Hello there!<br><br>our company (e-learning) is trying to roll out Mumble for our users. The quality is awesome but setting up the program is a pain for our users and we'd like to improve that for our users and for Mumble users in general.<br>
<br>The main issue our users have is with voice detection. Speaking with Mumble's community they informed me<br><div style="margin-left: 40px;">Now, one of the most twitchy aspects of the audio tuning is voice
activity detection for voice-activated transmission. For this we're
relying on speex, and fixing speex' VAD has been a long-overdue issue.
If you want to offer a bounty for that, it would probably be better
directed at the Speex forums though.
A good, long-term, high-precision VAD would also allow us to auto-tune
the AGC maximum and minimum levels automatically, at which point much of
the need for the audio wizard is gone.<br></div><br>We'd like to know if we can sponsor such a fix and if this can be improved for the larger Speex community.<br><br>Thanks in advance!<br clear="all"><div></div>Danilo<br>