Rangefinderproblems on new M9s

[jaapv]

Advertisement (gone after registration)

All the messages concerning bad focus have scared me from the M9.

 

Regarding accuracy - I've never in over forty years had an M body that failed to focus properly. That includes seven different M bodies - one M2, four M4 and two M7s) and several lenses, but mostly the early 35mm Summilux and first batch of 75mm Summilux.

 

So, why do you claim that film does not require as much focus accuracy as digital?

Because film has a thickness that covers many sins. And you didn't have the instant feedback, nor the possiblity to enlarge your photograph to 2x3 meters at the click of the mouse.

[Advertisement]

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

[lct]

...are you saying that once rangefinder is adjusted you still need to adjust it once or twice a year? how can that be?

There are alignment screws which rorate more or less depending on bumps, vibrations and the like. I had an interesting experience with my second Epson body in 2006. I live in France and purchased the body in Italy. The dealer is serious and knows well rangefinders so i asked him to check the RF alignments before shipping the body. He did it and the body was spot when it left Italy but when arriving in France it had a severe vertical misalignment. I thought this would never happen to a Leica then.

[jaapv]

The problem with the vertical alignment is that there is a little fork behind the red dot that controls it. It used to have an excentric screw, but on newer cameras this screw has been left out and the fixation relies on something much like nail polish. A bump or vibration may shift this fork. If this happens the technician will soften the fixation with Acetone and use an excentric tool to readjust, after which the glue will stiffen again. Actually this happened to me yesterday when I dropped the camera, but luckily I was 20 minutes from Will van Manen who did exactly what I described above, so I was on the road again with a perfect camera within the hour

[pico]

The question was why is focus more difficult in digital than with film.

 

Because film has a thickness that covers many sins. And you didn't have the instant feedback, nor the possiblity to enlarge your photograph to 2x3 meters at the click of the mouse.

 

The later point is appreciated but in that case the critical factor is the CoC required for a huge enlargement. I won't go there.

 

However, I do not understand why one would say that film covers a multiple sins (errors) because it is thick. Correct me if I am wrong, please, but a digital sensor well is thicker/deeper than film emulsion.

 

Typically, a film's single emulsion is ~0.00006 inches thick. Multiply X3 for color and modern B&W films to make ~0.00018 inches. A nominal digital sensor well is 0.00035 inches. IOW, the digital well is almost six times thicker than a films emulsion(s). Finally, the CoC for full-frame 35 is about six times larger than film is thick.

 

So it appears that due to film's _thinness_ that it's easier to focus.

[jaapv]

I fear that is a misconception. For focussing purposes a sensor behaves like a mathematical plane. And indeed, let's not get into COCs, those were defined for prints of 6x9 cm in the 1920ies and are of limited practical use nowadays.

[pico]

I fear that is a misconception. For focussing purposes a sensor behaves like a mathematical plane. And indeed, let's not get into COCs, those were defined for prints of 6x9 cm in the 1920ies and are of limited practical use nowadays.

 

Could you show the math and why it makes a difference between film and sensor?

 

CoC was of concern by Leica from the start. They have their recommendations based upon a given enlargement factor and viewing distance. It still matters to anyone who prints. Of course, if one presents his M9 images only on a monitor, then the camera is a waste of technology and money.

 

I was ready to buy an M9, but after all the posts concerning lens and rangefinder calibration, I dare not risk it. Of the six Ms I have, none fail to focus with my lenses. It would be tragic to adopt a camera that is fundamentally incompatible. A shame, really.

[jaapv]

Advertisement (gone after registration)

Sorry to say it, but those are all the wrong reasons not to get an M9.

 

As for film vs sensor and RF/lens adjustment, search the forums, this has been discussed ad nauseam

My post reflects the official position of Leica too.

 

The same goes for DOF and COC. It is fundamentally different on a sensor camera -in print as well. Search the forum.

 

 

It means older film lenses MAY have to be adjusted for use on digital as the tolerances for film were much wider. There are plenty of third party workshops that can do that for you - fast and not expensive. Nothing incompatible at all. If a lens is good on digital, it is fine on film. If you want to have some of your lenses coded, they can help you out too.

[pico]

Sorry to say it, but those are all the wrong reasons not to get an M9.

 

As for film vs sensor and RF/lens adjustment, search the forums, this has been discussed ad nauseam

My post reflects the official position of Leica too.

 

The same goes for DOF and COC. It is fundamentally different on a sensor camera -in print as well. Search the forum.

 

[...]

 

Show me the math or links. I think you are full of it.

[gwelland]

Show me the math or links. I think you are full of it.

 

Hmm, friendly ...

 

Try this: RF Accuracy Discussion

[bybrett]

Show me the math or links. I think you are full of it.

 

Film is full of it... layers. It's a sandwich with depth. A sensor is a plate with a surface.

[pico]

I have read the referred materials and still do not see any evidence that digital sensors have different depth-of-field or depth-of-focus metrics. Jaapv's description using triangles is unclear, and his explanation that it's 'math' is unfulfilled.

[jaapv]

Well, look further. You are querying well-established facts, you can't expect us to do your research for you.

[zlatkob]

I was ready to buy an M9, but after all the posts concerning lens and rangefinder calibration, I dare not risk it. Of the six Ms I have, none fail to focus with my lenses. It would be tragic to adopt a camera that is fundamentally incompatible. A shame, really.

 

That's a bit extreme. Photographers are making fine in-focus photos with the M9 every day, disproving this idea of "fundamentally incompatible".

[pico]

Well, look further. You are querying well-established facts, you can't expect us to do your research for you.

 

With respect, jaapv, I have searched far and wide and found nothing to substantiate the claim. If you found a novel contribution to the technical literature, can you please cite the sources?

 

I'm still looking.

[rosuna]

Typically, a film's single emulsion is ~0.00006 inches thick. Multiply X3 for color and modern B&W films to make ~0.00018 inches. A nominal digital sensor well is 0.00035 inches.

 

The well is just a hole. The photosensitive surface is just that, a flat surface of silicon under the wells. It is not a perfectly flat surface (it is a bit curved), but it is just a surface, a plane without thickness. Filters bring problems to this design (difraction, refraction, etc.), but filters are not light collectors....

 

Film has translucent layers (gelatine) and photosensitive layers, so that thickness "is" photosensitive. Flatness is another matter...

 

Sensors and film are very different. You cannot compare total thickness of a film and a sensor.

[SJP]

I would be surprised if there are technical sources (as in scientific articles). What is clear is that focus shift, rangefinder adjustment, matching of lenses to bodies etc. seems to be more prevalent with M8 and M9 than with film M's (based on the number of postings on such topics here and at RFF).

 

Some possible reasons:

1 immediate feedback from the LCD, it is easier to check the results and compare with intended focal point/plane. On film it would be hard to be sure that you didn't make a focussing mistake (after a few days/weeks for development & printing).

 

2 pixel peeping it is a lot easier to blow up images on screen to establish the exact location of focus than in the past (again with a delay), this has also led to the common belief that the old standard CoC is way too lenient in the digital age and that you need use 1-2 stops more on the lens barrel (ie. use f/1.4 marks at f/2.8 lens aperture for M8, f/2 marks for M9). This could also imply that we demand better focus accuracy now on digital than we did in the past on film. No new physics here only changed expectations. For normal size prints on normal viewing distance nothing much has changed on the M9 (the M8 needs approx. 1 stop due to crop factor) but we expect - and get - more.

 

3 film appears to me to be less "sharp" than the current digital sensors, this is my impression using my own comparison, not based on detailed analysis. Film is a multiple scattering medium so there must be some lateral spread of info, this would give a built in slight softness and would lead to slightly more lenient focus accuracy (via convolution of the image and the film "softness"). Probably depends on emulsion grain, film brand, ISO etc.

 

I hope this is of some use - all three factor make sense to me from a physicists perspective and point in the direction that Jaap suggested admittedly for different reasons. A minor difference of opinion is fine, and I'm sure Jaap will not jump down my shirt for this. I must admit I am not very impressed with the argumentative tone that 'mr. pico' seems to find neccesary.

[wlaidlaw]

For an M8/9 centre, would not the point of focus need to be the optical centre of the micro-lens not the actual surface of the sensor or the bottom of the Bayer wells? In diagrams I have seen, parallel rays strike the micro-lens and are then concentrated at the bottom of the well. I would guess that the optical centres are all pretty much exactly the same distance within a micron or so, of the sensor plane. Sensor surfaces will be much flatter than film. You only have to see how a slide behaves in a projector to see how bendy film can be. Zeiss used to be paranoid about film flatness, to the extent that their RTS cameras had a vacuum pump to suck the film onto a porous ceramic plate. Zeiss claimed that the primary reason for motor drive in all their cameras was film tension rather than convenience/fast shooting.

 

Wilson

[pico]

Thanks to all of you for your helpful comments. I simply needed a bit of gob-smacking to get thinking in a new way.

[Advertisement]

Hallo,

Look here: Leica M

[k-hawinkler]

For an M8/9 centre, would not the point of focus need to be the optical centre of the micro-lens not the actual surface of the sensor or the bottom of the Bayer wells? In diagrams I have seen, parallel rays strike the micro-lens and are then concentrated at the bottom of the well. I would guess that the optical centres are all pretty much exactly the same distance within a micron or so, of the sensor plane. Sensor surfaces will be much flatter than film. You only have to see how a slide behaves in a projector to see how bendy film can be. Zeiss used to be paranoid about film flatness, to the extent that their RTS cameras had a vacuum pump to suck the film onto a porous ceramic plate. Zeiss claimed that the primary reason for motor drive in all their cameras was film tension rather than convenience/fast shooting.

 

Wilson

 

Wilson,

 

I am convinced that a similar point can be made for flat panel displays. The image is created at or near the surface - but very uniformly across the screen. This should also provide a very plane surface for focus charts. I believe the first objection put forth by lars_bergquist in http://www.l-camera-forum.com/leica-forum/leica-m9-forum/111918-m9-lens-focus-setting.html#post1184428 post #2 is unfounded as I showed in post #7 and #38 of http://www.l-camera-forum.com/leica-forum/leica-m9-forum/120468-summilux-50mm-f1-4-asph-chrome.html#post1280242. Indeed, the Moiré pattern in an M9 image that indicates where a displayed focus chart is actually in focus on a flat panel screen is certainly of great help to me in deciding rangefinder focus accuracy.

 

I also believe it is really unfortunate that the M9 doesn't have LiveView or Leica doesn't have for sale another convenient tool so that folks with imperfect eyesight could easily determine whether a camera/lens combination operates within rangefinder accuracy specs.

 

The rangefinder aspect seems to offer great benefits but also causes tremendous frustration when not working properly.

 

K-H.

[jaapv]

Thanks to all of you for your helpful comments. I simply needed a bit of gob-smacking to get thinking in a new way.

Happy to oblige Don't think we do not understand your position. Most of us have a non-digital background in photographic matters, some, like me, for more decades than we care to remember, and we all had to unlearn a large number of seemingly unshakable truths in the process.