[MEncoder-users] mencoder finds duplicate start-frames in openRTSP MPEG4 captured avis

[ted morris]

I updated this post with some more info I learned very recently --
This might be slightly 'off-topic' for the mencoder group and I have
also cross posted in the live555 forum as well.

Per problems below with capturing mpeg4 through RTP/RTSP packets
with my Sony SNC-CH240 high res Net camera--

I discovered the issue is that there are vop start frames that are sent
within
the RTP payload immediately preceding vop I-frame group type frames.  I
looked at openRTSP code and what *might* be happening is that openRTSP
is assigning a time-stamp to the start-frame type, then assigning the
*next*
timestamp to the I-frame. The source code in openRTSP indicates it can only
handle one vop frame type per packet payload. :-?
There appears to possibly be a dropped frame between
the duplicate+ I-frame to preserve the desired framerate (15 fps). Also--
and
I now know this for a fact -- the camera does not send any vop
end-of-sequence
frames; only continuous (!) vop start-frames.

So I'm questioning if openRTSP can even handle streams like this or not.
Would anyone out there be able to suggest or recommend RTSP capture software
alternative that can run on linux to openRTSP?

 If there is a way to deduce if
a frame is actually there but being tossed out during decoding the MPEG4
with
mencoder, I would very much like to hear any suggestions from the group.
Does mencoder currently utilize openRTSP libraries?

thanks,
ted