tashley Posted June 17, 2013 Author Share #101 Posted June 17, 2013 Advertisement (gone after registration) My point of post was to outline the various aspects of the issue, and how it can lead to different conclusions. I see the points on all the sides, and can see differing conclusions that all have some plausibility. I certainly wouldn't choose to suggest someone (forum member 120) is an ignoramus and easy prey for propaganda and manipulation because they might have a differing viewpoint but that's just me. If you are "happy" to do so, so be it. Good for you. Ignoramus was certainly not where I was coming from. But 120 does keep misquoting me and I am sorry if that has made me short.... Link to post Share on other sites More sharing options...
Advertisement Posted June 17, 2013 Posted June 17, 2013 Hi tashley, Take a look here 3 months of M240 - am I keeping it?. I'm sure you'll find what you were looking for!
macjonny1 Posted June 17, 2013 Share #102 Posted June 17, 2013 Ignoramus was certainly not where I was coming from. But 120 does keep misquoting me and I am sorry if that has made me short.... Ah maybe there is more there then, I haven't read everything. Personally I'd agree with current users such as yourself if there is that much of a difference. You are obviously more advanced than 99.999% of folks out there in terms of testing and such. Also Leica is a strange company and certainly one cannot be sure about the lack of information. I'd say if it was Canon or Nikon I'd be pretty certain that a lack of information was better evidence. Link to post Share on other sites More sharing options...
IkarusJohn Posted June 17, 2013 Share #103 Posted June 17, 2013 I'm incredibly grateful to Tim for taking so much time to explain all this. I'm not even in the market for an M (typ 240), but I have learned a lot. I have similar, lingering concerns about the stability of the rangefinder mechanisms in my M3, M9-P & Monochrom. They have all been back to Solms at varying times over the last 3 years, so I am confident that they should be pretty good - but I wonder for how long ... So, I asked Solms if they could do an upgrade (as they did with the M6?). Here is the response: Unfortunately there is no way to upgrade the M9 / Monochrom with the Leica M rangefinder. But you should not be concerned about your rangefinder as we – Leica - will find a solution in case it might be necessary. Clearly, something is different in the new M rangefinder, and it is working well. I will just test my lenses and rangefinders from time to time to assuage my worries. Over time, I'm sure those concerns will fade! Cheers John Link to post Share on other sites More sharing options...
algrove Posted June 18, 2013 Share #104 Posted June 18, 2013 Very interesting reply from Solms AFAIK. Link to post Share on other sites More sharing options...
tashley Posted June 18, 2013 Author Share #105 Posted June 18, 2013 I'm incredibly grateful to Tim for taking so much time to explain all this. I'm not even in the market for an M (typ 240), but I have learned a lot. I have similar, lingering concerns about the stability of the rangefinder mechanisms in my M3, M9-P & Monochrom. They have all been back to Solms at varying times over the last 3 years, so I am confident that they should be pretty good - but I wonder for how long ... So, I asked Solms if they could do an upgrade (as they did with the M6?). Here is the response: Clearly, something is different in the new M rangefinder, and it is working well. I will just test my lenses and rangefinders from time to time to assuage my worries. Over time, I'm sure those concerns will fade! Cheers John And thank you John. That is fascinating. One of the things I love about Leica is that they aren't sufficiently monolithic to 'message' consistently. Unlike Nikon... One year in, and none of us really knows the truth behind the 'interesting' focus behaviour of the D800... Link to post Share on other sites More sharing options...
120 Posted June 18, 2013 Share #106 Posted June 18, 2013 My diagram is not labelled anywhere with the phrase Image Space. That phrase refers to the method of calculating optimal DOF and focus when making measurements between the front and rear standards of a technical camera during movements. I refer you to this article, which explains it clearly. I merely have noted, as a result of a large number of experiments, that the RF in the M240 has a better ability to ascertain focus so as to place the DOF favourably to a planar subject than does the EVF. The rest is my conjecture. It is unlikely to be happening purely by chance so your own conjectures are welcome but I would note that at least one other participant in this thread has suggested that your information is outdated. Any observations you have from using your own M240 would be most interesting! But the idea that something is not true until Leica tells you that it is true is patently unsound: most serious users of this camera have noted that the RF is a great deal better and whether that relates to design, or to improved tolerances, is of theoretical interest only. Something is different, and I am surprised that you have not noticed this in your own camera - maybe yours is not so well adjusted? Let me rephrase my comments, as both the new article and the old one are well-written and interesting; truly. Re the "misquoting," it is not correct to label the the plane of best focus "image plane," and you are way off on "image space". Re the rangefinder, take what I said at face value; I am only interested in how the rangefinder works, so that would be interesting to have a new rangefinder, or new standards for the adjustments. You may turn out to be correct, or not. The bet about the cams I will take. The part about talking to Leica is not because Leica is authoritative, but it's their lenses and rangefinders. Or ask DAG about the cams; I think that's his father who wrote an article on image topography, right up your alley. Even if you are wrong about both the novelty of the rangefinder and the grinding of the cams, you would have to, what, delete 2 or 3 sentences in your article? Link to post Share on other sites More sharing options...
120 Posted June 18, 2013 Share #107 Posted June 18, 2013 Advertisement (gone after registration) ... Mark, since this has turned into a thread about rangefinders, can you tell me what is the underlying standard for the rangefinder focusing? Maybe extension of a 50mm lens? And hopefully give a reference. I have not been able to find this anywhere, and remember you have posted something about it in the past. Link to post Share on other sites More sharing options...
marknorton Posted June 18, 2013 Share #108 Posted June 18, 2013 Yes, the standard is based around a 50mm lens and it's no coincidence Leica use a 50mm lens when aligning rangefinders and I have a chrome 50mm Summicron which I use only for the purposes of establishing which of my cameras is correctly aligned. This is illustrated by the humble 50mm f2.8 Elmar, the simplest of all (recently) current lenses. In this lens, there is just a single helicoid and the rangefinder roller bears directly on the back of the lens barrel which does not itself rotate. When the lens is set to infinity, the barrel is as far back as it goes and at this point, the rangefinder should be coincident for objects at infinity. As the lens is focussed at closer distance, the lens barrel moves forward and the roller with it. The amount of lens extension is proportional to the focussing ring angle turned back from infinity which in turn depends on the pitch of the helicoid. There's a classic formula in optics: 1/f = 1/u + 1/v which is valid to a first approximation for camera lenses, f is the focal length, v and u are the object and image distance; plug the values in and you get a lens extension of 0.251mm at 10 metres distance, 2.631mm at 1 metre and 3.846mm at 0.7mm which defines the range of movement of the rangefinder roller required. You can plot a graph (I have one somewhere, I will publish it if I can find it) of lens extension against focussing distance. Interestingly, you can also plot the circumferential distances of the markings on the lens distance scale back from infinity against the marking values themselves and you get the same (approximately) graph. This is then the simplest coupling between lens and camera. The camera never sees any different - there is a defined relationship between roller displacement and coincident distance set by this humble 50mm lens. Any "smarts" to handle the characteristics of particular lenses are built into the lenses themselves. The most basic characteristic which has to be handled is focal length. The extension from infinity to close focus is much greater for a 90mm lens than for a 50mm lens and the extension for a 21mm lens is much less. However, for the rangefinder still to work, there has to be a surface on the back of the lens which moves through the same range as the lens barrel for the "reference" 50mm lens. This is accomplished in two main ways. First, there can be a second helicoid which moves the surface against which the roller is pressing at a different rate to the lens barrel. At infinity, the lens barrel helicoid is set so that the lens focusses to infinity and the focussing helicoid is set so that the rangefinder is coincident at infinity. As the focussing ring is turned, the lens moves forward at one rate and the focussing surface at another. They are both still linked to the angular rotation of the focussing ring. The 75mm Summicron is a good example. The second option if you don't want to go to the trouble of having a second helicoid is to arrange for the roller to press on a surface linked to the focussing ring which then rotates and is profiled so that as the focussing ring rotates, the roller moves as required. Examples are the 18mm Super-Elmar and the 90mm Elmar. The problem now is that the rangefinder roller can make contact at a slightly different lateral position depending on its own alignment and throw the focussing off, especially when the surface is steeply profiled. It works in the 90mm Elmar because the profiling is quite shallow and with the 18mm because of the depth of field available. Things get tricky when you have a difficult lens like the old Noctilux or 75mm Summilux because optical variations meant not all copies behaved in the same way and the Noctilux is reputed to have had a hand-ground cam profile. In the most complex lenses, such as the 21mm Summilux, it is not sufficient to just have the two helicoids because the lens has different focussing characteristics to the reference 50mm; in this case, they use a second focussing helicoid to get it in the right ball park AND profile the cam to get it exactly right. Fundamentally though, the rangefinder characteristic which defines roller displacement against coincident distance goes back to that basic 50mm lens. It's cameras and lenses both adhering to that characteristic which is essential to allow lens interchangeability. Sorry for the long post, if you got this far, you deserve a medal! Link to post Share on other sites More sharing options...
pedaes Posted June 18, 2013 Share #109 Posted June 18, 2013 Mark -thanks, fascinating. Makes one appreciate the design complexity of something we take for granted. It was not at all onerous to get to the end! Link to post Share on other sites More sharing options...
Peter Branch Posted June 18, 2013 Share #110 Posted June 18, 2013 Mark, ..........Things get tricky when you have a difficult lens like the old Noctilux or 75mm Summilux because optical variations meant not all copies behaved in the same way and the Noctilux is reputed to have had a hand-ground cam profile........ A very good and technically correct explanation - thanks. My recollection is that the focus travel of the "50mm" lens is calculated using a focal length of 51.6 mm. This was the designed focal length of early "50mm" lenses. My current 50mm LUX ASPH is marked as being 51.4mm. If readers have older Leica 50mm lenses they may well find that the rear surface which connects with the roller is visibly profiled. My 50 LUX, perhaps because it is not exactly 51.6mm, shows clear signs of machining along that part of the rear of the lens which connects with the roller. Link to post Share on other sites More sharing options...
douglasf13 Posted June 18, 2013 Share #111 Posted June 18, 2013 Well, I have used extensively manual focus with Leica lenses on my NEX-5N with EVF and NEX-7.My experience is, probably due to my non-perfect eye sight, that for scenes with high micro contrast focus peaking gets me close to a good focus. However, unless I just got extremely lucky, if I then zoom in all the way I find that focus still can be improved. I also find that focus peaking typically seems to work best with 1x magnification. Zooming in can significantly reduce the strength of the focus peaking signal. But the magnification more than makes up for that. All, of course, on a tripod. I also find, being able to position the focus box where I want it within a frame, is an important feature that's reportedly missing in Leica's current implementstion. A very disappointing oversight for my photographic interests. Yeah, I was referring to focus magnification in my post. With just regular magnification and focus peaking, I'd say I'm pretty darn good, but not rangefinder good. Link to post Share on other sites More sharing options...
120 Posted June 19, 2013 Share #112 Posted June 19, 2013 ... Thanks! I appreciate the detailed answer, and especially the classification of the mounts and the examples. That is hard-to-find info; I wish there was a book or article on that topic. Already someone says the standard is 51.6, so that is another reason I was hoping for an official source. I think the detail I am really missing is the linearity of the roller travel with the turning of the telescope/prism arm. Hopefully that's not a requirement, because it would make the description of the rangefinder easier. If the travel of the roller is purely the extension of a 50 or 51-point-whatever lens, then you can show it's not linear. Link to post Share on other sites More sharing options...
jaapv Posted June 19, 2013 Share #113 Posted June 19, 2013 Are you going to build your own rangefinder? Link to post Share on other sites More sharing options...
Peter Branch Posted June 19, 2013 Share #114 Posted June 19, 2013 Thanks! I appreciate the detailed answer, and especially the classification of the mounts and the examples. That is hard-to-find info; I wish there was a book or article on that topic. Already someone says the standard is 51.6, so that is another reason I was hoping for an official source. I think the detail I am really missing is the linearity of the roller travel with the turning of the telescope/prism arm. Hopefully that's not a requirement, because it would make the description of the rangefinder easier. If the travel of the roller is purely the extension of a 50 or 51-point-whatever lens, then you can show it's not linear. Oh it's much more complicated that that! For a start the distance scale engraved on the lens is the distance from the object to the image plane. As Mark Norton has correctly pointed out, the 1/u + 1/v = 1/f formula he used is an approximation. So the 0.7m measurement is actually "u" + "v" + "δ" where δ is the distance between the two nodes of the lens, (which might be negative!). Then there is the tricky subject of what is the actual focal length. In any lens which exhibits focus shift, the clue is in the words, the "focal length" is a function of the f stop being used. There is more......... Link to post Share on other sites More sharing options...
Maarten Posted June 19, 2013 Share #115 Posted June 19, 2013 Mark, to your knowledge: in the lenses with the double helicoid is the adjustment at infinity being performed by moving the two helicoids in relation to each other while the lens head remains focused on infinity? I would assume that in the lenses with the 'profiled surface' the adjustment is performed with shims? What I can add to the discussion is that by bending the roller cam arm (with the 'bending tool' described in one of the repair manuals) the linearity of the roller operation can be optimized. This optimization however is to correct the rangefinder only so is independent of the lens being employed. Hope I could make myself clear... Maarten Link to post Share on other sites More sharing options...
Guest Marc G. Posted June 19, 2013 Share #116 Posted June 19, 2013 glad you're keeping it. I enjoyed reading the articles on your site but one thing that I always found annoying is that you're overly critical when it comes to lens performance. I suppose that goes with using a D800 Link to post Share on other sites More sharing options...
ALD Posted June 19, 2013 Share #117 Posted June 19, 2013 Is there anyway somebody could do a technical testing of the differences between the rangefinders on the m240 and the M9 please. The improved rangefinder is a game changer for me. Link to post Share on other sites More sharing options...
farnz Posted June 19, 2013 Share #118 Posted June 19, 2013 I suggest that the best technical testing that can be done is with your eye and the two cameras at a dealers because your vision is unique to you and normative technical testing is unlikely to provide you with much useful information in this instance. Pete. Link to post Share on other sites More sharing options...
120 Posted June 19, 2013 Share #119 Posted June 19, 2013 Oh it's much more complicated that that! For a start the distance scale engraved on the lens is the distance from the object to the image plane. As Mark Norton has correctly pointed out, the 1/u + 1/v = 1/f formula he used is an approximation. So the 0.7m measurement is actually "u" + "v" + "δ" where δ is the distance between the two nodes of the lens, (which might be negative!). Then there is the tricky subject of what is the actual focal length. In any lens which exhibits focus shift, the clue is in the words, the "focal length" is a function of the f stop being used. ...... I would agree with all of that. The formula is exact in this setting, just that the u and the v are measured from the principal points. I don't think there is anything else to use; you can see especially the US military papers on rangefinders. You would use one of the usual approximations to get the object distance from the object-to-image distance. You can show it will not matter. Note also the distance from the object to the rangefinder is not the same as the distance from the object to the image, in the Leica. Small differences in nominal focal lengths (50, 51.6, etc.) can be compensated by a cam. I do not think any rangefinder camera tries to compensate for focus shift. So I am only interested in how Leica did the rangefinder; if they made some approximation, great. It will be interesting to know about it. There is more......... more welcome Link to post Share on other sites More sharing options...
120 Posted June 19, 2013 Share #120 Posted June 19, 2013 I think that was the way it was done with the first cams,about eighty years ago. His source must be caught in a time warp . Nowadays they are machined to a narrow tolerance and I believe not by Leica themselves but by Uwe Weller. I have no +/- tolerances for the cams Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.