lens would also be unreliable because of the same
variations? Would the self-testing user not have to go through dozens of
different lenses (and several dozen rolls of film) to make sure that he got
one of the 'good" ones? How can you take their subjective opinion on the
quality of a lens as meaningful to YOU if they have given the lens a 'bad'
rating because they got a lemon out of a run of excellent lenses or did they
give a "first class" rating because they lucked out and got the one good out
of a bad batch?
It all then just a matter of the luck of the draw and you stand just as good
a chance of getting a totally fantastic lens for $100 as getting a
completely "lousy" lens for $1000.
At the other end of the extreme are those that put complete faith in MTF
charts published by the lens manufacturers. Very few manufacturers (and
certainly none of the 'cut-rate' ones) actually publish MTF graphs based on
tests of production lenses. Most of the charts you see are computer
generated "projections" of how the 'ideal' lens in that design should
perform. The lens designer could well have his 'perfect' design
specifications overruled by the marketing department and have design
compromises introduced in order to reach a certain price-point. The
production lens specs may be quite a bit different from the design specs
but, what do you think the manufactuer publishes when asked for the MTF
charts?... the 'idealized' specs and, if you are lucky, it may actually get
pretty close if it is not a complex design.
Those who have a slavish devotion to the MTF charts could find themselves as
frequently surprised as the MTF deniers.
What is the solution? MTF or no MTF? Photodo or Photodon't?
Well, my first piece of advice is to remember that "You get what you pay
for." Don't think that you are going to get optimum quality by scrimping on
the lenses you buy. Buy cheap and you will get cheap even though there are
many people with an enormous capacity for self deception and who will INSIST
that "You can't tell the difference!!" and it only cost him $12.69 at a swap
meet.
The manufacturers with world-wide reputations for excellent lenses did not
get there by producing inconsistent, variable-quality crap. For those who
want to hold on to the cherished belief in "lens quality variation" then, by
all means, stick with the low quality, third-pary lens manufacturers and
they will have all the validation of their beliefs that they crave.
For all the anecdotes about how they got this lens that was supposed to be
really crappy but was really surprised when turned out to be better than he
thought there are thousands and thousands of photographers who bought top
quality lenses and got exactly what they expected without having to worry
about unpleasant (or pleasant) surprises. Part of what you get with a top
quality lens is consistency across the board and, yes, you have to pay for
that assurance.
"brian" [email protected]> wrote...
> Tony:
> Your anti-MTF crusade is really tiresome. Which lens is it that you
> have a disagreement with? MTF is the most objective/least subjective
> method of testing a lens. Are you acusing Photodo of presenting
> falsified MTF curves?
>
> Brian
>
>
> T P please.reply@newsgroup> wrote...
> > [email protected] (DBaker9128) wrote:
> > >
> > > I don't think the burden of proof rests with me at this point.
> >
> > Hi Doug,
> >
> > You are quite right. You have posted a hypothesis on here that is based
> > on the facts as you see them. It's up to any one opposing you to argue
> > their case on the basis of better and/or more accurate evidence, if of
> > course such exists.
> >
> > I will happily side with you here. As you know, for some time now I
> > have had deep concerns about the methodology used by Photodo's lens
> > testers. I simply don't believe some of the test results I have seen at
> > the Photodo site, but of course any lens tester has the get-out clause
> > that is conveniently called "sample variation".
> >
> > > A serious
> > > challenge has been raised as to Photodo's MTF testing methodology by a
widely
> > > acknowledged optical expert, Mr. Erwin Puts. Photodo says nothing to
defend
> > > itself. Why is this?
> >
> > That's because Photodo, for all its web site's pretensions towards
> > offering highly consistent and objective reporting standards, is just
> > another photo mag, no more and no less. There's nothing objective about
> > tests that are based on such a highly subjective approach (even an
> > individual or a personal approach) such as Photodo appear to have
> > employed.
> >
> > As I said, Photodo is just another photo mag. For an indication of the
> > atrocities that get into print in such mags, just look at the BS that is
> > printed every month in a popular American photography magazine
> > (mentioning no names here!) but which masquerades as fact.
> >
> > Yet this dreadful magazine is still very popular. Go figure.
> >
> > ;-)
From: "SWB" [email protected]>
Newsgroups: rec.photo.equipment.35mm
Subject: Re: The top five dumbest Leica owner comments.
Date: 03 Oct 2001
"greg" [email protected]> wrote
> These are ongoing themes on this ng. The top five dumbest things that we
> continually hear from Leica owners are:
>
> 2) Leica outperforms medium format enlargements.
>
> Remember reading comments from the Leica owners who claimed the M6
> would outperform a Hasselblad. C'mon, get a life, your M6 can't
> outperform a hundred dollar YashicaMat 124G. It's a basic law of
> physics. Start with a larger negative and you'll get a sharper 11x14.
>
Well thats one of the stupidest things I have seen you write. Sharpness does
not equate to size. Take an area the same size as a 35mm neg out of a medium
format neg and you may or may not get the same level of sharpness. With a
Leica lense, it will (given the image is in focus, same film etc,) be
sharper than nearly all medium format negs. Erwin Puts, the noted lense
tester, has a new area on his Web site http://www.imx.nl/index.html called
Capita Selecta. In it he discusses Leica vs medium format, and states that
he is sure he can get medium format quality from a Leica generated 35mm neg.
Now before all the Nikon and Canon users get all hot under the collar I
think there is a proviso. Under normal circumstances, using the camera hand
held at anything below 500th sec pictures on any equipment all have a degree
of degradation. So saying that my Nikon is as sharp/contrasty as your Leica
is probably correct. Both are fuzzy, so you couldn't tell the difference.
Put them on a tripod with very slow fine grain film, optimize the exposure
and film development, and the differences start to show. Not many Leicas are
used like that, but the possibility exists according to Erwin. Anyway, trawl
through his site, and for once in your life greg learn something.
Steve
From: "Mxsmanic" [email protected]
Newsgroups: rec.photo.equipment.35mm
Subject: Re: quality - inherent vs observable? Re: Why is Leica so expensive?
Date: Thu, 04 Oct 2001
"Robert Monaghan" [email protected] wrote
> I believe they are deluding themselves if they
> believe they are not suffering severe reductions
> in lens resolution and other quality factors by
> such techniques.
In most cases, yes. However, I have occasionally obtained better than
100 lp/mm at roughly 50% modulation transfer or above on Provia shooting
handheld with a 90mm (Leica) lens, so it can be done, just not very
often or consistently (and probably not often or consistently enough to
make it a useful argument in favor of such lenses). My examples were in
normal daylight, though, and thus at about 1/250 sec., not 1/30. I have
obtained very sharp results at 1/30, but I haven't checked them under a
microscope, and I suspect they aren't anywhere near 100 lp/mm; still,
the major source of blurring in such cases seems to be movement of my
subject, not movement of my camera.
From: "Mxsmanic" [email protected]>
Newsgroups: rec.photo.equipment.35mm
Subject: Re: quality - inherent vs observable? Re: Why is Leica so expensive?
Date: Thu, 04 Oct 2001
"Robert Monaghan" [email protected]> wrote
> With Provia, on a tripod, timer on lens,
> mirror up, fixed lens test chart, with
> zeiss Hasselblad optics, I rarely reach
> 80 lpmm. I'm probably averaging under 40
> lpmm handheld too.
Unless you specify the degree of modulation transfer, these figures
alone are meaningless.
You get more than 80 lp/mm with just about every shot, but the MT is
very low.
I looked at my slides with a microscope, not just under a loupe, so I
could count the dye clouds. I was surprised by the achieved resolution
myself. Not every slide reached this level, but the fact that any of
them did was quite impressive.
From: Paul Chefurka [email protected]>
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Why are Un Leicas So Inexpensive ?
Date: Fri, 05 Oct 2001
"Max Perl" [email protected] wrote:
>I know we have some Leica users which also have e.g. Nikon
>equipment. I think it could be interresting if we could see a
>blow up of two identical images taken with a Leica and e.g. a
>Nikon. It should of course be one the e.g. Nikon lenses which
>a said to be very good. I could suggest the AIS 105/2.5.
>I have always belived if you select your lenses carefully most brands
>have made "super" lenses where it is purely the 35mm format which
>is the limiting factor.
>
>If a blow up is shown it should be possible to see the difference on
>the web?
>
>Max
I've tried it. My effort is at
http://members.home.net/chefurka/Photo/LensTests2/LensTests2.html
IMO it is not possible to conclusively demonstrate the differences between
Nikon and Leica lenses on the web, at least not with the test setup I used.
And given that I tried to be pretty rigorous in my methodology, I think
that if someone shows a photo on the web as an example of the superiority
of lens/camera/system X, the correct response should be amused indulgence.
Paul
From: [email protected] (Karl Winkler)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Why are Un Leicas So Inexpensive ?
Date: 5 Oct 2001
"Max Perl" [email protected]> wrote
> I know we have some Leica users which also have e.g. Nikon
> equipment. I think it could be interresting if we could see a
> blow up of two identical images taken with a Leica and e.g. a
> Nikon. It should of course be one the e.g. Nikon lenses which
> a said to be very good. I could suggest the AIS 105/2.5.
> I have always belived if you select your lenses carefully most brands
> have made "super" lenses where it is purely the 35mm format which
> is the limiting factor.
>
> If a blow up is shown it should be possible to see the difference on
> the web?
Let me start by saying that these threads seem to pop up every few
days, and the funny thing (to me) is that they're just like threads of
this type in other groups. One of the most similar is rec.audio.pro
where people argue about microphones, speakers, amplifiers... etc.
Another one is the rec.audio.high-end, where they argue about this
tube vs. that tube, $10,000 speakers cables vs. $1,000 speaker cables
(get it?).
So here I go, jumping into this one anyway....
I've used a number of different camera systems over the last 20 years
or so, including Pentax, Olympus, Nikon, and most recently, Contax
rangefinders. Based on shooting slides, making B&W prints, doing
scans, etc., I've reached the following conclusions:
1) The photographer still makes the most difference. Today I shoot
with the famous Zeiss rangefinder glass, and yet my images aren't as
good as the ones I see in National Geographic, which were mostly shot
with Canon & Nikon cameras. Those photographers just blow me away.
2) There is a difference between lenses. But what I've found is that
each manufacturer seems to have a few "magic" designs, then some very
good lenses, then some mediocre lenses, then a few dogs. As much as I
love my OM4 for the way it operates (and God I love that multi-spot
meter!), I've really not found the Zuiko lenses to be all that hot (28
f/2.8, 50 f/1.4). That is, with the exception of the 85mm f/2, which
is just a wonderful portrait lens.
3) The Zeiss lenses (and Leitz also) are really amazing. But I've seen
shots taken with Nikkor, Canon, Pentax, etc. that are stunning.
4) Although it's generally not possible to see the difference in
sharpness between lenses by posting pictures to the 'net, you *can*
tell the difference in color rendition, contrast, bokeh, etc., which
are of course the more important factors in the "look" of a lens.
5) Rangefinder systems seem to produce sharper images with better
color and excellent shadow detail. I've heard several theories as to
why this is, but I couldn't say for sure I know the reasons. But by
using less elements, putting them closer to the film plane, and
removing the mirror from the whole works, it begins to make sense. Of
course, rangefinders are a pain for some things, including using ultra
long lenses (like, anything over 90mm on the G1) and of course for
being able to see what is happening with the focusing... but I
digress.
6) There's no way to take "exactly the same picture" with two
different cameras and lenses (unless your subject is a lab chart - and
I for one don't really care to know if one lens reproduces lab charts
better. I'm interested in real world performance).
So I humbly suggest that the only cure for "is that lens better than
mine" is to just stop worrying about it and take lots of photographs.
Isn't that why we're all into this in the first place?
-Karl
From: Bill Tuthill [email protected]>
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Why are Un Leicas So Inexpensive ?
Date: Sat, 06 Oct 2001
Paul Chefurka [email protected]> wrote:
>
> I've tried it. My effort is at
> http://members.home.net/chefurka/Photo/LensTests2/LensTests2.html
>
> IMO it is not possible to conclusively demonstrate the differences between
> Nikon and Leica lenses on the web, at least not with the test setup I used.
> And given that I tried to be pretty rigorous in my methodology, I think
> that if someone shows a photo on the web as an example of the superiority
> of lens/camera/system X, the correct response should be amused indulgence.
Thanks again for your (new) effort! I had seen LensTests1 but not 2.
It *is* possible to see differences between lenses with your test.
But you're saying that getting the focal plane correct is one reason
to be suspicious of your results? And of course all the things
you didn't test:
Differences between center and edge sharpness
Closeup performance
Infinity performance
Flare resistance
Linear distortion
Coma and similar aberrations
Colour rendition
Tonal discrimination
Bokeh
The "Leica Glow"
The "Un Leica Non-Glow"
From: [email protected] (brian)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: which diff. limit rule? ;-)
Date: 11 Oct 2001
Hi Robert:
You're absolutely right about 50mm normal lenses. Interferometry,
aerial image testing or MTF testing might reveal slight differences at
f/8, but these differences would not show up on film or in electronic
format from a CCD. This is probably true for most prime lenses in the
50-135mm focal length range, where lens design is a very
straightforward business and has been for about 30 years now.
530lp/mm might be a stretch, and would definitely require the use of a
narrow band filter to eliminate all chromatic aberration. I did check
a Nikon design I have for a 50mm f/1.4, and it will resolve just
beyond 500 cycles/mm in monochromatic light (550nm) at f/2.8 near the
optical axis. This is not possible at f/4 or slower due to
diffraction. Darn.
You're also right about wideangles, where color fringing usually
prevents really good stopped-down performance. This is why the Nikon
17-35mm/2.8 zoom is actually better than most of the fixed focal
length wideangles that preceded it. The zoom has little or no color
fringing except for a small amount at the extreme wide end.
Brian
[email protected] (Robert Monaghan) wrote
> Thanks, Brian, for sharing these points and info ;-) My impression is that
> quite a number of 50mm OEM normal lenses reach the high 350 to 530 lpmm
> aerial resolution ranges at least centrally (not so hot on edges as you
> noted ;-) per table XVII in Skudrzyk's Photography for the Serious Amateur
> and calculations from known film resolution limits and observed on-film
> system resolutions. The slower medium format lenses costing kilobucks may
> be a bit of a compromise (due to coverage/cost..) but don't seem to leave
> much room for improvement either, at least for central resolution ;-)
>
> my original underlying point was that many 50mm normal 35mm SLR lenses
> tend to deliver reasonably similar resolution performance by the time you
> stop them down to f/5.6 or f/8, and that past f/8 the performance is more
> often limited by diffraction than the quality or cost of the lens might
> suggest. Now for many 35mm wide angles, I wouldn't make such a claim ;-)
>
> This is why I was not surprised that the pentax 50mm f/1.4 1974 screwmount
> lens when used at f/8 as in Popular Photography's comparison against a new
> Leica Summicron 50mm f/2 performed essentially at the same resolution
> levels (per Keppler's standards and eye). Even at the "sweet spot" of such
> lenses (f/4 to f/5.6), resolution performance is still rather good and
> similar. The real differences are more likely to be seen in other criteria
> than resolution (contrast, distortions...) and more easily seen wide open
> and in the corners.
>
> Many 35mm photographers believe that they would get much sharper (&
> better) photographs by buying the kilobuck lenses instead of the cheapy
> normal lenses they currently have. For general shooting, where you can use
> f/5.6 or f/8 "sweet spots", I doubt many folks could reliably tell the
> difference at f/8 (at least for resolution) between a good cheapy OEM lens
> like the Pentax 50mm and a more pricey 50mm OEM lens by Leica or
> Zeiss/Contax (bokeh, yes/maybe, but resolution?). That's my theory anyway.
>
> I'm running a blind lens test now in medium format comparing different
> lenses ranging from $3k US to $40 TLRs. It will be interesting to see just
> how many (or how few) people can reliably sort which shots were taken by
> which lens etc. I suspect a similar test for 35mm optics would be quite
> sobering to many folks who believe a particular brand of lenses is clearly
> superior to other lenses? ;-)
>
> grins bobm
From: "Malcolm Stewart" [email protected]>
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Leica-Konica incompatibility? Film plane gap precision
Date: Fri, 12 Oct 2001
Many years ago when I started using a wide angle lens (Minolta) on my
Minolta SRT101, I noticed that the plane of best focus was not where I'd
focussed on. So I set about checking out the depth of the film plane gap
using a depth micrometer and purpose made flats. The front register rails
were correctly placed but the lugs on which the pressure plate sat were way
too high. After masking off the shutter I ground these down, and results
improved dramatically - particularly on Kodak's Panatomic -X. (Film
transport remained fine.) Later bodies haven't shown this problem.
M Stewart Milton Keynes, UK
Gordon Moat [email protected]> wrote
> And now for something completely different . . .
snip
> http://www.cameraquest.com/nfspy.htm>
>
> To all those so incredibly concerned about film planes, film gates, and
> test charts, perhaps this Nikon body is the ultimate camera. I would
> guess that it cost new much more than any Leica.
>
> Ciao!
>
> Gordon Moat
> Alliance Graphique Studio
> http://www.allgstudio.com>
snip
From: Martin Jangowski [email protected]>
Newsgroups: rec.photo.equipment.large-format
Subject: Re: experimental result surprises Re: non-coated..coated...multicoated
Date: 12 Oct 2001
Robert Monaghan [email protected]> wrote:
> As you noted, the
> real gap is between uncoated and coated optics; the multicoated optics get
> better light transmission but not so much improved flare/ghosting unless
> they are more complex designs (e.g., zooms, more than 5 groups etc. ).
> That makes me feel better, at least, about continuing to use these older
> lenses and worry less about the flare performance I'm getting too ;-)
I have experiences with similiar lenses in MF format: a 3.5/75 lens from a
Mamiya 6 (multicoated, 6 lens Planar type) and a Rolleiflex E3 with a
3.5/75 6 lens Planar (single coated). The pictures made with both are
virtually undistingishable (sp?), the Mamiya lens is marginally more
contrasty. You'll readily see the effects of the much better film
guidance in the Mamiya, the two sharp corners in the Rollei can create
film flatness problems in the Rollei (I use a Stouffer resolution
target enlarged 10x in the center and all four corners of the picture).
It is nearly impossible to see the same high resolution in all four
corners on one negative. Even in high-flare situations it's very
difficult to see the differences between the coatings, Zeiss obviously
knew how to coat a lens even in the 1960s.
Martin
From: Bruce Graham [email protected]>
Newsgroups: rec.photo.equipment.35mm
Subject: help - lens test problem
Date: Fri, 19 Oct 2001
A few words of context. I have replaced my 20 yr old stolen Nikons (FM,
FE, 35/2, 85/2, 50/1.8, 28/2.8E, 100/2.8E, TC201, MD12, zooms, flash,
macro etc) with an Elan 7E with 28-105, 50 f1.4, 100-300 USM and flash.
I nearly bought an FM3A but was seduced by features or updated depending
on point of view.
I have not been really happy with the six rolls I've shot, even at 6x4,
so I decided to test the lenses. The measured performance seems pretty
awful, so I hope I have made an error with the test.
I made printed two letter size copies of the USAF 51 target spiral
downloaded from the web, taped one on the wall at the centre, one
radially near the corner about six feet away, taped some newspaper in
between and used post-it notes to record lens name, focal length and
aperture. I use a Manfrotto tripod, indoors on a concrete floor. I
used single shot autofocus, switched to MF, then used a remote IR
release with mirror lock-up. I used aperture priority, 1.5 stops
exposure comp for the white wall which gave properly exposed negs. Film
was Fuji Superia 100 consumer.
To provide a reference, I borrowed my 88 year old father and his M3 and
collection of old lenses and shot another roll on that (obviously
manually focussed, cable release). He has:
28mm f3.5 Nikon screw mount (scruffy - bent filter threads, cleaning
marks front element)
35mm f3.5 Summaron - nice
50mm f2 DR Summicron (some fungus), f2 screw mount Summicron, f3.5 Elmar
and a Summitar
90mm f4 Elmar
135mm f6.3 Hektor (1937 vintage and beat up looking)
My expectations were that 6x4 prints should look sharp to the naked eye
at all apertures and that I would only be able to really check
performance by buying a decent scanner (which I plan to do soon).
The reality was different.
The 28mm Nikon scruffy screw mount vignetted horribly wide open but was
tack sharp.
All the Leica lenses (except the Summitar which must be very faulty)
behaved were quite sharp even the DR Summicron with the fungus (contrast
was down a little).
The Canon 28-105 was obviously soft to the naked eye on the 6x4 prints
wide open at all tested focal lengths (28, 35, 50, 90 and 105) and when
checked with the loupe, still improving down to f16
The Canon 50 f1.4 was also soft to the naked eye at on the 6x4 at f1.4
and not really sharp until f2.8
The Canon 100-300 USM is not as bad. A slight softness wide open only.
Now from reading reports on the newsgroup, I had assumed that the 28-105
and especially the 50 f1.4 would perform very well.
Possible causes I can think of.
1. I bought from a pro shop, so I might have got rejects (but the
packaging seemed fresh).
2. I screwed up with the test. Could the USAF51 target with its black
white bars fool the autofocus? Should I use one of the side focus
sensors to focus off the bars? Any other traps you can see?
3. Maybe this new fangled plastic stuff really is just crap.
I'd appreciate your help before I go back to the shop. Unfortunately it
is two months since I bought it so I need to be prepared for the "its
normal" or "you broke it" responses.
Thanks - Bruce Graham
From: "Oliver Bryk" [email protected]>
To: "HUG" [email protected]>
Subject: [HUG] Grain etc.
Date: Thu, 25 Oct 2001
IMHO "Image Clarity: High-Resolution Photography" by John B. Williams, Focal
Press 1990, ISBN 0-240-80033-8, is the best book explaining the factors that
affect photographic imaging, their interaction and contribution to lack of
sharpness. AFAIK the book is out of print.
In my copy of the book I found a print-out of an e-mail from Hasselblad that
responded to my query about the CoC used for DoF markings. It is 60 microns.
"The indications on the lens barrels are theoretical distances for different
apertures for a lens without other lens errors. This means that diffraction
effects is (sic) also not included. Our recommendation is to take the
indicated depth-of-field with some care. If you need good quality we
recommend you use only half the indicated DOF. For example if you use
aperture f/11 you should consider DOF for aperture f/5.6 as good quality."
"good quality" strikes me as a bit too modest for some of the finest lenses
in the world...
Oliver Bryk
Date: Wed, 07 Nov 2001
To: [email protected]
From: Richard Knoppow [email protected]>
Subject: Re: [Rollei] Rollei XF35 and Voigtlander VF135
you wrote:
>Dear folks,
>
>I own a VF 135. It takes brilliant sharp photos, and the exposure meter is
>quite accurate. The price I paid of appr. =80 30 is a shame for the quality
>of the lens although the mechanics of the camera are quite junky. However I
>have a problem with the focus:
>
>The rangefinder meter was de-adjusted, so with the help of our beloved list
>members I found the two screws under the leatherette for vertical and
>horizontal adjustment.
>As I don't have an autocollimator available I did fix an object far away
>through the range finder and turned the screws in order to adjust the two
>split images to one image when the focus ring at the lens was at
>"infinity". So now I obtained a perfect range finder image exactly when the
>focus ring came to the infinity stop.
>
>However, after having looked to the first developed rolls of film I have
>the impression, that close objects (appr. 3,5 ft distance) taken with full
>open aperture are not as sharp as they could be. The impression is, that
>the plane of focus of these objects was a little bit more far away.
>However, after checking the infinity stop of the lens ring, that one
>complies perfectly with the range finder images. So, can someone give me
>advise how to readjust the rangefinder? Or is there another screw which
>adjuts the distance of the lens from the plane of film?
>
>
>Greetings
>
>Dirk-Roger Schmitt
>
The easiest way to test this is to photograph a newspaper sheet or
something similar at a slant to the camera. Focus on a particular plane and
indicat it by drawing a line on the paper. If the negative is not sharply
focused on the line you know somethign is wrong and in which direction the
error is.
Make a series of exposures at different distances, all at near the
largest aperture to reduce depth of field to minimum.
The rangefinder can probably be checked by examining the image using a
small bit of ground glass in the film gate (assuming the back comes off or
at least opens. But, you have to make sure the ground glass is oriented in
the true film plane. This is sometimes not so easy to figure out since
different cameras guid the film somewhat differently.
----
Richard Knoppow
Los Angeles, CA, USA
[email protected]
From: [email protected] (aekalman)
Newsgroups: rec.photo.equipment.35mm
Subject: Re: Printable Test Patterns? Where?
Date: 16 Jan 2002
Re:
> http://normankoren.com/Tutorials/MTF5.html
Just a follow-up ... I visited this site around X-mas, and while I
haven't yet had time to do a full film-to-printer MTF analysis, one
thing I _did_ do is print out a chart and then look at it through a
bunch of different lenses on a copy stand.
It was very enlightening to see that I could visually resolve much
higher l/mm at the center of the frame than at the edges.
Norman's site is _very_intersting_ (plus his stuff on tube audio,
too!) and I'd highly recommend it.
Regards,
Andrew
From: David Grandy [email protected]>
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Lens Sharpness
Date: Sun, 05 Aug 2001
The rule of thumb is that lenses - all lenses - are sharpest in their
middle apertures. Since many many large format lenses are one or two
stops slower than 35mm lenses the middle apertures should be a stop or
two smaller as well. I have a large format 300 mm Nikkor M. Its
maximum aperture is f9 (!) so for me to use this "rule" I stop it down
to f32 or so. That lens stops down to f128 by the way.
From: "Gianni Piccoli" [email protected]>
Newsgroups: rec.photo.equipment.medium-format
Subject: Lens Sharpness
Date: Sun, 5 Aug 2001
Hi, I have read that to obtain the maximum sharpness from the medium format
lenses you need to close them at f11 or f16, in comparison with the lenses
for the 35 mm format that do their best at around f8, is this true ?
Why this difference ?
Thanks, Gianni Piccoli
[email protected]
From russian camera mailing list:
Date: Fri, 07 Dec 2001
From: [email protected]
Subject: Soviet lenses resoloution
Lens Middle Edge
Vega 11y 70 40
Industar 50y 60 20
Industar 96y 60 24
Industar 90y 50 25
Industar 23y 50 13
Industar 100y 70 35 another document says 40
I also found a document that says the Volna 9 has a resolution of 47 middle and
30 edge and a second that says 42 middle and 30 edge.
I can find nothing on the Helios 103 1.8/53 for Kiev rangfinders.
Nathan Dayton
www.commiecameras.com
--
From russian camera mailing list:
Date: Fri, 07 Dec 2001
From: "Per Backman" [email protected]
Subject: Re: Soviet lenses resoloution
[email protected] wrote:
>Lens Middle Edge
>Vega 11y 70 40
I found 65/35 at 5X enlargement.
>Industar 50y 60 20
>Industar 96y 60 24
>Industar 90y 50 25
>Industar 23y 50 13
>Industar 100y 70 35 another document says 40
Vega 29u (50/2,8), has a resolution power of 75 l/mm in the center and 60
l/mm at the edge
>I also found a document that says the Volna 9 has a resolution of 47 middle
and 30 edge and a second that says 42 middle and 30 edge.
The factory specs say "not less than 42 in the center and 30 at the edge" at
infinity (at 24cm it is 42/20). This is by fully open aperture, in practice you
would not use it as the depth of field is to small.
It seems to have been a policy to set the specs so low, that no complaints
could be expected.
>I can find nothing on the Helios 103 1.8/53 for Kiev rangfinders.
Helios 103 53/1,8 has a resolution power of 55 l/mm in the center and 28
l/mm at the edge.
Per
From russian camera mailing list:
Date: Fri, 07 Dec 2001
From: "Per Backman" [email protected]
Subject: Re: Re: More than one version of Volna-9
....
Nathan posted the resolutions already, but it is important to remember,
that these figures can not be compared to figures published by companies
in other countries, the Soviet numbers are always low in comparison. Out
of paractice I can say, that at least Industar-23u is a good lens,
probably all Industars beat three-element cheap German or Japanese
enlarging lenses. The Soviet ones are hard to find, I tried to find some
Vega enlarging lenses, when I was in Riga, but without succes.
Helios 103 53/1,8 has a resolution power of 55 l/mm in the center and 28
l/mm at the edge.
Vega 29u (50/2,8), has a resolution power of 75 l/mm in the center and 60
l/mm at the edge
Sovietskoie Foto was published at least from the april 1926 as the journal
of the Journalists Union in the USSR. It was in Russian, but with a
resume in English.
Per
From russian camera mailing list:
Date: Fri, 07 Dec 2001
From: Kevin Kalsbeek [email protected]
Subject: Re: Soviet lenses resoloution
Hi Per,
David Anderson of the now very defunct Kiev Report believed that the Soviets
tested their lenses to line PAIRS per millimeter, rather than lines per mm.
This would effectively double the resolution. This makes more sense- at least to me,
considering the results they appear to be capable of.
Do you have any info about this?
Regards,
Kevin
Regards,
Kevin
From russian camera mailing list:
Date: Fri, 07 Dec 2001
From: Bob Shell [email protected]
Subject: Re: Soviet lenses resoloution
Kevin,
ALL lens tests are done in line pairs per millimeter. It's usually just
written as "lines per millimeter" in a sort of shorthand.
Bob
....
From Kiev88 Mailing List:
Date: Sun, 12 Aug 2001
From: "olivier" [email protected]>
Subject: free Chart for tests ---
Hi everybody , i just upload in the files two charts :
A) a Definition chart for lenses testing (definition, calibration of
groundglass and more)
B) a Color chart for emulsions testing and comparations ...
The def chart is an Adobe ilustrator file (+- 400kb) and have to be
printed in BW lazer in a good paper .
the color chart (80Kb ) is a JPEG file and can be printed in color lazer
...
good luck, Olivier .
From nikon mf mailing list:
Date: Tue, 13 Nov 2001
From: "Roland Vink" [email protected]
Subject: Re: Sharpnes and Resolution
> I am sorry if this is a bit offtopic. Can someone help me
> defining the difference betweeen Sharpness and Resolution?
Hi Francisco,
Often the terms are used interchangeably. More specifically, I'd say
Resolution is the amount of detail you can record on the film. Resolution
is measured in lines per mm. It determines whether you can see the
individual hairs on the picture of your pet cat, the fibres on clothing or
individual filaments on the feathers of an Owl.
I think "Sharpness" is a more subjective term which includes resolution
and contrast, and possibly other factors.
Back in the days before multicoating, manufacturers made lenses with as
few elements as possible, to reduce internal reflections. This gave their
lenses higher contrast, and they claimed their lenses were sharper.
Other manufacturers made more highly corrected lenses with more
elements. The contrast was lower, but the resolution was higher. They
also claimed their lenses were sharper.
They were both correct, depending on your definition of sharpness. It's
possible for a high contrast lens to appear sharper than a low contrast
lens with very high resolution. With modern multicoating, you get the
best of both worlds.
Roland
From leica mailing list:
Date: Mon, 02 Apr 2001
From: imx [email protected]
Subject: [Leica] Myth and anti-myth
It is remarkable that the idea that there is a significant trade-off
between high contrast and low resolution still rides high in Leica lore.
As far as I know no one who holds his view has ever presented demonstrable
evidence or corroboratable measurements to prove this point. Generally a
high contrast implies a high resolution and the other way around. It may
be that a shift in focus plane may change this relationship to a small
degree, but the general correlation is evident. More contrast is higher
resolution. And statements to the effect that a "slight" reduction of
contrast brings a "slight' improvement of resolution beg, nay scream for
evidence.
Now to kill two more myths. Sometimes I feel like Buffy the Vampire
Killer.
I have the Kodachrome films which I used as comparison for the 100 to
400ISO slide film test some weeks ago. Results will kill some preconceived
ideas. The King of all slide films is by now the Kodachrome 64, which
resolves easily 90 lp/mm, much more than the E100SW and even close to the
resolution of TP in normal circumstances. Especially noteworthy is the
excellent acutance, the great clarity of detail and the fine grain. A
disappointment was the K25 which at best was as good as the K64, with a
small gain in grain smallness, but not enough to offset the drop in speed.
The fading out of the K25 then is sensible. No added value. Sorry.
Big surprise the K200, which showed as expected a tight but visible grain
pattern, but a resolution that beats the Provia 400F at 70 to 75 lp/mm. So
the idea that fine grain supports high resolution is as false as the idea
that low contrasr supports resolution.
If you want to test the qulaity of your lenses, there is only one easy
way: use K64! and even K200 will show the defects of most lenses. Do some
actual testing!
I also had the opportunity to test the surfaces of filters on an
interferometer. Results will kill another myth. I used four different BW
filters in several colours (not relevant for testing, but to show that
there must be different batches).
Results? Take a deep breath: NO, absolutely NO image degradation by the
filter as all surfaces of the four filters were absolutely plane to the
highest possible degree. At worst only one interferometer stripe for the
experts.
Of course secondary reflections are possible. But the commonly held notion
that the addition of the filter adds two surfaces and by that fact should
degrade the image quality is simply not supported by measurements.
A well made filter in front of the lens will NOT make a drop of image
quality!
These results show that myths are fine if you wish to cling to stories that
seem sensible because they are repeated over and over again and even have
been 'explained' to some degree. But so the flatness of the earth had its
followers and scientifically based stories. But only measurements bring
the facts.
Erwin
From Rangefinder Mailing List:
Date: Tue, 2 Oct 2001
From: Joachim Hein [email protected]
Subject: RE: lens questions (lil' OT)
> I believe diffraction damage is dependent on absolute hole size, rather
> than relative size (f-stop), which is why longer lenses can stop down to
>
> a smaller relative number, and still not reach the absolute minimum hole
>
> where diffraction becomes a problem.
Nope, it is the f-stop. If you go through the algebra you will notice,
the increased magnification when using a longer lens, eats exactly what
diffraction decreases due to the larger opening. In the end it is the
ratio of diameter and focal length which matters. This ratio is the
f-stop number.
o underline this further, the British `Amateur Photographer' publishes
test results from fully open down to 16. At 16 all lenses perform about
equal, 100 lpmm for their high contrast target and 90 lpmm for their low
contrast target.
As pointed out by others, the amount of diffraction acceptable seems to
depend on the format. Many present digital cameras with 1/2 inch or
smaller CCD stop down to 8 only. Large format lenses commonly offer
values of 64.
With respect to the longer lenses and macro lenses (35mm format)
offering aperture 32 it seems motivated by DOF considerations only.
From Rangefinder Mailing List:
Date: Tue, 2 Oct 2001
From: [email protected]
Subject: RE: [RF List] lens questions (lil' OT)
Joachim -
You're only correct when you express resolution in lines/mm. However the
lpm metric ignores how different film sizes are enlarged differently. If
you stop way down so you're only getting 25 lpm and you plan on making a
16x20 print, it makes a big difference whether you start from a 24x36mm or
4x5 inch negative.
If you think in terms of the amount of information on the film (area times
resolution squared), then a diffraction-limited lens of a fixed absolute
size provides the film with a constant amount of information regardless of
its focal length, provided that you scale up film size in proportion.
Another interesting item is that holding lens aperture to a constant
absolute size, if you vary focal length and film size in proportion, depth
of field remains constant.
- Dave
From Rangefinder Mailing List:
Date: Tue, 2 Oct 2001
From: Michael Darnton [email protected]
Subject: RE: lens questions (lil' OT)
Oh darnnit--I wanted to stay out of it, but here I go again. My
understanding of diffraction was that it was completely independent of
format and magnification in the way Joachim suggests, in different ways.
First, an example: if you have waves in an infinite body of water,
there's no diffraction. If, however, you put two docks in that body,
from opposite sides, they induce diffraction. This is seen by watching
waves that pass the docks--as they move beyond the docks, they don't
continue on in a straight path as they would if the docks simply masked
the wave effect and nothing else--they wrap around the docks a bit and
are diffracted into the area sheltered by them.
Now here's where the discussion gets relevant to the current discussion:
if the water body is huge and the docks small, the proportional amount
of disturbance is very small, and irrelevant. If however, you move the
docks sufficiently close together, 100% of the water passing through has
its wave form disturbed. This has nothing to do with viewing distance,
magnification or any other relative phenomenom except one: the
wavelength of the waves and their proportion to the aperture diameter.
Bringing it back to cameras, *my understanding* of it (and I certainly
can be wrong) is that the critical factor is the wavelength of light vs
the diameter of the hole, not the diameter of the hole vs the distance
to film, as Joachim suggests. This ratio determines the percentage of
light that makes it through a give size of hole undisturbed, and thus
the quality of the image, at any focal length.
However, does anyone really care? :-)
--Michael
From: [email protected] (Frank Loeffel)
Newsgroups: rec.photo.equipment.large-format
Subject: Re: Resolution vs Fstop
Date: 22 May 2002
[email protected] wrote...
> In a recent issue B&W magazine ran an feature on the outstanding work of
> Clyde Butcher. In the B&W interview with Mr. Butcher he is quoted as saying
> " the more you stop down the more quality you lose because there less
> information per millimeter as you stop down. If you could shoot at F16, then
> you getting as much definition as you have available on the film" Is he
> referring to iris diffraction I wonder?
For someone not minding a bit of simple math: I just went through
this, on a theoretical basis.
I created a simple model that gives resolving power in lines/mm (same
as line pairs/mm in my opinion) based on defocus on the film plane,
aperture, diffraction and a model of lens aberrations.
If you e-mail me I'll e-mail you the Excel spreadsheet.
[Ed. note: thanks to frank, you can now download this Excel spreadsheet]
The formula in Excel giving resolving power in lines/mm I use is:
=SQRT(1/((1/$B$3^2)+(1/($B$6+$B$7*$B32)^2)+(1/($B$5/$B32)^2)+(1/(2*1/(C$9/$B32))
^2)))
Where:
$B$3 is film resolution (estimated 147 for Velvia at a 1:32 contrast)
$B$6+$B$7*$B32 is a simple model for lens aberrations: $B$6 is a base
resolution (set to 50 lines/mm), $B$7 is a factor by which lens
performance improves when stopping down, set to 3.5, $B32 is the
relative aperture.
$B$5 is the diffraction factor, set to 1500, $B32 again is the
aperture.
$C9 is the amount of defocus, as measured on the film plane, $B32
again is the aperture.
I think the model is useful for educational purposes. It does not yet
distinguish center and corner sharpness. I think it's a very good
model actually and I intend to use it!
For example, you see that by stopping down further from f/32 gains you
very little sharpness at the limits of your near and far focus range
but it sacrifices a lot of sharpness in parts of the image that are in
the middle of the focus range. UNLESS your defocus is huge, but then
you're in the range of smaller than fifteen lines/mm, something you'd
not want to enlarge much more than 2x.
The table has tab characters in it. I hope this shows up OK on your
screen.
The table gives resolving power dependent on defocus (horizontal) and
aperture (vertical).
defocus 0 0.33 0.5 0.66 1 1.33 2 4 8
5.6 61 30 21 16 11 8 6 3 1
8 65 39 29 23 16 12 8 4 2
11 66 47 37 30 21 16 11 5 3
16 63 53 45 39 29 22 16 8 4
22 56 51 47 43 35 28 20 11 5
32 43 42 41 39 36 32 26 15 8
45 32 32 32 31 30 29 26 18 11
64 23 23 23 23 23 22 22 19 13
Date: Wed, 18 Sep 2002
From: "SWB" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Hasseblad or Mamiya??
A selection of lens tests can be seen here
http://www.photodo.com/nav/prodindex.html
I think most people will acknowledge there isn't much in it between Mamiya
and Hasselblad on resolution terms, but they haven't found a way to measure
bokeh yet! And IMO there isn't much difference in terms of bokeh between the
two.
Steve
"Austin Franklin" [email protected] wrote in message
> Depending on how esoteric you are as
> a photographer, I buy cameras for the lenses, not for the camera body...as
> long as it does it's job.
[Ed. note: Special thanks to Charles R. Batishko for sharing these files!!!]
Date: Thu, 19 Sep 2002
From: "Batishko, Charles R" [email protected]
To: "'[email protected]'" [email protected]
Subject: RE: request for scan for photogs, thanks! Re: USAF Test Bar Target
Bob...
The [.png] file I referenced is attached. As you'll see, it's not a high
quality image since it wasn't generated for that reason, but rather just to
communicate the text and numerical information. I assume this will be useful
as you already have very nice files of the target itself.
Nice web page. I've bookmarked it for future reference.
Optiker
Chuck
(810) 958-0819 (fax)
[email protected]
-----Original Message-----
From: [email protected] [mailto:[email protected]]
Sent: Thursday, September 19, 2002
To: Batishko, Charles R
Subject: request for scan for photogs, thanks! Re: USAF Test Bar Target
greetings!
Would you be so kind as to share a copy of your scan of these USAF
related materials? I have a website on photography, and provide USAF
test charts for photographers wanting to do lens resolution tests
(see http://medfmt.8k.com/mf/resolution.html) I suspect many would be
interested in this information and resource too!
thank you for your efforts in making this information available!
regards bobm
[Ed. note: thanks to Bertho Boman for sharing this info, and please let him and us know
if you have access to the original or updated 1951 USAF test target specs -thanks!!]
Date: Thu, 19 Sep 2002
From: Bertho Boman [email protected]
To: Robert Monaghan [email protected]
Subject: Re: USAF Test Bar Target
Parts/attachments:
1 Shown 21 lines Text
2 133 KB Image
Hello Robert,
I remember your site! I spent a lot of time reading there about a year ago.
Lots of good info.
I will add a link shortly to your site from my camera info page at:
http://www.vinland.com/_Misc.html
If you find a real USAF-1051 specification, (not chart), I would like a
copy, or I can have it copied if it is a lot of pages. I like to know the
details of the specifications. For example: Reflectivity of the white and
black bars, font type and size. Also for historical reasons, it would be
fun to have.
Here is the AutoCAD file for the test target.
Let me know if you are having any trouble or suggestions.
Best Regards,
Bertho
From zeiss interest group mailing list:
Date: Fri, 25 Oct 2002
From: Edward Meyers [email protected]
Subject: Re: Re: Zeiss Triotar 100mm/
Good, better best. Describing useful lens quality
brings in to play (as you know) the camera. If film is not
positioned properly a great lens may perfom badly.
The Zeiss Contarex (I'm told) has a very good film channel.
This takes advantage of the wonderful lenses made for it.
Experts in this area questioned the design of smaller
SLR cameras, such as the Olympis OMs, which had short
film channels. Film needs to be stabilized before reaching
the exposure position, I am told.
Testing for resolution doesn't tell you about contrast.
And testing for contrast alone doesn't tell the whole story
on lens performance. You gotta take pictures. Then the
enlarging lens and printing or projection comes into
play.
Funny story. When I was the lens tester (resolution, only)
at Modern Photography magazine in the early 1960s, I tested
the current enlarging lenses in actual use. Rodenstock did
not test out as best, at that time. The importer of the
Rodenstock lenses threatened to pull their advertising unless
I reported a "correction". I agreed to retest the rodenstock
enlarging lenses in the importer's darkroom along with their
expert, Rudy Simmons, one of the famous Omega enlarger
brothers. I showed up at his darkroom in Long Island City,
New York. We went into the darkroom and immediately the
room began to shake. What is that? Rudy explained that there
was a printing company next door and when the presses
were running (most of the day) the darkroom vibrated.
How were we supposed to test the enlarging lenses then?
Rudy looked at his watch and said, "let's go to lunch".
We did, and the case was closed. Ed
From: "Q.G. de Bakker" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Are used medium format camera prices dropping?
Date: Tue, 3 Jun 2003
Q.G. de Bakker wrote:
> > Did they test this using lenses or some scientific device that has
> > absolutely nothing to do with photography?
>
> They used microscope lenses capable of producing more than 1000 lp/mm
> images.
I tell a lie. That's what Fleischer/Mueller says manufacturers of film use
to test their products.
Zeiss, he says, used typical photographic conditions: sunlight outdoors;
exposures using normal camera shutters; focussing done using normal
focussing devices found on cameras; normal film processing in a regular lab;
and using normal photo lenses, made by Zeiss, of course. The only thing not
found in normal photographic practice was the Carl Zeiss Stereomicroscope SV
11 Apo they used to examine the film.
Fleischer/Mueller mentioned how manufacturers of film say (as one reason why
they don't make more high resolution films) they don't believe that lenses
are good enough to show what a high resolution film is capable of.
Fleischer/Mueller, being a Zeiss man, of course rebukes by pointing to the
1996 Photokina Zeiss exhibition, in which they displayed photos, made from
Ektar 25 (capable of 200 lp/mm) negatives, and made using Zeiss lenses, that
show lenses in fact can use the film's high resolution. Every last bit of
it, in fact.
He also mentions that they (Zeiss) are considering staging another such
demonstration. When they do, David should make sure he attends. ;-)
From: "Q.G. de Bakker" [email protected]
Newsgroups: rec.photo.equipment.medium-format
Subject: Re: Are used medium format camera prices dropping?
Date: Tue, 3 Jun 2003
David J. Littleboy wrote:
> (I strongly object to twits who claim film resolves 100 lp/mm, but it
> certainly resolves 40 lp/mm, even in actual practice