Practical lens tests.

[fidget]

I have a small collection of FEDs and Zorkis. Some of these work very well, some are yet to be fixed. With the LTM shooters came the standard lenses of the day, Jupiter 8, Ind61L/D, Ind61 Panda, Ind50, Ind26.
I had performed a practical test on some of these lenses. The subject matter was a few items at about 6 feet with detail and texture with a part of the frame showing buildings far away. Each lens was mounted on my FED 3b (& tripod) and two frames taken. Both frames focused on the objects, one at f8 and one at f2.8 (or the max aperture if less than 2.8). I used Ilford Delta100, printed in my darkroom and compared the results of a section of the picture at 8”x10”.
At first I was pleased with the results which showed that none of the lenses was particularly bad, but that the best were from the J8 and the Industar 26 (surprisingly!). The Industar 61 L/Ds shots were not as well defined, and seemed no better than the 61 Pandas.
After reading more of these lenses and cameras on this site, I realize that my testing was probably a waste of time. If any of the lenses “tested” were not matched to the body or the RF was off, the results from them may be degraded. (although the f8 shots could overcome errors in collimation?)
So, I will make a further attempt at some practical comparison tests. The body is set up a well as I can ensure, although I can’t measure the mount to film plane distance to the accuracy I’ve seen quoted (+/- 0.02mm). I intend to arrange some test cards in the picture, one at focus distance in the centre of the frame, one at the edge and two or more spaced around a foot in front and behind the centre target at around 8 feet. I hope that this will show up errors in the focus plane at open aperture. I will again include some distant object in the frame and shoot at f2.8 and f8.
To make this attempt more meaningful than the last, I would be grateful for your input to what these simple tests could include, or what problems could be spotted.
Anyone know of a test card which might suit the task? (I saw one a while ago which had a series of sets of numbered horizontal and vertical bars, decreasing in size.)
Would it be better to shoot in a diffuse light?
What effect might a lens with plenty of light coating surface scratches have?
Purely for the sake of interest, I thought that I might fire off a few of the first frames in my nikon F70 + Nikkor 50mm. I could include a few from my Konica Auto S2.

Regards......

 

[Sonnar Brian]

The first step is to make sure each lens works well with the camera used in the test. For that, a camera with a removable back will be helpful. Make a film strip with some translucent paper or plastic taked to it, to be used a screen. I use the paper from some negative sleeves. Use the RF to focus the lens, then check the image at the film gate with a loupe. I found one with a reticle on glass that fit in nicely over the guide rails. It acts like a pressure plate. If you have to adjust the lens to acheive focus, you'll know that it does not agree with the camera's RF. I've been doing this to re-shim and adjust Jupiter-3's to work with my Canon 7 and other LTM cameras. Six out of Six had to be re-shimmed and adjusted. One required the helical to be re-assembled and a 4mm(!) shim added. It was originally off by 12' at 15'. Others were off froma few inches to 1' at 4'.

 

[fidget]

Wow, I thought that my FSU tinkering might be limited to the bodies, seems there's lots more variables to contend with!
Part of the purpose of the test is to determine which lenses are worth working on to match a body (hopefully a "standardised" body with the nominal flange to film plane measurement). I had followed Matt Denton's guide on relubing lenses and did this on a few I61's and a I26. I didn't see how or where one would shim a lens to match a body. Is there a guide to this somewhere in wwwborough?

 

[Nickfed]

results which showed that none of the lenses was particularly bad, but that the best were from the J8 and the Industar 26 (surprisingly!). The Industar 61 L/Ds shots were not as well defined, and seemed no better than the 61 Pandas.
After reading more of these lenses and cameras on this site, I realize that my testing was probably a waste of time. If any of the lenses “tested” were not matched to the body or the RF was off, the results from them may be degraded. (although the f8 shots could overcome errors in collimation?)
So, I will make a further attempt at some practical comparison tests. The body is set up a well as I can ensure, although I can’t measure the mount to film plane distance to the accuracy I’ve seen quoted (+/- 0.02mm). I intend to arrange some test cards in the picture, one at focus distance in the centre of the frame, one at the edge and two or more spaced around a foot in front and behind the centre target at around 8 feet. I hope that this will show up errors in the focus plane at open aperture. I will again include some distant object in the frame and shoot at f2.8 and f8.
To make this attempt more meaningful than the last, I would be grateful for your input to what these simple tests could include, or what problems could be spotted.......

I think the short answer is not many, particularly as you have already determined that none of the lenses are particularly bad, and thus the tests you made were probably not a waste of time. With that in mind, and noting that all the lenses are 50mm, I respectfully submit the results from the tests you propose, as far as focus is concerned, are going to be totally subjective, i.e. you are only going to see what you want to see, and therefore an utter waste of time. For starters, at eight feet at f8, you can expect a useful depth of field of about, say, four feet and well over a foot even at f2.8. Need I go on?
And as for matching the lenses to the body, I would suggest you do the maths first, before you get too excited about this prospect. You are likely to find that is an utter waste of time too.
People can get pretty unlucky, vide brians in the current thread "variable focussing" where he appears to have a very unfortunate mismatch. I suspect there may be (merely) a rangefinder follower problem.

 

[pangkievrange]

Better to use a glass plate with a marked cross (there's a thickness of the mark that needs to be taken account of. I place the part of the marked cross onto the film rail. With the glass plate, you don't need to worry about the bowing of film strips that can introduce significant errors in collimating.

pangkievrange

The first step is to make sure each lens works well with the camera used in the test. For that, a camera with a removable back will be helpful. Make a film strip with some translucent paper or plastic taked to it, to be used a screen. I use the paper from some negative sleeves. Use the RF to focus the lens, then check the image at the film gate with a loupe. I found one with a reticle on glass that fit in nicely over the guide rails. It acts like a pressure plate. If you have to adjust the lens to acheive focus, you'll know that it does not agree with the camera's RF. I've been doing this to re-shim and adjust Jupiter-3's to work with my Canon 7 and other LTM cameras. Six out of Six had to be re-shimmed and adjusted. One required the helical to be re-assembled and a 4mm(!) shim added. It was originally off by 12' at 15'. Others were off froma few inches to 1' at 4'.

 

[fidget]

Thanks for your comments. I agree that a more down to earth check could be made, as in checking the focus of the lens/body using a screen and loupe. Nickfed is correct in suggesting that I do the math before jumping in and spending time on this. Quite rightly Nickfed points out that the DOF even at f2.8 at a focus distance of 8 feet is nearly a foot each side. My planned test would show very little. Many thanks for this.
I borrowed a vernier gauge (accurate to +/-0.02mm!) and played with a few measurements. My Fed2 mount to film plane distance was 28.70 to 28.80, Fed3b 28.90. In the light of the practicalities involved I checked what this meant in terms of focus. My J8 cam ring extends 0.14mm as a marked distance is moved between the two f2 marks on the DOF scale. At f2.8 this is 0.32mm and 0.70mm at f8. So it seems that (for 50mm lenses at least) the latitude allowed at smaller apertures swamps small errors in the film plane/mount distance. At f2 this is more significant and I assume that fast lenses might have a problem with the discrepancies in my examples. I need to rethink what a test could provide.
As a coincidence, a couple of processed films arrived in the post today, one of them from my Fed2, which I thought was a worker. In the photos of bright scenes, where I would have selected a fast shutter speed, I see the effect of shutter tapering.
Oh, the joys of FSU cam ownership!