[pulseaudio-discuss] Underruns when changing sinks on a 'live stream'

[Alejandro Gonzalez]

On Mon, 2011-06-06 at 10:26 -0500, Lin, Mengdong wrote:
> Hi Alejandro，
> 
> Have you check the whether PA_STREAM_ADJUST_LATENCY is set by the application? If yes, you may have the same problem as me.
> http://lists.freedesktop.org/archives/pulseaudio-discuss/2011-June/010146.html

Hi Amanda, 

Seems like you are right, if I comment the PA_STREAM_ADJUST_LATENCY flag
on the gst component pulsesink, I can switch between the sinks, without
hitting all the underruns I was seeing previously.

But honestly I don't quite get it since on my previous logs the latency
actually seem to be increasing, but maybe not quick enough, and on your
other mail-thread you mention yours is actually decreasing. So, it
confuses me.

This is one of my previous logs: 

I: alsa-sink.c: Increasing minimal latency to 1.00 ms
D: protocol-native.c: command_get_playback_latency 
D: protocol-native.c: SINK_INPUT_MESSAGE_UPDATE_LATENCY
D: protocol-native.c: command_get_playback_latency 
D: protocol-native.c: SINK_INPUT_MESSAGE_UPDATE_LATENCY
I: alsa-sink.c: Underrun!
I: alsa-sink.c: Increasing minimal latency to 2.00 ms
...
D: protocol-native.c: SINK_INPUT_MESSAGE_UPDATE_LATENCY
I: alsa-sink.c: Underrun!
I: alsa-sink.c: Increasing minimal latency to 4.00 ms
...
D: protocol-native.c: command_get_playback_latency 
I: alsa-sink.c: Underrun!
I: alsa-sink.c: Increasing minimal latency to 8.00 ms
...
D: protocol-native.c: SINK_INPUT_MESSAGE_UPDATE_LATENCY
I: alsa-sink.c: Underrun!
I: alsa-sink.c: Increasing minimal latency to 16.00 ms

And now without that PA_STREAM_ADJUST_LATENCY flag, I only hit a couple
of underruns each time I switch the sink, but audio do not get
distorted, at least not until now, but I'll test it thoroughly
tomorrow. 

Thanks and Regards, 

Alejandro.