Question on pipeline

[Tim Müller]

On Sat, 2015-05-02 at 10:10 +0200, Sebastian Dröge wrote:

> > > gst-launch-0.10 filesrc location=/home/Amazon_1080.wmv \
> > >  ! fluasfdemux name=d ! queue ! fluwmvdec ! ffmpegcolorspace ! queue !
> > > theoraenc \
> > >  ! rtptheorapay ! udpsink  d. ! queue ! fluwmadec ! audioconvert ! queue !
> > > vorbisenc ! rtpvorbispay \
> > >  ! udpsink auto-multicast=true host=239.168.3.10 port=5000
> > > 
> > > Is this the correct pipe line for that?

In addition to what Sebastian said: if you send vorbis or theora over
RTP/UDP, you need to somehow (e.g. via RTSP or by copy'n'pasting caps or
somesuch) transfer set up data to the receiver out of band. Without that
extra data the receiver won't be able to decode the vorbis/theora.

The easiest way to do this is by using gst-rtsp-server, which supports
multicast as well. You can find examples here:
http://cgit.freedesktop.org/gstreamer/gst-rtsp-server/tree/examples

On a side note, gst-libav (in 0.10 gst-ffmpeg) contains windows media
audio/video decoders as well, and there's also an asfdemux in
gst-plugins-ugly.

 Cheers
  -Tim

-- 
Tim Müller, Centricular Ltd - http://www.centricular.com