[pulseaudio-discuss] Does PulseAudio compensate for clock drift?

[Tanu Kaskinen]

On Fri, 2012-10-19 at 18:19 +0300, Ansis Māliņš wrote:
> Sorry to trouble you again, but I couldn't find any info in the
> documentation or the mailing list archives on this.
> 
> 
> I'm trying to write a simple Windows app that would stream audio to a
> remote PulseAudio. How do I do adaptive resampling in this scenario?
> If the RTP module does it, what are the requirements of a PulseAudio
> compatible RTP stream? Or maybe I can implement the native protocol?

Sorry, it seems that I never answered your last mail, and it got buried
in the pile of unread pulseaudio-discuss mail.

The RTP module does do adaptive resampling, and it's probably the only
way you can avoid writing any adaptive resampling code yourself.
Unfortunately, I don't know the details of what is required from RTP
streams for pulseaudio to be able to receive them. I'm not very familiar
with the RTP stuff. Someone else on this list might know those details.

(Re-)implementing the native protocol doesn't really make sense - it's
much easier to just use libpulse. But that doesn't help with the
resampling anyway. When you use the native protocol, the stream clock is
the sound card clock, so you can't push stuff at your own rate and
expect pulseaudio to adapt to that rate.

-- 
Tanu