[FFmpeg-devel] xvYCC Conversion to Wide Gamut RGB with FFmpeg

[Thomas Worth]

On Sat, Apr 2, 2011 at 7:33 AM, Michael Niedermayer <michaelni at gmx.at> wrote:
> On Sat, Apr 02, 2011 at 04:32:00PM +0200, Michael Niedermayer wrote:
>> On Fri, Apr 01, 2011 at 09:28:16PM -0700, Thomas Worth wrote:
>> > On Fri, Apr 1, 2011 at 8:45 PM, Kieran Kunhya <kieran at kunhya.com> wrote:
>> > >> Are you sure Canon DLSRs use xvYCC? I thought this was a
>> > >> Sony thing.
>> > >> If you want to check you can download our tool, 5DtoRGB,
>> > >> and convert a
>> > >> clip using "none" as the decoding matrix. This will show
>> > >> you raw pixel
>> > >> values for Y, Cb, and Cr copied straight from the H.264
>> > >> file prior to
>> > >> being decoded. I don't know if Canon use values outside
>> > >> 16-240 for
>> > >> chroma, but it's possible. Some tests with a color chart
>> > >> and jacking
>> > >> up the saturation in a Picture Style may help determine
>> > >> this.
>> > >
>> > > There's an SEI if I remember rightly that has all this information. The H.264 full_range_flag is usually incorrect.
>> >
>> > Does the full_range_flag setting affect the actual decoding of the
>> > stream? In other words, will pixel values be different off the decoder
>> > if this is 1 vs 0? I'm just talking about the decoder, not any chroma
>> > reconstructing or matrix decoding.
>>
>> ffmpegs h264 decoder returns whats stored thus theres no change in the
>> returned YUV values from that flag but the decoder might choose to
>                                             ^^^^^^
> i meant of course player
>
>> interpret them differently

Ok, so just to confirm, the H.264 decoder doesn't care what
full_range_flag is for its own decompression, it will just return
whatever YUV values are stored in the file. full_range_flag is only
useful to the scaler / player. Correct?