[Bug 739868] rtpmanager: rtpjitterbuffer fixes and improvements

[GStreamer (GNOME Bugzilla)]

https://bugzilla.gnome.org/show_bug.cgi?id=739868

--- Comment #57 from Miguel París Díaz <mparisdiaz at gmail.com> ---
(In reply to Sebastian Dröge (slomo) from comment #55)
> (In reply to Miguel París Díaz from comment #54)
> > (In reply to Sebastian Dröge (slomo) from comment #51)
> > > Comment on attachment 300185 [details] [review] [review] [review]
> > > 0004-gstrtpjitterbuffer-not-calculate-jitter-for-packets-.patch
> > > 
> > > Wim said on IRC that we should handle all packets the same here
> > 
> > I don't agree, because of the case of handling RTX packets(the common
> > problem that we are discussing in other patches).
> > If this is not a good solution, we should look for another one.
> 
> How does this patch relate to any of the others?
I mean that jitterbuffer has problems if manage RTX packets as the "original"
packets.
We have been discussing about how to avoid this case using RFC4588, using other
PT and/or SSRC in the RTX packets, but we have to take into account that there
are clients (as Firefox) that do no do this correctly. So:
 - Could we do something to do GStreamer works fine in these cases?
 - Should we assume this case and keep the current behaviour?

> Wim's argument basically
> was that there is nothing else we can do than to consider RTX packets also
> for the jitter calculations. And it might also be just jitter in the end.
> What exactly would your patch solve? What's the problem you're observing?

This patch aims to avoid using RTX packets for jitter calculation because it is
done basing on the last received packet, and this is not real.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.