What distance?

[puderse]

After working on a camera one needs to make a test, especially if a lens has been exchanged or the camera is not a view camera. I have a large black piece of foam core with a US Air Force test target, a large sheet test target with 3 bars running in every direction and size, a Color Checker, a Kodak step gray scale and an 18% grey card on it.

The question is this: Is there an optimal distance from the lens to test performance for comparitive reasons? I think I have read that 100 times the focal length of the lens is standard. Close and filling the frame certainly reveals barrel and pincushion but doesn't tell much about real life situations.

What I really want to decide is, after all my modification is "this" better than "that" and unless the "test" compares only one different thing, ie: the lens etc. then I'm back to subjective comparisons.

Discuss, shoot holes, badmouth, call me a fool, advise, etc.

 

[Mike Wilde]

I'm not up on fiddling with the internal affairs of a lens, so most of my work comes to fitting a whole lens assembly back onto the camera.

In this case I usally just prop the shutter open, if it lacks a B setting, and fitting a piece of velum onto the rails of the mf or 35mm cameras as a surrogate ground glass to allow the infinity focus to be set.

I usually walk out the front door in the evening , and work on focussing onto a streetlight three or four lights down the street, with the focus scale set on infinity.

The test chart you have may be useful of there is some element of the lens that you have apart, and are not sure which side should be mounted which way when putting it back together.

 

[puderse]

More of what I had in mind is looking at the resolution gains/loss by modifying/exchanging elements or lenses with or without changing shutters.

Resolution testing

 

[keithwms]

The question is this: Is there an optimal distance from the lens to test performance for comparitive reasons?

Well, I don't think the distance will have much effect on center sharpness, but sure, corner sharpness might well suffer noticeably if you are too close.

I suppose that when one runs the calculations to design a lens, a particular reproduction ratio (and therefore focusing distance) is assumed. If the lens is to be used at very small working distances then additional corrections are warranted (e.g. a floating element).

Why don't you just try it and see? I'd be curious what numbers you get at close focus versus far. Obviously a macro lens should shine at ~1:1, whereas a tele should suck there. But I'd be interested in the real numbers. My own anecdotal experience suggests that apo-corrected lenses shine at all distances, whereas non-apo/uld/ed lenses tend to have issues at close focus. But I do not shoot charts....

 

[Ralph Javins]

Good morning;

This may be related to the size of the lens resolution test chart you are using, and the interpretation of the lp/mm numbers from the photographed image of that test chart. For the ones I have, the readings are designed to be accurate when photographed at a distance of 101 times the focal length of the lens from the film plane. For a 50 mm lens, this will be 5,050 mm, or 5.05 Meters, or 16 feet and 6.82 inches. Making a test chart of a precise accurate size is one of the reasons why they tend to be expensive. The best ones probably are the ones made on a glass plate. They do not change with variations in humidity as is possible with some of the paper ones.

 

[puderse]

Over the week-end I read on the internet 20:1! Thats what I need, advice to wing-it and conflicting 'experts'.

 

[Q.G.]

Your question was flawed.

Performance always changes with distance.
So you can test a lens' performance at any distance you like, and the test will then reveal how the lens performs at that particular distance.

So what do you want to know?
If it is how the lens performs at a given distance, test it at that distance.

If you want to know at what distance the lens performs best, you have to go through a series of tests to find out.

If you want to compare two lenses, again, only a series of tests will provide an answer (one lens may be better close up, the other at infinity, or vice versa, or ... etcetera)

 

[puderse]

distance

The gist I think I am hearing here is that the valid tests of lenses is by comparison only and subjective. I like 'A' better than 'B' etc.

The tables one sees in print of resolution in lines/mm are only usefull in comparison. "good, better, best" or "that sucks, OK, 'I like', or WOW!" lens is all in the eye of the beholder.

 

[keithwms]

Well, it's not like the lens shines at one distance and falls apart at others. There will be some distance at which a lens shows optimal performance across the frame, that's all.

 

[Q.G.]

It depends.
It depends on the lens design.

Some are great, but lose greatness rather rapidly with changing distance.
Others are far less distance/scale dependent.

Some tests do not take that into account.
Photodo, for instance, tested macro-lenses at infinity. Probably because they test all lenses at infinity.

And some people's appreciation of lenses does not either.
For instance, the Zeiss Planar f/3.5 100 mm is widely regarded as the bee's knees of MF optics. Yet its extremely good performance is rather distance dependent. And when not used as an photogrammetry-lens, taking pictures at very large distances, of subjects that from that distance are as good as flat, the 'humble', often poorly regarded f/2.8 80 mm Planar is just as good, or perhaps even better.