Ask Your Question

Wireshark 3.6.1 is hiding an audio issue

asked 2022-02-09 19:33:41 +0000

Gilmar gravatar image

updated 2022-02-09 21:23:30 +0000

Hello Community,

I am troubleshooting a synthetic/robotic voice issue and I noticed that Wireshark player is fixing the voice.

If I play or export the audio to a .wav file the audio is perfect.

However, when I exported as .raw, I was able to hear the synthetic/robotic voice with Audacity.

I also tryied to play the stream in a previous portable version (3.4.11) and the wireshark player is not fixing the audio.

Someone know a way to change this behavior in the version 3.6.1?

Thank you

edit retag flag offensive close merge delete


Hi, Can I ask you to share captured packets? I will check it with the code... Jirka

Jiri Novak gravatar imageJiri Novak ( 2022-02-10 07:39:36 +0000 )edit

Yes, of course.

I upload it the link bellow:

In this file we have only the affected stream, but you will need to use the option "Decode As..." and set the "Current" as RTP.

Gilmar gravatar imageGilmar ( 2022-02-10 18:24:21 +0000 )edit

How was the capture made? I see every RTP frame twice, but with different values, so they seem to be sent on the network twice, rather that only captured twice. I did not check the audio yet, just wanted to check the capture setup first...

SYN-bit gravatar imageSYN-bit ( 2022-02-11 21:42:06 +0000 )edit

Hi. I'm confirming that every packet is duplicated, but with own IP ID so it is weird. Nevertheless, audio replay in Wireshark 3.6.1 is fine. It shows duplicates in graph and analysis and audio sounds OK to me.

Jiri Novak gravatar imageJiri Novak ( 2022-02-12 09:09:03 +0000 )edit

Hello guys,

I believe that your analysis is right.

I didn't any special setup in the capture and I really believe it was a bug in the softphone.

I only posted this question because when I tried to listen to the audio, Wireshark 3.6.1 reproduced it perfectly while in the real world and in the version 3.4.11 it was robotic.

Gilmar gravatar imageGilmar ( 2022-02-23 21:19:57 +0000 )edit

1 Answer

Sort by ยป oldest newest most voted

answered 2022-02-09 20:59:05 +0000

Jaap gravatar image

You'll probably have to more carefully tune the RTP player to match the performance of your RTP endpoint. Choose Playback Timing and Jitter Buffer to match your expectations. The robotic voice perception is an effect seen when the RTP endpoint experiences a significant amount of packet loss.

edit flag offensive delete link more


You might be right about the packet loss since there are a lot of packets with the status "Wrong sequence number".

Since it's affecting only outpulsed packets and the robotic audio was captured in the source PC, we are considering things like CPU, drivers or maybe a softphone bug.

Regarding the player settings, I unsucessfully tryied a lot options, and, for now, our workaround is export the .raw or install the version 3.4.11.

Gilmar gravatar imageGilmar ( 2022-02-09 21:21:20 +0000 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Question Tools

1 follower


Asked: 2022-02-09 19:33:41 +0000

Seen: 74 times

Last updated: Feb 09