[Bug 732266] vaapidecode: h264: add support for decoding SVC base layers only
GStreamer (GNOME Bugzilla)
bugzilla at gnome.org
Tue Aug 22 21:13:31 UTC 2017
https://bugzilla.gnome.org/show_bug.cgi?id=732266
--- Comment #36 from sreerenj <bsreerenj at gmail.com> ---
(In reply to Orestis Floros from comment #34)
> Created attachment 358103 [details] [review]
> libs: decoder: h264: decode SVC base view only
>
> (In reply to sreerenj from comment #33)
> > Review of attachment 358041 [details] [review] [review]:
>
> > @@ +1941,3 @@
> > GST_DEBUG ("decode subset SPS");
> >
> > + /* Don't replace. Could overwrite needeed SPS's. */
> >
> > Bit confused with this comment. I understood that in base-only case, we
> > don't have to store the subset_sps. But when can this "Overwrite needed sps"
> > happen?
>
> spec says:
>
> > NOTE 7 ? A decoder must be capable of simultaneously storing the contents of the sequence parameter sets and subset sequence
> parameter sets for all values of seq_parameter_set_id. The content of the
> sequence parameter set with a particular value of
> seq_parameter_set_id is overwritten when a new sequence parameter set NAL
> unit with the same value of seq_parameter_set_id is
> received, and the content of the subset sequence parameter set with a
> particular value of seq_parameter_set_id is overwritten when
> a new subset sequence parameter set NAL unit with the same value of
> seq_parameter_set_id is received.
>
Here the spec clearly saying that "SPS should be replaced by another SPS with
the same id" and "SUBSET_SPS should be replaced by another SUBSET_SPS with the
same id".
Am I misinterpreting something here? :)
> Eg in SVCHS-2.264:
> nalu type=7 has seq_parameter_set_id=0 and next nalu type=15 has
> seq_parameter_set_id=0.
Hm, I am afraid we need some serious changes in
codecparser/videoparser/decoders since the GstH264NalParser is not keeping
separate array for sps and subset sps.
Is it the only sample which exhibiting this behavior?
--
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.
More information about the gstreamer-bugs
mailing list