|
List Info Thread: Re: A question or two about jack and usb-soundcard
[1-2]
|
||||||||||||||||||||
|
about | contact Other archives ( Real Estate discussion Medical topics ) |
||||||||||||||||||||
Re: A question or two about jack and usb-soundcard |
|
List Info Thread: Re: A question or two about jack and usb-soundcard
[1-2]
|
||||||||||||||||||||
|
about | contact Other archives ( Real Estate discussion Medical topics ) |
||||||||||||||||||||
lists.sourceforge.net
Let me give you an example of what I can do: I can play a
mp3 over
audacious-jack and then I start Hydrogen (which makes the PC
get
completely sluggish, 100% cpu usage... and actually takes
over a minute
to have hydrogen launched), but: there is not a single xrun
even when it
is connecting to jack! then I can put hydrogen to play
something while
there is still the mp3 playing, and I hear both outputs and
although the
computer gets even more slow and I can hardly move the
mouse, I only
got like one or two xruns once and output is perfect. So I
think I got
it, right?
I also managed to solve the problem with ardour: it was a
strange issue
with realtime priorities, somehow ardour didn't get them,
and that's why
it was so sluggish and crashing... now it's working
perfectly. There is,
however, one serious thing I discovered, and although I
suppose that it
is not a problem in jack, I thought I'd better report it
anyway, because
the way jack reacts to it does seem pretty wrong to me: ;)
if I happen to connect the soundcard to my USB 2.0 PCMCIA
Card, then
first everything appears to be working, BUT: if the system
load gets a
bit higher (f.ex. starting hydrogen while playing a mp3
over
audacious-jack, like I described above) , there comes a
point where (it
happens every time) audio output stops totally, I see dozens
of xruns
happening, and after a few seconds jack crashes completely:
it just
vanishes, no error message, NOTHING!. I think that's pretty
weird, no?
The newly launched application then obviously complains
about missing
jack... this happens under many other combinations of
software, so I am
completely unable to use the soundcard on the usb 2.0
interface if I
want to use jack. On the notebook's 1.1 USB port everything
works great
as I just reported above, though. So I wonder what's the
problem... the
usb interface actually uses the same IRQ (which I obviously
gave RT
priority, as I have the RT kernel patches from Ingo) as the
internal USB
controller, so that can't be the reason.) And USB 2.0 has a
much larger
bandwidth, so why does this happen? Any ideas? Yes, I did
pass the
parameter nrpacks=1 to the snd-usb-audio module.
Not that it would be really important to me to get this
working over the
usb 2.0 interface, but it's still very weird... and if you
wonder: the
interface is from D-Link and is perfectly supported in linux
(never
giving me any trouble). So I don't know what it could be...
but if you
think that's something related to jack and you want to
investigate, I'll
be available if you need anything.
So, the bottom line is: I'm finally really happy with Jack,
yes it was
very hard to get this all working, but it was worth it. So I
think I
won't bother you guys anymore with my reeeaaally long
messages lol
Am Freitag, den 02.02.2007, 10:45 +0000 schrieb John Rigg:
> On Fri, Feb 02, 2007 at 12:26:25AM +0000, Elias Jahn
wrote:
> > I am using Zenwalk linux, a distro based on
slackware, so there is no
> > PAM. Instead I use the set_rtlimits approach
(which, by the way works
> > very well). There is only this memory lock
issue... seems a permission
> > lack, because as root it doesn't appear... any
ideas?
>
> set_rtlimits can't set memlock limits AFAIK. It has
been superseded by
> set_rlimits, which can set memlock:
>
>