darylgo Posted January 27, 2016 Share #21 Posted January 27, 2016 Advertisement (gone after registration) Do you mean the focus variation is within 2 meters with tendency to front focus? Or that Leica Shuttle reads different values each time focusing for the same distance...? Yes to the first question. Second question: Leica Shuttle reads accurately, the distance to the tree is the same, the readout of distance as set on the lens is different because the camera/lens choose a different distance with AF. Was this with many lenses the case? Iirc, all tested this way. Link to post Share on other sites More sharing options...
Advertisement Posted January 27, 2016 Posted January 27, 2016 Hi darylgo, Take a look here S 006 focus error. I'm sure you'll find what you were looking for!
ZHNL Posted January 27, 2016 Share #22 Posted January 27, 2016 Many thanks for posting your findings! I am very disappointed because the accuracy of the AF system is very important for me. I did not notice problems so far with the size of the AF spot. I noticed that the AF does not always re-focus, so my usual habit is to de-focus manually to make sure AF is re-invoked. On the 180mm the focus sometimes does not fall into the DOF area even at F8... In theory if in near distance you have correct focus it should be possible to compensate far distance focus errors by firmware or adjust the AF system mechanically. Please let me know once you have a clear answer from Leica about this! The S lenses have individual focus calibration data stored in their chips IIRC, and Leica was very proud of the resulting accuracy of the S system's AF. And that's the reason why the S bodys do not offer focus fine tuning settings. I hope that the factory can overwrite the focus compensation data of the S lenses with new measured values! The center spot is too large for Portrait: say you want nail the focus on eyes. How you know the camera is focus at eye slash or eyes ball? Of course it will also confuse me in field for WA shooting that deep DOF is under covered by center AF cycle. I even think this is one of reason there is no super fast S glass because of AF technical issue. Sure most studio user won't care much WO performance. I did a test with 100 cron at Leica store with my S006 and S7. My keep rater is not stellar on people unless I have a very well defined subject like huge sign on street, big build door frame etc...or Leica logo in front of Leica store SF. Hence, I put AF as bad point in 'to S or not' thread. TBH, I am quite tolerant with Leica with quirks like these over the years. I still love the company and want be their loyal customer. and there is no other system I used can produce file as beautiful as Leica S. and I am quite skilled on manual focus so most complain here is in need I want AF occasionally. The reason I am holding ship back my whole system is Leica may not be able to fixed by firmware or whatever if it is simple limitation of S AF engine as indicated by Darylgo's post. I really hope my guess is wrong and these are solvable. As for Leica reply, I don't have any technical reply as promised (I was told will be updated by Leica Germany). I think my question may not easy to answer and some information may not want be shared with regular user like me. So, I decide I will give them a try by sending my whole system soon. Link to post Share on other sites More sharing options...
ZHNL Posted January 27, 2016 Share #23 Posted January 27, 2016 My S photography is in studio so I am using closer focus ranges of course. Bu focus is much more critical and DoF much less there even when stopped well down. Any system AF problem there would be much more noticeable than at infinity just on DoF alone. I was really talking about AF for long distance say 20 M and above. It is not about how critical AF and how shallow DOF. I think it is more than that. Since I use manual focus a lot, I can tell you what is my guess here. Take 70mm Summarit as example, due to extreme short focus throw, from 30M(a guess) to infinity, you only require turn focus ring 1~2mm. That is very high requirement for fine adjust for AF engine to take it for Lens performance like S glass and 37.5M sensor. Canikon seems do fine here and getting better and better over each upgrade even with short focus throw. These is definite trade off between focus throw and focusing speed and accuracy. Having them all is not technical trivial I guess. The focus error is not huge if you look Jpk's example. May not show up for 10M file. but based on his description on 180mm, even f8 f11 won't clear the error, I guess he is seeing the same thing as I do, the longer the distance, the worse the results.(meaning front focus) Link to post Share on other sites More sharing options...
albireo_double Posted January 27, 2016 Share #24 Posted January 27, 2016 I've never had any issues with AF accuracy with my S2 and now with the S007. I haven't done any detailed tests but the consistent AF accuracy of the system, no matter which lens (Leica S from 24 to 180, Hasselblad H 300 via adapter) and what shooting situation (macro, portrait, landscape), has been a key strength of the system for me. I suggest that the OP should get the camera body tested and serviced by Leica. My experience has been a 1-3 week service turnaround time for the S system, door to door, which would make the decision easier for me if I had any doubts about the performance of my camera. Link to post Share on other sites More sharing options...
jpk Posted January 28, 2016 Author Share #25 Posted January 28, 2016 the readout of distance as set on the lens is different because the camera/lens choose a different distance with AF. Iirc, all tested this way. This is bad news indeed. I was really talking about AF for long distance say 20 M and above. It is not about how critical AF and how shallow DOF. I think it is more than that. Since I use manual focus a lot, I can tell you what is my guess here. Take 70mm Summarit as example, due to extreme short focus throw, from 30M(a guess) to infinity, you only require turn focus ring 1~2mm. That is very high requirement for fine adjust for AF engine to take it for Lens performance like S glass and 37.5M sensor. Canikon seems do fine here and getting better and better over each upgrade even with short focus throw. These is definite trade off between focus throw and focusing speed and accuracy. Having them all is not technical trivial I guess. The focus error is not huge if you look Jpk's example. May not show up for 10M file. but based on his description on 180mm, even f8 f11 won't clear the error, I guess he is seeing the same thing as I do, the longer the distance, the worse the results.(meaning front focus) The motor drives the focusing helical, and the focusing helical drives the lens barrels focus ring. Hence the motor driven gear can make the focus throw look short from the users point of view: 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! This is the situation for the phase detection AF in the close range: And this is how the phase detection sees far distances: So accuracy may be more noticeable in the close range due to the shallow DOF, but is at least as relevant in the far distance. The length of the focus throw is only a result of the gear which is driven by the focusing helical... Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! This is the situation for the phase detection AF in the close range: And this is how the phase detection sees far distances: So accuracy may be more noticeable in the close range due to the shallow DOF, but is at least as relevant in the far distance. The length of the focus throw is only a result of the gear which is driven by the focusing helical... ' data-webShareUrl='https://www.l-camera-forum.com/topic/256088-s-006-focus-error/?do=findComment&comment=2977657'>More sharing options...
ZHNL Posted January 28, 2016 Share #26 Posted January 28, 2016 This is bad news indeed. The motor drives the focusing helical, and the focusing helical drives the lens barrels focus ring. Hence the motor driven gear can make the focus throw look short from the users point of view: SCAN0004-3.jpg This is the situation for the phase detection AF in the close range: SCAN0004.jpg And this is how the phase detection sees far distances: SCAN0004-2.jpg So accuracy may be more noticeable in the close range due to the shallow DOF, but is at least as relevant in the far distance. The length of the focus throw is only a result of the gear which is driven by the focusing helical... Thanks for the explanation, in first picture, you basically saying inside should have finer control than what outside feel. I agree it might be just fine. it might not. depend on gear ratio. but you do see my point why I speculate about its problem at infinity. For example, for achieve critical focus at 40m away to differentiate 30m away, the focusing turn less than half mm, (I wish I have LV to verify my number, but I know it is around that) What about the gear control have finite turn resolution, hence one of two mis step will cause a lot focusing error at long distance. Hope I explain clearly here for everyone about my thought. This is the type of question I will never hear from Leica. This is my wild guess why we see accuracy issue here from three different users all from long distance focusing. For anyone claim they have no problem what so ever, I really want to hear is how you test it, how you inspect the results, have you shoot WO for subject 30M away, how about 40m away, how about infinity. Have you carefully inspect the DOF window for the shots? The focusing error in my case can range from small to big depend on focusing target size, contrast etc... Link to post Share on other sites More sharing options...
darylgo Posted January 28, 2016 Share #27 Posted January 28, 2016 Advertisement (gone after registration) This is bad news indeed. It is not confidence building to see this, however in real world shooting the camera performs very well and I rarely see af issues. For anyone claim they have no problem what so ever, I really want to hear is how you test it, how you inspect the results, have you shoot WO for subject 30M away, how about 40m away, how about infinity. Have you carefully inspect the DOF window for the shots? The focusing error in my case can range from small to big depend on focusing target size, contrast etc... +1, methodology is critical.... and well done by the OP. . Link to post Share on other sites More sharing options...
jpk Posted January 28, 2016 Author Share #28 Posted January 28, 2016 you basically saying inside should have finer control than what outside feel. I agree it might be just fine. it might not. depend on gear ratio. but you do see my point why I speculate about its problem at infinity. For example, for achieve critical focus at 40m away to differentiate 30m away, the focusing turn less than half mm Yes, I see your point! For anyone claim they have no problem what so ever, I really want to hear is how you test it, how you inspect the results, have you shoot WO for subject 30M away, how about 40m away, how about infinity. Have you carefully inspect the DOF window for the shots? The focusing error in my case can range from small to big depend on focusing target size, contrast etc... I just double checked my S2P files and got the feeling that there might have been similar issues. I only took always some extra shots to make sure there are sharp versions among them and deleted the soft versions, but some near infinity shots are not as sharp as they should be. +1, methodology is critical.... and well done by the OP. Thanks! I have experience with focus errors, see this thread which was the reason for me to move to the S system... It is not confidence building to see this, however in real world shooting the camera performs very well and I rarely see af issues. In my real world shooting I have often used AF between 20m and infinity and ended with files which I didn't buy the S system for... In the near range the AF was perfect for me though. Link to post Share on other sites More sharing options...
hoppyman Posted January 28, 2016 Share #29 Posted January 28, 2016 The center spot is too large for Portrait: say you want nail the focus on eyes. How you know the camera is focus at eye slash or eyes ball? Of course it will also confuse me in field for WA shooting that deep DOF is under covered by center AF cycle. I even think this is one of reason there is no super fast S glass because of AF technical issue. Sure most studio user won't care much WO performance. SNIP ........................................................ My experience with the S2 and S (Typ 007) focusing for portraits is very good. If I had a focus error it was due to my technique. Camera movement is a significant cause of softer images as always. If you focus and then recompose the distance changes a little. Now I tend to keep the eye near the centre of frame in the composition and crop later. I have not yet put the microprism and split image screen in my S. I have made over 4000 exposures with it so far and am still happy to use the rear button for AFS mode in manual focus only. I have a thread with some results here: http://www.l-camera-forum.com/topic/247391-studio-beauty-portrait-s2/ About the first 40 images here are with S2 and S (Typ 007) http://www.pbase.com/hoppyman/cp Link to post Share on other sites More sharing options...
Manolo Laguillo Posted January 28, 2016 Share #30 Posted January 28, 2016 This thread is highly interesting. Thank you! :-) It makes me remember some instances where the end result was not optimal. That happened always when focusing at infinity. Therefore: if at infinity or near infinity the auto focus can miss the hit, would it be better to focus manually? Answer: yes. Link to post Share on other sites More sharing options...
ZHNL Posted January 29, 2016 Share #31 Posted January 29, 2016 The center spot is too large for Portrait: say you want nail the focus on eyes. How you know the camera is focus at eye slash or eyes ball? Of course it will also confuse me in field for WA shooting that deep DOF is under covered by center AF cycle. I even think this is one of reason there is no super fast S glass because of AF technical issue. Sure most studio user won't care much WO performance. SNIP ........................................................ My experience with the S2 and S (Typ 007) focusing for portraits is very good. If I had a focus error it was due to my technique. Camera movement is a significant cause of softer images as always. If you focus and then recompose the distance changes a little. Now I tend to keep the eye near the centre of frame in the composition and crop later. I have not yet put the microprism and split image screen in my S. I have made over 4000 exposures with it so far and am still happy to use the rear button for AFS mode in manual focus only. I have a thread with some results here: http://www.l-camera-forum.com/topic/247391-studio-beauty-portrait-s2/ About the first 40 images here are with S2 and S (Typ 007) http://www.pbase.com/hoppyman/cp I really like post #11&12 from that linked thread, the skin tone are gorgeous , one of big reason I want use S over any other system. First of all, like I said, it is show no obvious problem to me at close focus distance. even there is an error, I can hardly blame S since I have no idea if the center cycle focus at eye slash or eye ball like I mentioned before. Your portraits are mostly shoot after f8 to me if not f11, even DOF is still shallow at that distance, it is not a challenge to any AF system right now in market I would say. For close distance portrait, say 0.5~0.7M with 70mm WO, the AF with that size center spot is useless to me if I want nail focus at eyes consistently. good thing is S has wonderful VF so manual focus in that situation is a blast. I really wish someone can chime in claim he has perfect infinity focus WO consistently or consistent most of time. It at least give me hope so that i can ship my whole system to Leica to have them fully checked. For now, if I need shoot infinity WO occasionally, I use focusing mark on lens body to do that. the difficult is for people 10 meter away, I will have to rely on some luck here for either AF or MF. 36M is merciless even for f2.5 Lens at that distance. Link to post Share on other sites More sharing options...
jpk Posted January 29, 2016 Author Share #32 Posted January 29, 2016 First of all, like I said, it is show no obvious problem to me at close focus distance. I can also confirm very accurate focus in the close range. The AF spot was for me fine to get the focus on the eylash not because it's small enough but becaus the viewfinder is so good that I can see if focus is where I want to have it, so with refocus or slight body movement I can compensate if necessary (as I am used to from the M system...) I really wish someone can chime in claim he has perfect infinity focus WO consistently or consistent most of time. It at least give me hope so that i can ship my whole system to Leica to have them fully checked. Absolutely!!! So: QUESTION: Who has perfect focus near infinity wide open with AF on the S2/006? And can someone confirm that these problems also exists with the 007? (Dear admins: do you think it's useful to change the thread title to something like "AF accuracy near infinity"?) Link to post Share on other sites More sharing options...
mgrayson3 Posted January 29, 2016 Share #33 Posted January 29, 2016 Not sure this helps, but I used an 007 with 120mm lens for a few days and AF at 1000m wide open was perfect. Branches at only 100m were blurry, and I could change focus to them accurately. I did not try it at the 30m-40m range. Best, Matt Link to post Share on other sites More sharing options...
xiaubauu2009 Posted January 31, 2016 Share #34 Posted January 31, 2016 This is bad news indeed. The motor drives the focusing helical, and the focusing helical drives the lens barrels focus ring. Hence the motor driven gear can make the focus throw look short from the users point of view: SCAN0004-3.jpg This is the situation for the phase detection AF in the close range: SCAN0004.jpg And this is how the phase detection sees far distances: SCAN0004-2.jpg So accuracy may be more noticeable in the close range due to the shallow DOF, but is at least as relevant in the far distance. The length of the focus throw is only a result of the gear which is driven by the focusing helical... jpk, is S's focus system a phase detect system??? or is it a contrast detect... Link to post Share on other sites More sharing options...
John McMaster Posted January 31, 2016 Share #35 Posted January 31, 2016 I often got my S2-P with 120mm or 180mm to focus on a hill crest ~3km away when shooting sun-downs. john Link to post Share on other sites More sharing options...
jpk Posted January 31, 2016 Author Share #36 Posted January 31, 2016 I checked focus in the near range again and it seems that focus is not 1/3 in front of and 2/3 behind the in-focus area, more likely it's about 50:50... jpk, is S's focus system a phase detect system??? or is it a contrast detect... I think phase detect... Link to post Share on other sites More sharing options...
jpk Posted January 31, 2016 Author Share #37 Posted January 31, 2016 I often got my S2-P with 120mm or 180mm to focus on a hill crest ~3km away when shooting sun-downs. And focus is perfect...? Link to post Share on other sites More sharing options...
John McMaster Posted January 31, 2016 Share #38 Posted January 31, 2016 And focus is perfect...? Yep john Link to post Share on other sites More sharing options...
alan.y Posted February 15, 2016 Share #39 Posted February 15, 2016 On my 007, my Summicron-S shows the same symptoms as described by the OP and ZHNL. It appears to front-focus consistently, at any distance, in phase detection AF--but I seem to have had accurate AF occasionally. Contrast-detection and manual focusing (I have the microprism screen) are both accurate and yield results different than phase detection. My other lenses (30, 45, 70) focus accurately. The 100 yields results like the attached 100% center crops. Tested another Summicron at a Leica store on my S007. It focused accurately. 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/256088-s-006-focus-error/?do=findComment&comment=2990160'>More sharing options...
alan.y Posted February 15, 2016 Share #40 Posted February 15, 2016 This is a crappy photo and slightly blurry from shake, but I happen to have it on my computer, as opposed to the many other test shots I took. It shows how the Summicron front-focuses at distances much closer than infinity, in this case around 5 meters away. The AF point is around the center, at the postings. But what's in focus is the man's face. 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/256088-s-006-focus-error/?do=findComment&comment=2990172'>More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.