On Sun, Apr 5, 2009 at 6:04 PM, Hadmut Danisch <had...@danisch.de> wrote: > Hi, > > just a proposal for an improvement:
This lets get a few things clear. UFRaw's "Color Matrix" is in fact a very small camera specific color profile. This color profile was made by Adobe. They are usually of good quality. User made (downloadable) profile are of greatly varying quality. Quality in the sense of color profiles is pretty weird to talk about in the first place. Accuracy and Prettyness, and likeness to camera JPEGs are often three different and conflicting goals. > Pictures get significantly better if a specific icm file is used to > convert the camera color > space to e.g. sRGB. Unfortunately, you have to select the icm file, set > the required gamma > and linearity parameters every time. Well, first this is not always true. Color Matrix is usually pretty good. People are often disappointed because default gamma (0.46) and linearity (0.10) are not "right" for their camera. Well at least when comparing to in camera generated JPEGs. I've often given the example of my 400D, which looks pretty close to camera JPEGs when using Color Matrix with gamma 0.35 and linearity 0.08. Everybody can obtain similar results by fiddling with the gamma/linearity while comparing with an in camera generated JPEG or the raw file in question. > Would be nice if ufraw would support those settings on a per camera base > (where the camera > model is given in the raw file), i.e. to allow different default > settings for different cameras, > e.g. when processing pics from a Sony, then automatically apply the > settings chosen for the > sony, and when processing pics from a nikon, use different settings. The problem with this is that gamma/linearity is not camera specific. Gamma and linearity are color profile specific (with each "Color Matrix" counting as an individual profile). The only feasible solution, would be to have database of gamma/linearity settings for all the Color Matrices... But there are practical issues. First there are a lot of camera's. Next, who to trust about these values? What would we do with camera's where no better default has been determined? Keep them at 0.45/0.10? For ICC profiles, there is no way to "know" what gamma/linearity to use except for the author's documentation. In essense, a "good" profile should have been built for use with gamma 1 (at which linearity doesn't matter anymore). Profiles which are built for a gamma other than 1, are (technically speaking) second rate to begin with. Creating a database of gamma/linearity settings for all camera ICC profiles out there would be insurmountable... > Basically, would require a table in the default settings file, or, > alternatively, have a different > ~/.ufrawrc for each kamera model. The ~/.ufrawrc file is for user settings. When you load an ICC file, and (if needed) assign it a gamma/linearity (and save a raw). The profile with it's settings are kept for future quick and easy use. So once UFRaw plus the profiles for you camera are setup, you should never have do bother with this again. Regards, Pascal de Bruijn -- http://blog.pcode.nl/ http://photos.pcode.nl/ ------------------------------------------------------------------------------ _______________________________________________ ufraw-devel mailing list ufraw-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/ufraw-devel
