What to do with front focusing Kobalux?

[rvaubel]

My recently procured Kobalux 21mm f2.8 seemed to be a little soft at f2.8 on my R-D1. From my pictures, it seemed to be focusing a little bit in front of the spot I was focusing on. I decided to check the focus accuracey by shooting a shot of a yardstick sloping away from the camera at a distance of three feet. Focusing at a mark at 1 meter, I checked the results @ 100%. The results confirmed that the focus point was indeed 4 or 5 inches closer than indicated by the rangefinder.
None of my other lenses do this including my 40mm f1.4 Nokton.
I checked the Cameraquest adapter rings and all of them are the specified 1.0 mm.
I hope that their is some way to adjust the lens. I notice that there appears to be a ring on the back of the lens held in place by 3 screws. I wonder if there are shims under the ring for adjustment?
Anybody have any experience with this problem?

Rex

 

[Ergo]

sell it to me,haha

 

[peter_n]

When you put the lens on the cam, where does the index mark on the top of the barrel line up? Dead center (12 o'clock) or a wee bit to the left or right?

 

[rvaubel]

peter_n
A little bit to the left.
Rex

 

[peter_n]

Turn the lens a bit more into the adapter if you can. Most of those lenses tend to line up a bit to the right (around 1 o'clock or so). Then see if that makes any difference to the focus accuracy.

 

[rvaubel]

If I would turn the lens clockwise in the adapter, it would become loose. Also, I need to REDUCE the lens to focal plane distance (I think) since it is close focusing.
Rex

 

[jlw]

1.0 mm thickness of the adapter is actually a bit too thick. The correct thickness (as measured from genuine vintage Leitz adapters) is 0.98-0.99 mm. (See tediously pedantic note at the end if you want to know more about this.)



Since a wide-angle lens requires a very short amount of travel to move from infinity to close focus, this alone might be enough to make your lens front-focus on your R-D 1. I'd suggest you try to buy or borrow a genuine Leitz adapter and see if this corrects the problem.

If it turns out that this is the case, then your options are to buy a correct-spec adapter for use with this lens, or to take a bit of thickness off the adapter you already have. Since you need to remove only 0.02 mm at the most, this should be feasible by using abrasive paper to wet-sand material from the front face of the adapter. The way that worked for me was to place the abrasive paper on a flat surface and then rub the adapter against it, changing my grip on the adapter frequently so my hand motions would randomize themselves. Use very fine-grit paper (I used 600) so you're not taking off much material per stroke, and check frequently with your caliper to make sure you are removing it evenly aroundl the circumference of the adapter and that you're not taking off too much. It's a tedious job, so if you can find a genuine Leitz adapter (beware of fakes on eBay!!) at a reasonable price, that's probably a better investment of your time.

If the adapter isn't the problem, you've got a bit more of a head-scratcher. I don't know which version of the Kobalux you have (I believe there were at least three generations) but mine doesn't have any shims behind the rear flange. Also, remember that your problem isn't strictly the spacing distance per se -- it's the relationship between the position of the lens and the position of the rangefinder coupling cam that moves the coupling arm inside the camera.

On my Kobalux, this relationship is adjustable to some extent by removing the coupling cam (which sits in its own little helical) and repositioning it by one groove to the left or right; this changes the relationship between its movement and the movement of the lens in its focusing mount. Note, though, that it's touchy to disassemble, and if you mess it up (by losing track of the original entry point, for example) it will take hours and hours of trial and error to get it reassembled correctly! So, I suggest you pursue the adapter-thickness option first.

Good luck!

Tediously pedantic end note: Even though the nominal thickness difference between screw and M bodies is indeed 1.0 mm, I believe Leitz deliberately machined its adapters 0.01-0.02 mm thinner to provide clearance for easy installation and removal. If the adapter were machined 1.0 mm thick, the lens would have to be torqued onto it hard enough to eliminate ALL slack; this would put a lot of stress on the body mount and make it difficult to remove the lens from the adapter. I base this belief on what I learned when rebuilding my vintage car's transaxle: The bearings holding the gearshafts in place had to be installed with a specified amount of clearance, then the retaining nuts torqued to a specified setting in order to produce the desired final bearing clearance.)​

 

[OldNick]

I recall a thread a few days ago that said that Leica-made adapter rings will measure slightly under 1mm. If you can find one, you might give it a try.

Jim N.

 

[rvaubel]

jlw
I remember your post last week on adapter thickness. At that time I read it solely for its academic interest as I had no problem with my three Camerequest adapters or my R-D1. Since I am the owner of an R-D1, shoot wide open, and am rather compulsively anal, I am well aware of most of the rangefinder/lens/R-D1 focus issues. I have tested all of my lens carefully, including telephotos at maximim apeture and various fast lens. So far everything has been spot on. I'm 99% certain the camera body/rangefinder is within factory specs (please, no rude comments about Epsons )
All three of my adapters are within +-,.005" from the nominal 1.000 mm. this may be a lttle thick, but nevertheless, I have experienced no problem with reqards to these adapters or any of the lenses I have. I haven't tried to mix and match them yet, but my hunch is it won't make much difference.
As to your suggestion about sanding dowm the front of an adapter, I thought of that because of last weeks post. However, that will result in the lens's registrations marks rotating to the 11:00 position (looking from the top, behind the camera). No biggy, but not ideal. I like the idea of reengaging the thead one turn as it seems like the more elegant solution. Wish me luck
Oh, by the way, I am getting a little overly anal about this as various DOF calculaters indicate that the depth of field of a 21mm f2.8 lens on a 1.5 crop sensor at 39" is +-, 5". My test indicate that it is front focusing by 5" so I am just on the edge of being acceptable. At greater distance or smaller F# I would truly be nit picking. But I CAN clearly see the discrepency. And what else are hobbies for if not incessently worrying about meaningless and insignificant bits of trivia?

Rex

 

[jlw]

I'm not going to make any rude comments about Epsons, since I have one too and have never had any QC problems with it!

I agree that the seating position of the lens would be affected by sanding down the front, but if you're only removing 0.01 - 0.02mm it won't be affected by much.

As to your observation that your other lenses work fine with your adapters, I'm not disputing it. One thing to remember, though, is that the wider the lens, the less focusing movement it requires -- look at how little your 21 extends as you focus it from infinity to 1m, vs. how far a 50 or a 75 extends. What this means is that a small amount of "extra" extension caused by a too-thick adapter would be a very small percentage of the total extension for a longer lens -- but would be a very significant percentage of the total extension for a 21.

To put some numbers to this, based on a spreadsheet I made up using a formula in one of the Kingslake books:

Focusing a 50mm lens at 1 m requires an extension of 2.969 mm vs. what's needed at infinity.

For a 75mm lens, the comparable figure is 6.618 mm.

For a 100mm lens, it's 12.500 mm.

For a 21mm lens, it's 0.460 mm!

What that means: Suppose your adapter is a mere 0.01 mm too thick. That would be only 0.34 % of the extension required to focus a 50mm lens at 1 meter.

But for a 21mm lens, it would be 2.2 %!

Still... If your lens is front-focusing by 5 inches at 1 meter, then your error is in the range of 12.7 %, so that's more than would be explained by a too-thick adapter (especially after you take into account the effect of focusing-cam displacement, as explained in tedious length in one of my other posts.)

So, I agree that experimenting with the coupling helical might be a better solution for you. Just BE CAREFUL!! I screwed up one of these, so I know how tedious it is to try to rediscover the correct position via trial and error!

 

[peter_n]

If I would turn the lens clockwise in the adapter, it would become loose. Also, I need to REDUCE the lens to focal plane distance (I think) since it is close focusing.
Rex

Looking from the front of the camera, if you turn the lens into the adapter, it goes from 12 o'clock toward 1 o'clock, so it would be tighter, not looser.

Many of the old Leica LTM lenses would tighten up in the camera at around 1 o'clock - Leica made them that way so you could see the index mark from the top easier if you had a viewfinder sitting in top of the cam.

 

[rvaubel]

jlw
Which Kingslake book where you refering to? I was looking for the formula you found to determining the actual longitudinal correction required to bring the lens in focus but could not find it in "Lens design fundamentals".
But thanks for the calculations. Short focus extension makes critical focus very difficult in astrophotography where .0001" may make an image totally unusable on a fast telescope.
Your comment about not screwing up, I take to heart. This is a mint chrome 2nd version that I have been looking for a long time. Do you have any hints on disassembly? The obvious place to start is the 4 philips on the back ring. Also the heliod seems pretty coarse. Moving it over to the next thread looks like it could be to much. But then I can always change it back.
Thanks for any more info.

Rex

 

[rvaubel]

peter_n
Thanks for your comment also. I was looking at the camera from behind, so your right. Certainly sanding the adapter down would be a lot easier and less risky than diassembling the lens. I am not worried about sanding in terms of parrallelism,etc as I have done optical flats =-, 1/16 wave. If the index is supposed to be slightly off 12:00 thats cool. But I don't want it over at 9:00.
I'll have to figure out just exactly how much I need to take off the adapter and then convert that to degrees of rotation
Fun in the afternoon
Rex

 

[jlw]

The Kingslake book from which I got the lens-extension formula was "Lenses in Photography." I used it to derive a chart in the form of a spreadsheet, which I'll try to attach to this message as a zip file.

The formula is built into the chart, so if you want to plug in other values, you should be able to do it.

The part of the chart that should be of most interest to you is the bottom section; it's a chart I worked out to see how much material I'd need to take off an adapter to correct a front-focus or rear-focus condition on a lens of a particular focal length. It takes into account the fact that when you focus, the coupling cam always moves in and out by the amount appropriate to a 50mm lens (actually 51.6mm) but the optical section itself has to move in or out by a different amount corresponding to the lens' actual focal length.

The chart is set up on assumption of doing a test like the one you did: photographing a meter stick at a 45-degree angle to the camera, allowing you to calculate the difference between the intended best focus at 1 meter and the actual best focus. (Remember that you can't just read the sharpest number off the meter stick, because it's at a 45-degree angle; you have to use the Pythagorean theorem to convert this to the actual "undershoot" or "overshoot" distance.)

For example, looking at the chart line for 21mm lenses, if your intended best focus plane was at 1 meter but your actual best focus was at 988 mm, you need to move the lens 0.01 mm closer to the camera to correct for this. (If I recall your original post correctly, your actual undershoot distance was much larger than this, far more than could be accounted for by a too-thick adapter -- but this will give you an idea of how far off you are.)

As to Kobalux disassembly, I don't think I can help you there -- mine's a third-generation lens and I believe it has a different mount. If you're not totally confident about working on it, I'd suggest sending it to Essex or some other place that works on lenses -- as you know, these Kobalux 21s are not easy to find nowadays, so it might be worth some investment to have a professional work on it for you.

 

[rvaubel]

jlw
Yikes!! That spreadsheet you sent indicates that I would need to remove about .1mm from the face of the adapter. Thats about 10% of the total thickness! I would do this, if necessary, leaving the adapter dedicated to the Kobalux. Although, it does seem to be "off the wall" tolerance wise. More probably the focusing helix skipped a thread in assembly.
Anyway, the test I did was pretty informal in that I handheld the lens, the target was roughly at 45%, etc. In other words I did the whole test while watching TV and not moving from my chair. Still, I did the test a number of times with results +-,20%
What I haven't done is any test at the other end of the distance spectrum i.e. infinity. Plus I'll redo the first test more formally before I attack the lens with a screwdriver or start grinding on an adapter.
I haven't read "lenses in photography" . Is that "History of lenses in photography"? I find Kingslake to be more accessable to the non-proffessional than most. I'll have to Amazon myself a copy.
Thanks again. This kind of help could only be found on RangefinderForum. I don't think I'll be getting a spreadsheet based on some Rudolf Kingslake optical formulas emailed to me soon from the Fred Miranda forum!

Rex