01af Posted March 3, 2013 Share #1 Posted March 3, 2013 Advertisement (gone after registration) The M's new MAX sensor from CMOSIS is said to be much flatter than the M9's CCD sensor, hence less sensitive to oblique rays near the frame's borders with wide-angle lenses. Many are wondering how the new M (Type 240) will fare with old-style super-wide-angle lenses, such as the Super-Angulon 21 mm or the first version of the Elmarit 28 mm, i. e. those with particularly short exit-pupil distances. Well, I don't have a Super-Angulon ... but today I tried a Minolta W.Rokkor-PI 21 mm 1:4.5 lens on a Novoflex LEI/MIN adapter with both my M9 and a borrowed M. This lens was meant for the Minolta SR and SR-T line of SLR cameras from the 1960s and supposed to be used with locked-up mirror and an accessory finder in the camera's shoe. The shots were taken at f/8 and processed with Adobe Camera Raw 7.4 Beta. I tried other f-stops as well but the result is always the same (slightly more vignetting, generally, at full aperture). 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! As you can see, the lens vignettes badly due to the short exit-pupil distance, but in the M9, the manually selected Elmarit-M 21/2.8 profile mostly takes care of that. Some minor magenta cast remains mainly along the frame's bottom and lower left. In the M, however, the vignetting is slightly less to begin with, but a really bad magenta cast along the frame's right-hand edge (plus a lesser cast at the left) totally ruins the pictures. I also tried selecting the "Elmarit-M 21/2.8" code from the M's lens menu—but unlike the M9, this doen't make any discernible difference in the M. While this looks a bit disappointing, please note this is an extreme test scenario. With the Leica Tri-Elmar-M 16-18-21 mm Asph, for example, the M behaves perfectly benign, at least as well as the M9, if not even slightly better (and the new firmware 1.1.0.2 wasn't even applied yet). Do not extrapolate this result to other old-style wide-angle lenses, such as the Super-Angulon 21 mm! With those, the results may be similar or totally different ... well, uh, probably similar but we cannot know for sure before someone actually tried. Further testing required. By the way, the W.Rokkor 21 mm lens is excellent at the frame's center, even at full aperture, even by modern standards, but both illumination and sharpness fall off badly towards the corners. At f/8 and f/11, the corners get pretty good—but still, never match modern 21 mm lenses such as Tri-Elmar-M Asph, Elmarit-M Asph, or Super-Elmar-M Asph. Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! As you can see, the lens vignettes badly due to the short exit-pupil distance, but in the M9, the manually selected Elmarit-M 21/2.8 profile mostly takes care of that. Some minor magenta cast remains mainly along the frame's bottom and lower left. In the M, however, the vignetting is slightly less to begin with, but a really bad magenta cast along the frame's right-hand edge (plus a lesser cast at the left) totally ruins the pictures. I also tried selecting the "Elmarit-M 21/2.8" code from the M's lens menu—but unlike the M9, this doen't make any discernible difference in the M. While this looks a bit disappointing, please note this is an extreme test scenario. With the Leica Tri-Elmar-M 16-18-21 mm Asph, for example, the M behaves perfectly benign, at least as well as the M9, if not even slightly better (and the new firmware 1.1.0.2 wasn't even applied yet). Do not extrapolate this result to other old-style wide-angle lenses, such as the Super-Angulon 21 mm! With those, the results may be similar or totally different ... well, uh, probably similar but we cannot know for sure before someone actually tried. Further testing required. By the way, the W.Rokkor 21 mm lens is excellent at the frame's center, even at full aperture, even by modern standards, but both illumination and sharpness fall off badly towards the corners. At f/8 and f/11, the corners get pretty good—but still, never match modern 21 mm lenses such as Tri-Elmar-M Asph, Elmarit-M Asph, or Super-Elmar-M Asph. ' data-webShareUrl='https://www.l-camera-forum.com/topic/199544-new-m-and-old-super-wide/?do=findComment&comment=2258817'>More sharing options...
Advertisement Posted March 3, 2013 Posted March 3, 2013 Hi 01af, Take a look here New M and old super-wide. I'm sure you'll find what you were looking for!
macjonny1 Posted March 3, 2013 Share #2 Posted March 3, 2013 Thanks for taking the time to post this! Link to post Share on other sites More sharing options...
farnz Posted March 3, 2013 Share #3 Posted March 3, 2013 Does your borrowed M have the firmware update 1.1.0.2 installed that improves vignetting correction for wides and super-wides? Pete. Link to post Share on other sites More sharing options...
UliWer Posted March 3, 2013 Share #4 Posted March 3, 2013 O1af's explanation and the example he shows say it was firmware 1.100. Nonetheless this is astonishing and disappointing news. The sensor's description in the brochure and what was said by Leica spokesmen when the camera was announced implied that colour shift would not be an issue any more. It was said that only vignetting correction was applied by the camera's software and that non-Leica-lenses could be used without restrictions (if one didn't mind vignetting). We know better now. And that the new firmware which is hoped to do better was introduced on last call is no good sign either, since the 'red-corner-problem" should have been on the agenda for the sensor's design from the very beginning. Link to post Share on other sites More sharing options...
algrove Posted March 3, 2013 Share #5 Posted March 3, 2013 I'm just glad the FW update didn't list the 28/1.4 prematurely. Link to post Share on other sites More sharing options...
k-hawinkler Posted March 3, 2013 Share #6 Posted March 3, 2013 Well, it seems backward compatibility with the existing lenses has some negative consequences, It would be interesting to see whether the newer firmware helps with this issue. Also, it appears to me that none of the half dozen or so reviews pointed out this issue. Did I miss one that did? Thanks to 01af for demonstrating this feature. Link to post Share on other sites More sharing options...
rscheffler Posted March 3, 2013 Share #7 Posted March 3, 2013 Advertisement (gone after registration) I don't think you missed one. None of the pre-release reviews I've come across talked all that much about super and ultra wide angle lenses, let alone legacy optics of this nature. Jono's review on his site has some examples from the WATE, 28 Cron and 24 Lux, with great looking results. I think Jono also hinted at getting good results with the CV15. It could be the CMOSIS sensor is better optimized for these newer wide designs and most 'typical' legacy glass, but still suffers under extreme conditions. If you think about the physics a lens such as this Rokkor presents, it would be amazing for any FF sensor to not have problems with it. It will be interesting to see M results with the CV15 and ZM21/4.5, which are probably two of the most colour-shift prone currently available lenses for use on the M9. Link to post Share on other sites More sharing options...
scott kirkpatrick Posted March 3, 2013 Share #8 Posted March 3, 2013 My recollection is that Jono knows that nobody ever got the CV15 to work well on the M9, and he didn't feel it was a fair test of the M. I doubt that he still has one (but I do!). That and the SuperAngulon are the worst known cases for short back-focus. The WATE is quite telecentric, a piece of cake. scott Link to post Share on other sites More sharing options...
sandymc Posted March 3, 2013 Share #9 Posted March 3, 2013 Pretty much all of the CMOS based cameras, e.g., the Sony Nex's, have color shift problems with short rangefinder lenses. No surprise that the M has the same problems. Sandy Link to post Share on other sites More sharing options...
jaapv Posted March 3, 2013 Share #10 Posted March 3, 2013 I guess the best we can -and should- expect is that all of Leica's current lenses and the vast majority of their legacy lenses will work well. Imo it is hardly reasonable to think that really extreme designs from the past and third parties were part of the setting up of the camera. After all, these rare cases can easily be dealt with by flat field correction in post processing. It will be interesting to see the reports coming in from now whether this is the case in real life. Link to post Share on other sites More sharing options...
k-hawinkler Posted March 3, 2013 Share #11 Posted March 3, 2013 Pretty much all of the CMOS based cameras, e.g., the Sony Nex's, have color shift problems with short rangefinder lenses. No surprise that the M has the same problems. Sandy That's a pretty broad brush you are painting with. The NEX-7 (also NEX-6) is indeed much more problematic than the NEX-5N. The NEX-5N, although not perfect, actually does remarkably well with 21, 28, and 35 mm FL. Both NEX-5N and NEX-7 seem to do pretty well with the WATE 16-18-21/4 lens. At least, that's my experience. Nothing is perfect though. I agree with that. Link to post Share on other sites More sharing options...
macjonny1 Posted March 3, 2013 Share #12 Posted March 3, 2013 Big difference with a crop sensor and FF Link to post Share on other sites More sharing options...
k-hawinkler Posted March 3, 2013 Share #13 Posted March 3, 2013 Big difference with a crop sensor and FF Thanks. I agree. Link to post Share on other sites More sharing options...
adan Posted March 3, 2013 Share #14 Posted March 3, 2013 My bet is that the M240 sensor - being touted as the "R Solution" via the live-view and R adapter - required microlens offsets that would be more compatible with the expected use of long (> 200mm) lenses (and thus less compatible with short-focus superwides). However, it could just be that the firmware still needs adjustment (it took - what? - 18 months or more to get the M9 FW dialed in for the 21 Elmarit pre-ASPH, which as 01af points out, is the most extreme correction for red-edge used). Link to post Share on other sites More sharing options...
rramesh Posted March 3, 2013 Share #15 Posted March 3, 2013 I wonder if we need to carry 3 camera bodies for the long, the wide and the B&W? Link to post Share on other sites More sharing options...
UliWer Posted March 3, 2013 Share #16 Posted March 3, 2013 Well, the CV 12mm doesn't seem to be too weird: http://www.l-camera-forum.com/leica-forum/forum-zur-leica-m9/274421-praxisberichte-und-bewertungen-der-m.html#post2332472 Link to post Share on other sites More sharing options...
01af Posted March 3, 2013 Author Share #17 Posted March 3, 2013 My bet is that the M240 sensor—being touted as the "R Solution" via the live-view and R adapter—required microlens offsets that would be more compatible with the expected use of long (> 200 mm) lenses (and thus less compatible with short-focus superwides). In a way, you're right—the MAX sensor has basically no microlens offsets at all, and so it's more compatible with longer exit-pupil distances (which typically come with longer focal lengths). It was written that the sensor's flat architecture plus the high-refractive microlenses make microlens offsets obsolete. Anyway, meanwhile I tried more legacy Minolta Rokkor lenses, including the MC W.Rokkor-NL 21 mm 1:2.8 and the MD W.Rokkor 17 mm 1:4, which both are retro-focus lenses for regular SLR use—and they have no issues at all. No undue local colour casts or colour shifts. So I guess it's safe to assume the new M, with an adapter, really is a good digital full-frame solution for the vast majority of legacy 35-mm SLR lenses (Minolta Rokkor lenses in my case), including retro-focus super-wides. In fact, the old non-retro-focus W.Rokkor-PI 21 mm is the only lens I've found so far which has problems on the new M ... and, by the way, updating the firmware to 1.1.0.2 didn't make any difference for this lens. Link to post Share on other sites More sharing options...
Ecar Posted March 3, 2013 Share #18 Posted March 3, 2013 Just fired a couple of shots of a white wall with the M, the CV15 and the latest FW (1.1.0.2) The red edge is indeed pretty nasty... My CV15 is coded as a WATE. Manually setting the Elmarit 21 code actually seems to make things worse in terms of vignetting, with red edge being only marginally better. Another feature I noticed is that, even if you set lens detection to "Off", the camera seems to pick up the code and apply the correction. In other words, if the lens is coded, this overrides the "Off" menu setting. Must be a bug, as I don't see the purpose of the Detection "Off" menu setting. Could someone please confirm this behaviour? "Manual" lens selection, on the other hand, works as expected and overrides the physical lens code. Addition: no problem whatsoever with the WATE. I'll try the CV21 (a known offender on the M9) next. Link to post Share on other sites More sharing options...
marknorton Posted March 3, 2013 Share #19 Posted March 3, 2013 Thank you, looks as though the CV 15 will remain happiest on the M8. The size of it may rule it out but I am looking forward to trying the Nikon 14mm f2.8. Link to post Share on other sites More sharing options...
Ecar Posted March 3, 2013 Share #20 Posted March 3, 2013 Did a quick and highly unscientific test with the CV 21/4 against the usual white wall trying various codes. The one that seems to work best is that for the WATE. Still quite a bit of red edge, but much better than with the first iteration of the M9 firmware. Whilst we shouldn't expect Leica to come up with a solution for 3rd-party lenses, perhaps a future FW will improve things further. Once I'm back in front of my computer, I will also check the effects of the lens settings in LR 4.4 RC. And there's always CornerFix. Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.