Film SLR lens mount paralellism to film plane vs image quality (Nikon)

[Sierra Club]

[Edit: I am aware of this short thread - https://www.rangefinderforum.com/forums/showthread.php?t=163129]

Is there any way to calculate how inaccuracy in lens mount warping may impact image quality (in terms of resolution), especially for Nikon film SLR cameras?

I understand that lens mount is precisely shimmed and screwed in to fairly strict tolerance at the factory to ensure that it's as much parallel to the film surface/plane as possible.
For Nikon F cameras I believe the FFD is exactly 46.50 mm.

To rephrase the question - after DIY service dissassembly and reasssembly without dynamometric tools or depth measuring digital calipers, how flat and in-plane would camera-side lens mount need to be (in mm) to ensure that any variance would be imperceptible?

I understand that aside from lens mount warping, lens mount distance from the film plane also plays a role here. Anyone knows what are the tolerances there? Would lens aperture setting help to minimize any potential lack of planarity of the lens mount surface and shifts in the flange focal distance?

 

[phofseth]

Depends on the lens; long focus requires big displacement to move the focus-plane image while wideangles are quite sensitive. Also if you depend on close range correcting trickery, small fractions of a mm can be essential. Also, larger light openings require higher precision.



p.

 

[former member]

46.5mm +/- 0.02mm for the Nikon F2, as per the repair manual.

For digital cameras, probably less because of the microlens array and pixel-peepers.

For a limited time- I'll leave the download link in place for the manual:

https://www.dropbox.com/s/358tbjkjef28n08/NIKON_F2_SERVICE.pdf?dl=0

For DIY: Using my Digital Leica cameras to set shims I have to get it within 0.01mm for a 1930s 5cm F1.5 Sonnar lens before the difference cannot be seen using the back-screen. I cannot focus the lens better than that on the M9 using a 1.25x finder.

 

[Sierra Club]

46.5mm +/- 0.02mm for the Nikon F2, as per the repair manual.

For digital cameras, probably less because of the microlens array and pixel-peepers.

For a limited time- I'll leave the download link in place for the manual:

https://www.dropbox.com/s/358tbjkjef28n08/NIKON_F2_SERVICE.pdf?dl=0

Perfect, that's exactly what I needed!
I assume this applies to all Nikon film SLRs across the board.
Thank you very much.

 

[former member]

You are welcome.

I would feel comfortable using this for all Nikon film cameras- interesting that the SP and F repair manuals just state "Proper Thickness". The F2 manual puts a number on it, and it is the same that Leica used for their film cameras. I remember Leica using the spec in an advertisement in the early 1970s "less than a strand of human hair" or such.

 

[zuiko85]

And then there could be other out of spec tolerances that affect image resolution. What about the angle of the mirror in a SLR? Or the position of the focusing screen?
If you focus carefully on the focusing screen, but that focus is slightly different than focus at the film plane...
I mention this because it happened to me with a brand new 35mm SLR (not a Nikon though). Admittedly an inexpensive camera it was a Fujica ST701 that was new to the market. I was getting negatives that were not quite sharp enough. I began to suspect screen focus did not agree with film plane and a check with a repair person confirmed it.
Long story short, the mirror down position was adjustable and was off. After adjusting it the problem was solved to any tolerance my (then 22 year old) eyes could detect.
Have also seen even name brand lenses after being well used get enough slop in the focusing helical that the screen image shifts when you change focusing direction.
Just thinking that the single metric of flange to image plane is part of a whole chain of mechanisms that have to work right to obtain the best resolution the camera can deliver.

 

[Sierra Club]

This is an interesting subject as - one the one hand - it has critical impact on optimal resolution of any optical system, but in reality, with all of the related and interacting mechanical parts, it's so hard to maintain this precision throughout the camera's life...

One interesting quote I found on the net (Kiev related site):

(...) I asked a noted expert, Henry Scherer, what he thought.

The correct answer is that there isn't any correct answer. The value cited in Peter Tooke's book which is the Zeiss factory specification is an ideal which, in practice, was almost never achieved.

The reason for this is that back in those good old days there was no such thing as any modern method of quality control practiced by anyone in Germany or Russia. In actual practice the actual dimensions of the lens mount as well as the camera body would vary by such a large degree it is my experience that even if one has a camera where the lens flange to film plane measurement is within 35.00mm +0.01/-.03 it is highly unlikely that the rear lens surface will be within the optimal distance from the film plane when it is mounted on the camera.

Zeiss lenses have considerable range of adjustment for the lens cartridge within the lens barrel. And there were a great number of lenses, perhaps most of them, set incorrectly at the factory.

The mounting of the lenses in the rangefinder is also something with a lot of variation. So the rangefinder result has a fair amount of variation in it.

So my overall answer is that it doesn't matter much. There's no point in trying to adjust the position of a lens mount on a camera body when it's likely that the inner machining that sets the position of the lens has a lot of variation.

I think it would have been better for Zeiss to have specified the position of the flanges that are inside the lens mount and to which the lens tabs couple than it was for them to specify the distance from the external tabs. But even if they had done this the variation of everything around the lens mount and which affects the focus is so great this is also rendered relatively meaningless.

 

[former member]

The Nikon F uses shims under the plate that the focus screen sits on, proper thickness is critical. Position and thickness of the mirror is critical. SLR's only work of the distance of the image formed on the focus screen is the same as that falling on film, or the sensor. Probably another reason why manufacturers are moving to mirrorless designs. The trade-off: power required and latency to take the picture. Global shutters in sensors: much more electronic noise in the image. This is known as "Conservation of Inconvenience".

Once an SLR is calibrated, it tends to take major mechanical failure to knock out of calibration. A rangefinder is more easily knocked out of calibration.

 

[zuiko85]

“Conservation of Inconvenience”
LOL, I like it. Although I have a Oly mirrorless I’ve often wondered just how much mechanical monkey business has to go on every time the shutter is pressed. Doesn’t the open shutter have to close, then open and close, and then open again so I can see?
Almost as bad a leaf shutter SLR which were very complicated little monsters in the film era.

 

[former member]

Although I have a Oly mirrorless I’ve often wondered just how much mechanical monkey business has to go on every time the shutter is pressed. Doesn’t the open shutter have to close, then open and close, and then open again so I can see?
Almost as bad a leaf shutter SLR which were very complicated little monsters in the film era.

Most mirrorless cameras use mechanical shutters, and it is just like the old Kodak Retina Reflex cameras, shutters have to close from live-view mode, open/close to take the picture, then open again for liveview.

The "Global Shutter" on a sensor is all electronic. The penalty is noise level in the image, some of the u43 sensors that had a global shutter were 10x as much noise when used in that mode.

 

[Oren Grad]

This is a major issue with high-resolution digital cameras - not just FFD mismatch but also swing/tilt in the mounts.

See, for example:

https://www.zsystemuser.com/nikon-z-system-news-and/my-tolerance-for-tolerances.html

https://diglloyd.com/blog/2020/20201018_1332-SonyA7R_IV-sensor-parallelism-evaluated.html

For most users and most uses, film will be less demanding of such ultrafine tolerances. Have you seen evidence of a problem with your equipment?

 

[zuiko85]

You know, I think there was a article by Roger Cicala on the LensRentals blog back in June about flange to sensor variations in photo cameras.
As Artie Johnson used to say on Laugh In ‘very interesting’.

 

[Oren Grad]

You know, I think there was a article by Roger Cicala on the LensRentals blog back in June about flange to sensor variations in photo cameras.
As Artie Johnson used to say on Laugh In ‘very interesting’.

https://www.lensrentals.com/blog/2020/06/the-great-flange-to-sensor-distance-article-part-ii-photo-cameras/

 

[Sierra Club]

I was always worried when I saw (film) cameras for sale which had visible damage to screws holding the lens mount ring.
Saw that on Nikons and Leicas.
I guess as long as lenses can be mounted and don't fall off, it's fine 😀

Film is such a forgiving medium.

 

[Sarcophilus Harrisii]

Post Deleted