Re: Digital vs 35mm camera optics

From: David Littlewood (david_at_nospam.demon.co.uk)
Date: 08/30/04


Date: Mon, 30 Aug 2004 17:37:33 +0100

In article <s0isb0hcp3.fsf@diesel.graphics.cornell.edu>, Stephen H.
Westin <westin*nospam@graphics.cornell.edu> writes
>David Littlewood <david@nospam.demon.co.uk> writes:
>
>> In article <s0n00ggtdn.fsf@diesel.graphics.cornell.edu>, Stephen
>> H. Westin <westin*nospam@graphics.cornell.edu> writes
>> >>
>> >
>> >Over what area? And where are the tests showing this? Let's see. A
>> >20"x30" print at 8 cycles per millimeter would be over 150 cycles/mm
>> >on the negative. Not only is it a wonderful negative you're talking
>> >about, but a most remarkable enlarger lens. Remember that the
>> >theoretical diffration limit at f:8 is 200 cycles/mm. For example, the
>> >Schneider Apo-Componon HM 60mm f:4 seems to be down around 60% MTF at
>> >40 cycles/mm at f:8; see
>> ><http://www.schneideroptics.com/photography/photo_enlarging/apo-componon
>> >_hm/pdf/apo-componon_40_60.pdf>.
>> >
>> This is surely not the relevant comparison, since the purpose of an
>> enlarger lens is to make a print, for which 10 lp/mm is sufficient and
>> 20 lp/mm is near-perfection. However, the magnification must be taken
>> into account; at 10x magnification, f/8 is effectively f/88
>> (f-no*(1+M)) as seen from the paper. This takes you to about 20 lp/mm
>> for the diffraction limit - which a top class lens like this should
>> achieve at f/8 - and on track for a good print (subject, of course, to
>> the contributions from the negative). It may however give better
>> results at f/5.6 or f/6.3, you would need to test it.
>
>But we're talking cycles per millimeter at the focal plane. To get 10
>cycles on the print at 10x magnification, you need 100 cycles at the
>focal plane. But we only have information for 40 cycles, and that
>looks to be attenuating by about 40%. At that sort of spatial
>frequency, a good color film wil attenuate by about 50%, and then
>there is attenuation from the MTF of the camera lens. Even assuming
>perfect focus on both the camera and the enlarger, perfect film
>flatness, no camera motion, etc. to get 100 cycles per millimeter on
>the film and then onto the print seems like a tall order. Especially
>to get it at a contrast that would actually make it visible. To quote
>Erwin Puts,
>
>"Even allowing for all of these effects the goal of 100 lp/mm on the
>negative is hard to reach. On the other hand, this study has shown
>that the 100 lp/mm are the limit of what we need (or can use) for
>high-resolution imagery with 35mm equipment."
>
My apologies; I think I must have misread your earlier post; I think we
are saying the same thing. Certainly, 35mm is not a medium in which it
is even remotely possible to get "ultimate" quality; nor is there any
point in wasting effort trying, it is much easier to get the desired
quality in larger formats.

However, there was a valid additional point in my post, namely that most
people use enlarger lenses at smaller than optimum aperture, thus
guaranteeing less print sharpness than they could get. The effects are
much more pronounced than diffraction in camera lenses.

David

-- 
David Littlewood