[xiph-rtp] My comments

Ramón García ramon.garcia.f at gmail.com
Sat Oct 30 13:34:21 PDT 2004


Hello, I am new to this list. I am developing RTP support for the
Gstreamer media framework. I sent mail to most of you in this list but
received no replay. I have just found this list. Here are my comments
about the proposal that I have just read on this list.

The idea of Tor-Einar of using a separate data stream for codebooks,
though smart and interesting, is too complicated. Most software
frameworks will assume that a single RTP session is all necessary to
support an audio (or video) stream. Thus this method would make it
difficult to support Vorbis in DirectShow and other multimedia
architectures.

With regard to the problem of chaining, I think that this is beyond
the scope of a codec format. Different streams should go in different
RTP sessions, or in the same session with different SSRCs. The Vorbis
packet format should not act as a framework to play different streams,
that is the job of the OGG container for disk files, or of RTP/RTSP
for streams. If one wishes to adapt the stream to bandwidth changes,
as Aaron suggests, that can be done by changing the encoding quality, 
and there is not great advantadge in changing the sample rate. For
files chained by the user, that should be handled by the server like
when the server contains a playlist. This has the inconvenient of not
offering a global seek across the OGG file, but that is good, because
it would be surprising for the user to see a change of title, ...
after seeking.

Finally, I am against including fragmentation of Vorbis frames. The
UDP protocol can fragment packets and reassemble, thus there is no
need to duplicate that functionality.


More information about the xiph-rtp mailing list