Problems while converting RTP streams in to audio files

asked 2020-06-17 07:57:04 +0000

updated 2020-06-17 10:39:59 +0000

grahamb gravatar image

Lately i´ve been having problemas while converting RTP streams into audio files. When wireshark shows me the audio it appears to be right (sorry i´m new and i cannot attach image files, so ill link you my images)

image 1

But when the RTP stream is converted into audio with stream analysis option some noises appears in the beggining of the recording (I use audacity to open the audio files created by wireshark)

image 2

This noises are part of the call, but they shouldn´t be there. They shoulb be later during the call, not in the very beggining.

Does anyone know why this happens or any possible solution? Is this a bug?

Thank you for your help

edit retag flag offensive close merge delete


Linking to images like this is fine.

Not sure what you mean by "they should be later during the call". In any case, things like this are hard to interpret without the capture file itself and precise description of the test scenario.

Jaap gravatar imageJaap ( 2020-06-17 09:29:12 +0000 )edit

I mean that the noises that appears in the image are noises that in the original phone call exists. But they are later in the call, and they should not be in there. I can add the file and test scenario later ( i need to cut part of the call because it may contain personal details)

Thank you for your help

Samuel Moya gravatar imageSamuel Moya ( 2020-06-17 10:38:56 +0000 )edit

What you see in the capture and what you hear in a VoIP endpoint are not necessary the same thing. VoIP endpoints can do a lot of processing to cover up troubles in the RTP streams. It's impossible to emulate this perfectly in Wireshark due to the nonstandard and often proprietary way endpoints do this. There are some common features, like tuneable jitter buffer, which can emulate certain VoIP endpoint processing features.

Jaap gravatar imageJaap ( 2020-06-17 17:28:15 +0000 )edit