RTP headers & data channels question
71 views
Skip to first unread message
israel.r...@gmail.com
Apr 25, 2021, 1:36:43 PM4/25/21
to kurento
Hello,
I have a few cameras that send telemetry data along side with the video feed.
Can I make kurento forward them to the client using data channels?
If im using "useEncodedMedia", do I need to? Will kurento wrap the whole thing or only the video feed?
Since I believe the answer is no (I couldn't find any data about rtpheaders in kurento) since many times rtp headers are very camera specific, and since I don't have a way to explain kurento's Gstreamer how to read the headers and do something about them - is it even possible to use kurento for this?
(I guess I need to read the stream in another process with ffmpeg\gstreamer and deal with the data channel? Though I will lose data-video frame sync)
I don't have any idea how to approach this :P
Thanks!
Juan Navarro
May 14, 2021, 1:10:27 PM5/14/21
to kur...@googlegroups.com
Did you mean that the additional telemetry data is sent inside the
RTP packet headers?
If so, I think that's something difficult to do with Kurento. Because the PlayerEndpoint will establish a reception endpoint, which is a GStreamer RtpBin element, and this will probably extract only the video data, discarding any unknown headers. So I'm afraid these extra RTP headers would get lost upon receiving at KMS.
"useEncodedMedia" will still remove the RTP encapsulation of incoming packets. What this parameter does is that the encapsulation ends right there, at the codec level, instead of also trying to fully decode the media. But, the RTP encapsulation will be removed nevertheless, because it cannot be used for the sending side.
I'm curious though; what kind of software is this camera expected to work with? I can only think that 99% of media players would do the same: extract the media inside RTP packets, and discard any unknown RTP header. So only a specific-purpose software would be able to understand those headers and do "something" with the data they contain.
If so, I think that's something difficult to do with Kurento. Because the PlayerEndpoint will establish a reception endpoint, which is a GStreamer RtpBin element, and this will probably extract only the video data, discarding any unknown headers. So I'm afraid these extra RTP headers would get lost upon receiving at KMS.
"useEncodedMedia" will still remove the RTP encapsulation of incoming packets. What this parameter does is that the encapsulation ends right there, at the codec level, instead of also trying to fully decode the media. But, the RTP encapsulation will be removed nevertheless, because it cannot be used for the sending side.
I'm curious though; what kind of software is this camera expected to work with? I can only think that 99% of media players would do the same: extract the media inside RTP packets, and discard any unknown RTP header. So only a specific-purpose software would be able to understand those headers and do "something" with the data they contain.
--
You received this message because you are subscribed to the Google Groups "kurento" group.
To unsubscribe from this group and stop receiving emails from it, send an email to kurento+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/kurento/a36bc032-30dc-48f6-81a0-bcd36d5e101fn%40googlegroups.com.
israel.r...@gmail.com
May 25, 2021, 6:45:59 AM5/25/21
to kurento
Yeah,
Some cameras producers choose to send data in RTP headers.
If the data arrives as a klv channel (meaning ffmpeg in this case will recognize the data channel), will kurento forward it as a webRTC data channel to the browser?
I wrote my own player, and I need to try making it work with each new annoying camera my customers buy. (And trust me, they are annoying)
Big thanks Juan you are the man :P
Reply all
Reply to author
Forward
0 new messages