In article <33474978.233F@foto.no>, nikon@foto.no says...
>Bob Neuman wrote:

>> Hmmm, unless the lens has some seriously decentered elements, I would
>> not expect performance inferior at f11-22 to that provided by another
>> similarly high-quality lens. A really poor lens could be worse at these
>> stops, but it would be worse yet at wider apertures... Smaller stops are
>> the great lens-levelers, since diffraction degradation is the same for all
>> lenses for a given format at the same small stops. There may be some
>> differences in corner performance even at small stops with super-wides
>> and wide-angle zooms, but I would expect most of the frame to look pretty
>> much similar in good lenses by f16, anyway...

>It is colour aberration, the achilles heel of wide-angle zoom lenses
>that comes into play here. Many modern designs, including the 20-35
>f/2.8 Nikkor which used for photojournalism at fairly large apertures is
>an extremely good lens, have sacrificed off-axis colour correction in
>order to solve other optical problems. It is all a matter of computer
>optimising criteria, I think.

Yes - if I understand what you are saying (not having used the 20-35,
but having used several 25-50mm Nikkors [and LOTS of other Nikkors...;-]),
you are saying that the 20-35 is rather poor at the edges/corners, even
well stopped-down (though good in the center at wide stops), and therefore
is more suitable for the less edge-sharpness critical photojournalism use
than for landscape work than the better-on-the-edges-but-not-as-sharp-in-
the-center-at-wide-apertures-25-to-50mm-zoom? ;-)

>The reason why 20-35 f/2.8 cannot match the old 25-50 f/4 for landscape
>photography is quite simple: At f/11-f/22, settings that are
>indispensable for landscape photography use, colour fringing in the DOF
>zone makes the image perceptably fuzzy. This effect can be clearly seen
>in the corners and using a good 20X loupe, also near the axis. I like to
>stress this results from the lens design and not from a poor sample of
>the 20-35 f/2.8. Each and every one of these lenses I have tried
>exhibits the same pattern of colour aberration. The old 25-50 f/4,
>whilst not coming up to the sharpness of 20-35 f/2.8 on-axis, is not
>troubled with colour aberration and this is evident in landscape
>photography when the old lens gives high-contrast, high-definition
>images at f/11-f/22. These are lenses with different designs and
>therefore, giving different photographic results.

Hmmm, with one Nikkor design, the 35-105mm f3.5-4.5 MF, I did encounter
a relatively high percentage of poor samples (4 out of 5!), with most
showing poor edge performance - but two out of 10 showed amazingly
good corner performance at all stops (even beating short FL primes!).
Perhaps you have encountered a sub-standard run, or the barrel design
cannot keep elements centered and straight (not uncommon with zooms,
and I suspect that tolerance requirements get tighter with short FL
lenses...).

>Finally, I would like to challenge the statement that lenses poor at
>small apertures would be worse yet at larger apertures. This simply
>isn't true. Take a 200 f/2 Nikkor as a good example. This lens performs
>magnificently wide open and down to f/5.6. Beyond f/5.6, its optical
>quality declines rapidly and at f/22, it's nearly unuseable. The same
>relationship holds for the Nikkors 300 f/2, 300/2.8, 400/3.5 and 600/4,
>all of which I own and have tested in depth.
Bjorn Rorslett

Hmmm, while I respect the fact that you actually observe results from
lenses you use, and have used a lot of lenses (and samples), I cannot
agree with the above, from my own experience with MANY Nikkors (I own
about 25, and have checked hundreds, including the 300mm f2.8 and
400mm f3.5, and have not observed what you have. We have a mystery.
I sent your posts to David Jacobson for comments (he is much more
knowledgeable about lens theory than I am), and his reaction was
the same as mine... (we both thought other factors might be involved,
such as problems with steadiness at the shutter speeds necessary for
small-aperture testing of long lenses, etc. - though it is hard to come
up with a good all-inclusive guess...). I must return to my observation
that, unless there is a physical problem with the lens, stopping down
to mid apertures should improve the image of most good lenses (a VERY
few may peak in the center at f4, or even slightly wider...), and
stopping down further will cause a slow decline as diffraction
limits are reached (though edges and corners of super-wides and zooms
may improve with small stops) - but good performance should be had
even at f16 with lenses designed for 35mm (and I have seen many a good
photo taken at f22 - though for me, the diffraction-caused further
decline in performance by f22 is a bit too much for my taste...). With
good lenses, checked with sharp film (which would exaggerate the
aperture performance differences), and with high magnification of the
image center, one might observe really excellent performance at f4-5.6,
slowly declining to so-so by f16 or so... (though the corners of the
20-35mm zoom should not follow this pattern, unless it is one VERY
exceptional lens!). We may simply have a difference in interpreting
the results. I would be VERY surprised to find very noticeable
differences in center performance at f16 among good samples of most
Nikkors on the same film shot (and processed) under similar conditions...
Hope This Helps
(If all goes well, "HTH" and "Bob Neuman" will become "David Ruether"
- with the advent of a web page and a new e-mail account...;-)