Closed olebowle closed 8 years ago
There was a difference in the values calculated by wireshark and voipmonitor. After this fix the values of both applications agree.
Hi Ole,
nice catch! Do you have pcap sample which I can see the problem?
2016-05-27 10:34 GMT+02:00 Ole Ernst notifications@github.com:
There was a difference in the values calculated by wireshark and voipmonitor. After this fix the values of both applications agree.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/voipmonitor/sniffer/pull/29#issuecomment-222091383, or mute the thread https://github.com/notifications/unsubscribe/AGt_pxyVZ8fP3SYJlPg-SMf7-_bj6X5Uks5qFqyqgaJpZM4IoSOS .
Best regards Martin Vit
I created one with simulated jitter of 75ms both ways using tc, see jitter.pcap
Thank you, I have manually merged the code. I'm curious how it is possible that this was broken for such long.
due to strong jitter, the last RTP timestamp might be smaller than the current one the subtraction of u_int32_t variables will than underflow leading to incorrect transit value fix this by casting the variable to (double) earlier