Le lundi 20 juillet 2020 à 10:40 +0200, Philipp Zabel a écrit : > On Fri, 2020-07-17 at 11:56 -0400, Nicolas Dufresne wrote: > > Le vendredi 17 juillet 2020 à 10:14 +0200, Philipp Zabel a écrit : > > > On Fri, 2020-07-17 at 00:49 -0300, Ezequiel Garcia wrote: > > > > From: Nicolas Dufresne <nicolas.dufresne@xxxxxxxxxxxxx> > > > > > > > > The CODA960 manual states that ASO/FMO features of baseline are not > > > > supported, so for this reason this driver should only report > > > > constrained baseline support. > > > > > > I know the encoder doesn't support this, but is this also true of the > > > decoder? The i.MX6DQ Reference Manual explicitly lists H.264/AVC decoder > > > support for both baseline profile and constrained base line profile. > > > > Hmm, double checking, you are right this is documented in the encoding tools > > sections, not the decoding. But there is extra buffers that need to be passed > > for ASO/FMO to work, I greatly doubt you have ever tested it. > > And you are correct, I don't think I use any test streams that have > ASO/FMO enabled. > > > This is not supported by GStreamer parser, or FFMPEG parsers either. > > Again, we need to make sure in V2 that encoding and decoding > > capabilities are well seperated. > > > > As for advertising ASO/FMO, I can leave it there, but be aware I won't be > > testing it. I can provide you links to streams if you care (they are publicly > > accessible throught the ITU conformance streams published by the ITU). > > That would be welcome. https://www.itu.int/net/ITU-T/sigdb/spevideo/VideoForm-s.aspx?val=102002641 Notably, if you download the AVCv1 series, there is at least FM2_SVA_C.zip that uses FMO (slice groups). I haven't looked them up, I barely started harnessing it for GStreamer. You can find the link to everything else here, including HEVC. https://www.itu.int/net/ITU-T/sigdb/spevideo/Hseries-s.htm > > > But as for GStreamer and FFMPEG, this is not supported anyway. > > Ok, how about changing the commit message to say that this is > unsupported for the encoder and untested for the decoder because there > is no userspace support? Agreed. > > regards > Philipp
Attachment:
signature.asc
Description: This is a digitally signed message part
