Aeolus seems to have a bug which causes it to be
kicked off by jackd under certain circumstances.
After restarting it while jack didn't get restarted, I
get glitches in my speakers. The glitches cannot
be heard when recording from Aeolus directly
using ťqarecord --jackŤ and connecting it to
Aeolus'
output directly, but can be clearly heard through
the speakers.
After restarting both Aeolus and Jack, everything is
fine again.
Sure, Aeolus shouldn't crash, but asking Fons
Adriaensen we agree that there also seems to be
a bug in JACK itself.
I use JACK via qjackctl on top of a cheap Terratec
Aureon 5.1 USB with the following command:
/usr/bin/jackd -v -R -P89 -dalsa -dhw:2 -r48000
-p512 -n4
I use a vanilla 2.6.10 kernel with the LSM-RT-patch
applied. Latency is about 42,7msec (according to
qjackctl). Yes I know that's much :) .
Alsa is
Advanced Linux Sound Architecture Driver Version
1.0.6 (Sun Aug 15 07:17:53 2004 UTC).
Compiled on Jan 26 2005 for kernel 2.6.10ce
(SMP).
OSS is completely disabled in the kernel.
Steps to reproduce:
* Start JACK
* Start Aeolus
* Connect a MIDI keyboard to the pedal of Aeolus
* Play around (maybe moving the hand up and
down on the keyboard speeds up Aeolus
crashing :) while also changing
some pedal stops from time to time
* Aeolus will crash
* Restart and reconnect Aeolus
=> You'll get glitches in the output.
Included there's an ogg file where you can hear
the result. It is recorded redirecting the output of
the Aureon to its inputs and using qarecord.
Interestingly, when recording the output of Aeolus
directly to the harddrive (without using the
soundcard) then I do not get the glitches.
For my needs, this is a minor bug because Aeolus
does not crash that often and if so I can restart
both JACK and Aeolus. But if there's a bug
perhaps it could be fixed.