I have a capture and it appears if the Timestamp within the RTP packet is too large that correct delay between frames is not kept intact when converting to .au or when using Wireshark's playback feature. Normal voice does not seem to have this problem, but pre-recorded speech playback in an automated system seems to have this issue. I have a small file I can share to illustrate what I'm talking about, but I have an 8second capture. Between the 2nd second and the 7th second, there is a time-gap of 4,856 ms (4.8 seconds) that the RTP stream analysis properly detects. However, when the file is converted to an .au file or the playback feature is used, the resulting audio and RTP-Player window shows the entire length as only 4 seconds long (rather than just over 8 seconds). Is there anyway to have Wireshark properly account for the gap in between frames using either the timestamp w/i the packets or the timestamp of the received frame? asked 16 Nov '10, 13:16 TaddyOH |