On Fri, 2012-05-04 at 07:23 +0300, Tanu Kaskinen wrote: > On Thu, 2012-05-03 at 18:17 +0200, Fr?d?ric Dalleau wrote: > > Hi, > > > > This is follow up to the previous patch, using Tanu's comment. > > http://lists.freedesktop.org/archives/pulseaudio-discuss/2012-May/013454.html > > > > So i've manager to suspend the loopback using the first 2 patches. > > After that I found that the A2DP source didn't suspend itself when the > > stream was suspended, so i added patch 3. > > I don't know if this affects the rate adjustment logic. Maybe not, at > least much, because it seems to base its calculations on the received > and sent byte counters, and both those should stop and continue pretty > much in sync when the loopback is stopped and started. > > Another thing is that when the loopback stops, there will be data left > in the buffers. When the loopback continues, it will play the old data > first, which will probably be audible as a glitch. I think this and also > the rate adjustment problem would be solved if the loopback state would > be reset when stopping to what it was at module load time. And yet another thing: the time event should be disabled while the loopback is stopped to avoid unnecessary wakeups. -- Tanu
