On 12/22/18, Steinar H. Gunderson <steinar+ffm...@gunderson.no> wrote: > On Sat, Dec 22, 2018 at 09:53:16AM +0100, Paul B Mahol wrote: >>> FFmpeg doesn't have a good understanding of gamma (it rarely actually >>> converts between different gamma ramps), but that's not a problem with >>> PhotoCD per se. I can't find a good reason why FFmpeg could not be >>> extended with conversions from the PhotoCD color space to sRGB, at least >>> not if clipping out-of-gamut colors is acceptable. >> I'm all ears. > > I happen to have a library that does all of this stuff... :-) > (https://movit.sesse.net/) > > I don't think FFmpeg really wants to link in Movit for a variety of reasons, > and in this case, PhotoCD is nominally Rec. 709, so you don't actually need > a colorspace transform. This means that the only steps you really need would > be: > > 1. Decode the YCC to RGB. Allow for out-of-0..255 (ideally float, but > FFmpeg probably wants to use int16 or something similar instead). > 2. Convert to linear gamma (either float, or 16-bit fixed point) using the > inverse of the function mentioned in Wikipedia. > 3. Convert back from linear gamma to sRGB or Rec. 709 gamma. > 4. Clip to 0..1, then scale to 0..255. > > Step 2, 3, 4 can probably be collapsed into a LUT that can be applied three > times (the channels are independent, since the RGB color space is the same). > > If you're not happy with blown highlights (colors that clip), it's a much > harder problem.
I can not accept internal conversion to RGB. This is subsampled format after all. _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
