fiftyonepointsix Posted August 31, 2015 Share #21 Posted August 31, 2015 Advertisement (gone after registration) If your three cameras don't agree with each other why are you looking for some advanced technological explanation? In order to understand the assumptions made for best focus with each camera. I have 75 or so lenses in Leica mount. If the focus shift due to spherical aberration of a lens exceeds the calibration tolerances of the Rangefinder then you need to know which F-Stop the lens is optimized for. Or you need to set the lens for which F-Stop you want it to be optimized for. If the focus shift due to chromatic aberration of a lens exceeds the calibration tolerances of the Rangefinder then you need to know which portion of the spectrum the agreement is best. Or you need to set the lens for which portion of the spectrum you want it to be optimized for. With most lenses- this can be done by changing a shim, or building up the RF cam, or changing the distance between lens groups. I do all three. Lenses made to the Zeiss standard: you have the additional issue of the focal length being a nominal 52.4mm. I believe the later Amedeo adapters use an indexed cam to correct the error. I have one late West German Zeiss 50/1.5 Sonnar "T" that was given to me as it never focused correctly. Shim and element spacing were off, much better now: set for the S-Mount. Manassas Battlefield by fiftyonepointsix, on Flickr Nikkor 8.5cm F2, Yellow filter, wide-open on the M Monochrom. Focus is accurate enough for me. Link to post Share on other sites More sharing options...
Advertisement Posted August 31, 2015 Posted August 31, 2015 Hi fiftyonepointsix, Take a look here Does the Monochrom M246 have a different register from the M240?. I'm sure you'll find what you were looking for!
tobey bilek Posted August 31, 2015 Share #22 Posted August 31, 2015 As we are dealing with mechanical parts with a accuracy tolerance rather than a fixed exact point and optics where the acceptably in focus range at different distances and apertures varies, there is a whole range of variables which on some occasions cancel each other out and give you the happy with situation with your 50/1.5 + 246 and a different result with your M240. Add in complex lenses with floating elements and you will add another variable that may help or confound, depending on circumstances. If you have 4 well calibrated M bodies (which I have) ...... and a dozen or so lenses which are also allegedly calibrated to Leicas tolerances you will still find some lenses with some bodies will have issues wide open and at various distances ...... not enough to be a problem with everyday photography ..... but certainly obvious if you test carefully. Even with Leicas rigorous attention to detail the unforgiving nature of digital sensors combined with the mechanical optics and focussing of a rangefinder makes 100% accuracy with every combination of components impossible. ....... and almost certainly the driving force behind a possible digital RF focussing mechanism discussed elsewhere in this forum. One reason I am waiting for 240+. I want an electronic option/assist without using live view which is accurate, but cumbersome. My 8 and 9 are spot on. The 8 was serviced at some point to remove a red line . early M8 and perhaps all were made to film camera specs. Problem is those specs were set for a film channel with depth and a digital sensor has no depth and needs to be set right on. Zeiss had it right with a vacuum back that made the film flat with one Contax SLR. Link to post Share on other sites More sharing options...
tobey bilek Posted August 31, 2015 Share #23 Posted August 31, 2015 In nominal terms, the flange/sensor distance has NEVER changed from M8, and in turn M8 sensor/flange distance is the same as the film/flange distance of all M Leicas from M3 onwards (including CL), which in turn is 1 mm less than the film flange distance of all Leicas from Standard onwards. Then lot of considerations on tolerancing do enter, very different from mechanical to digital Leicas, in which you have a sensor assembly to mount, and a sensor' surface which is covered by glass layers of different size in the different digital Ms... then a rangefinder (with its own tolerancing) and finally the beholder eye... . I doubt that any discussion on this topic can bring to a definitive conclusion : it's a matter followed with high attention at Leica, with surely many compromises carefully balanced between the basic concept of backward compatibility and the needs (and possibilities) of the evolution of the technologies they do implement. What is sure (and the number of posts on this topic are a sign of this) is that current sensors and current lenses do approach the limits of the whole system... at film times none would discuss on such issue... if a Leica should had gone out of register (uneven, but did happen for heavy users) the fact emerged clearly, and it was easily adjusted by a good lab, period. No surprise about the rumors on something new from Leica on the focusing environment... I know of one M8 that did not work with a visoflex and many adjusted the screen to match there camera. Well the M8 went in for service for focus errors. The service dept was convinced after some discussion the sensor was in the wrong place. The camera came back with sensor at correct depth, the viso now does not need shinmng, and the lenses were recalibrated so all works perfectly. All my visos work without shims to screen on both 8 and 9. Link to post Share on other sites More sharing options...
Amedeo Muscelli Posted August 31, 2015 Share #24 Posted August 31, 2015 Sometimes, when I am making adapters and must discard one because is out of tolerances in 0.03mm or more, I remember than a photo films have thickness of 0.20mm, then my question, where Leica has located the focal plane? In the surface of the film? in one of the color layers? The digital cameras have the same focal distance of film bodies? Regards, Amedeo Link to post Share on other sites More sharing options...
fiftyonepointsix Posted August 31, 2015 Share #25 Posted August 31, 2015 I have found that all of my Nikkor LTM lenses are perfect with my M9 and perfect with my M Monochrom with a yellow filter. I have the 5cm F1.5, 8.5cm F2, and 13.5cm F4 all made in 1949- all were perfect. I think the assumption was that film was perfectly flat, which worked out for the M9. The same is not true of the Canon 50/1.5: it was in best focus at F2.8, so I changed the shim. I would "guestimate" that the difference between my M9 and M Monochrom is ~0.015mm. I can set a lens "in spec" for both cameras. An F1.5 lens will have just enough DOF for both camera. I set the lens to slightly front focus on the M Monochrom with no filter in use. When used with a Y52 and deeper, it works out for most lenses. Link to post Share on other sites More sharing options...
dante Posted August 31, 2015 Author Share #26 Posted August 31, 2015 51.6, in Sonnars (and similar lenses), what direction does the focus drift with yellow-orange filters? Thanks, Dante Link to post Share on other sites More sharing options...
luigi bertolotti Posted August 31, 2015 Share #27 Posted August 31, 2015 Advertisement (gone after registration) Sometimes, when I am making adapters and must discard one because is out of tolerances in 0.03mm or more, I remember than a photo films have thickness of 0.20mm, then my question, where Leica has located the focal plane? In the surface of the film? in one of the color layers? The digital cameras have the same focal distance of film bodies? Regards, Amedeo Right question... to pose in a different way, "the 27,8 mm is the nominal distance from M mount flange (its outer surface) to... WHAT exactly ?" I seem to remember to have read somewhere (E. Puts ? Forum ?) that in film Leicas 27.8 is flange to film pressure plate while in digitals is flange to sensor's front surface (to say, sensor "hard" surface PLUS filter(s)) : but am not 100% sure... some other Forum member surely can be of help ; if so this would mean variations in the effective position of focus plane from one model to another... but, of course, means also that, simply, any model has one (theoretical) factory RF calibration of its own (not all, clearly... M8 is like M8.2 etc...) Link to post Share on other sites More sharing options...
dante Posted August 31, 2015 Author Share #28 Posted August 31, 2015 It is also about sensor alignment and as far as I know Leica does not use mechanical measurements. Nor could it be a simple mechanical measurement applicable to all - because almost every digital M has a different filter pack from the next, with a different refractive index, sitting in front of the CCD. Dante Link to post Share on other sites More sharing options...
120 Posted August 31, 2015 Share #29 Posted August 31, 2015 Right question... to pose in a different way, "the 27,8 mm is the nominal distance from M mount flange (its outer surface) to... WHAT exactly ?" I seem to remember to have read somewhere (E. Puts ? Forum ?) that in film Leicas 27.8 is flange to film pressure plate while in digitals is flange to sensor's front surface (to say, sensor "hard" surface PLUS filter(s)) : but am not 100% sure... some other Forum member surely can be of help ; if so this would mean variations in the effective position of focus plane from one model to another... but, of course, means also that, simply, any model has one (theoretical) factory RF calibration of its own (not all, clearly... M8 is like M8.2 etc...) In the case of the film cameras the question "from where to where" is clearly answered in the service manuals. Regarding a Puts claim for digital, I have never seen anything in writing from Leica. M.N. has speculated here on the M8 distance since the camera first came out. For the question where exactly does one locate the image plane for a digital sensor, I found a dearth of information and have referenced a Stanford paper in a previous thread. Link to post Share on other sites More sharing options...
fiftyonepointsix Posted August 31, 2015 Share #30 Posted August 31, 2015 51.6, in Sonnars (and similar lenses), what direction does the focus drift with yellow-orange filters? Thanks, Dante The focus moves towards the infrared index, on the Nikkor 5cm F1.4 and Summarit 5cm F1.5 (which have marks) towards infinity. This means moving the lens slightly closer to the image plane. So- giving myself a head-ache about setting shims for the M9 and M Monochrom: that makes sense. I need a thinner shim when setting the lens for the M9 than setting it for the M Monochrom when using the lens without a filter. A lens shimmed perfectly for the M9 works on the M Monochrom with an Orange filter. The lens must sit closer to the image plane than it would if chromatic aberration were not a factor. The Contax IIIa does not have an IR index. When the original M Monochrom debuted, Leica stated that pictures taken with Orange and Red filters would back-focus. This is consistent with my experience. I would expect the lens to slightly front-focus on the M Monochrom used without a filter. Of course with a Sonnar: field curvature makes a big difference on where you are in the image. I've found this factor also exceeds RF calibration. L1010911 by fiftyonepointsix, on Flickr 1934 5cm f2 Sonnar, wide-open on the M9. nikki_f2_1934_sonnar by fiftyonepointsix, on Flickr 1934 Sonnar 5cm F2, wide-open, orange filter. I think setting up a Tripod and using the Ultra-Achromat on the M9 and M Monochrom via scale focus would give a visual indication of the actual register was different between the two cameras. This would take chromatic aberration and RF calibration out of the equation. Calibrating RF's for focus shift for F-Stop is bad enough, add in chromatic aberration and Field curvature: headache. Link to post Share on other sites More sharing options...
120 Posted August 31, 2015 Share #31 Posted August 31, 2015 I think you all are missing the notion that the 27.8 mm standard is also from the point of view of the lens--it's something the lens has to match. It's sometimes called the working distance. In the case of the film cameras the rails were located at 27.95 mm to account for the film. The pertinent part of the M8 sensor will be located at some other distance, the M9 sensor at some other distance, and so on. The "register distance," however, is still 27.8 mm--that's what a lens has to match. We presume that Leica has picked the appropriate distance for the film rails and different sensors since the cameras can usually be satisfactorily adjusted. Customer lenses and bodies are still adjusted by Leica using a reference "body" or "lens" incorporating the 27.8 mm standard (ignore the fancy giant rulers, displays, etc.--it is same as it ever was). Thus for any two customer bodies that don't agree in a valid test, at least one of the bodies is by definition out of adjustment. Link to post Share on other sites More sharing options...
120 Posted August 31, 2015 Share #32 Posted August 31, 2015 The focus moves towards the infrared index, on the Nikkor 5cm F1.4 and Summarit 5cm F1.5 (which have marks) towards infinity. This means moving the lens slightly closer to the image plane. So- giving myself a head-ache about setting shims for the M9 and M Monochrom: that makes sense. I need a thinner shim when setting the lens for the M9 than setting it for the M Monochrom when using the lens without a filter. A lens shimmed perfectly for the M9 works on the M Monochrom with an Orange filter. The lens must sit closer to the image plane than it would if chromatic aberration were not a factor. The Contax IIIa does not have an IR index. When the original M Monochrom debuted, Leica stated that pictures taken with Orange and Red filters would back-focus. This is consistent with my experience. I would expect the lens to slightly front-focus on the M Monochrom used without a filter. Of course with a Sonnar: field curvature makes a big difference on where you are in the image. I've found this factor also exceeds RF calibration. L1010911 by fiftyonepointsix, on Flickr 1934 5cm f2 Sonnar, wide-open on the M9. nikki_f2_1934_sonnar by fiftyonepointsix, on Flickr 1934 Sonnar 5cm F2, wide-open, orange filter. I think setting up a Tripod and using the Ultra-Achromat on the M9 and M Monochrom via scale focus would give a visual indication of the actual register was different between the two cameras. This would take chromatic aberration and RF calibration out of the equation. Calibrating RF's for focus shift for F-Stop is bad enough, add in chromatic aberration and Field curvature: headache. Nice pics... Link to post Share on other sites More sharing options...
dante Posted September 1, 2015 Author Share #33 Posted September 1, 2015 For infrared, there was actually a published universal correction (well, for most achromat lenses) that was expressed in absolute barrel extension (i.e., focusing closer) as a fraction of focal length. Wouldn't work for lenses with IF, but you get the point. It might be my imagination, but the #13 green seems to make some lenses focus a couple of inches closer. Dante Link to post Share on other sites More sharing options...
fiftyonepointsix Posted September 1, 2015 Share #34 Posted September 1, 2015 The Pop Photo June 1976 edition evaluated thirty-two 50mm lenses and gave the focus shift for spherical aberration and chromatic aberration. If I get snowed in, would make an interesting excel spreadsheet for focus shift at F-Stop and wavelength... For the Zeiss Opton: this is a very easy lens to change the focus. Simply take the rear retaining ring off, pull the optics module out, and screw the variable-stand-off ring in slightly- on the order of 1/6th turn. This will move the lens closer to the sensor and correct the front-focus for the M240. For the M246, using it with a deep yellow or orange filter should match it up to the M240 assuming that you are using it without a filter for the original post. Link to post Share on other sites More sharing options...
luigi bertolotti Posted September 1, 2015 Share #35 Posted September 1, 2015 I think you all are missing the notion that the 27.8 mm standard is also from the point of view of the lens--it's something the lens has to match.... I don't think that here someone misses this point... : all is based on the assumption that any lens manufactured by Leitz is correctly machined and assembled to have a working distance of exactly 27,8 mm... to say when set at infinity focuses a distant object on a plane 27,8 mm far from its back flange... Wouldn'it be so, we would come back to the times of the "pre-standard" Leica Ic... Link to post Share on other sites More sharing options...
120 Posted September 1, 2015 Share #36 Posted September 1, 2015 I don't think that here someone misses this point... : all is based on the assumption that any lens manufactured by Leitz is correctly machined and assembled to have a working distance of exactly 27,8 mm... to say when set at infinity focuses a distant object on a plane 27,8 mm far from its back flange... Wouldn'it be so, we would come back to the times of the "pre-standard" Leica Ic... This thread suggests 1. Leica uses different register distances for different M models, or suggests 2. Leica biases the rangefinder adjustment for different M models, which misses the point of the standard. If, rather, the posters would say that Leica made some MISTAKE in locating the distance of the film or digital sensor in some particular type of camera, that would at least make sense. Link to post Share on other sites More sharing options...
jaapv Posted September 4, 2015 Share #37 Posted September 4, 2015 The definition of the register distance has basically nothing to do with the lens. It is a value of the camera, the distance of the flange to the film/sensor. The lens has to match the standard of the body. For instance an LTM lens has a longer register distance than an M lens which will be corrected by the adapter. The register was set in the film days and must be matched by cameras using present-day sensors. The discussion here is not about the standard as such, but about the defining of the measuring points. The flange is easy. That is a fixed point. But the receptor? Film is easy: it has a huge tolerance. It is not absolute where the image is formed due to the thickness of the emulsion, the film is by no means flat, so the tolerance in setting the standard can easily be 0.10 mm without affecting precision. On a sensor it is different. The location of the microlenses and the receptors can be calculated into a precise plane. So if one is 1/100th out, in the defining of the register distance it will already be visible in the focus, even on a lower resolution sensor. The calculating of the register distance must be within a tolerance zone of 1/1000th mm nowadays, as we get even higher pixel counts in the future it gets tougher, at least when pixel-peeping. We judge focus at 100% magnification, as the pixel count increases magnification increases. This introduces questions like the colour of the light and spectral focus shift. Where exactly does one position the focus point under which image-forming conditions? I think that this is the thing being discussed here, not the standard register distance as defined by Barnack and the M3 designers. And it certainly is not about mistakes, it is about a moving target as technology progresses. Link to post Share on other sites More sharing options...
120 Posted September 4, 2015 Share #38 Posted September 4, 2015 The definition of the register distance has basically nothing to do with the lens. It is a value of the camera, the distance of the flange to the film/sensor... The existing M lenses had a working distance that the new M8 had to match. I am sorry for the unfortunate terminology but I have never seen any other. The lens has to match up with the body and vice-versa so the distance has as much to do with the lens as the body. The flange to film/sensor distance is a fiction for many reasons including those that you gave...again, what is done is to measure 27.95 mm to the film rail and subtract 0.15 mm for good measure. This is the same for every M film camera and well-documented. In the case of the digital cameras you would have to ask Leica. The calculating of the register distance must be within a tolerance zone of 1/1000th mm nowadays, as we get even higher pixel counts in the future it gets tougher, at least when pixel-peeping. Well I know what you mean but the register distance is not "calculated," it's a standard. And where does this number 1/1,000 mm come from? On a sensor it is different. The location of the microlenses and the receptors can be calculated into a precise plane. It's no different because the image is not planar, etc. Link to post Share on other sites More sharing options...
menos I M6 Posted September 20, 2015 Share #39 Posted September 20, 2015 This is an interesting thread and I would not think that there is an easy answer in that there is a fixed, precise flange distance Leica adjusts for in the factory but rather one theoretical one (which excludes deviation from the theory due to sensor flexibility, filter thickness, sensor design, tolerances of components and tolerances in assembly and calibration to factory standards). Personally I have found that none of the digital Leica cameras I have purchased over the years have had a perfectly aligned rangefinder from the factory. Quite contrary, both my M Mono and M9 received a substantial in the field rangefinder calibration in the hotel room after having picked them up and a proper fine calibration once I was back at the work bench. I have started to calibrate many vintage lenses back when my only digital has been my M8.2 and those lenses do behave substantially different with the M9 sensor ( but I always discounted this to be a substantial design difference with the very thick filter glass on the M9 sensor, resulting in much lower acuity compared to the M8 and therefore also a different behavior with the same lenses on both cameras). Now here is the interesting part. I regularly re-calibrate my rangefinders across all camera bodies based on a handful of reference lenses and I am very, very picky about perfecting those rangefinder settings across the entire distance and across many lenses tested thereafter on the calibrated cameras. On every single occasion I had a digital M into Leica service (including a RF check up), the camera was returned exactly with the same RF synchronistic to my lenses as I have adjusted them. My assumption: The initial factory calibration + transport and handling up to the customer receiving a new digital camera will have a rather wide spectrum of calibration. A specific custom calibration either by a technician or at the factory will render a much more precise outcome. There are differences between M8, M9 (different filter glass), MM (much closer back to M8 acuity) and M10 (entirely different sensor design) APART from the rangefinder calibration. Btw - I do tune my RFs and lenses WITHOUT filters on the lens and rather do adjust for the use with filters in the field. This is rather logical to me, as I mostly do remove filters once the light gets low and I tend to shoot with larger apertures. During lighting conditions my contrast filters work at greatest effect (full sun, no clouds, outdoors), the stopped down aperture will mask the shift from yellow and orange filters I use. Only at the rarest moments do I use red filters and if so, I do adjust for the focus shift. Verdict: There is a purely theoretical register used for engineering. There are factors as sensor design, micro lenses, filters and protective glass on the sensor, filters on lenses, lens design, optical glass used, sensor flexibility, manufacturing and assembly tolerances etc … that have significant impact on how focus behaves apart from the exclusively mechanical linkage of the rangefinder. I want to strongly believe that what difference we see between these cameras is rather a function of the variables above than a change in register distance. Link to post Share on other sites More sharing options...
fiftyonepointsix Posted September 21, 2015 Share #40 Posted September 21, 2015 The cover glass of the M8 is 0.5mm, made by Kyocera. Looks to be the equivalent of some Schott group 1 coverglass. The cover glass of the M9 is 0.8mm, S8612 glass. The new cover glass is BG-55, the thickness has not been announced. For the M Monochrom, I set my lenses with an idea of what filter will be used with it most of the time. Usually deep yellow or orange. Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.