I'm trying to play a stream from an RTSP server, which has elementary audio and video on separate rtp channels.

The stream plays fine in all cases when using a static pipeline, like this one:

gst-launch-1.0 rtspsrc location=rtsp://127.0.0.1:53923/211790253373785 name=src
  src. ! rtph264depay ! h264parse ! avdec_h264 ! videoconvert ! xvimagesink
  src. ! rtpmp4gdepay ! aacparse ! avdec_aac ! audioconvert ! pulsesink

However playback does not start until audio is received when using playbin. The problem is that this stream has sparse audio, sometimes the server does not stream any audio, while audio is still advertised in the SDP. rtspsrc only calls gst_element_no_more_pads until all pads are exposed in the rtpmanager element, which only happens after data has been received on all input pads. Playbin does not preroll until no_more_pads has been called.

Is there an easy way to fix this ? I've tried calling no_more_pads very early in rtspsrc but if I do that no video nor audio is played.

Thanks