Hi guys,
I'm very new to Kurento but have got a few demos up and running and also studied the docs a fair bit.
I'm however stuck trying to get a custom RTP stream going into Kurento via the RTPEndpoint.
I'm using the
rtp-receiver as this mimics very closely what I want to achieve. But when I switch to an RTP source other than the demo GStreamer code I'm hitting brick walls.
In case of unidirectional connections (i.e. only one peer is going to send media), the process is more simple, as only the emitter needs to process an SDP. On top of the information about media codecs and types, the SDP must contain the IP of the remote peer, and the port where it will be listening. This way, the SDP can be mangled without needing to go through the exchange process, as the receiving peer does not need to process any answer.
Does this imply I can just blast RTP at the endpoint and it should work?
I switched on debugging for the logs and my logs are filling up with these for every packet i send:
2019-01-11T17:45:18,173903 26295 0x00007f180d7fa700 warning rtpsource rtpsource.c:983 calculate_jitter() cannot get clock-rate for pt 100
2019-01-11T17:45:18,173988 26295 0x00007f180d7fa700 warning rtpjitterbuffer gstrtpjitterbuffer.c:2833 gst_rtp_jitter_buffer_chain() <rtpjitterbuffer20> No clock-rate in caps!, dropping buffer
I then
found this bug from over 10 years ago which has left me rather confused as it looks to be the same bug I'm facing. It implies specifying the payload type for h264 should work (which I don't mind doing) but I've really no idea how to do so. The documentation only really shows broadcasting RTP from GStreamer and I don't know what it's doing behind the scene to establish the connection.
Has anyone achieved using a custom RTP connection connecting to an RTPEndpoint? If so, I'd much appreciate a simple outline of the process. Been stuck on this for three days now!
All the best!
John