Hi--

>>Since a point light source photographed against a dark background is an
>>ideal way to photograph the effects of diffraction, you can clearly
>>see those effects. Aperture selection and diaphragm shape will alter the
>>amount and character of the effect, but the effect is inherent in optics
>>and does not change with different lenses.

>Actually I considered the possibility of diffraction (the more
>so because I find six `legs' in the stars of a lens with six
>diaphragm blades), but since the aperture is so much larger than
>the wavelength of the light, I dropped it. Nevertheless you are
>probably right, because of the extreme brightness difference
>between the light source and its dark surroundings. It makes me
>wonder what the effect would be of a lens with so many blades
>that the aperture is next to circular. If there's is time,
>I'll have to calculate the Fraunhofer diffraction pattern.

I think it would be the classic concentric ring pattern....

>I have had another reply saying that the effect I describe
>is known as `coma' and relates to the internal structure of
>the glass, with the index of refraction being different in
>different directions, and that it could be minimized in
>expensive lenses such as the noct-Nikkor and a certain
>Leica lens.

Blather....;-) Well - there is coma, and it CAN be reduced...,
but this is not what you have seen (it appears progressively
greater as you approach the edge of coverage, and it is not
rotationally symmetrical - it looks more like seagull wings,
and it gets smaller with stopping down, not bigger).

>Because of the symmetry, and because it sounds very likely,
>I would choose for your diffraction answer. The only problem
>that remains is that my neighbour here at work tells me that
>he also observes stars when he looks at such bright light
>sources with his glasses on. To his best knowledge, the diaphragm
>in his eyes does not consist of a number of blades!

Hmmm, vertically oriented bright lines are possible with squinted eyes
(horizontal for cats at small stops...;-), but my guess is that fine scratches on his lenses are causing the diffraction pattern.
Uh, remember "diffraction gratings"...? ;-)

>>BTW, this effect is also clearly
>>visible when photographing bright sun reflections on water at small stops.
>>Makes one wonder about how wonderful a lens could be if optimised at maybe
>>f2.8-4, removing much of that diffraction-caused point spread for each
>>subject point rendered by the lens...

>Oh, but most f1.4 lenses have their optimum performance at f4-f5.6!

Um, um, um...! ;-) I find that one of the myths extant is that faster
lenses reach their peak (abberation reduction hits diffraction increase)
at wider apertures - but most lenses for 35mm format peak just about f8,
but for the few slightly better ones that may peak in the center at wider
apertures. A lens that peaked at f2.8 would be VERY, VERY sharp at f2.8!

>It appears to me that the more tele a lens is, the less problems
>diffraction at the aperture edges gives, since it is the absolute
>size of the diaphragm that matters. Wide-angle lenses should behave
>worse in this respect.
> Paul van Walree

Hmmm - judging by the size of the hole, this should be true, yet it
doesn't seem to work this way. Good lenses from 8mm to 400mm seem to
peak at about the same aperture, and perform about equally well.
When all are well into diffraction-limited performance, they perform
very similarly at the same relative aperture, so what you say cannot
be true, logical as it seems...
David Ruether