Archived posting to the Leica Users Group, 2011/07/22
[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]Paul, I'm told that there is now some noise reduction applied in camera with the DNGs. The black point set by the camera for the DNGs is shifted as the ISO increases. Uncompressed DNGs do better in that regard than the compressed ones. Of course the CMOS sensor output benefits from well developed and sophisticated noise reduction in camera from the firmware in the Nikon and Canon dSLRs too. Cheers Geoff *Australia, paying for the world's carbon sins, one tonne at a time**.* On 23 July 2011 10:26, Paul Roark <roark.paul at gmail.com> wrote: > Michiel Fokkema <michiel.fokkema at gmail.com> wrote: > > > Do you think the M9 high iso perfomance is any good? > > > I find it ironic that Leica uses CCD technology, which excels at its > native ISO, but is not as good at high ISO as CMOS technology. For a > landscape shooter like me, CCD is ideal. But for street/available > light shooters or wherever extremely high speed is needed, CMOS might > be a better solution. > > CCD must do its amplification after the signal is transferred off the > sensor, and that transfer is where lot of noise seems to be acquired. > So, I wondered how much difference there would be between a neutral > gray that was amplified in camera compared to one that is amplified in > Photoshop. To explore this, I set the M9 exposure manually for > neutral gray at 2500 ISO and took a shot of an frosted/opal glass over > the lens (totally smooth, out of focus image). I then moved the ISO > back to 160 and took a shot at the same exposure settings. The 160 > ISO image was near black when initially opened (ACR 3.x, with black > slider all the way to the left). But when curves were used to take > the 160 ISO gray up to the same level as the 2500 ISO gray, the noise > levels in the images were essentially the same. See > http://www.paulroark.com/BW-Info/CCD-iso-v-curves.jpg > > While the in-camera amplifier does ultimately help the image quality > with even darker values, the message from the experiment is, I think, > rather important for those of us who shoot M9s. I don't bracket much > any more. Rather, I set the exposure for the highlights like I used > to do with slide film. I manually "expose right", checking the > histogram often, and just let the low values fall where they may. > Amplification in PS is, over the ranges of values I've recently run > into, good enough that HDR is not needed and would not accomplish all > that much anyway. This is very different than the style than is > needed for CMOS, and I prefer it to HDR and bracketing. > > Paul > www.PaulRoark.com > > _______________________________________________ > Leica Users Group. > See http://leica-users.org/mailman/listinfo/lug for more information >