helged Posted December 13, 2015 Share #21 Posted December 13, 2015 Advertisement (gone after registration) It does not disappear with the 280 Apo Telyt, though. Correct. Attached are three images, SL+280 APO f4 (shoot at f5.6, +2/3 exposure compensation, sky at infinity, directly from LightRoom), in order: Lens code ACTIVATED and with Leica R-M adapter Lens code OFF and with Leica R-M adapter Lens code OFF and with Novoflex R-M adapter Also note that the dark corners in the image with the Novoflex adapter are slightly less pronounced than those with the Leica adapter. Please resolve, Leica! 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/254259-r-m-l-adapters-stacked/?do=findComment&comment=2947845'>More sharing options...
Advertisement Posted December 13, 2015 Posted December 13, 2015 Hi helged, Take a look here R-M-L Adapters stacked. I'm sure you'll find what you were looking for!
scott kirkpatrick Posted December 13, 2015 Share #22 Posted December 13, 2015 I rechecked the white wall test with apo 90 summicron and this time used an exposure shift of +1 (a Grey wall test, not a white wall). This time I do see an overcorrected effect at the edges. I'll post pictures, using in-camera JPEGs, reduced in size. I"m using the stacked Leica adapters. Here f/2 L1000207 by scott kirkpatrick, on Flickr next f/4 L1000209 by scott kirkpatrick, on Flickr and f/8 L1000211 by scott kirkpatrick, on Flickr scott Link to post Share on other sites More sharing options...
ramarren Posted December 13, 2015 Share #23 Posted December 13, 2015 This is what I'm seeing with the pre-APO Summicron-R 90mm f/2. I made three exposures (ISO 200, manual exposure to +1.7EV) at f/2, f/5.6, and f/16 with the Lens Code 11254 enabled and then without. For this lens, and with this lens code, it appears that the SL's processing reduces the edge darkening when wide open and otherwise does little at the other two lens openings. Note that for the f/5.6 exposures, I increased exposure by .3EV when processing the raws to obtain the same white value on the central clouds as I was slightly off on my metering. No LR "Lens Correction" processing was added for any of these images. Link to post Share on other sites More sharing options...
scott kirkpatrick Posted December 13, 2015 Share #24 Posted December 13, 2015 Shooting the sky allows the clouds to move around from one picture to the next. Shooting an interior wall does defocus the lens but still features of the wall can distract you. The best way to do this kind of test is with a very uniform sheet of diffuser material, such as Phase One kits include for lens cast corrections with their digital backs, and a uniform external light source. I'll see if I can find my diffuser sheet. I just used jpegs out of the camera to avoid introducing any manipulations. When Sean Reid does these tests he uses the raw files, and color corrects each file to be 180,180,180, a standard neutral value, at the center. I believe each lens in the list gets its own table, since the area in the firmware is not expensive, and table-lookup is fast, so what is working on one lens might not be correct for another very similar model. Still checking... scott Link to post Share on other sites More sharing options...
ramarren Posted December 13, 2015 Share #25 Posted December 13, 2015 Shooting the sky allows the clouds to move around from one picture to the next. Shooting an interior wall does defocus the lens but still features of the wall can distract you. The best way to do this kind of test is with a very uniform sheet of diffuser material, such as Phase One kits include for lens cast corrections with their digital backs, and a uniform external light source. I'll see if I can find my diffuser sheet. I just used jpegs out of the camera to avoid introducing any manipulations. When Sean Reid does these tests he uses the raw files, and color corrects each file to be 180,180,180, a standard neutral value, at the center. I believe each lens in the list gets its own table, since the area in the firmware is not expensive, and table-lookup is fast, so what is working on one lens might not be correct for another very similar model. Still checking... I've shot walls, diffusers, et al. I like using the sky for long lenses like this ... I want the lens at infinity and whatever it's imaging to be sharp. I find diffusers give me incorrect readings. A clear blue sky is best, but I haven't seen one of those in this neighborhood for a while. The photos I posted, to me, show how this lens behaves with and without the lens code processing very clearly. It's a real target with real features. Link to post Share on other sites More sharing options...
IkarusJohn Posted December 13, 2015 Author Share #26 Posted December 13, 2015 As I wrote before, on the R lenses profile there are two 180 APO Elmarit f2.8 listed. One with model number 11273 and the other with number 11357. My model is the 11273, first series, which does not accept the 1.4X converter. The model 11357 Should be the one that does accept the converter. I also had the same halo problem that however disappeared when I chose the correct profile for my APO 180 Elmarit first series model 11273, which does not accept the converter. I frankly do not understand why nobody is trying to choose the correct model of lens. Cheers Massimo I will give that a try, Massimo - thank you. Link to post Share on other sites More sharing options...
Likaleica Posted December 13, 2015 Share #27 Posted December 13, 2015 Advertisement (gone after registration) I rechecked the white wall test with apo 90 summicron and this time used an exposure shift of +1 (a Grey wall test, not a white wall). This time I do see an overcorrected effect at the edges. I'll post pictures, using in-camera JPEGs, reduced in size. I"m using the stacked Leica adapters. Scott, this is very similar to what I saw with the apo 90. I also saw a lot of coma near the periphery and am wondering if you have noticed this as well. Link to post Share on other sites More sharing options...
FMB Posted December 14, 2015 Share #28 Posted December 14, 2015 It seems to me, after having read all these posts, that my first conclusion about the problem was really in some manner a software issue and belongs to Leica update the old profiles for the R lenses, clearly surpassed when we have used in a massive way the SL camera with these ancient lenses. Francisco. Link to post Share on other sites More sharing options...
scott kirkpatrick Posted December 14, 2015 Share #29 Posted December 14, 2015 I did a careful test today and found problems with the profiles for the apo 90 (as already reported), the apo 100 macro, and the super-elmarit-R 15. With new firmware coming out today or shortly, I'll have to check to see if these have been corrected, but will post their status either way. For the moment, lens ID should be set to OFF in firmware 1.1 for those three lenses. scott Link to post Share on other sites More sharing options...
helged Posted December 14, 2015 Share #30 Posted December 14, 2015 It would be helpful if we could collect easily accessible information regarding those R-lenses (or other Leica lenses) that have a working/non-working profile on the SL (with the new firmware, if any changes there). Perhaps a table organised by the moderators (?), in the first entry of one/this thread (?), like: APO-Telyt-R 280mm f/4: Lens profile OFF. Link to post Share on other sites More sharing options...
scott kirkpatrick Posted December 14, 2015 Share #31 Posted December 14, 2015 It would be helpful if we could collect easily accessible information regarding those R-lenses (or other Leica lenses) that have a working/non-working profile on the SL (with the new firmware, if any changes there). Perhaps a table organised by the moderators (?), in the first entry of one/this thread (?), like: APO-Telyt-R 280mm f/4: Lens profile OFF. The firmware developers really have to do their own measurements, as they probably have programs to generate the necessary tables, and have some standard setup to generate images from which to drive the programs. We can just flag the problem lenses. I'll post some test frames to show what goes wrong with the three lenses I mentioned. I cannot see any differences in firmware 1.2 from 1.1. Also with the 35-70/4 Vario-Elmar-R, the focal length still displays as 0 mm, so I think this update deals with older and more serious problems (like fixing the log-C profile for 8 bit video recording). I will be looking to see what is meant by "R lenses displaying properly in the EXIF". scott Link to post Share on other sites More sharing options...
helged Posted December 14, 2015 Share #32 Posted December 14, 2015 Sure, lens corrections is the work of Leica. But it would be good to have a list of the good and problematic lens profiles so that we know - up front - whether to activate the embedded lens profiles or not. Its frustrating to use embedded lens profiles if it turns out that these deteriorates the image (with a halo, for instance). We all can, of course, test on an individual basis, but as a user community we can help each other - and possibly also put some gentle, coordinated pressure on Leica... Link to post Share on other sites More sharing options...
scott kirkpatrick Posted December 14, 2015 Share #33 Posted December 14, 2015 That's the idea. Here are some results: Here's the grey wall (actually indirect daylight seen through a sheet of diffuser material) for the APO-90 R at f/2, with no corrections, and then at f/8: L1000226 by scott kirkpatrick, on Flickr L1000228 by scott kirkpatrick, on Flickr If I turn lens ID on (so that corrections are applied), the same two images become L1000223 by scott kirkpatrick, on Flickr L1000225 by scott kirkpatrick, on Flickr Also, someone asked about the APO-100/2.8. Without lens corrections there is a little bit of vignetting wide open with this lens, and from f/5.6 upwards the illumination is very even. There are no color cast problems. But when you turn lens id on and correct in the SL, things get very strange. First f/2.8 and then f/11: L1000247 by scott kirkpatrick, on Flickr L1000249 by scott kirkpatrick, on Flickr The macro Elmarit-R 60 is bulletproof. It does fine with and without lens corrections. enough for now. scott Link to post Share on other sites More sharing options...
jonoslack Posted December 14, 2015 Share #34 Posted December 14, 2015 HI There Scott This is interesting - very - and suggests more work needs doing. What difference does distance make? It's often ignored during these tests, but in my experience it's really important. Link to post Share on other sites More sharing options...
scott kirkpatrick Posted December 14, 2015 Share #35 Posted December 14, 2015 HI There Scott This is interesting - very - and suggests more work needs doing. What difference does distance make? It's often ignored during these tests, but in my experience it's really important. Yes, There is a lot of work in this, and I can see that it might come late in the development schedule. I've kept the lenses focused at infinity. I would expect that focusing closer (moving the lens further from the chip) might reduce this kind of problem. But the SL doesn't know what distance you are focused at (or the focal length you have an R zoom set at -- at least not yet). These are over-corrections, as if someone wanted to remove vignetting when the lens if focused at infinity wide open. There is design data from which you could even calculate the parameters without measuring anything, as can be done for the lens distortion corrections. That would be a mistake, as was found with the M's in camera lens profiles. Here, with the apo-90, it looks as if the correction tries to remove vignetting at the corners wide open (they are the same brightness as the center of the frame) but applies the correction too late, leaving a dark band further in. Then if that correction is also used as the lens is stopped down, the overcorrection leaves the corners brighter than the center. BTW, the nice little 35-70/4 Vario-Elmar benefits from the lens corrections, although the SL could do better if it could be told the focal length in use. When this lens is at 35 mm, the uncorrected image is a bit green, although vignetting is not too bad. The corrections take that out, but probably leave longer focal lengths a little warmer than is natural. Did you see some of this during the summer, when camera and firmware were pretty immature? And feed back corrections on the lenses that you used? scott Link to post Share on other sites More sharing options...
scott kirkpatrick Posted December 14, 2015 Share #36 Posted December 14, 2015 I do have a day job but tomorrow I will try to see what I can find out about the R15/2.8. Its front element and fixed petal lens shade make it impossible to work with a diffuser, but some pictures I have gotten with it show light patches around the edges -- could also be flare. And the APO 180/2.8 and 280/4. My 180 is also the early one -- I wonder why they behave so differently. scott Link to post Share on other sites More sharing options...
ramarren Posted December 15, 2015 Share #37 Posted December 15, 2015 Scott, These problem examples are of a diffuser screen set, presumably, some few feet from the camera while you're focusing on infinity. Have you tried shooting the diffuser screen with the focus set to the screen's distance? What I've found when doing other kinds of testing like this with white walls was that even if the camera doesn't know anything about the focus distance, the corrections gave false "bad" results if there was a major discrepancy between the actual distance to target and the set focus distance. This is why when I test at infinity for lens corrections, I point the camera at the sky; if I use a diffuser screen or white wall, I set focus to that distance. Link to post Share on other sites More sharing options...
scott kirkpatrick Posted December 15, 2015 Share #38 Posted December 15, 2015 Godfrey, I am using the procedure that Phase One recommends for lens cast calibration with technical cameras and digital backs. I put a sheet of translucent plastic (1/8" thick, about 5x5 inches in extent) in contact at the end of the lens shade, so that the light entering the lens is a directionless even glow, diffused from the daylight outside. scott Link to post Share on other sites More sharing options...
ramarren Posted December 15, 2015 Share #39 Posted December 15, 2015 Godfrey, I am using the procedure that Phase One recommends for lens cast calibration with technical cameras and digital backs. I put a sheet of translucent plastic (1/8" thick, about 5x5 inches in extent) in contact at the end of the lens shade, so that the light entering the lens is a directionless even glow, diffused from the daylight outside. I'm not sure what the point of doing that is. Light passing through a lens might form a different image when radically out of focus vs in-focus, and the corrections presume an in-focus image I would think. Regardless, I did a quickie test set up as you described and my pre-AP0 90 images with exactly the same qualities, both with and without lens code, as it does at infinity focus on the clouds. I'm not going to worry about using the lens code for my lenses, all of which seem to have excellent processing with the lens code enabled. Link to post Share on other sites More sharing options...
scott kirkpatrick Posted December 15, 2015 Share #40 Posted December 15, 2015 I'm not sure what the point of doing that is. Light passing through a lens might form a different image when radically out of focus vs in-focus, and the corrections presume an in-focus image I would think. Regardless, I did a quickie test set up as you described and my pre-AP0 90 images with exactly the same qualities, both with and without lens code, as it does at infinity focus on the clouds. I'm not going to worry about using the lens code for my lenses, all of which seem to have excellent processing with the lens code enabled. I'm comfortable with the physics argument for using a diffuse light source at the mouth of the lens if the question is with what pattern the lens delivers light to the imaging chip. For Jono's question -- does it matter if the lens is set at infinity or closer -- I think the answer is yes, but it will take some more work when I have the time. In the mean time, here is a sky shot series with the APO R 90/2, at f/2, 4, and 8. I think it is clear that this lens is being handled differently than your pre-APO 90 R. L1000276 by scott kirkpatrick, on Flickr L1000277 by scott kirkpatrick, on Flickr L1000278 by scott kirkpatrick, on Flickr So for now, I recommend not using the lens ID for the two APO teles, the 90 and the 100, until Leica fixes this. scott Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.