[Libav-user] Parsing RTP packets to decode the payload

Denis Gordenin pointbox at mail.ru
Fri Feb 10 14:49:14 EET 2017

>2017-02-10 4:46 GMT+01:00 Denis Gordenin < pointbox-at-mail.ru at ffmpeg.org >:
>> 2017-02-09 18:39 GMT+01:00 Denis Gordenin < pointbox-at-mail.ru at ffmpeg.org >:
>>> I want to decode video info that coming decoding with a codec. No matters
>>> which codec is. The problem is parsing the RTP packets that contain the
>>> video. My application connects to a camera using RTSP
>> Why don't you feed libavformat with the rtsp url?
>> In this case ffmpeg will connect with a camera using RTSP?
>I hope so, if not there may be an (unknown) bug.
>> The RTSP exchange already made using low level sockets for
>> optimising speed of our application.
>Apart from the fact that rtsp should have a negligible impact on
>video decoding time: Do you have any indication that FFmpeg's
>rtsp implementation is slow?
Of course, ffmpeg's implementation shouldn't be slow because there are few RTSP packets in a session.
But we want to control the RTP traffic. I hope it's implementation is not slow, but we could make it more fast for our specifics. In this case we have to have our low level sockets and make their infrastructure. But if I use RTSP in ffmpeg I have to use RTP sockets also there.

>Carl Eugen
>Libav-user mailing list
>Libav-user at ffmpeg.org

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://ffmpeg.org/pipermail/libav-user/attachments/20170210/10032ac3/attachment.html>

More information about the Libav-user mailing list