Jamie Roberts Posted April 25, 2011 Share #41 Posted April 25, 2011 Advertisement (gone after registration) How did you test that? There is no perspective. Anyway, with a rangefinder you need to test on a surface that is parallel to the sensor plane, as the rangefinder patch will show different focus between top and bottom with a test ruler at 45 degrees, which is the normal test rig. Jaap--I'm also a bit confused by this. Surely in the examples above you're focusing on the zero mark? How can having the focus point on an angle confuse the rangefinder exactly? When I was in Kindermann's, any of the focus checks beyond close focus were also taken at an angle. Link to post Share on other sites More sharing options...
Advertisement Posted April 25, 2011 Posted April 25, 2011 Hi Jamie Roberts, Take a look here problem with front focusing. I'm sure you'll find what you were looking for!
Guest #12 Posted April 25, 2011 Share #42 Posted April 25, 2011 Jaap--I'm also a bit confused by this. Surely in the examples above you're focusing on the zero mark? How can having the focus point on an angle confuse the rangefinder exactly? When I was in Kindermann's, any of the focus checks beyond close focus were also taken at an angle. We had this same thread last year, and Jaap is right, there is some error from not shooting the ruler at exactly 45 deg. And as Malcolm Taylor told someone, it even makes a little difference which direction you slope the ruler. The error has nothing to do with the rangefinder really; it's not huge, and it means you should take some care to make sure the ruler is 45 degrees. It doesn't mean you just throw out testing altogether. All the pictures I have seen of Leica setups in N.J. or Germany have always been sloping targets; the same in textbooks. For 2 or 3m distance I like a parallel target better, like Jaap suggested, for legibility. Link to post Share on other sites More sharing options...
Guest #12 Posted April 25, 2011 Share #43 Posted April 25, 2011 also, I think from the ad this target has a bubble level, so level your camera and there you go. Link to post Share on other sites More sharing options...
hoppyman Posted April 26, 2011 Share #44 Posted April 26, 2011 There is a large difference between the way Solms tests rangefinder accuracy and using one of these test targets at home. Home testing ought best to be interpreted within its limitations including the target, technique and understanding how the rangefinder works differently to the autofocus systems for which these home targets are primarily designed. The Solms jig to test this does have (multiple) precisely angled targets from .7m out to about 6 or 8m from memory. They are not set at anything like 45 degrees from vertical in fact; much less (at least the close ones). They are centred perfectly vertical relative to the lens axis by both rig height variations and adjustable target heights. They are centred horizontally relative to the lens axis by means of staggering the targets laterally and sliding the camera position to precise locations laterally too. Solms use a benchmark standard lens (a 50 I think), They plot close focus and infinity variations at each distance on a projected laser dot and analyse the images in a highly magnified form with a specialised video camera too. Link to post Share on other sites More sharing options...
Guest #12 Posted April 26, 2011 Share #45 Posted April 26, 2011 ...Home testing ought best to be interpreted within its limitations including the target, technique and understanding how the rangefinder works differently to the autofocus systems for which these home targets are primarily designed.... It's just a ruler! The "rangefinder" model is the same as the "autofocus" model. ...They plot close focus and infinity variations at each distance on a projected laser dot and analyse the images in a highly magnified form with a specialised video camera too. "...close focus and infinity variations" means... ? Rack the lens from either direction, maybe. And I guess they look for fringes in the magnified image. There is nothing qualitatively different here than what you can do with the plastic target...it's still a ruler test. Testing is just a part of rangefinders, I think. If we were talking about a $30 soviet lens...there wouldn't be any issue with checking it yourself. I think there is room for both kinds of testing. Link to post Share on other sites More sharing options...
hoppyman Posted April 26, 2011 Share #46 Posted April 26, 2011 #12 you have mis-interpreted what I contributed. For those interested the testing by Leica Camera is documented in a dpreview article. We weren't allowed to photograph this earlier. Leica Factory Tour 2009: 1. Introduction: Digital Photography Review Here's the test result screen photo German terms translated. 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! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/147442-problem-with-front-focusing/?do=findComment&comment=1655851'>More sharing options...
Guest #12 Posted April 26, 2011 Share #47 Posted April 26, 2011 Advertisement (gone after registration) ...They plot close focus and infinity variations at each distance on a projected laser dot and analyse the images in a highly magnified form with a specialised video camera too. sorry if I misinterpreted you, but how so? That's why I asked you, what does "close focus and infinity variations at each distance" mean? I'm the one who gave you the link btw. I asked at the time about the white and blue graphs, "broken tolerance," and what the "40" means. Link to post Share on other sites More sharing options...
Guest #12 Posted April 26, 2011 Share #48 Posted April 26, 2011 what is "Ramitek wedge test"? Link to post Share on other sites More sharing options...
hoppyman Posted April 26, 2011 Share #49 Posted April 26, 2011 .................I'm the one who gave you the link btw. I asked at the time about the white and blue graphs, "broken tolerance," and what the "40" means. The article dates from 2009 and some of the members at the Forum meeting saw this tesing prior to that and we've been discussing it since. You are a couple of years too late to discover the link and give it to us Link to post Share on other sites More sharing options...
Guest #12 Posted April 26, 2011 Share #50 Posted April 26, 2011 ...Solms use a benchmark standard lens (a 50 I think), They plot close focus and infinity variations at each distance on a projected laser dot and analyse the images in a highly magnified form with a specialised video camera too. this is what I'd like you to please clarify: "they plot close focus and infinity variations at each distance..." thanks Link to post Share on other sites More sharing options...
Guest #12 Posted April 26, 2011 Share #51 Posted April 26, 2011 The article dates from 2009 and some of the members at the Forum meeting saw this tesing prior to that and we've been discussing it since. You are a couple of years too late to discover the link and give it to us no, I mean you and I discussed this several weeks ago--so I appreciate the reference, but it's not going to answer my questions. I don't understand how the test works, that's why I'm asking about it. Someone here described it as interferometry, someone said there was one in England...that is about all that has been said about it here. Link to post Share on other sites More sharing options...
Guest #12 Posted April 27, 2011 Share #52 Posted April 27, 2011 this is the screen shot here: http://www.dpreview.com/articles/leicafactory2009/images/massembly/focuscheck4.jpg You can see the results for the six distances plotted. My guess is the two graphs are just the usual backlash test, but I could be wrong. Someone should just ask Leica what the tolerance units are. I would be interested if anyone knows what the previous setup was...I heard it was something entirely different. Link to post Share on other sites More sharing options...
Guest #12 Posted April 27, 2011 Share #53 Posted April 27, 2011 The picture is for a testing error when the ruler is at the wrong angle (dotted line). Here you are focusing on the center of the ruler, but there is some backfocus. Segment "b" is the error measured along the ruler, and segment "a" is what it should be. Then b/a = Cos Alpha / Cos Beta, and from this you can get the percent error. This holds for back or front focus, and however you slope the ruler. I don't think you would be off by more than a couple of degrees with that plastic target if you level, so that's no more than a 3-4% testing error in this case. The error from misalignment in space, not just a plane, would be in the ballpark. Another error that came up recently was misalignment of a (perpendicular) target to check the parallax compensation...that is more involved, but apparently doesn't count for much, either. corrections welcome 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! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/147442-problem-with-front-focusing/?do=findComment&comment=1657766'>More sharing options...
Guest #12 Posted April 27, 2011 Share #54 Posted April 27, 2011 as a p.s., I don't think the 1cm error the OP found is much to bother about. Link to post Share on other sites More sharing options...
Guest #12 Posted April 28, 2011 Share #55 Posted April 28, 2011 ...You can use an oblique target - if you manage to test focus distance exactly in the optical axis of the RF system i.e. the middle of the RF patch... I just don't think this right, or that you have the right diagram. You can focus on the center of the ruler with the center or edge of the patch, and you will get the same result as long as you are 45 deg. to the target. If you focus with the edge of the patch and then slide the camera laterally, now you are back in the center of the patch, and the focus distance hasn't changed. Now if the center of the ruler is still in the center of the patch, and you focus on a point of the ruler at the edge of the patch, you will get a different result...but that is not an error. Link to post Share on other sites More sharing options...
Guest #12 Posted April 28, 2011 Share #56 Posted April 28, 2011 And just to clarify, +x degrees error in alignment does not count for the same as -x degrees. So if the error is from rotating the camera in the same direction to compensate for the viewfinder displacement, that is when it matters which direction you slope the target. (Malcom Taylor) Link to post Share on other sites More sharing options...
Guest #12 Posted April 30, 2011 Share #57 Posted April 30, 2011 ...Better to use a parallel target and use the whole patch. Sorry to beat this topic to death, but I think it is important enough. I think the error you are hunting for is your visual acuity. In other words, when you focus on the center of the ruler there is always some visual uncertainty in setting the coincidence in the patch. And you will have the same error whether the target is sloped or parallel. You can find this error described in the rangefinder literature; the formula is usually a u^2 / (m where "a" is visual acuity or resolution, u is the target distance, m is finder magnification, and b is the (0.06925m) baselength. The value for "a" is really just a wild guess, but two examples I've seen in the context of rangefinders are 0.000291 and 0.00005. So at 1m distance with the stock 0.68X magnification the error would be 1.1mm to 6.2mm. But this is just a wild guess, as mentioned. If you use a 2X magnifier to test focus, you can see that this particular error will be quite reduced. This is why Leica uses a laser to set the coincidence. Link to post Share on other sites More sharing options...
Guest #12 Posted April 30, 2011 Share #58 Posted April 30, 2011 ...They plot close focus and infinity variations at each distance... so you posted this; what does it mean? Thank you for any info. Link to post Share on other sites More sharing options...
ho_co Posted April 30, 2011 Share #59 Posted April 30, 2011 a u^2 / (m Does that mean auu/mb ? You call it a formula, but there's no "=" sign, so to me it seems meaningless as a formula. What is auu/mb equal to? Link to post Share on other sites More sharing options...
Guest #12 Posted April 30, 2011 Share #60 Posted April 30, 2011 Does that mean auu/mb ? You call it a formula, but there's no "=" sign, so to me it seems meaningless as a formula. What is auu/mb equal to? Apologies for the typesetting, but u^2 = u u. The error range is E = a u^2 / (m , plus or minus. In other words, if the camera and lens is perfect, and you focus on something at a distance u, the focus plane will be within u+-E. Don't worry, this is not mine, but the usual textbook formula. Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.