[Freeswitch-users] Changing device on one leg improves the audio recorded from another leg?

[David P]

We had this bizarre experience today: A user on this device...

Dell Latitude 5580
i7-7600U CPU
Windows 10 Education OS 19042.685

...made verto calls in which he heard clicks and pops every few 100ms from
our virtual agent on another leg, and those audio defects also appear in
the mp4 recording. When he then switched to this device *on the same desk*
using the same network...

Dell Latitude 5590
i5-8350U CPU
Windows 10 Education OS 17763.1821

...there were no audio defects. Playing audio from other sources on the
problematic device revealed no problems with its speaker. At least one of
the failed calls used srflx candidates, not any relaying.

How could changing the device on leg A affect the audio recording that
apparently comes from leg B?

If it's relevant:

   - We use  "jitterbuffer_msec=5p:100p"
   - xml_cdr logs for both legs show seemingly high values for some aspects
   like these...
   - <skip_packet_count>187</skip_packet_count>
   <jitter_packet_count>1620</jitter_packet_count>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freeswitch.org/pipermail/freeswitch-users/attachments/20211029/3dd76efb/attachment-0001.html>