Lenshacker Posted February 26, 2015 Share #101 Posted February 26, 2015 Advertisement (gone after registration) ^^^ With web sized images, it is not possible to analyze high frequencies and IR contamination defocusing. You must mean the resized-images, as just about any sized image can be processed. All I've done is look at the compressed JPEG's to see which required more space. I would rather have the raw files, but they are not available. The IR contamination problem throws off the full-sized image which is used to generate the interpolated image. I guess we need a set of images with/without IR filters and shot with a non-APO lens. Might try that with my M8. I guess we'll know when the answers are posted. Link to post Share on other sites More sharing options...
Advertisement Posted February 26, 2015 Posted February 26, 2015 Hi Lenshacker, Take a look here CCD vs CMOS: Can you tell which is which?{merged}. I'm sure you'll find what you were looking for!
CheshireCat Posted February 26, 2015 Share #102 Posted February 26, 2015 In that case you are the only one....This has been discussed extensively on this forum over the years and is a common complaint. By M240 users.... An extensive discussion is not a proof. Can you point me to the IR response graphs for both cameras ? Also, how do you explain the much better skin tones in artificial light on the M240 ? Link to post Share on other sites More sharing options...
jmahto Posted February 26, 2015 Share #103 Posted February 26, 2015 .... and the M9 files show more frequency content. But without two images to compare, it is impossible to determine what is different about the image content from the two cameras. .... Could that be because of higher noise (high frequency) in M9 file? Link to post Share on other sites More sharing options...
elmars Posted February 26, 2015 Share #104 Posted February 26, 2015 An extensive discussion is not a proof. Can you point me to the IR response graphs for both cameras ? Also, how do you explain the much better skin tones in artificial light on the M240 ? Thank You for Your questions. Your scepticism is legitimate. Due to my comparisons there ist no difference in the sensor output. The differences are mainly caused by a different white balance or the interpretation of the white balance in LR. See here: http://www.l-camera-forum.com/leica-forum/leica-m9-forum/361314-400-leica-photographers-agree-we-love-2.html#post2889875 Elmar Link to post Share on other sites More sharing options...
thighslapper Posted February 26, 2015 Share #105 Posted February 26, 2015 Thank You for Your questions. Your scepticism is legitimate. Due to my comparisons there ist no difference in the sensor output. The differences are mainly caused by a different white balance or the interpretation of the white balance in LR. See here: http://www.l-camera-forum.com/leica-forum/leica-m9-forum/361314-400-leica-photographers-agree-we-love-2.html#post2889875 Elmar Yes Elmar ....... and I did the same 2yrs ago and posted loads of images with the X-Rite colour chart included ..... and when WB is corrected the results are the same as yours. I agree entirely with Ricks post earlier ...... we argued about the M9/M240 differences at great length whilst doing the beta testing for the firmware updates ....... and essentially the M240 files are superior and that gives you facility to emulate the M9 look ..... whatever that is meant to be...... You can argue as long as you like about the difference between CCD and CMOS output ..... but the fact remains that Leica has engineered the processing so the DNG's end up looking similar and with the best ultimate image quality they can get ....... otherwise there would be even more grumbling when they bring out a new model ...... If Dave's experiment proves anything at all ..... it is how futile this whole exercise is and how little it impacts on the production of real-world images. Link to post Share on other sites More sharing options...
IWC Doppel Posted February 26, 2015 Share #106 Posted February 26, 2015 Harder to tell with just one image of each scene. Two of the images had a unique M9 image size ratio, the rest had the M240 ratio. But- this parameter could easily be modified in post-processing. I just didn't really see familiarity which could be processing, so I made a large assumption Link to post Share on other sites More sharing options...
ndevlin Posted February 26, 2015 Share #107 Posted February 26, 2015 Advertisement (gone after registration) This is a fun thread and thanks to David for the work. However, it proves both everything and nothing. Even in the web jpegs I can see differences in saturation and contrast, as well as overall brightness on a few frames. But the pictures from both are excellent and the differences minimal. On the second part of the test, with single images, there is really no way to tell which camera produce the file. Since this (the web) is the format in which most images are seen today, the answer is that the difference is irrelevant and really doesn't matter. Thus the test is conclusive But the test also proves nothing, because post-processing technique is infinitely variable. Between myself and my retoucher, there is no difference shown on these files that could not be overcome - and seen to be overcome - through skilled and judicious work on the image. Moreover, what any one person likes or feels looks 'right' is also significantly variable. Even on 30" graphics monitors, or large prints, a really well treated final file from either of these camera (and indeed from many of their competitors) will be hard to identify. For people who aren't highly skilled in post-processing, or don't otherwise treat their files in a way to truly maximize them, this whole discussion is kind of moot. It becomes like judging film cameras based on 4x6 prints from a one-hour photo lab. The bottom line is that people simply like, and prefer to work with, the file from one camera or the other. Or, they prefer the physical response and handling of one over the other. Right shall not be proven. Except that Diet Pepsi is better than Coke. Of that there can be no dispute. - N. Link to post Share on other sites More sharing options...
pgk Posted February 26, 2015 Share #108 Posted February 26, 2015 The bottom line is that people simply like, and prefer to work with, the file from one camera or the other. Or, they prefer the physical response and handling of one over the other. Right shall not be proven. Actually ..... I'd prefer the files from either camera to respond in the same way to the same workflow and not have to relearn post processing workflows because of different physical responses and handling as you put it. What I'd really like is consistency from one model of camera to the next, CCD or CMOS. Link to post Share on other sites More sharing options...
BerndReini Posted February 26, 2015 Share #109 Posted February 26, 2015 Sorry, but that's a pipe dream. It just isn't possible unless you use the same sensor. Link to post Share on other sites More sharing options...
peggers Posted February 26, 2015 Share #110 Posted February 26, 2015 By chance I came across this article today, and it rang true about how hard it is to be truly objective, not matter how hard we try... 'You Are Not So Smart': Why We Can't Tell Good Wine From Bad - The Atlantic Link to post Share on other sites More sharing options...
uhoh7 Posted February 26, 2015 Share #111 Posted February 26, 2015 You can dumb down an M240 file to look like an M9 file, but you can't up-res, up-dynamic range or up the color depth of an M9 file to look like the M240. Rick Not my experience or anyone else who's paying attention. You really think you can make the 50 Lux or 28 cron from the 240 look like it came from an M9? I think many would love to hear the steps for that, as untold anguish has ensued from many new 240 owners on the subject. Others like the 240 and I think that's fine. You can quote all the paper numbers you want, I and many others prefer the colors and rendering of the CCD to the 240 CMOS. The so called lack of DR is moot as the shadows from RAW M9s respond very well without making alot of noise. We don't want to edit from one look to the other anyway, even if it could really be done. Do you really understand how different CCD is from CMOS? Or do you just count pixels and read DR numbers? You seem to assume Leica went to CMOS because "it's better". More logical is to assume "it's cheaper". The M9 CCD is a very good one, and I bet the CMOS in the 240 is less expensive. Link to post Share on other sites More sharing options...
tookaphotoof Posted February 26, 2015 Share #112 Posted February 26, 2015 Do you really think Leica buying people wouldn't spend 500 euro extra on a camera body if that would be the result of putting a ccd sensor in it? Link to post Share on other sites More sharing options...
thighslapper Posted February 26, 2015 Share #113 Posted February 26, 2015 Do you really understand how different CCD is from CMOS? Or do you just count pixels and read DR numbers? Pardon? You have photoreceptor with a bayer filter in front ..... with the output as a digital number and the colour value calculated/interpolated by the response of the receptors under the respective colour filters ..... it's the same for both. EVERYTHING ELSE is a result of processing of the data ....... and that data has calibration applied to it to get the final image output that the sensor/camera manufacturer WANTS. This whole CCD/CMOS difference cr*p is a bit like saying CD's of identical recordings pressed by different manufacturers sound different ........ the output read from the disc is digital ..... and does not vary ........ it's what is done by the DAC, amplification and reproduction by the speakers that is responsible for 99% of the perceived differences. Link to post Share on other sites More sharing options...
jaapv Posted February 26, 2015 Share #114 Posted February 26, 2015 Not my experience or anyone else who's paying attention. You really think you can make the 50 Lux or 28 cron from the 240 look like it came from an M9? I think many would love to hear the steps for that, as untold anguish has ensued from many new 240 owners on the subject. Others like the 240 and I think that's fine. You can quote all the paper numbers you want, I and many others prefer the colors and rendering of the CCD to the 240 CMOS. The so called lack of DR is moot as the shadows from RAW M9s respond very well without making alot of noise. We don't want to edit from one look to the other anyway, even if it could really be done. Do you really understand how different CCD is from CMOS? Or do you just count pixels and read DR numbers? You seem to assume Leica went to CMOS because "it's better". More logical is to assume "it's cheaper". The M9 CCD is a very good one, and I bet the CMOS in the 240 is less expensive. Huh? The colour comes from the dyes in the Bayer filter and subsequent interpolation, not from the sensor, which is a monochrom device. http://www.anpdm.com/newsletterweb/43465E437243425B4471454159/47425C4A72444A50407348435143/wn Link to post Share on other sites More sharing options...
Lenshacker Posted February 26, 2015 Share #115 Posted February 26, 2015 Could that be because of higher noise (high frequency) in M9 file? I thought about that- The images that David posted are scaled to 1800x1200, should be enough to smooth over the noise. Again- would be nice to have the DNG files. I've looked at DNG files for the M8, M9, and M Monochrom using a HEX editor, shows numerical values. Link to post Share on other sites More sharing options...
Lenshacker Posted February 26, 2015 Share #116 Posted February 26, 2015 An extensive discussion is not a proof. Can you point me to the IR response graphs for both cameras ? http://www.onsemi.com/pub_link/Collateral/KAF-18500-D.PDF Page 14 for the M9 response curves. Page 15 for the QE curve as a function of illumination angle. I don't know what cover glass was used for the CMOSIS sensor, nor is a data sheet posted. Taking a picture of a Wii light bar would "shed light" on the issue. Link to post Share on other sites More sharing options...
jaapv Posted February 26, 2015 Share #117 Posted February 26, 2015 An extensive discussion is not a proof. Can you point me to the IR response graphs for both cameras ? Also, how do you explain the much better skin tones in artificial light on the M240 ? Could you Dropbox one of the offending files? Link to post Share on other sites More sharing options...
jonoslack Posted February 26, 2015 Share #118 Posted February 26, 2015 You can quote all the paper numbers you want, I and many others prefer the colors and rendering of the CCD to the 240 CMOS. I and many others prefer the colors and rendering of the 240 over the M9 . . . fine but it's not the point What you cannot demonstrate is that this is an empirical difference between CCD and CMOS - it's only a difference between the M9 and the M240 - and it's only a matter of personal preference at that. Link to post Share on other sites More sharing options...
BerndReini Posted February 27, 2015 Share #119 Posted February 27, 2015 I am starting to believe that anybody from either camp that says one camera is significantly better than the other is dillusional. I don't think there is that much difference, for me neither enough to reject the M240 nor enough to upgrade. One and a third stops better high ISO performance in the M240 (2500 usable compared to 1000 on the M9 for me). More IR pollution with the M240. Nicer shutter with the M240. Better operation and battery life but more weight. More post processing at base ISO in daylight with the M240. Dynamic range, detail, sharpness? Completely insignificant difference either way. I know that a lot of people from both sides of the argument will disagree, but it is the kind of hairsplitting that sells more product. Link to post Share on other sites More sharing options...
lct Posted February 27, 2015 Share #120 Posted February 27, 2015 Never got skin tones like this with the M240. Was an early M9 pic though. http://lctphot.smugmug.com/photos/3900877352_wJTGf3J-D.jpg (7 MB file) Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.