2000.10.01 15:43 "Gamma Correction for LogLuv Format, why?", by Stefan Maass
there is a gammacorrection in the XYZ->RGB conversation when reading a tifflogluv file and asking for RGB values (file tif_luv.c). What is the reason for that? I could't find a gammacorrection in the writing case.
I have a second question about the LogLuv format: Converting from XYZ -> RGB (file tif_luv.c) the used transformation matrix is the same as in the article from Greg Ward Larson: "LogLuv Encoding for Full-Gammut, High-Dynamic Range Images" (journal of graphics tools, Vol3, Nr.1) But the maticies XYZ->RGB and RGB->XYZ are not fully inverse each to another. So if a have a rgb-image as as result of global illumination computation, and writing it as a Tiff-LogLuv file with the proposed RGB->XYZ matrix and open it again with libtiff, i will make a little error. Is this true?
Many Thanks in advance