I wonder if the confusion comes from the option for compressed NEF as the raw format. The D200 default is uncompressed & lossless but it's easy to change to the just barely lossy compressed option. Compressed in-camera squeezes the 12 bit, 4096 native analog RAW value scale into 683 values, ~9.4 bits but differently allocated. Nearly all the compression loss is in more highly gradated high values. In uncompressed RAW, the top 4 stops use 3840 (2048+1024+512+256) values of the 4096. The remaining 256 values cover the rest of the 8 bits. Compressed NEF's allocate 251 values to the lower 256 (8 stops) and the remaining 432 to the top 4 stops (3860 raw values). The result is just a little less recoverable highlight data -- on average more values per f-stop than the lower range. The loss seems to empirically provable but hardly ever meaningful in the image. I still shoot uncompressed NEFs just in case. The D80, D70, D50 & D40 bodies only have compressed NEF.
http://www.photography-forums.com/t80862-drawbacks-of-compressed-nef-in-d200 .html Bob G -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Berry Ives Sent: Saturday, July 14, 2007 8:05 AM To: [EMAIL PROTECTED] Subject: [filmscanners] Re: color bit depth and digital cameras Dpreview.com's review indicates that it is a 12-bit raw format. ~Berry On 7/13/07 11:27 PM, "David J. Littleboy" <[EMAIL PROTECTED]> wrote: > > From: <[EMAIL PROTECTED]> >>>>>>>>>>>> > > I was just playing with my new Nikon D200 and discovered > something that surprised me. Unless there is some quality > adjustment setting I missed, it's color bit depth apparently is > only 8 bits in NEF Raw. By comparison, my Polaroid SprintScan > 4000 scanner has a color bit depth of 12 bits, and other scanners > have much higher color bit depths than this. While color bit > depth is a commonly cited specification for scanners, I've seldom > seen it cited for digital cameras. Does the lower bit depth for > the D200 imply lower quality color rendition than my 12 bit scanner? > <<<<<<<<<<<<<<<<<<<<<<<<<< > > I suspect you've done something wrong. This reference > > http://www.clarkvision.com/imagedetail/digital.sensor.performance.summary/in de > x.html > > Shows the D200 doing very well indeed at ISO 100. I'm quite sure it uses a > 12-bit A/D converter. > > Note that just because a camera or scanner has X bits in its A/D converter > doesn't mean you have X bits of valid data in the output files. > > David J. Littleboy > [EMAIL PROTECTED] > Tokyo, Japan > > > ---------------------------------------------------------------------------- -- > ---------- > Unsubscribe by mail to [EMAIL PROTECTED], with 'unsubscribe > filmscanners' > or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or > body ---------------------------------------------------------------------------- ------------ Unsubscribe by mail to [EMAIL PROTECTED], with 'unsubscribe filmscanners' or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or body ---------------------------------------------------------------------------------------- Unsubscribe by mail to [EMAIL PROTECTED], with 'unsubscribe filmscanners' or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or body
