On Sat, 15 Aug 1998 12:19:39 GMT, zinyflikop@sprintmail.com wrote:

>When I tested lenses by using USAF targets I came to interesting observation
> for which I don't have explanation yet. I checked lenses resolution for two
> different cameras in two ways.
>
>1. I took pictures of the targets on B&W 100 Kodak negative film. Lenses were
> mounted on the cameras that have been installed on tripod. To prevent camera
> shake I use remote release and shatter speed above 1/125 since cameras don't
> have a mirror lock. After I have film developed I checked negatives via 20X
> lope and defined resolution.
>
>2. I used the same lenses at the same distances from targets however at this
> time I used them without cameras. I just mounted them on tripod and analyzed
> targets looking via lenses at camera distanses and using the same 20X lope.
>
>In second case (direct observation) I received about twice higher resolution
>then by studying negatives. I checked cameras focusing at camera film planes
>with a half transparent screen. Focusing looked correct. However, I was
>unable to define resolution properly because of screen granularity.
>
>
>Does anybody have an explanation?

On-film resolution is a function of both lens resolution and film
resolution, and will never be as great as either alone. As a result,
though, on film resolution is not "limited" by either the film or
the lens (except maybe in extreme cases...;-), so improving
either will result in improved on-film resolution.
BTW, I've never liked test-pattern testing of lenses, since the
results are often different from what one finds in practice,
using lenses in the real world. Many lenses show considerable
variation in performance level at different focus distances;
many lenses show moderate field curvature which can go
unnoticed in most shooting conditions; a few lenses look
sharper at the edges in color compared with (unfiltered)
B&W; and it is very difficult to focus accurately on a flat
test target (or to get correct alignment with it). I use
familiar land/city-scapes for checking lenses (the great
distance of the subject material removes some of the
difficulties encountered with close test-target shooting).
This doesn't produce numerical results, but comparitive
results are what I'm after, anyway.