[pulseaudio-discuss] Using module-combine-sink with intermittent devices

Matt Feifarek matt.feifarek at gmail.com
Tue Jan 19 16:50:12 PST 2016


On Mon, Jan 18, 2016 at 8:25 PM, Arun Raghavan <arun at accosted.net> wrote:

> > In short, I have two run-states: one with one soundcard, one with two; I
> > want to support both transparently, playing the same source streams,
> without
> > having to log into the server and run pacmd. Is this possible?
>
> Are there more sound cards on this system that you do not want added
> to the combine sink? If not, when you don't specify the "slaves"
> argument to the module, it will automatically combine in all new
> sinks.
>

Actually, sortof, yes.

I have *3* cards in the system:

1. Motherboard, used for audio output
2. USB device, used for audio output
3. Card, which is only there for SPDIF input

Motherboard and USB are combined into one sink for playback.

I use a loopback to push the incoming bits from "Card" onto the combined
sink.

Now that I've learned of how combine works without the slave clause, I'll
try that, but I'm worried about ALL of the various devices (motherboard has
like 12 alsa sinks, because of 2ch, 4ch, 5.1ch, etc). Also, it seems that
the more sinks that PA has to keep track of, the harder it is to do latency
matching? I have no need to pipe sound to all of the outputs that I'm not
using, and since it seems to make PA unstable, it's not "free" to just let
it send everything everywhere.

I keep getting crashes in the log like this (which seem to be related to
latency):
E: [alsa-sink-ICE1724 IEC958] sink.c: Assertion
'!s->thread_info.rewind_requested' failed at pulsecore/sink.c:1323,
function pa_sink_render_into_full(). Aborting.
Aborted (core dumped)

But I'll try it tonight. Thanks.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/pulseaudio-discuss/attachments/20160119/bfbab199/attachment.html>


More information about the pulseaudio-discuss mailing list