On the distance film to lens mount and how to measure it

[sanmich]

Hi All
I recently started to mess around with basic cleaning of the few barncaks I have and so far things are not very complicated and, well, I'm enjoying the process.
The first time I cleaned a '33 Black leica III and frightened everyone home by shouting "IT'S ALIVE" will always be a good day to remember J
The only thing I understand is to be seriously checked before jumping with these marvels from nice kids pictures to serious shooting is the …Drum rolls….

….Distance from the film to the lens mounting flange….

Well I have access to fairly accurate measuring equipment and I found the Leica spec for this distance BUT:

1- The tolerance for this measure, according to Leica is 28.80 +0.015 -0.005. This requires a quite crazy level of precision. Even the micrometers around here or a digital caliper will give a measure precision of the order of 0.01mm.

2- The film is maintained in place by a plate which is mounted on springs. How the heck can I measure something elastic?? Even quite low loads from the caliper/measure instrument will bring a deflection of the plate.

Has someone here performed such a verification?
How?

Thanks and happy new year to all.
Michael

 

[Bryce]

If you've got a granite table, height gage, test indicator, a 1-2-3 block and gage blocks you can test it the following way-
Support the body on a 1-2-3 block, lens mount up, with the shutter held open and its back off/ open by its film rails. Set the height gage up so the test indicator reads "0" when its probe is resting on the lens register surface.
Now move the height gage away from the camera and replace the camera with 28.8mm worth of gage blocks, and slide the height gage over the blocks. Deviation from "0" indicates error in film to flange distance.
A micrometer should also work if you can get it to reach around the camera. Micrometers are readable to about .002 mm.

 

[sanmich]

Hi Bryce and thanks for your answer.
I'm talking about barnacks, no back door, no way to reach the film rails...

 

[jlw]

A depth micrometer isn't all that expensive (if you get a cheap-but-serviceable Chinese one; I think mine cost about US$ 35) and will be a good investment if you want to work on these cameras on an ongoing basis. If you're willing to spend more you can get one with a digital readout, which reduces the chance of reading errors; still, since you're just trying to see whether the cameras are close to 28.8 or not, the conventional type should work.

This type of micrometer has a flat base with the measuring rod extending downward from the base. You can sit it on the front of the lens flange and measure the depth directly. The measuring rod is extended by a knob that has a spring-loaded clutch on it, so you can apply a consistent amount of pressure when taking repeated measurements.

I'm not sure where you got your measurement and tolerances; I've heard the 28.8mm figure before, but also read somewhere (Dante Stella's site?) that it isn't actually documented in any Leica literature! I do know that different manufacturers used different values for film thickness and film-gate thickness, so you're probably on your own to some extent here.

What I guess I'd do in your situation would be to load the camera with a scrap roll of the film I intend to use most often, and use the depth micrometer to measure from the flange to the actual film surface. You'd want to measure at several points to make sure the lens flange is parallel to the film surface.

As to the issue of depressing the spring pressure plate -- I can see how that might be a problem if the spring clutch on your micrometer lets you exert too much pressure. Do you have a camera with a known-good flange-to-film distance that you could measure and use as a reference?


The other way of tackling this problem is to disregard mechanical measurements entirely and use an autocollimator, an optical instrument which lets you examine the plane of focus directly. It consists of a backlighted focusing target and an eyepiece which lets you view the target (via a beamsplitter.) The target/eyepiece/beamsplitter assembly is positioned at the back end of a small telescope, which positions the assembly at "optical infinity" as seen from the front of the telescope.

You use it by mounting a lens on the camera and inserting a small first-surface mirror into the film gate, then pointing the camera/lens into the front of the telescope so the lens "sees" the target positioned at optical infinity. The lens focuses the image of the target on the first-surface mirror; the mirror reflects the target back through beamsplitter so the reflection can be seen through the eyepiece. If the target looks uniformly sharp when the lens is focused at the infinity position, then you know your camera's back focus is correct. If the target isn't sharp, you can adjust the lens to get an idea of which direction you need to move the lens flange and by how much.

Commercial autocollimators are expensive because of the cost of the telescope, but I've heard of people making their own out of a lens and eyepiece from a pair of binoculars. There's an explanation in one of Thomas Tomosy's camera-repair books; let me know if you're interested and I can look up which one.

Good luck...

 

[payasam]

The Leitz method, I think, involved the use of a front-surface mirror at the film plane, and that is probably what JLW describes. JLW, I do not see the relevance of film thickness, since the emulsion surface will be at the film plane, regardless of the thickness of the film base.

 

[sanmich]

Thanks a lot for the extensive answer.
I think I will try the micrometer way as it seems more simple than the autocollimator. About the leica requirements, attached is a sheet fromm the IIIf I found posted here with the tolerances. I don't understand the need to put a film in the camera as it should give the same reading as the pressure plate. The film is pushing back the pressure plate isn't it?
I am thinking of pushing a thick plate of plastic instead of the film in order to increase the pressure on the springs and lower the chances of depressing them during measurement. What do you think of the idea?
Michael

 

[Bryce]

Michael-
O.K., I admit it- I've never owned a screw mount Leica and had forgotten the back is permanently attached.
Anyhow, the depth micrometer should work and I can't see why film is necessary. As for the pressure plate springing away from the mic., a shim could work or just very gentle pressure at the thimble.

 

[John Shriver]

If I read Canon's service instructions for their bottom-loader LTM cameras correctly, they slid a glass plate into the film gate, and then put a 28.80mm thick block of metal on top of it. Then you used a dial indicator (properly supported) to compare the height of the block to the height of the lens mount.

 

[payasam]

Michael, you want to be careful: those springs shouldn't lose their tension.

 

[varjag]

I should just mention that difference in films curvature (and pressure plate doesn't make film 100% flat) makes these alignments less critical than it is often believed.

 

[Xmas]

Plastic in place of film would need to be 35mm in width to stay between the outer railway tracks in (i.e. inside) which the 35mm film runs.

The pressure plate rides on the outer rails the film on the inner the inner rails are lower than the outer i.e. the film is neared the lens mount then the pressure plate by the difference in the rail 'heights'.

The tolerance is probably asymetric because the depth of focus error effect is asymetric.

The film will bow away from the pressure plate - in most circumstances, which is also an asymetric effect.

If you are using a Noctilux f1 (on an M) leitz recommended that you have the body registered, if you only have a f3.5 this registration is less critical.

You also have to register the rngfdr at infinity, 10m and 1m...

Hearsay has it that the FSU lenses are made to a different focal length standard...

I accept this is everything short of real help sorry...

Noel

p.s. sORRY eUGENE WE DOUBLE POSTED...

 

[ZorkiKat]

The Leitz method, I think, involved the use of a front-surface mirror at the film plane, and that is probably what JLW describes. JLW, I do not see the relevance of film thickness, since the emulsion surface will be at the film plane, regardless of the thickness of the film base.

Leitz used a depth measuring device, and such had been seen once on ebay. This was a modified depth gauge. A dial is found on top, mounted on a very precisely milled crossbar which sits on the lens mount. The depth sensor drops into the body cavity and strikes the pressure plate. This sensor is powered by a weak spring (or at least weaker than the springs which support the pressure plate). The values are read off the dial, which can show the actual distance down to 0.02 or 0.015 values required by Leitz.

For a camera with a non-removable back and whose focus accuracy depends on set measurements (like lens working distance/register, lens helical movement)- as opposed to real images seen on a focussing screen, -a depth gauge may be a more sensible method focus calibration.

 

[ZorkiKat]

My measuring method

My measuring method

This is the method I use for measuring lens flange to film plane distance:

http://jay.fedka.com/index_files/Page433.htm

I use a Mitutoyo vernier caliper. This caliper is accurate to 0.02mm, which is the tolerance standard for Soviet cameras and lenses. The 0.005 difference (as against Leitz's 0.015) shouldn't matter much. I've been able to calibrate my cameras using this method- and so far the results have been good- good enough to at least have Jupiter-3 (50/1.5) lenses focus correctly.

 

[edodo]

The body shell has to be removed to clean the shutter and have access to the back plate. Then a ground glass at the film plane should be used for the lens flange.

 

[ZorkiKat]

Note that only IIIc and later bottom loading Leica can be measured lens flange to film plane with the body shell off.


With earlier Leica, bottom-loading Zorki, FED, and Canon, the lens flange to film plane distance cannot be measured with the shutter crate off the body shell. The lens mount is attached to the body. The only way by which the value can be determined is with the camera properly assembled.

It is also impossible to use a groundglass to view the focused image on the film plane if there isn't anything on the front of the crate to put the lens in.
To have the lens mount over the film plane, the shutter crate must be put back in the body shell. With the shell in place, the film plane can no longer be seen.

Substituting glass in an assembled body may not work with less precise camera bodies - the value obtained with a substitute pressure plate in place may not be the same as when the real pressure plate is replaced. Disassembling the body-again- to remove glass and put the actual plate back in would mean another disturbance in the calibrations and cause variations in actual flange to film plane distances again.

 

[payasam]

Here's a shot from the factory manual for the IIIf.

 

[Limpovitj]

Plenty of good tips already! The only time I had to check focus (this was after flattening the front of a dented body shell of a Leica II that had been dropped, lens down - don't ask!) , this is what I did:

First people often assume that you can't back-sight a barnack Leica because of the closed back, but you can. At least, I could. ;-) I found a tiny hole at the bottom of the shutter crate, and with the camera upside down (and bottom plate removed of course) could position a small pen light shining down through the hole onto the film plane. With a discarded film that I drew some lines on, rewound and then loaded as usual, and wound until the lines showed up in the film gate (opened at 'B'), I could back-sight it. No problem. There was plenty of light to see the lines.

Alternatively, to overcome the problem with the closed back, you could - in theory, I haven't tried it out - place a half-mirror (like the ones used in a rangefinder) at an angle between the Leica and the SLR used to check focus, and thus, with an extra lamp, provide illumination at the film plane on the same axis you are viewing through the lens to check focus. I can't see why it shouldn't work.

Now, if this wasn't understandable (I should have been asleep several hours ago), or if you're not familiar with back-sighting, let me know!

I like the idea of doing this with film, which in itself will - so to speak - compensate naturally for any curvature, and thus - again, in theory - be more of a real world test than using something perfecly flat at the film plane.