Leica M3 CLA Expectations

[SteadyAim37]

Advertisement (gone after registration)

Hello everyone!

First post – I should warn you that this might be a dumb question, but it's one that I can't seem to find an answer for. 

Recently, I had an M3 serviced by Youxin Ye (YYe Camera) via the rush option, and it was returned to me quite quickly. At first inspection, everything seemed good. The camera had been experiencing shutter capping, which was no longer an issue, and the rangefinder was clean and adjusted appropriately. The camera itself, a double-stroke with buddha ears, is in pretty nice condition, with almost all of its original vulcanite.

More recently, I got my hands on a TTL shutter tester (PulseHPT 3-in-1 Shutter Speed Tester, configured with the light sensor). Based on the shutter tester's readings, the 1/1000, 1/500, and 1/250 speeds are at approximately 1/300, 1/200, and 1/140. All other speeds are accurate, within 1/4EV or less. I've also tested the shutter speed tester with a 500CM and electronically controlled cameras like the Pentax LX, and those cameras do not exhibit the same issue of slow top speeds when used with the shutter tester. 

I emailed Mr. Ye, and he stated "The speed for a Leica mechanical camera is an approximation, not exact," and that "Even a new Leica MP, due to Leica adjusted speeds in too tight tolerance, we have to readjust the speeds after a few months the camera left the factory, to prevent capping problems."

 

Is it true that the tolerances for the top speeds of Leica cameras, even a brand new one, are this flexible? I thought that adjustment of the shutter speeds, as part of the CLA, is supposed to bring them into spec, not just get it to some degree of 'functional.' Otherwise, what would be the point?

 

(Photos of said camera attached for eye candy)

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[105012]

If your shutter speed testing is accurate, and your technique in using the tester is good, 1/1000 = 1/300 (1.6ev discrepancies) [etc], is a much larger inaccuracy than I think one should reasonably expect. Using a Reveni Labs shutter speed tester, I see very good timings at 1/1000 on two of my M3's (third has film in it, will test when I get a chance). 

Again, I would start by carefully checking my shutter speed testing procedure/apparatus before coming to any conclusion regarding the shutter tensioning/adjustment...

[Anbaric]

A note in the KS-15(4) service manual (version of the M2 made for the US military) says:

Permissible shutter speed tolerances:

1/1000 through 1/250-second + or - 20%
1/125-second + or - 10%
1/60-second + 25% - 10%
1/30-second + 20% - 10%
1/15 through 1-second + or - 10%

[spydrxx]

Anbaric hit my numbers over the years on the head. Generally I look for things to be within 1 f/stop. Higher shutter speeds, in my experience are more often off than slower speeds. For years it was general knowledge that Leica's SL2 (the old film one built like a tank) which had a stated 1/2000 shutter speed was usually much closer to 1/1000. I've used Youxin Ye for a number of years, and generally his explanation of variances is more detailed. It may have been with your camera that due to wear over its lifetime that he adjusted the speeds as best as he could. Replacement of worn parts isn't always an option. I'd recheck your findings....I usually took 4-5 readings and averaged them as after all we're dealing with aged mechanical machinery.

[SteadyAim37]

Thank you for the responses, everyone! The figures I provided are averaged across the frame (four corners + center) over multiple firings. The procedure I used was to place a 5600K light panel at maximum brightness in front of the camera, shielding it so that light isn't spilling over the body of the camera. I then used a jig to mount the photometer at the film plane. For corner testing, I offset by about 3-4mm from each edge. The readings I got were always pretty consistent, and given the good readings I got when using the tester with cameras utilizing an electronically controlled shutter, I'm not inclined to think there's a major issue with the tester. 

I do think it's reasonable to expect that the camera will have variance beyond the 'when new' parameters kindly provided by @Anbaric, but the confirmation of the service being completed included a statement that the shutter speeds were adjusted. That statement gave me the expectation that they would have been brought within spec, and that if there was an issue in doing so, the statement instead would have said something like "The camera shows excessive wear of internal parts, replacements required to bring within correct timing."

In any case, this helps clear things up. I'll probably think about it a bit more, but I think a partial refund would be fair, so I'll ask for that.

[spydrxx]

Or ask him to readjust the timings so they're within acceptable tolerance. Enjoy using your "new" camera by the way!

[PCPix]

Advertisement (gone after registration)

As a suggestion, try a different type of testing altogether… Put a roll of Fujichrome Velvia or Provia through the camera and do a set up on a tripod photographing something in steady light using all the different shutter speeds… Transparency film will reveal exposure errors of one third of a stop or more. If your images look to be acceptable, don’t worry about it. However, if the higher shutter speeds are showing noticeable over exposure then your shutter is running slow at those speeds and should go back for further  adjustment. 

[Fotoklaus]

If you measuring method is accurate I would not accept that.

A mechanical cloth shutter will never be as accurate as a electronic controlled shutter, but 1/300 instead of 1/1000 is way off.

Well adjusted the M-Shutter is good enough for slide film which will not accept more than 1/2 stop off.

 

 

[250swb]

On 6/20/2024 at 8:35 AM, SteadyAim37 said:

 

Is it true that the tolerances for the top speeds of Leica cameras, even a brand new one, are this flexible? I thought that adjustment of the shutter speeds, as part of the CLA, is supposed to bring them into spec, not just get it to some degree of 'functional.' Otherwise, what would be the point?

Speeds are never going to be exact, but closer than the bullshit Mr Ye has told you. I’ve had plenty of new Leica cameras and they’ve never needed to be readjusted after six months, and if you are dealing with a well used Leica the components are already bedded in.

Edited June 22, 2024 by 250swb

[250swb]

On 6/20/2024 at 12:31 PM, Anbaric said:

A note in the KS-15(4) service manual (version of the M2 made for the US military) says:

Permissible shutter speed tolerances:

1/1000 through 1/250-second + or - 20%
1/125-second + or - 10%
1/60-second + 25% - 10%
1/30-second + 20% - 10%
1/15 through 1-second + or - 10%

But it’s a US government service manual, tolerances may be different from Leica tolerances. The manual is meant to keep cameras working, not make exposures Ansel Adam’s would be proud of.

[Anbaric]

2 hours ago, 250swb said:

But it’s a US government service manual, tolerances may be different from Leica tolerances. The manual is meant to keep cameras working, not make exposures Ansel Adam’s would be proud of.

'Close enough for government work' ? 🙂

But I strongly suspect this manual was prepared mostly or entirely by Leica themselves. It was apparently the textbook for a 120 hour course given over 3 weeks at Leitz NY, intended for technicians who were already 'experienced camera repairmen'. I don't think they were teaching GIs how to bodge it in the field. My guess would be that this is similar to Leica's own internal training material, but we happen to have access to this version because it was supplied as part of a contract with the US Government. The description of how the shutter speeds were tested appears to reference Leica's own instruments developed for this purpose.

[Michael Geschlecht]

Hello Everybody,

Leitz's advertising for the later M3's in the middle 1960's said that the shutter speeds of M3's were within 10% of the marked speeds.

These later M3's also had accurate variable speeds from 1 second to 1/1000 EXCEPT between 1/8 & 1/15 and between 1/30 & 1/60. 

The lightening bolt between 1/30 & 1/60 is an accurate 1/50.

Best Regards,

Michael

[Fotoklaus]

vor 21 Stunden schrieb 250swb:

But it’s a US government service manual, tolerances may be different from Leica tolerances. The manual is meant to keep cameras working, not make exposures Ansel Adam’s would be proud of.

But 1/300 to 1/1000 is about 200% off. This is just completely unadjusted.

Even the US government would not accept that...

[250swb]

2 hours ago, Fotoklaus said:

But 1/300 to 1/1000 is about 200% off. This is just completely unadjusted.

Even the US government would not accept that...

Is it different from the Leica M2 manual, I don’t know, are the technicians it was meant for better or worse, I don’t know. I do have the KS-15 manual.

[Mikep996]

"As a suggestion, try a different type of testing altogether… Put a roll of Fujichrome Velvia or Provia through the camera and do a set up on a tripod photographing something in steady light using all the different shutter speeds…"

 

Yep.  Worry about if the pics show there is a problem.  Otherwise it's not worth any concern at all.

[davidmknoble]

The mechanical nature of the high shutters speeds is such that setting the 1/1000 and 1/500 to the exact speeds means that they will degrade much faster to a slower speed.  Setting them to about 10% to 20% less will help them stay within that tolerance longer.  It has to do with the tension and the springs.  I set mine to a 10% less.

Given that this post is about an M3, that is likely 60 years old or more, there may be issues with the shutter system that they cannot hold the setting well or that the measurement equipment is faulty at the higher speeds (noting that the slower speeds seemed to be closer).  I agree with @Mikep996, shoot some film in the higher shutter speeds and lower shutter speeds and see how well they come out.

I would try a series of the same image on a tripod that goes from 1/30 second on up to 1/1000 second by changing the f/stop.  Then you could see what shutter speed, if any, start to show a real change in brightness.  If none, then everything is within tolerance and the measuring device is off.  If not, you’ll still learn where the problem is. 

Just some thoughts.