Related Local Links:
Lens Test (35mm vs. 645/6x6) by Fernando Carello
Lens Testing
Lens Resolution Charts and Testing

Why conduct tests?

My concerns

We intend to go back and island hop some more. This time I will take the digital, a better 35mm and probably a folding 120 to record some pictures on a larger negative. My SLR and lenses represent a ball and chain that I would rather leave at home. My Leica 50mm Elmar is supposed to be a good lens. I also have a Russian set of lenses of 35mm, 50 mm (f2.0), a 85mm plus a Canon 100mm. These are much more portable than the SLR. Is the quality any good? I have never used them for anything more demanding than snap shots, so wanted better data.

In addition, my camera cabinet is full of other cameras that I wanted to evaluate. My Minox EL has to be the most portable 35mm that ever was. How does it compare to my slightly larger, somewhat heavier and less user-friendly Rollei 35 SE?

An Internet article by Darrell Young ( Medium Format In Your Pocket! ) extolled the virtues of the large negative and portability of the folding 120 camera. This seemed like a possibility for an additional travel camera. Most of these cameras can be purchased on e-bay for less than $100. Indeed, my two top performers cost less than $25 each. In the course of four months I purchased eleven folding cameras. About half of them turned out to be 100% functional with good or better results. But which was the best one to take? Hence the idea of a test.

Resolution targets

Of course, this is only valid if you shoot all the lenses/lenses under similar lighting and exposure.

None of this is rocket science. If I can do it, anyone can do it. It took some experimenting to develop a reasonably valid comparison.

Set-up

If this seems like a long way away to shoot a sheet of paper, it is. However, it helps minimize minor focus errors and helps the lenses that use a moving front element to achieve a sharp focus.

Chart Setup on Steps

I put the target on a cardboard box and set the box on a staircase, about four feet above the ground. I could have taped the target to a wall. In fact, this is what I did for the first test. Some cameras appeared to be slightly out of focus, so for the next test I put up three targets, one on each step. Each target was about one foot apart, front to back. The measured distance was to the middle target. The others were one foot in front and one foot behind, in case the camera focus was not spot-on. Something like this:

	|
>	     |
                |

Exposures

In 35 mm I shot one exposure wide-open and then one shot each at 5.6, 8 and 11. f5.6 was supposed to be the �optimum� opening. In 120 I shot only 5.6, 8 and 11. f8 was supposed to be �optimum�.

The target should be shot at two stops of under exposure. This tuned out to be critically important and not as easy as it sounds. A �normal� exposure for a scene results in an over exposed white sheet of paper on which the black tribars are invisible. I got the best result by metering the general scene and then shooting at 2 stops under. For the automatic cameras this was achieved by setting an ASA two steps faster than what was in the camera.

I used Fuji Velvia (ASA50) in 35 mm and recommend this film highly for its very fine grain and excellent resolving ability. In 120 I used Fuji Astia slide film (ASA 100) because this was what I had in the fridge.

ASA 100 turned out to be too fast! By underexposing by two stops, the film is exposed as ASA 400. On a sunny day, an exposure at f4 requires 1/8000th of a second!! Since the folders had only 1/500th at best, should have used ASA 50 film. As it was, I waited until the targets were in shadow. An overcast day probably would have been best.

Test problems

I anticipated that manual re-wind cameras should be tested before the automatic re-winders. This avoids unnecessary film fishing out of cassettes. What I did not anticipate was the problem to clicking off the correct number of shots to get to the fresh film each time the film was loaded in a new camera. In a manual camera this is easy. Just put on a lens cap, set the shutter to 1/000 and f22 and start clicking. On an automatic focus and flash camera you can sometimes select �no flash� and set focus to infinity before clicking. If the camera has a shutter lock to prevent taking a picture when it is too dark (lens cap on) then it will have to be the first camera in the test.

The Minox EL does not have a lock, but its electric shutter tries to take a 20-minute exposure because it is so dark with the lens covered. I found I could slide a lens cap over a bit to block the lens but expose the light sensor.

Film handling

On my first test I made up a set of Post-It notes numbered 1 to 11 to show which camera was in use. For each shot I also put up a Post-It showing the f stop. This is the foolproof system, but makes for a lot of walking aback and forth. I need the exercise, but the quicker these tests are done, the more uniform the light.

I think the second test used a better system. I just used a film roll number Post-It note on the target and used the same numbers on my Excel spreadsheet prepared in advance. After shooting the 4 f-stops with a camera I set the exposure for the correct exposure for the scene and shot a picture of my back yard. When the slides came back, the scene picture marked the end of each different camera test and also provided a picture to help evaluate lens colour and contrast - something that cannot be seen in underexposed target pictures. set the exposure to the correct level and shot a picture of my backyard. When the film came back, the backyard scene marked the end of each camera�s test. By noting the roll number (which was marked on the cassette, the film developing envelope and on a Post-It stuck beside the target, it was easy to see where each particular camera or lens started and stopped on each roll.

Evaluating results

If you shoot different lenses and or cameras the same day with the same film at the same subject you will be able to make comparisons. If you shoot at the USAF resolution target at the correct distances you will be able to make easier comparisons with numerical results.

A 40X stereo microscope is used to evaluate tests...

The first step is to number all the slides consecutively. The next step is to label the Excel spreadsheet print out (or paper notes of the test) with these numbers so that each lens and f stop is identified with the matching slide number. Then the slides are carefully examined to see what is the smallest tribar set that can be seen. E.g. 2-4 (Set 2, tribar pair #4). Later these numbers are converted to line pairs per millimeter based on a table for the USAF target. 2-4 is actually 53 line pairs per millimeter.

Conclusions

Comparing contrast is a problem. I put sets of similar slides in the projector and convinced my son he should help me evaluate. He did not know which slide was which, so was less prejudiced than I was. We got to a few slides where he said �Wow! That�s like 3-D!� He was reacting to contrast, not resolution.

Use of Dual Grey and White High Contrast Charts

In discussing my desire to have numerical rather than subjective results, my friend Col. Charles Williams, USAF (retired) suggested a target of less contrast. Charles� career included a lot of photo-recon work - both flying and in the lab. My targets were printed on ordinary photocopier paper. The paper was bright white, giving maximum contrast.

My business stationary is a light gray. A bit lighter than 18% gray. For my last test, involving cameras that needed re-testing, I put up a regular white target and a gray paper target, too. A lens of good contrast results in the same or almost the same resolution reading on both targets, whereas the cheaper lenses have 3 or 4 steps difference between the two targets. I suggest that for the most useful results your targets not be printed on white paper.

[Ed. note: A very interesting observation and tip for evaluating lens contrast, thanks to Col. Charles Williams and Jim Bullock for sharing these ideas and tips!!]

Postscript

I now have more slides back from the last test where I used a white target and a lower contrast target. The cameras usually produced a lower resolution on the gray target. Lower by one or two steps. The Zeiss Ikonta (with 3.5 Tessar) scored the highest resolution and most frames showed the same resolution on both targets. !! Of course, Zeiss makes contrasty lenses.

My advisor, Charlie Williams tells me that he found a contrast ratio of 5:1 produced the most relevant test results. I have no way of measuring a contrast ratio (do I?) but I think if the target is printed on any coloured paper other than white the contrast would be reduced a bit and would result in more useful information - which is a consideration of both lens resolution and lens contrast.

The only other comment is that for camera to cameras results to be useful the exposure has to be the same in both cameras. Since old cameras have inaccurate shutter speeds it may take a test or two before you know how to set the shutter.

jim