Comparing Two Identical Lenses; One With Rear Scratch; Can't Detect A Difference!

Help me understand this one because I have always been under the

impression that ANY scratch on the rear element would have some

effect, even if minute: I have two 100mm Linhof select Planars, one

is mint while the other would be mint EXCEPT....on it's rear element

there is a 1/4" scratch by an estimated 1/10 mm. The scratch starts

around 1/8" from the edge and heads straight in towards the center

(1/4" in length). The scratch has been carefully filled with flat

black paint.

 

For the fun of it (I was bored what can I tell you!), I did side by

side comparisions using wide open, f/5.6 and f/11 looking at a

backlit scene and a set against a low contrast subject. The camera

was mounted on a very heavy tripod with fluid head, cable release,

bean bag over the camera (in otherwards, all steps to minimize

vibration). Much to my surprise, I cannot find ANY detectable

difference when using a quality 15X loupe.

 

Do I need to go back to Lens 101? Shouldn't there be a slight

difference?

There's probably a slight difference: you probably have an area with maybe

one percent less exposure spread over the image. I can't see that kind of

difference.

Your experiment tells the tale, regardless of your expactations!

 

I used to have a Commercial Ektar that had a nice ding in the rear element. I painted it in with black ink and could never tell the difference between the before and after pictures.

An unrepaired scratch should degrade the image with more flare, lowering the image contrast. Filled with paint as you describe, it decreases the apparent aperture in proportion to the image area covered by the repair. Either would be difficult to discern on the ground glass, and likely very difficult to measure with a densitometer.

I have probably the ugliest Caltar 210mm in existence. The rear element has a large chip that covers about 20% of the surface. I used some black construction paper to cover that part. The front filter ring is badly dented.

 

Takes some great pics. [shrug]

In my introductory physics class in undergrad we did an "experiment" where we had big bi-convex lenses and projected a light on to the wall through them. Then we made some sort of big mark on it with a sharpie marker. The instuctor then had us conclude that all those photographers out there that worry so much about dust and other such things on their lenses are full of it, because none of us could detect any change in the light projected on the wall.

 

Well, I was only beginning in photography at time and I thought he was wrong in his overly simplistic conclusion, but I could not say why. Now-a-days I think that there was infact a difference, but we were merely unable to discern it. Gunk on the lens should merely cause an increase in the softness in the waves passing through that area and a decrease in exposure (neither of which would be readily discernable when only projecting undifferentiated light). Sctratches in the lens will actually break up the waves as they exit, probably resulting in the same things, but to a greater extent. No doubt, any disturbance in the waves will effect the image being burned into the film, the question is "at which point does it become significant?"

 

If you shot a very tight grid pattern with high contrast, it may become apparent and therefore significant much faster than when shooting more amorphous subjects.

 

So to try and finally answer your question: undoubtedly, there is a difference, but that doesnt mean that you (or anyone) is nessecarily able to detect it in the image.

 

todd schoenbaum

With the scratched area blocked by black paint, there shouldn't be any detectable difference in the images, save for the tiny loss of light due to the blocked area. You can try removing the paint and doing the experiment again, but IMHO this whole "any defect on the rear element matters" thing is way overblown. If you're really curious, put a filter on the backside and start sprinkling crud on it. It might surprise you how much (or little) starts to be a problem.

I have an otherwise beautiful Voigtländer APO-Lanthar 150mm/f:4.5 with an unfilled 2mm scratch on the rear. I put it on a 5x7" camera, stopped down all the way to f:45, and made a careful note of where on the image the scratch would be.

 

No sign of image degradation at all - in fact, I couldn't begin to guess where the scratch was! I stopped all the way down assuming this would maximise the effect of the scratch, I usually never stop down more than f:16 or so. I also used 5x7" film to ensure that the scratch would be inside the film area - I've oriented the lens so that it falls outside most of my 4x5" photos.

Ole, remember that with the lens open, light rays pass through every part of the lens area to contribute to every point on the film. There really isn't any one place you'd expect to see evidence of a scratch. I could add some qualifiers to that (like a huge rear element close to 35mm film), but as a practical matter I believe it to be true. Thus, a reduction in contrast is all you'd expect to see, especially in large format.

Sometimes I feel that a lot of our 'beliefs' about lenses are overblown.

 

I found a link about a year ago where someone was testing a group of lenses. One cheap lens he was testing performed very well, but when he checked the lens, he found that 2 of the three elements were broken into 4 pieces! Strange, but apparently true.

 

See:

http://www.exaktaphile.com/tests/instruct.pdf

 

-- Bill

Now to apply the question to a front element:

 

Would it make sense that a front element scratch will disturb the impinging ligh tmore as far as flare etc goes? Does filling in black matter here?

 

The lens I have in mind is just a big barrel eBay find of unknown origin and could possibly be reversed anyway.

 

It's an Autologic (now owned by Agfa) lens, f.l. 165 mm or so f/2.3-8, 5" diameter rear, 4" diameter front, focusses about 1" behind lens, very narrow field (as guessed by light bulb projection onto wall). Might be a monochromatic lens, either copy, process or medical. All reflections are bright yellow. I removed the front scratched element with great difficulty (bad pitting from impact- lucky it didn't break). F.l. of front element is about 16" with scary color fringing - almost rainbow-like but mostly bright blue. Interestingly, the back f.l. is now about 5"...I might try & shoot something with it now, but don't know whether to point forward or backward...I guess whatever works.

 

Murray

Mr. Hoffmann (?), you'll see that I chose to test the lens at f:45 precisely so that an affected area of hte negative could be defined. I even noted down where it would be, with a sketch of the scene as seen on the ground glass. looking back through the lens (through the front), it was easy to see the scratch. The scene was a field with lots of long, dead grass straws - any significant loss of contrast should have been easy to see.

 

At larger apertures the loss of contrast will be less localised, and even less noticable. This suits me fine, as I prefer not to stop this fine lens down beyond f:8 (except for very specific testing purposes).

Ole, I bet diffraction is a bigger factor at f/45 than the scratch! Not to say that scratches never matter, but not to the degree that some worry about them. You said you viewed from the front. Actually, I'd be curious what you see if you remove the back (or gg) and look from the film plane at f/45. I bet the scratch still projects to a fairly large area of film. Not to go test crazy, but one could put a piece of clear glass or plastic in place of the gg, and mark the area where the scratch is within the aperture with a marker or grease pencil. Unless the scratch is near the center, I wouldn't expect to see much of it. Anyway, fun stuff to explore.