I'm not talking about the cam (roller) in the body, but its ground counterpart on the lens. Are these hand finished; matched to that particular lens' focal lenght? Does Leica actually place each individual lens in a jig to check focal length to 0.1 mm and machine the cam to match? If they do, I'm glad they show so much care. What other "cookie-cutter" company does this hand fitting.

My chrome lenses separate. I guess they test the heads, scratch the true length on the inner barrel, tell the person engraving tthe mounts of the number, which he/she engraves/cleans/paints/dries, then the head gets screwed to its match mount.

I ask this because when I test focus one lens on something, defocus and refocus, I seldom match the original point as marked on the barrel. It happens when I compare different lenses too; they show slightly different distances. Don't flame me; my rangefinders are fine: one M3 CLA w/DAG, M4 w/Krauter, infinity lines up. I've only had blurry results with the 90 TE wide open, up close, low light

-- chris chen (chrischen@msn.com), March 14, 2002

Answers

my fat thumb hit the touch pad.

...low light. Therefore, Leica builds in more precision than most of us will ever need/use/push the envelope.

-- chris chen (chrischen@msn.com), March 14, 2002.

I know that some lenses (older ones) had a choice of serveral different helical thread pitches to compensate for the minor variation in actual focal length. Others have the cam ground differently - but I have no idea if its done for each lens, or if they just make up 3 or 4 standard variations and use the one that best matches the actual focal length.

I can envision a fixture (my machine background comes into play here) that would be a cinch to cut the cams on lenses just by entering a value representing the actual focal length into the CNC controller - so maybe thats how its done on modern lenses.

Could also be that is just doesn't matter as much on some lenses as opposed to others. The 50/1.0, 75/1.4 and 90/2.0 would be better candidates for that sort of setup.

I'm still trying to get the flange to arm-bearing distances for infinity and for some closer distance (1 meter?) to check RF accuracy regardless of the lens used (ie, no lens used, just a depth mic)... Saying that in case anyone has such handy information available!

-- Charles (cbarcellona@telocity.com), March 14, 2002.

Chris, your intuition is pretty much right on the money. For the lenses you mention, the ones with the measured focal length engraved on them, they do, or at least did, have a variety of mounts available with graded helical thread pitches; say, five of them, two for lenses shorter than normal, one for right on the nose, and two for lenses longer than normal. They have the focal length recorded in a file cabinet, by lens serial number. If I lost the mount for my 90mm Elmarit, they could look it up and supply another of to match my individual lens. Or so I'm told. I've never had to put it to the test.

-- Bob Fleischman (RFXMAIL@prodigy.net), March 14, 2002.

The other issue you raise is, I think, a separate one. It is one of the repeatability of focusing. Technique could play a part here, because we may not arrive at quite the same point when approaching the point of focus from two directions; this could be partly mechanical, I think, if there is any backlash in the system. And I would not expect human vision to be entirely consistent (since the rest of me isn't consistent either). I recall reading, I think on Erwin Puts talk about rangefinder accuracy, that the limitations of a given finder magnification are determined not only by the rangefinder's optical specs, but also by the limitations of the eye. According to this, a fresh eye can perhaps achieve accurate focus with a 0.72 finder with, say, a 75 Lux; but a tired eye might require a 0.85 finder to achieve the same accuracy. In a similar way, it might be that one can achieve more consistent focus with a 50mm f/2 using the 0.85, than with a 0.72 or 0.58, and this might be as much due to the human factor as to anything about the lens or camera.

-- Bob Fleischman (RFXMAIL@prodigy.net), March 14, 2002.

1). When it comes to 90s, you really do need to match the lens to the body - I have a 90TE and a 90 'cron. The 'cron focuses dead-on with the M4-P and is iffy with the M4-2 - the TE is dead-on with the M4-2 and is occasionally iffy with the M4-P. The 'error' shows up something like this on the focusing scales: at 3 meters one camera/lens combo will have the 10 ft mark right on the index - with the other it will be just outside the edges of the "10" one way or the other (1.5mm of barrel movement?) The modern 90 Elmarit-M I tried was very iffy on both bodies. Canadian vs. German calibration? =8^o

The TE BTW is a "05" lens - the 'cron is a "00".

2) It's funny - I always imagined that RF 'cams' really were cams - with all the movement coming from a sloped ramp (which I guess is true(?) of the CL 90 f/4). But in fact the rear surfaces of my lenses are nearly plane (except for the 'corrective' machining) and most of the movement comes from a 2nd helical thread.

-- Andy Piper (apidens@denver.infi.net), March 15, 2002.