On 18 January 2015 at 09:40, Nicolas George <geo...@nsup.org> wrote: > Le nonidi 29 nivôse, an CCXXIII, Hendrik Leppkes a écrit : >> nvenc should behave the same as libx264, or any other video encoder, if >> this patch makes it do that, then it should be applied. >> >> If bt601 needs special handling not yet present in avcodec, it should be >> implemented in such a way that all encoders can benefit from it. > > That was one of the points I was trying to make, thanks for explaining it > more clearly. > > Two additional points: > > According to Philip's tests, nvenc adds a 45/44 = 720/704 factor, not > 720/702 = 40/39, so it does not conform to BT.601 anyway.
Why is it so hard for you to realise that the SARs are not arithmetically derived? > I have downloaded BT.601 from <URL: http://www.itu.int/rec/R-REC-BT.601/ >, > both the most recent and the oldest versions, and I did not find any > explicit reference of this 702 business, nor could I decipher indirect > references, so I still wonder where it comes from exactly. You can make the appropriate calculations *and simplifications* for BT601. There are plenty of books with the appropriate derivations. Kieran _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel