[gstreamer-bugs] [Bug 609658] [rtph264depay] doesn't mark output frames as keyframes correctly

GStreamer (bugzilla.gnome.org) bugzilla at gnome.org
Fri Apr 16 03:30:41 PDT 2010


https://bugzilla.gnome.org/show_bug.cgi?id=609658
  GStreamer | gst-plugins-good | 0.10.16

Mark Nauwelaerts <mnauw> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mnauw at users.sourceforge.net

--- Comment #7 from Mark Nauwelaerts <mnauw at users.sourceforge.net> 2010-04-16 10:30:40 UTC ---
(In reply to comment #6)
> Review of attachment 158802 [details]:
> 
> Is this something we should get into the release as well?
> 

I would not mind, probably does not hurt (unless some decoders or so turn out
to be very picky w.r.t. flags).

> ::: gst/rtp/gstrtph264depay.c
> @@ +418,3 @@
> +    return FALSE;
> +
> +  nal_unit_type = (GST_BUFFER_DATA (nal))[4] & 0x1f;
> 
> Doesn't this mean we should check for GST_BUFFER_SIZE (nal) < 5 above?

Right.

> 
> @@ +422,3 @@
> +  /* non-IDR VCL layer NAL considered DELTA */
> +  if (nal_unit_type >= 1 && nal_unit_type <= 4) {
> +    GST_BUFFER_FLAG_SET (nal, GST_BUFFER_FLAG_DELTA_UNIT);
> 
> Can we assume the buffer metadata is always writable? (Do we know the return
> value of _take_buffer() is always metadata-writable?)

Probably not in general, but one would hope/expect so in this case, since we
are taking everything that is in the adapter (and have put _new_and_alloc
buffers into it), but a gst_buffer_make_metadata_writable can be added for good
measure.

-- 
Configure bugmail: https://bugzilla.gnome.org/userprefs.cgi?tab=email
------- 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