[Live-devel] testRTSPClient does not assemble frames correctly

[BONNEAU Guy]

Gord,

We haven't used the live555 library to parse the mpeg2 capture. We use our own parser implementation. Thus I am afraid that I cannot help much regarding if the live555 could have any issue parsing the incoming Mpeg2 data. Here is some suggestion to help you with debugging your issue.

1- Make sure your network setup doesn't "lose" any IP packets.
2- Use an Mpeg2 stream analyzer to check the conformance of the capture stream and identify what is wrong. I use Elecard.
3- Use VLC to capture the IP stream simultaneously with your own application that uses live555. If VLC capture is fine then it might be an issue with either your application or the live555 library. If the capture is not fine with VLC then the issue is with the encoder (unless packets were lost).
5- To my knowledge a variable GOP size should not be an issue for an Mpeg2 decoder. If it is an issue for the decoder you use then you should switch to another decoder because the one you use is not conformant to the ISO spec.

Hope this help
Guy

|-----Original Message-----
|From: live-devel-bounces at ns.live555.com [mailto:live-devel-
|bounces at ns.live555.com] On Behalf Of Gord Umphrey
|Sent: Friday, April 27, 2012 9:14
|To: live-devel at ns.live555.com
|Subject: [Live-devel] testRTSPClient does not assemble frames correctly
|
|Thank you Guy for the good work with "testRTSPClient heap corruption".
|
|When testRTSPClient successfully receives a stream, we are finding that the
|received frames are not assembled correctly.
|
|For most (if not all to my knowledge) streams, you should have an I frame,
|followed by a fixed amount of P frames, followed by an I frame, etc, etc.
|
|I.e. you should see this:
|
|I frame
|32 P frames
|I frame
|32 P frames
|etc, etc
|
|
|However, we are finding a variable number of P frames.  Sometimes you will
|get an I frame every 10th frame, sometimes every 16th frame, etc.
|I think what's happening is that sometimes multiple P frames are being
|combined into a single P frame.
|
|This variable GOP size is a problem for a lot of decoders.
|
|
|So far we have tested this on 3 different RTSP sources with the same result.
|
|
|Any ideas?
|
|Thanks,
|Gord.
|
|
|_______________________________________________
|live-devel mailing list
|live-devel at lists.live555.com
|http://lists.live555.com/mailman/listinfo/live-devel