Folks,

Regarding the "missing repair person" thread yesterday, I have also worked with Sherry and found her very helpful and willing to discuss things if she has the time. The fourty year old M I got from her is beautiful.

My question regards my first M lens, a 35mm Summicron. It's the second version (right?) with the aperture ring tab. It has a strange spotlight effect, where the center of the frame is bright, but the exposure gradually falls off tward the outer part of the frame, about a one stop difference, a bit more on the left than on the right. Very consistant roll to roll and in all lighting.

Wouldn't call it vignette (SP?) exactly - it's not just a corner effect. Not using filters or a hood. Happens at all stops - not better closed down. Most weird - the images look sharp all the way to the corners. The shadow curves into the corners, so don't think it's a straight line effect, like a curtans moving slow. (The shutter had been CLAed.)

I know fall off is commonly found in bench tests, but this is very noticable in regular shooting. Not evident with my early 70s 35/2.8 Nikkor, for instance. Is this lens out of spec? Or, is this a characteristic of Summicrons of this vintage? I don't like it, but otherwise the image quality is great - really does make my Nikons look soft.

Any ideas? Somebody want to lend me another lens I can try? Just kidding.

-- Carl (cpultz@earthlink.net), June 06, 2002

Answers

If you hadn't said it happens at all apertures I'd say it was ok. The lens should be checked by a pro.

-- Jay (infinitydt@aol.com), June 06, 2002.

what hood are you using? where do you place your hands when you shoot? do you allow gulls to perch atop the camera/lens while you are shooting? any of these can be a source of mechanical vignetting. otherwise you have a serious decentering problem. have the lens checked. (i am assuming you don't have this problem with other lenses. you should certainly mount another lens on the camera and take a few shots -- even in a camera store -- to make sure the body is functioning properly.)

-- roger michel (michel@tcn.org), June 06, 2002.

It is 'natural' (as opposed to mechanical) vignetting. It is common to all "true" wide-angle lenses - and occurs less in retrofocus designs (like your 35 f/2.8 Nikkor) for simple optical/math reasons.

Draw a horizontal line 1.61 inches long. This represents the 35mm film (corner to corner). Draw a dot 35mm (1 and 3/8ths inches) above the middle of the line. This represents the aperture of your 35 summicron.

Now - draw diagonal lines from the dot to the ends of the 'film' line and measure them - they are substantially longer than the 35mm distance in the center (I get about 42mm).

So the corners of the film are substantially further from the light source than the center. Since light falls off as the SQUARE of the distance - (I'll let you do the intervening math) the center gets about 1.4x the exposure of the corners, simply because the light has had more distance to spread and 'thin out'.

Now if you go back and draw another dot above the first, at about 50mm from the film - that represents the aperture of your retro Nikkor 35, which has to leave room behind it for the SLR mirror. Do the same measurements and math for the corners, and you'll find the natural light fall-off to be only about 1.2x.

It IS called vignetting, but you're right, it is not the same thing as a lens hood getting into the image area, or even the lens barrel getting into the image area. THAT's mechanical vignetting.

Look through the back of any lens and move your eye from the center to the side. At the point where the circular/oval aperture starts to take on a 'cat's-eye' shape it is starting to get 'mecha' vignetting (to borrow from the movie "AI"). The edges of the various elements are starting to crop into each others field of view - which further reduces the amount of light reaching the corners.

This is common to most lenses regardless of whether they are retro, tele or 'normal' designs - but varies substantially from lens to lens. purely based on the way the designer drew it.

-- Andy Piper (apidens@denver.infi.net), June 06, 2002.

Thanks, Andy! That makes sense to me, a know-almost-nothing about optics. Wish I could post an image today so you can see the effect. Maybe Roger and Jay are right if the falloff is more than one would expect. If y’all will indulge me, I’ll get one to the net for evaluation tomorrow. I don’t want to bug a repair person over something that isn’t broken.

Hmm. So there are practical limitations imposed on designs by the physical size of the lens barrel (smaller on rangefinder cameras) and a cost to not using retrofocus. I gathered from the lit that everything would be better with these things, so you can understand my surprise.

I suppose that when you use an enlarger (all they had in 1970), this effect is compensated by the falloff of that system. No need to burn in the corners with this lens! These days, I'm using a scanner and there's no compensation for this effect.

No offense to Leicaphiles. The lens has clarity, richness and crispness that is unlike anything I've used before. I'll have to learn to dodge with Photoshop, if this baby can’t be fixed.

I’ll get that photo up tomorrow.

Carl

-- Carl (cpultz@earthlink.net), June 07, 2002.

Just to put a close to this, and for anyone finding this thread in archive, I've found another factor that was effecting the images. My scanner software was creating an exaggerated contrast that caused the falloff to be more severe than is apparent when a more normal gamma is used. I was working out a problem with shadow detail (scanning negatives) and discovered this effect. Spent all weekend recalibrating the workflow. Found out that Vuescan rocks!

Anyway, while there is this falloff effect on the Summicron, it's not as nasty as it first appeared, and does diminish as the lens is stopped down. It's a nice lens - small, quick, sharp.

-- Carl (cpultz@earthlink.net), June 10, 2002.