15 Dec
2009
15 Dec
'09
5:31 a.m.
On Tue, 2009-12-15 at 10:16 +0100, Atte André Jensen wrote:
Ng Oon-Ee wrote:
You may want to check what qjackctl does in Ubuntu then. The script may
be calling pasuspender, for example. The behaviour I was talking about
is consistent with my experience with latest jack2/pulseaudio in Arch
Linux, and indeed with intended behaviour (by the dev).
The other thing I just realized, if you're using jack1 instead of jack2
in Ubuntu the above behaviour would not apply (no dbus in jack1), so
pasuspender may be used as a 'workaround'.
When qjackctl opens, jack is usually started
(AFAIK).
Not necessarily...
Pulseaudio, for
better integration with JACK, gives up control of the soundcard to JACK
whenever JACK starts. This is not an error, its by design. The idea is,
if you're using JACK, you want it to have the soundcard, not pulse.
AFAIK, however, pulse only gives up a soundcard if jack takes it. This
should not happen if qjackctl doesn't start jack (I believe it does
start jack though).
First: I setup qjackctl to *not* start jack, and indeed it isn't, at
least jack clients started up with qjackctl open, but jack not started,
complains that jack is not running.
Second: Firefox (let's take that as an example) uses the buildin
soundcard, and qjackctl/jack is using the firewire card.
If jack is running I'm of course not expecting non-jack apps to be able
to access the soundcard that jack is using, but this is not the case here...