PDA

View Full Version : Lens Shimming Experience



ic-racer
29-Sep-2008, 19:51
I just shimmed my 8x10 lens today and thought I would share the experience.

INDICATIONS:
1) Ever since I started using the lens I could tell that the edges, at infinity, did not focus exactly on the same plane as the center (http://www.largeformatphotography.info/forum/showpost.php?p=293638&postcount=96). The error was small and easily subdued with f45, but I wanted to 'fine tune' the lens for optimum performance.
2) The lens was an e-bay special, where the lens came in a shutter with a non-matching f-stop scale (http://www.largeformatphotography.info/forum/showpost.php?p=285641&postcount=91). There were no shims.

DEFINITIVE TEST:
First I set the back swing to zero (no front swing available). I chose a distant, easily focused, object and rotated the camera on the leveled tripod so the object fell on either edge of the ground glass. I adjusted the back swing until the object was in focus when placed on either side of the ground glass.

Then I rotated the camera to place the object in the center and confirmed that it was slightly out of focus. Indicating a curved field.

I loosened the front element and progressively unscrewed it and re-checked edge and center focus of the distant object. Eventually I got to the point where I got good simultaneous focus at the edges and the center.

FINDING THE SHIM THICKNESS:
I measured the depth of my threads and divided by the number of threads and got 0.714 mm per thread. I needed to rotate the lens about two revolutions to get the flat field and this translated to about 1.4 mm (I was expecting less, but it is what it is...)

Erik Larsen
29-Sep-2008, 19:55
Dang it, now you are gonna make me test my lenses:) Ignorance was bliss!!
erik

ic-racer
29-Sep-2008, 19:56
I had a washer that was about the correct thickness but the hole was too small. After some work with the Dremel, I achieved a 40mm opening.

Skorzen
29-Sep-2008, 20:00
Hmm I have a 150mm Kowa in a shutter with no shims, but I believe it is supposed to have some.... might give this a try.

ic-racer
29-Sep-2008, 20:04
After some fine shaping with hand tools and sandpaper I painted the shim semigloss black. There is a place in the front of the shutter where the shim fits nicely.

This picture of a similar Symmar-S 210mm from the internet shows either a trim ring or spacer near the base of the lens that was missing on my lens.

ic-racer
29-Sep-2008, 20:07
Put it all back together and it looks great. Too dark to try it today, but I'm anxious to shoot some pictures with it tomorrow.

ic-racer
29-Sep-2008, 20:28
Thanks to all the previous posters on this topic for giving me the idea to try it. With this 8x10 format and this lens, the effect was clearly visible on the ground glass.

Doing this with my 4x5 would be more difficult. One reason is most of my 4x5 lenses have big image circles, so the 'edge' of the field is not going to be in view, so some more complicated camera gymnastics would probably be required to see an effect.

Don Hutton
29-Sep-2008, 21:26
I've never seen a shim on a factory shuttered lens which was remotely close to that thick - every one I have ever measured has been between 40 and 70 thousandth. I have seen thicker shims on barrel mounted lenses like Kowas and even those are considerably thinner than 1.4mm.

I presume you did this "testing" at shooting apertures, not wide open? Wide open, you'd have all sorts of issues going on at the edge of the image circle due to lens design which would explain why you need a monster shim to "correct" what you're seeing. I'd strongly advise that you make test images at shooting apertures (say f16 to 32) and very closely examine the resulting negs with a 10X loupe.

Don Hutton
29-Sep-2008, 21:29
After some fine shaping with hand tools and sandpaper I painted the shim semigloss black. There is a place in the front of the shutter where the shim fits nicely.

This picture of a similar Symmar-S 210mm from the internet shows either a trim ring or spacer near the base of the lens that was missing on my lens.That "trim" ring is purely cosmetic - it does not even contact the barrel of the front element. They are often missing as they are easily broken when you try to remove them. Spacers fit right between the barrel and front of the shutter (or between the rear element and shutter) and are pretty much never visible until you unscrew the elements. I'd seriously doubt that you would ever need a shim remotely close to that thick on a lens which was sold shutter mounted from the factory. FWIW, if you're also trying to squeeze 8x10 coverage from this lens, you will substantially reduce the coverage of the lens with a "slice of toast" shim stuck in it - check it out with and without... You want the edge of the front barrel and the edge of the rear barrel as close togther as possible for least mechanical vignetting.

Paul Fitzgerald
29-Sep-2008, 22:25
ic-racer,

"Then I rotated the camera to place the object in the center and confirmed that it was slightly out of focus. Indicating a curved field."

Why did you not use rear shift? Rotating around on the tripod only indicated rotating around on the tripod. I agree with Don, I have never seen any shims that thick, something is wrong.

just a thought.

Dan Fromm
30-Sep-2008, 02:06
Hmm. So the OP has a 6/4 plasmat type lens.

A while ago I moved the cells of a 105/5.6 6/4 plasmat type from their factory barrel into a Copal #1 Press shutter that came from a Polaroid CU-5. As best as I can measure, in that shutter the lens total length is ~ 0.3 mm longer than in a standard cock-and-shoot Copal #1 or Compur #1 or Prontor Press #1, also 0.3 mm longer than the cells' spacing in their barrel. So I'd done the equivalent of putting in an 0.3 mm shim.

It happens that my friend Eric Beltrando has the prescription for my lens and has written ray tracing software. See his site www.dioptrique.info . So I asked his opinion. His reply was to the effect that 6/4 plasmats' performance is not very sensitive to cell spacing and that I had to think in terms of "error"/focal length, not in terms of absolute "error." My little lens is a 105, 0.3/105 rounds to .003f; not enough, he said to worry about. And the lens shoots very well.

The OP changed his lens cell spacing by .005f, a hair more than I did. If Eric's right, we're all bobbing around in a teapot tempest.

Note that Eric's remarks were about 6/4 plasmats; other designs can be much more sensitive to spacing.

Cheers,

Dan

Struan Gray
30-Sep-2008, 02:11
Then I rotated the camera to place the object in the center and confirmed that it was slightly out of focus. Indicating a curved field.

Executive summary: your method is correct, except for this bit where you rotate the camera to re-frame the scene. If instead you use front shift you will get the right shim thickness.

Let's say you have rotated the camera to put your test object at the edge of the ground glass, and focussed so that it is sharp. Assuming you have applied no swings or tilts, the plane of sharp focus will lie perpendicular to the lens axis at the right distance to pass through your object off to one side. Ten seconds with the back of an envelope should suffice to see that the plane crosses the lens axis at a distance which is closer to the lens than the object you have focussed on. So when you rotate the camera to place the object on the lens axis, the plane of focus is too close. In smaller formats the difference is usually too small, so the common focus-and-reframe technique causes no problems, but in LF the depth of field is often shallow enough to make the focus error obvious.

What you then did when you unscrewed your lens elements was to introduce field curvature so that the focus 'plane' became roughly spherical, with the right radius to match the distance to your object. Ironically, you weren't getting everything in focus by eliminating field curvature, but by increasing it. That's why you needed such a thick shim.

If you re-do the experiment with front shift instead of rotations the changes in focus should be more subtle, and you should end up with a much thinner shim. As a side benefit, the end result should also be much sharper across the whole field of view. If you don't have enough shift, find a large wall with texture and set yourself up perpendicular to it.

PS: rear shift won't work because it moves the edges of the frame onto the optic axis. You need to compare focus on-axis with what you get off axis.

mikec
30-Sep-2008, 05:25
Don, 40 thou = 40/1000*25.4 = 1.016mm
70 thou = 1.778mm
looks like the ice-racer is still in the ball park

Don Hutton
30-Sep-2008, 05:28
Don, 40 thou = 40/1000*25.4 = 1.016mm
70 thou = 1.778mm
looks like the ice-racer is still in the ball park
Just checked my notes - I meant 4 and 7 thou! I've also seen them on all sorts of different designs - my 210 Apo Sironar S has a very thin shim (4 thou) - I can't believe that Rodenstock would bother if it makes little to no difference to move the elements apart by that much/little.

ic-racer
30-Sep-2008, 07:28
o see that the plane crosses the lens axis at a distance which is closer to the lens than the object you have focussed on.

Actually my method is sound. The object focused on is essentially at infinity. It is at infinity from both the center and edge of the image plane. Imagine that I used three stars, one in the center, and one at each edge. Then imagine that I rotated the camera to place a single star at the center and at each edge. The results are the same as all three stars are at infinity focus.

ic-racer
30-Sep-2008, 12:09
Thanks for the comments. I'll certainly check it at working aperture, I can always take the shim out if something is amiss.

If anyone has the overall dimensions of a 210 Symmar-S Multicoated that would be appreciated. Schneider lists specs for the non-MC lens (which I think is a different lens, because of the 294mm image circle; my lens has a circle of 310mm).

The place where the info should be is in this pamplet: http://www.schneiderkreuznach.com/archiv/pdf/symmar_s_90.pdf

There is a picture of the 210 Symmar-S Multicoated on the cover but....the lens is not listed in any of the tables!!:eek:

Don Hutton
30-Sep-2008, 12:20
Thanks for the comments. I'll certainly check it at working aperture, I can always take the shim out if something is amiss.

If anyone has the overall dimensions of a 210 Symmar-S Multicoated that would be appreciated. Schneider lists specs for the non-MC lens (which I think is a different lens, because of the 294mm image circle; my lens has a circle of 310mm).

The place where the info should be is in this pamplet: http://www.schneiderkreuznach.com/archiv/pdf/symmar_s_90.pdf

There is a picture of the 210 Symmar-S Multicoated on the cover but....the lens is not listed in any of the tables!!:eek:
I'm pretty certain that the multi and single coated versions are optically identical. I'm not sure how you are "calculating" the image circle - I suspect your 310mm is the circle of illumination whereas Schneider's image circle figure is likely defined by decent optical performance - sort of scientific "soft and muddy" don't count. FWIW, I've never seen one of these that covered 8x10 in a useful manner. The most recent Apo Symmar L just hits the corners at f22 (properly - so that you can see a complete un-obscured aperture at each corner), but I didn't find it at all "useful" on 8x10 due to bascially zero room for movement, and that has 75 degree angle of coverage.

ic-racer
30-Sep-2008, 13:00
I've never seen one of these that covered 8x10 in a useful manner.

Here you are! Sharp up to each corner, 310mm film diagonal. Assessed with 8x loupe on the light table and in 16x20 projection prints. Additional 5mm of 'light' available by watching the clipped corners of the ground glass.

Someone else out there has this lens, I'll try to find the link...

Struan Gray
30-Sep-2008, 13:28
Actually my method is sound. The object focused on is essentially at infinity. It is at infinity from both the center and edge of the image plane. Imagine that I used three stars, one in the center, and one at each edge. Then imagine that I rotated the camera to place a single star at the center and at each edge. The results are the same as all three stars are at infinity focus.

True. However, in your original post you said "a distant, easily focused, object". My humble experience is that in LF infinity tends to be further away than you expect.

I'm happy to be wrong.

Like Don though, I've never seen a factory shim that thick.

Don Hutton
30-Sep-2008, 13:41
Here you are! Sharp up to each corner, 310mm film diagonal. Assessed with 8x loupe on the light table and in 16x20 projection prints. Additional 5mm of 'light' available by watching the clipped corners of the ground glass.

Someone else out there has this lens, I'll try to find the link... Those are both a lot closer than infinity which is generally where one speaks of definitives in coverage terms except with macro lenses. Also, you probably want to asses vignetting in the corners on even mid to light tones rather than dark to black bits of forest!

ic-racer
30-Sep-2008, 14:25
Those are both a lot closer than infinity which is generally where one speaks of definitives in coverage terms except with macro lenses. Also, you probably want to asses vignetting in the corners on even mid to light tones rather than dark to black bits of forest!

You don't think the lens that made those images covers 8x10???

I have been using this lens as my sole 8x10 lens for almost a year and I know what it can and can't do. If this lens has fooled me into thinking it covers 8x10, it's a pretty clever lens and I'll gladly admit to being duped :)

Many of the posts I have found on the internet seem to indicate the Symmar-S MC 210mm is similar to the APO-Symmar 210mm which has a listed circle of 321mm and this would be consistent with my observations. The Schneider data for the APO-Symmar states 5-7mm of movement is available and that about what I observe with my lens.

ic-racer
30-Sep-2008, 14:34
True. However, in your original post you said "a distant, easily focused, object". My humble experience is that in LF infinity tends to be further away than you expect.

I'm happy to be wrong.

Like Don though, I've never seen a factory shim that thick.

Remember that thread on 'infinity.' Someone posted 1500 x the focal length would suffice, so I was shooting for that. So, not infinity, but perhaps close enough.

Once it stops raining here, I'll shoot some pictures and post some followup.

Struan Gray
30-Sep-2008, 14:47
Remember that thread on 'infinity.' Someone posted 1500 x the focal length would suffice, so I was shooting for that. So, not infinity, but perhaps close enough..

My take on that is archived here:

http://www.largeformatphotography.info/forum/archive/index.php/t-15134.html

Don Hutton
30-Sep-2008, 15:18
You don't think the lens that made those images covers 8x10???I'm not arguing with you - I'm just pointing out that the images proving it "covers 8x10" are not lose to inifinity - more like 5-20 feet. There's a big difference. If you're happy the lens works for you on 8x10, great.

Many of the posts I have found on the internet seem to indicate the Symmar-S MC 210mm is similar to the APO-Symmar 210mm which has a listed circle of 321mm and this would be consistent with my observations. The Schneider data for the APO-Symmar states 5-7mm of movement is available and that about what I observe with my lens.
This information is simply incorrect regarding the coverage of the Symmar-S and Apo Symmar 210mms. The coverage per Schneider of the Apo Symmar 210mm is 305mm at f22. Angle of view at f22 is 72 degrees: http://www.schneiderkreuznach.com/foto/sr_ap1/sr_ap1.htm The latest current design if the Apo Symmar-L - a significant departure from the previous plain Apo Symmar - coverage has been increased to 75 degrees and Schneider's data states an image circle of 321mm. See here http://www.schneideroptics.com/pdfs/photo/large_format_lenses.pdf

The Symmar-S was designed with 70 degree angle of view which equates to only 294mm at f22 according to Schneider, so not only a little narrower than the older Apo Symmar, but considerably narrower angle of view than the latest Apo Symmar-L. See http://www.schneideroptics.com/info/vintage_lens_data/large_format_lenses/symmar-s/data/1,5,6-210mm.html

ic-racer
30-Sep-2008, 15:56
http://www.schneideroptics.com/info/vintage_lens_data/large_format_lenses/symmar-s/data/1,5,6-210mm.html

Where is your data on the Multicoated Symmar-S that shows said angle of view and image circle? That info in the link is for the NON-MC which I don't own.

I pointed out earlier that the info for my lens should be in this: http://www.schneiderkreuznach.com/archiv/pdf/symmar_s_90.pdf
but it's not there in spite of a picture of the lens on the cover! If you know where the info is please post it :)

Don Hutton
30-Sep-2008, 16:31
Where is your data on the Multicoated Symmar-S that shows said angle of view and image circle? That info in the link is for the NON-MC which I don't own.

I pointed out earlier that the info for my lens should be in this: http://www.schneiderkreuznach.com/archiv/pdf/symmar_s_90.pdf
but it's not there in spite of a picture of the lens on the cover! If you know where the info is please post it :)
The info is exactly the same for the multi and single coated Symmar-S lenses - they didn't change the design, they simply started multicoating them instead of single coating them - the 150 Symmar S was in fact the first lens made by Schneider which was multicoated (in 1977 - some years into it's production) and the second was the 210mm Symmar-S ( http://www.schneideroptics.com/info/faq/photography.htm#q8 )- they continued production for many years after they started multicoating the same design. There is no difference. For proof that particular brochure in fact refers to the 210mm Multicoated version - refer to the first page where it lists the features of the Symmar S lenses and refers to the multicoating! It would also appear from the printing subscript that the brochure was published in 1987 - ten years into the production of MC Symmar S lenses... You still think it's for "the NON-MC which I don't own"?...

Keith Pitman
30-Sep-2008, 16:33
In Thalmann's View Camera article from 2003, he presents a table for the Symmar S series lenses that shows the 210 with an image circle of 294mm/70 degree angle.

Interestingly, my 8x10 was sitting on a tripod across the room. I put my 210 Symmar S (MC) on it and took it outside. Focused at infinity, it sure seemed to be bright and clear right out to the corners. I was suprised.

My lens doesn't have a shim, and the front element doesn't contact the "trim ring" referenced as missing in an earlier post.

ic-racer
30-Sep-2008, 16:42
The info is exactly the same for the multi and single coated Symmar-S lenses - they didn't change the design, they simply started multicoating them instead of single coating them - the 150 Symmar S was in fact the first lens made by Schneider which was multicoated (in 1977 - some years into it's production) and the second was the 210mm Symmar-S ( http://www.schneideroptics.com/info/faq/photography.htm#q8 )- they continued production for many years after they started multicoating the same design. There is no difference. For proof that particular brochure in fact refers to the 210mm Multicoated version - refer to the first page where it lists the features of the Symmar S lenses and refers to the multicoating! It would also appear from the printing subscript that the brochure was published in 1987 - ten years into the production of MC Symmar S lenses... You still think it's for "the NON-MC which I don't own"?...

The brochure link I posted does not list any 210mm lens! Where does it say they did not change the design?

Don Hutton
30-Sep-2008, 16:53
The brochure link I posted does not list any 210mm lens! Where does it say they did not change the design?Here's a link to what appears to be the same brochure with the 210mm included - note the multicoating reference on the first page as I mentioned and also the specs for the 210 later on clearly showing that the lens you have has an image circle (according to Schneider) of 294mm at f22 or field of view of 70 degrees - sunstantially less than the 210 Apo Symmar and the 210 Apo Symmar-L: http://www.schneideroptics.com/info/vintage_lens_data/large_format_lenses/symmar-s/Symmar-S_Group_Data.pdf

Anyway, I've provided some enlightening info for you in this thread - it appears now that your main interest at this point is to have an argument, so I'll pass on filling you in with any more facts - but the irrelevant information on the identical design of single and multicoated Symmar S lenses is out there....

ic-racer
1-Oct-2008, 06:47
Here's a link to what appears to be the same brochure with the 210mm included - note the multicoating reference on the first page as I mentioned and also the specs for the 210 later on clearly showing that the lens you have has an image circle (according to Schneider) of 294mm at f22 or field of view of 70 degrees - sunstantially less than the 210 Apo Symmar and the 210 Apo Symmar-L: http://www.schneideroptics.com/info/vintage_lens_data/large_format_lenses/symmar-s/Symmar-S_Group_Data.pdf

Anyway, I've provided some enlightening info for you in this thread - it appears now that your main interest at this point is to have an argument, so I'll pass on filling you in with any more facts - but the irrelevant information on the identical design of single and multicoated Symmar S lenses is out there....

Thank you, I have never seen that document. It does indeed indicate the image circle is less than 310, however, the specification is grossly underrated. That is not an argument, it is an observation based on measurement and experience. If you contend that the lens does not cover 8x10, please post your measurements and pictures so that we can compare and contrast to that which I have posted :).

It just irks me that someone would start an unprovoked rant claiming the lens I use does not cover the format on which I use it, especially in a thread that is not about lens coverage! What was your point?

ic-racer
1-Oct-2008, 06:58
ic-racer,

"Then I rotated the camera to place the object in the center and confirmed that it was slightly out of focus. Indicating a curved field."

Why did you not use rear shift? Rotating around on the tripod only indicated rotating around on the tripod. I agree with Don, I have never seen any shims that thick, something is wrong.

just a thought.

Good point, but I don't own a camera with rear shift :D

Since my focal point for testing was not quite at an 'infinity equivalent,' the error in my method is as follows:

Object distance I measured today was about 500 meters away.
Camera angle of swing was about 55 degrees from side to side.

This swings the focal plane in a radius of 500 meters with a cord depth of about 56 meters calculated from a 55 degree wedge of view.

So my error is about 56 meters at 500 meters away. That is roughly 10% which compares quite favorably to the about 30% focal error prior to shimming.

To my eye on the ground glass, with the lens at f5.6, objects at both 500 meters and 556 meters were both in acceptable focus.

ic-racer
1-Oct-2008, 07:18
CONCLUSION:

Shimming the lens improved flatness of field at the far edge of the image circle, and thus has effectively enlarged the useful image circle beyond the manufactures claim.

Struan Gray
2-Oct-2008, 00:52
ic, I don't want to be argumentative, and I don't want to tell you how to enjoy your own photographic equipment and activities, but your conclusion is entirely wrong.

Please note that I am not saying that you don't see what you see, but that your interpretation of your results in terms of optical science is at odds with the conventional use of the technical words you are using.

Had you really flattened the field, you would have objects in focus at 500 m at the edges of the frame and 443 m in the centre, not 556. If you have really done as you say, you have bent the focal 'plane' even further from where it was: it is now even more bulgy than the spherical surface you claimed to have created in your original posts.

Factory shims tend to be much, much thinner than 1.4 mm. If your lens needs such a thick shim something is very non-standard about either the lens cells or the shutter. Some wide angles that were originally mounted in the Sinar DB mount have lens cells that are incorrectly spaced when placed in a normal shutter. Some shutters have had the locating surface machined down to accomodate such a lens, or an adapter of some sort - for example to put Copal 0 lenses into Copal 1 shutters. But a 210 Symmar-S seems an unlikely candidate for any of these explanations.

ic-racer
4-Oct-2008, 15:51
Please note that I am not saying that you don't see what you see, but that your interpretation of your results in terms of optical science is at odds with the conventional use of the technical words you are using.

.

Thanks Struan on that pickup. Indeed the math on the slight field curvature is all correct. I just used an 'example' of not being able to tell a 10% difference (500m vs 565m) and this was a math error. It should have read 500m vs 443m.

Just to clarify the math. Relative motion lets us imagine my 500m-away lamp post was revolving around the camera in an arc of 500m (same as swinging the camera horizontaly on the tripod). The chord of the arc (at a 55 degree swing angle) is indeed 56 meters CLOSER when the lamp is direct center. (as you pointed out).

During the test only the 500m-away lamp was focused upon.

The next day I tried to see if a 10% focusing error would be visible at 500m, and this was a separate test from the lens shimming test.

As to the size of the shim, well, as I pointed out it is-what-it-is. Anyone out there can unscrew their lens and check it next time they are setting up a shot.

Daniel: I didn't mean to start an argument and apologize if it came across that way.