large format lens testing (method confirmation question)

df cardwell said:

Ballparking rayleigh criterion... 1760/f

Here' the Zeiss take on the subject:
http://www.dantestella.com/zeiss/resolution.html

Click to expand...

I've read the article addressed here (from Camera Lens Magazine No. 2, Fall 1997) and I will disagree with the application of some the results.

Mainly, there is the necessity of remembering that the subject lens in question was, in fact, a perfect lens, and at the same time, acknowledging that such a lens does not exist. Offshore, I've learned that this article was intended as a sort of rough primer describing the effects of diffraction for those not familiar with optics and lens design.
That "roughness" leads to my disagreement. The only way to determine just how "rough" it is would be to compare the results cited here with the traditionally accepted formulae, which if bruised memory serves me, was discoverd by Carl Zeiss, himsef in the late - or not so late 1800's.

I've been searching for copies of my work with that formula, but so far I haven't been able to find them.

If anyone has any information about these formulae, pleas post it here - it would save a great deal of time. Until then, my search will continue.

Now, the article itself i not completely wrong, but it is in danger of misappication.
As an example in opening the aperture of a PERFECT lens one stop, the resolution in L/mm will invarably be increased. In a lens designed for use by us mere mortals, resolution at the larger apertures is already limited by other factors; the design itself, manucacturing errors and compromises, so diffraction has no effect. It is a grave errror to consider "perfect lens data' and extrapolate it indiscriminately to "ordinary" lenses.

More when I recover the formula.

may application

"It is a grave errror to consider "perfect lens data' and extrapolate it indiscriminately to "ordinary" lenses."

Both the Abbe formula and the Rayleigh criterion were used as a reference to establish a limit, or ceiling, to what could be expected from testing a photographic lens. While I haven't looked too hard, I've yet to come across a photographic lens that can perform better than Abbe or Rayleigh predicted.

If a perfect lens could perform no better than X, the limit of an imperfect lens would therefore be less than X.

This was done to back up the notion that using film and a camera lens is so packed with buffering elements that the 'controlled tests' were both unverifiable and incapable of proving whether or not a good field lens was better than another.

It is what a teacher called the Theophilus Principle:

Theophilus Thistle, the successful thistle sifter,
While sifting a sieve full of unsifted thistles,
Thrust three thousand thistles through the thick of his thumb.

Now, if Theophilus thistle, the successful thistle sifter,
While sifting a sieve full of unsifted thistles,
Thrust three thousand thistles through the thick of his thumb,
See that thou, while sifting a sieve full of unsifted thistles
Thrust not three thousand thistles through the thick of thy thumb.

The point being that a perfect film and perfect lens would break down in the OP's field test. If a perfect lens and a perfect film could not yield magical results, what hope was there for a merely excellent lens and film to not fall apart ?

Which is what I think I hear you saying.

As for Zeiss, if the object of the exercise is to demonstrate why stopping down too far is undesirable, Zeiss seems a pretty good advisor. I'm sure there is somebody in Oberkochen who could make your head melt if you wanted to argue optics. I have no interest in it, I’m just a photographer, and a damned poor one at that. But this about trying to establish expectations for testing a LF camera and lens in the field.

And Carl Zeiss was a machinist.

The other two guys were Otto Schott and Ernst Abbe.

Schott made glass. Abbey was the numbers guy.

d