New JLM jig for focus adjustment

[rwfreund]

Advertisement (gone after registration)

I was wondering if John might be interested in machining a version of his lens adapter that would engage the roller bearing at precisely the correct infinity setting for us in field infinity adjustments.

It seems that this might be more useful than the trial and error approach or the approach of finding an old lens mount for locating the infinity roller position.

anyone with an opinion?

John?

thanks

-bob

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[jlm]

the adapter would be relatively easy to make, but figuring out the roller positon to be nuts on infinity would require...a test jig

[marknorton]

Interesting idea. There's a defined relationship between the position of the rangefinder roller (measured relative to the lens bayonet flange) and the resulting distance at which the rangefinder is coincident. I say, "defined" but it's not published. Yet, if we were able to find out this relationship, it would be possible to set up the camera without reference to particular lenses. The great problem is that if there's a focussing problem, it's tough to know if it's the lens or camera at fault. You just don't know.

 

So what I'd like to see John do is to produce a lens mount with a micrometer depth gauge which presses against the roller. You could set the camera-target distance and then adjust the micrometer until the images were coincident. Do that for 100 cameras and the relationship would be clear. John could then produce a lens mount which you mount on the camera and select 0.7, 1, 2, 5, 10, infinity to check your camera against this independent standard.

 

A few of the gauges circulated among people here with the results collated on this site would be very interesting.

[Shootist]

the adapter would be relatively easy to make, but figuring out the roller positon to be nuts on infinity would require...a test jig

 

Wouldn't the distance from the lens mount on the camera to the roller position for infinity focus be the same for every lens made?

Otherwise the infinity setting would show double images for each focal length lens when looking at a star or planet, or at a very long distance subject.

Taking one of your mounts and attaching a micrometer adjustable ram on it to push the roller in and then looking at a star and adjusting the ram until the images coincide would give you the distance for for proper infinity focus.

 

To be honest I'm not sure why Leica themselves don't have such a device and other fixed mount jigs for 1-5-10 (whatever distance) meters.

[jlm]

one would have to measure the flange plane to roller distance eyeballing through the rangefinder for an object at infinity. (dim 1)

then measure from the flange plane to the ramp on each lens when each lens is at infinity (dim2)

if enough lenses were done, the dim2's could be compared to see if they are standardized.

then you could set your camera's dim1 to match.

 

sounds like a lot of trouble

[marknorton]

No! This would be done without any reference to any lenses. As soon as you introduce lenses into the picture, you introduce an extra element of imprecision. This adjustment is for cameras only.

 

I'm quite sure Leica have an adapter which does just this, for adjusting rangefinders, but they do not make it available.

 

If your goal is lens interchangeability, you have to be able to adjust the camera to an independent standard, without reference to any lenses.

[Shootist]

Advertisement (gone after registration)

Mark I agree with you. The wheel is in the same position for any given focusing distance independent of what focal length lens you have attached to the camera.

The only part of the rangefinder that is coupled to the lens is the frameline selector and to my knowledge from taking apart a M3 it has nothing to do with how or how much the rangefinder moves when the wheel is pushed back into the body of the camera.

So whether you have a 21mm or a 135mm lens attached to the camera and you focus on something at 5 meters the wheel would be pushed in X amount from the lens mount.

 

It is then up to the lens cam, and Leica when setting up that lens cam, to move the wheel for that X distance for any given focusing distance.

 

That is why I find it so hard to think that Leica needs both your body and ALL your lenses to do a proper RF adjustment on the BODY. ALL the BODIES should be set to a standard of "Focus at 5 meters" "The wheel move in X amount from the lens mount".

[jlm]

my was to that to determine the correct distance woulkd require looking at many lenses to find the standard; then use that number to set the body dimension

[wlaidlaw]

John,

 

Wouldn't you just need one perfect lens, the longer the focal length the better, ideally say a 90mm Summicron. If someone had a Summicron 90 that was spot on at all focus distances wide open on a focus chart and the M8 was good on other lenses, that might be OK as a standard. By bringing other lenses into the picture, and averaging, you might end with more of a compromise. However I defer to you - you're the expert.

 

Wilson

[marknorton]

Key to understand is that you do not need any lenses at all to do this. You just measure the position of the roller of your M8 for each distance and publish the result. From that, the "standard" relationship of roller position (relative to the lens mount) to focus distance is established and this should work for film Ms as well as the M8.

 

If John was to make one of these, I had it in mind to circulate among M users for them to use, publish their results and pass on. So, Me, Jaap, Carsten, Wilson, Scott and so on, building up the data so that we really get to understand how to calibrate a body, without any reference to lenses. Then, with a perfectly aligned body, lenses can be evaluated against that known reference.

[rwfreund]

Mark,

I agree that this might be one way to establish an as-built or in-the-wild distribution, but there must be a specification for infinity as well as other distances, otherwise, third party lenses would not rangefinder couple. Not knowing what errors that might be resident amongst the in-the-wild population along with measurement errors and wear, I would think it better that the theoretical value be used an adjustment jig.

Your idea for a measurement jig goes well beyond the idea for an adjustment tool/secondary standard.

Perhaps someone (Dante Stella?, Leica?) can tell us what it is supposed to be relative to the mounting flange?

thanks

-bob

[wlaidlaw]

Mark,

 

If I have got this right you are thinking of something like a metal body cap with a micrometer depth gauge running through it, with the micrometer base fixed to the outside of the cap. Once you have the reference measurement of the micrometer base to flange depth, one could then turn the micrometer adjustment collar and the anvil, which is resting against the RF cam would then push it in. One would look through the RF and use the micrometer collar to adjust so that the images coincided at various distances, noting the MC reading at each distance. By subtracting the micrometer base reading, one would then arrive at the flange to cam measurements for various RF distances.

 

Wilson

[jaapv]

What Mark is saying is basically a piece of reverse engineering, which will work well. We have four M8 bodies in my near family, so I would be happy to contibute to the statistical universe.

[marknorton]

Wilson, yes, that's exactly what I had in mind and I think it's what CV and Zeiss will have done when they wanted to know how to profile the focussing cams on their own lenses. Over a large population of cameras, calibration and experimental errors will likely cancel and we'll be left with the relationship we're looking for.

 

Look at any lens, set to infinity and measure by eye the distance from the bayonet flange to the focussing cam. What is it, about 6.8mm? Set the lens to 0.7m and what is is now, about 2.1mm?

 

I've always been concerned about users adjusting the roller to make the camera work with one lens because that will inevitably change the adjustment for other lenses. The goal, of course is to adjust both cameras and lenses so that any lens can work on any body.

 

With this knowledge, it would be possible for John to produce an adapter which bayonets into the camera with click stops for various distances to allow users to check their camera. That would then answer the question: Is it the lens or the camera?

[wlaidlaw]

There is a depth gauge for sale in the USA on eBay at the moment for $11 - item number 160152342684

 

Wilson

[marknorton]

I've modelled the M8 rangefinder geometry in Excel, so this shows the sort of results a test gauge like this would give us:

 

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!

 

This graph shows, for a given focussing distance in metres (X-Axis) how far the focussing cam should project behind the lens bayonet flange in mm (Y-Axis).

 

This next graph shows the effect of adjusting the roller eccentric, the so called infinity adjustment. Not that the vertical distance between the lines is constant.

 

 

Finally, this graph shows how adjusting the lever arm length changes the calibration. In this case, the vertical distance between the lines increases with focussing distance and confirms that the two adjustments interact.

 

[wlaidlaw]

Should we now ask John to make us our 'calibration gauge'. I am happy to join with others over the cost of this. It might be sensible to get two made initially, one for either side of the pond to keep transit costs down. Once we have them made and have arrived at an agreed table, these gauges could be rented out for a modest Paypal fee, which would be paid back to the original investors proportionately, until the investment was amortized and then would become a forum resource.

 

Sadly, without an optical bench, we can't make the reverse tool to check lenses, because we would need to detect the focus point, as well as measuring the distance from the flange to the focus ramp. We did make our own optical bench in 1966, with a microscope body, an ultra-fine diffraction grating and various scrapped bits of scientific apparatus from the university labs but I was never too sure how accurate it was. It was basically for measuring MTF's not focus accuracy.

 

Wilson

[marknorton]

Yes, it would be interesting to know from John how much these would cost, I'd be happy to pay for one outright.

[Advertisement]

Hallo,

Look here: Leica M

[rwfreund]

Yes, it would be interesting to know from John how much these would cost, I'd be happy to pay for one outright.

 

+1

[rwfreund]

Mark,

Based on th sensitivity of the roller adjustment, do you have a feeling for the required precision for both the gauge and the micrometer for meaningful results?

 

thanks

-bob