On Wed, 8 Jan 2003 22:33:00 +0000, David Littlewood <david@nospam.demon.co.uk> wrote:

>In article <3e1c9f04.1130298@newsstand.cit.cornell.edu>, Neuman -

>Ruether <d_ruether@hotmail.com> writes

>>On Wed, 8 Jan 2003 18:59:53 +0000, David Littlewood

>><david@nospam.demon.co.uk> wrote:

>[snip]

>>

>>>Vignetting (whether internal or external) is normally at its worst at

>>>maximum aperture. Particularly with a wide-angle lens, the image of an

>>>external obstruction may become visibly sharper on stopping down, which

>>>may create the appearance of it being worse.

>>Yes - but at most middle to wide stops, the visible

>>effect of the vignette is minimal, and generally

>>acceptable; only at the smallest stop is the vignetting

>>shade really objectionable. In reality, virtually all

>>lenses (at least the short and medium FLs...) vignette

>>at their widest stops due to just their physical

>>structure, and adding a bit more with a shade is

>>generally not very important...

>Well, it can make "just acceptable" become "quite unacceptable". Maybe

>we don't really disagree. At wide apertures you get more vignetting, but

>it is spread out across a wider area with no sharp edge. At smaller

>apertures, the area affected by vignetting has a harder edge, but is

>further from the axis and may be outside the image area. Thus, you

>either escape the effect or it looks worse. My view is that I would want

>it not to occur in the worst case (wide open); if it does not occur

>there, it doesn't occur at all.

Yes, we agree on this (or at least on the theory...;-).

In practice, I find that shades that *just* avoid the

sharp-edged vignetting at the smallest stop are quite

acceptable at the widest stop, though in theory, they

should, as you said, show a larger, softer-edged vignette

at the largest stop. Since in practice this effect

appears to be minimal, I'm guessing that the (normally,

alas...) insufficient clearance of the edges of the optics

to avoid vignetting at wide stops are also hiding the

potential vignette due to the shade (the lens doesn't

"see" much of the shade that would otherwise cause some

vignetting...) - but mebbe I'm "talking through my hat"

on this...;-) In any case, I often use shades that cover

rather exactly at small stops (checked with an F3 VF,

that is close to 100%), without visible ill effects...

>Many tele zooms also show significant vignetting; my Canon 100-300

>f/4.5-5.6USM (now deceased after a swim in a rock-pool) vignetted quite

>badly wide open at the long end.

Some vignette is present with almost all lenses - but

with teles, the vignette is often so soft that it just

effectively slows the lens at the widest stops...

>>>In my experience checking for vignetting by looking through the

>>>viewfinder is much more difficult than seeing it on film. This would be

>>>even more the case if the lens were stopped down.

>>No - when well stopped down, the abrupt darkening of

>>the viewfinder corners caused by a vignetting shade

>>(or even lens structure) is quite evident with short

>>FL (for the format) lenses when viewing a bright

>>subject area... BTW, more awkward, but the technique

>>used for large format for checking coverage can also

>>be used with 35mm: with the lens stopped down to

>>minimum, look through the front of the lens for

>>the VF screen corner (this takes a light shining into

>>the VF and illuminating the corner of the VF to

>>work) - if the corner is visible from the front, there

>>is no sharp vignette; if not, there is...

>YMMV; I have often found visible signs of vignetting on slides when no

>sign was noticeable through the viewfinder. Maybe I just did not look

>hard enough...

100% VF coverage is best for judging this, and only with

the lens stopped down, viewing a large, even, bright area;

wide-open, no-one can see any but the worst vignetting

in the VF...

>Your tip about looking through the lens on a view camera is not one I

>have heard before. I must try it; I use a Lee variable lens hood and

>it's a bit fiddly to get maximum shading without going too far.

With view cameras, it is usually done from the rear,

looking at the lens aperture through the (clear) cut-off

corners of the viewing screen - but the reverse also

works. Notice that in both cases one rarely sees the

whole aperture circle if the lens is wide-open (due to

lens structure cut-off, which also results in wide-stop

vignetting...).

>>>Also, the best check for vignetting should IMO be at infinity focus.

>>>When focussed close up, the image on film (or focussing screen) is only

>>>a central fraction of the normal image circle* and you are likely to be

>>>missing the obstruction.

>>>* A 35mm lens may serve as a decent macro lens for larger formats

>>In this you are (generally) quite correct - I must have

>>been asleep when I wrote the reverse (though this is not

>>always true - some lenses [mostly zooms...] show a wider

>>field of view at closest focus rather than at infinity

>>focus...).

>These must (presumably) be internal focussing lenses where they focus by

>changing the focal length otherwise I don't think it would be possible.

Many zooms show the field of view increasing noticeably

with close focus (some that I recall: the Nikkor 17-35mm

f2.8, 25-50mm f4, 28-70 f3.5-4.5, and 35-105 f3.5-4.5

[AF] - though not the 70-210 f4 and 80-200 f2.8).

With those that show this effect, vignetting should be

checked at closest-focus rather than at infinity-focus...