ubuntu-audio-dev team mailing list archive

[Luke Yelavich <luke.yelavich@xxxxxxxxxxxxx>]

On Thu, Jun 13, 2013 at 11:30:17PM EST, David Henningsson wrote:
> On 06/12/2013 08:39 AM, Luke Yelavich wrote:
> >On Wed, Jun 12, 2013 at 10:00:12AM EST, Luke Yelavich wrote:
> >>Yes agreed. I recently updated the ALSA test cases, but I am going to see about getting users to run through the pulse test cases as well, since that stress tests ALSA rather well.
> >
> >Spoke too soon, the problem has reared its ugly head again, and I suspect its pulse based. I need to find some time to debug it, but I am still ok with pushing ALSA 1.0.27.
> 
> Anything I can help out with to speed up the process?

It probably has something to do with the pulse simple API, since thats what is currently being used for speech synthesis output via pulseaudio. I just ned to capture some logs at the time of the problem, as it comes and goes. Some of the audio buffer gets lost or muddied every time something gets spoken, and sometimes I get a rather unpleasant crackling as well. I'll hopefully get a chance to get some logs today.

Luke