On Sun, Sep 15, 2024 at 7:41 PM Terje J. Hanssen via Cin < [email protected]> wrote:
Den 15.09.2024 18:02, skrev Andrew Randrianasulu via Cin:
On Sat, Sep 14, 2024 at 1:08 PM Andrea paz via Cin <[email protected]> wrote:
Sorry Phyllis, I have tried to follow the various posts about ffmpeg color problems, but I can't figure out how it works. I get the impression that even the ffmpeg developers are not clear on how to handle color in their program. I tried reading Poynton's excellent book, recommended by Andrew, a couple of years ago but stopped early because it was beyond my capabilities. there is simpler colorful intro from around 2006
https://tech.ebu.ch/docs/techreview/trev_301-gierlinger.pdf
There are too many variables, beyond the problems of ffmpeg: lack of color management (which I think all reduces to an initial transcode to the absolute X,Y,Z color space from which then the program could do all the transformations it needs without too much loss of color info), lack of support for ICC profiles, not knowing how each and every function and plugin of CinGG treats colors, interaction between CinGG and ffmpeg, and who knows what else. As Adam said, “There is no color management workflow.” It is no coincidence that colorimetry in editing programs is found only in the big, rich, specialized companies. -- Cin mailing list [email protected] https://lists.cinelerra-gg.org/mailman/listinfo/cin
Just in case something here is not already found: https://trac.ffmpeg.org/wiki/colorspace
https://medium.com/invideo-io/talking-about-colorspaces-and-ffmpeg-f6d0b037c...
https://kdenlive.org/en/project/color-hell-ffmpeg-transcoding-and-preserving... In this specific case I found following sentence helpful^ ======== n 1928, the mineralogist Siegfried Rösch described a body (Fig. 6), whose surface is constructed of optimum colours. Optimum colours are those among the colours (trichromatic), which are the brightest of the same chro- maticity and the most satu- rated of the same brightness. The base of this body is the diagram published by the CIE in 1931. It was constructed as the result of expansive tests with emmetropes. This diagram enables calculation of the position of every colour in relation to its primary colour. *This two-dimensional diagram is only valid for onerespective colour temperature,* and it is a section of a funnel-shaped body (Fig. 7) ===== bold by me. I do not think I've encountered this bit of info/history before.
-- Cin mailing list [email protected] https://lists.cinelerra-gg.org/mailman/listinfo/cin