26 Mar
2005
26 Mar
'05
6:17 p.m.
Hi,
it's easter time, so I play a bit with aeolus ;-) .
Anyway, I have a Dell Inspiron 8200 with a vanilla 2.6.10
kernel. The only patch applied is the LSM-RT-patch.
The Inspiron offers a USB 1.1 port. A USB 2.0 hub is connected
to it.
On the hub, there are currently two USB devices:
* A Midisport 8x8
* A Terratec Aureon 5.1 USB
On top of the card, I start JACK. I tried various settings,
currently I use
/usr/bin/jackd -v -R -P5 -p16 -t2000 -dalsa -dplughw:2 -r48000
-p256 -n3 -P
As soon as I play through JACK, I get notches, regardless if
it was alsa modular synth, Aeolus or xmms. The log shows
lines like the ones attached below.
So I stopped JACK and tried to use the device in ALSA only
mode.
* xmms now plays fine on the device hw:2,0
* ams and aeolus produce a lot of xruns and I get no audio
output using hw:2,0
* when starting ams as
ams --soundcard plughw:2,0 --rate 48000
I get a lot of xruns again and the sound is worse (more
notches) as when running ams or Aeolus on top of JACK
* notches also appear as soon as I try to record audio input
using qarecord connected to JACK
* when using plughw:2,0 and a samplerate of 48000 (44100 will
not work) in the aeolusrc, I get proper output at least by
Aeolus
I could need a helping hand by making the card working
properly. Anyone an idea where I can put the screwdriver?
I did also have a look into the modinfo of snd-usb-audio, but
I cannot detect a parameter to tell it, mybe, to use a
samplerate of 48000 or similar settings:
modinfo snd-usb-audio
author: Takashi Iwai <tiwai(a)suse.de>
description: USB Audio
license: GPL
parm: index:Index value for the USB audio adapter.
parm: id:ID string for the USB audio adapter.
parm: enable:Enable USB audio adapter.
parm: vid:Vendor ID for the USB audio device.
parm: pid:Product ID for the USB audio device.
parm: nrpacks:Max. number of packets per URB.
parm: async_unlink:Use async unlink mode.
vermagic: 2.6.10ce SMP PENTIUMII 4KSTACKS gcc-3.3
depends: snd-usb-lib,snd-pcm,snd
Every wee small hint (even RTFM ;-) is much appreciated.
BTW: Is this the right place, or should I post it to LAD?
Best regards
ce
JACK log:
client qjackctl-13201: start_fd=5, execution_order=0.
client aeolus: in subgraph after qjackctl-13201,
execution_order=1.
client qjackctl-13201: wait_fd=18, execution_order=2.
client alsa_pcm: internal client, execution_order=3.
-- jack_rechain_graph()
23:10:59.394 Audio active patchbay scan...
load = 11.6229 max usecs: 636.000, spare = 4697.000
load = 11.3899 max usecs: 595.000, spare = 4738.000
load = 11.2547 max usecs: 593.000, spare = 4740.000
load = 11.5246 max usecs: 629.000, spare = 4704.000
load = 11.4345 max usecs: 605.000, spare = 4728.000
delay of 5962.000 usecs exceeds estimated spare time of
4728.000; restart ...
23:11:05.205 XRUN callback (4).
load = 11.9895 max usecs: 669.000, spare = 4664.000
load = 12.4077 max usecs: 684.000, spare = 4649.000
load = 11.8479 max usecs: 602.000, spare = 4731.000
load = 11.7931 max usecs: 626.000, spare = 4707.000
load = 11.8219 max usecs: 632.000, spare = 4701.000
load = 11.9676 max usecs: 646.000, spare = 4687.000
load = 12.2936 max usecs: 673.000, spare = 4660.000
load = 12.2128 max usecs: 647.000, spare = 4686.000
delay of 6203.000 usecs exceeds estimated spare time of
4686.000; restart ...
23:11:13.187 XRUN callback (5).
load = 12.2755 max usecs: 658.000, spare = 4675.000
load = 12.1850 max usecs: 645.000, spare = 4688.000
load = 12.3273 max usecs: 665.000, spare = 4668.000
load = 12.1640 max usecs: 640.000, spare = 4693.000
load = 11.9792 max usecs: 629.000, spare = 4704.000
load = 11.7837 max usecs: 618.000, spare = 4715.000
load = 11.8641 max usecs: 637.000, spare = 4696.000
load = 12.0824 max usecs: 656.000, spare = 4677.000
load = 12.1260 max usecs: 649.000, spare = 4684.000
load = 12.5321 max usecs: 690.000, spare = 4643.000
load = 12.4915 max usecs: 664.000, spare = 4669.000
load = 12.4899 max usecs: 666.000, spare = 4667.000
delay of 5249.000 usecs exceeds estimated spare time of
4667.000; restart ...
23:11:25.203 XRUN callback (6).
load = 12.1516 max usecs: 630.000, spare = 4703.000
load = 11.8511 max usecs: 616.000, spare = 4717.000
load = 11.6165 max usecs: 607.000, spare = 4726.000
load = 11.6868 max usecs: 627.000, spare = 4706.000
load = 12.0031 max usecs: 657.000, spare = 4676.000
26 Mar
26 Mar
6:34 p.m.
On Sat, 2005-03-26 at 23:17 +0100, Christoph Eckert wrote:
The Inspiron offers a USB 1.1 port. A USB 2.0 hub is
connected
to it.
On the hub, there are currently two USB devices:
* A Midisport 8x8
* A Terratec Aureon 5.1 USB
On top of the card, I start JACK. I tried various settings,
currently I use
Are you sure this is supposed to work? With USB 1.1 no less? Please
check the hardware docs for the Aureon. AIUI many USB audio devices do
not like being connected through a hub, and will not work this way.
Lee
7:15 p.m.
Are you sure this is supposed to work? With USB 1.1
no
less? Please check the hardware docs for the Aureon. AIUI
many USB audio devices do not like being connected through
a hub, and will not work this way.
thanks for the pointer, Lee.
Unfortunately, it makes no difference at all. Sorry, I forgot
to write that I have tested it without hub some time before.
Thanks & best regards
ce
27 Mar
27 Mar
5:42 a.m.
Christopher,
I have extensively tested kernel 2.6 with my USB soundcard (Terratec
PHASE 26) without success, the 2.4 kernel seems to be a better choice
for USB soundcard for the moment...
Regards
Philippe
On Sun, Mar 27, 2005 at 12:15:38AM +0100, Christoph Eckert wrote:
Are you sure this is supposed to work? With USB
1.1 no
less? Please check the hardware docs for the Aureon. AIUI
many USB audio devices do not like being connected through
a hub, and will not work this way.
thanks for the pointer, Lee.
Unfortunately, it makes no difference at all. Sorry, I forgot
to write that I have tested it without hub some time before.
Thanks & best regards
ce
9:23 a.m.
I have extensively tested kernel 2.6 with my USB
soundcard
(Terratec PHASE 26) without success, the 2.4 kernel seems
to be a better choice for USB soundcard for the moment...
well, if the 2.6 has problems with this cards, and 2.4 not,
this is an issue to be solved. What about a bug report?
Best regards
ce
10:48 a.m.
Hallo,
Christoph Eckert hat gesagt: // Christoph Eckert wrote:
I cannot reproduce this: I run my USB card with 2.6 the same as with
2.4, I actually prefer 2.6 now, and I didn't find any differences in
my past tests (except a known and solved M-Audio issue.) Also the ALSA
drivers are very similar same for both kernels.
Sorry, but I cannot really help with your Aureon problem. My Aureon
works well on both my machines, laptop and big one, Intel and Via,
using various 2.6 kernels. I actually suppose, it's some USB
controller issue, but I'm guessing blindly with that so you might
ignore it.
Ciao
--
Frank Barknecht _ ______footils.org__
_ __latest track: "scans" _ http://footils.org/cms/show/41
I have
extensively tested kernel 2.6 with my USB soundcard
(Terratec PHASE 26) without success, the 2.4 kernel seems
to be a better choice for USB soundcard for the moment...
well, if the 2.6 has problems with this cards, and 2.4 not,
this is an issue to be solved. What about a bug report?
11:57 a.m.
Sorry, but I cannot really help with your Aureon
problem.
My Aureon works well on both my machines, laptop and big
one, Intel and Via, using various 2.6 kernels. I actually
suppose, it's some USB controller issue, but I'm guessing
blindly with that so you might ignore it.
Thanks a lot for the info anyway. It clearly shows that it is
no general problem.
Maybe I should try one of the live distros to check if it is
related to my (probably misconfigured?) Gentoo system.
Thanks a lot,
ce
2:25 p.m.
On Sun, Mar 27, 2005 at 04:56:55PM +0200, Christoph Eckert wrote:
Maybe I should try one of the live distros to check if it is
related to my (probably misconfigured?) Gentoo system.
Perhaps it is a gentoo specific issue, nevertheless, I have tested the
Planet CCRMA 2.6.10 kernel with bad results.
My test reference was to load a wav file in TerminatorX during playback
I had Xruns with 2.6 CCRMA and Gentoo dev-sources
The best result was obtained with gentoo-sources 2.4.28 even better than
with the 2.4.27 kernel from the DeMuDi distro.
I have heard that 2.6.11 series are better for audio, I will certainly
give a try soon.
Regards
Philippe
28 Mar
28 Mar
11:13 p.m.
New subject: [linux-audio-user] improvisational music
Some improv done recently, presented here for your listening pleasure.
These are *mostly* ambient tracks.
http://ruinaudio.com/media/sleep.ogg
http://ruinaudio.com/media/easter.ogg
http://ruinaudio.com/media/inc-one.mp3
http://ruinaudio.com/media/inc-two.mp3
Be fowarned, easter.ogg is pretty big.
I used jack-rack, timemachine, ices (to stream easter.ogg), an Ibanez
"sg rippoff" electric, and a circuit bent casio PT-1.
Visit http://ruinaudio.com/ for more music and some photos and art.
Constructive feedback is welcome.
-ry
27 Mar
27 Mar
7:36 a.m.
Hi,
On Saturday 26 March 2005 23:17, Christoph Eckert wrote:
On the hub, there are currently two USB devices:
* A Midisport 8x8
* A Terratec Aureon 5.1 USB
On top of the card, I start JACK. I tried various settings,
currently I use
/usr/bin/jackd -v -R -P5 -p16 -t2000 -dalsa -dplughw:2 -r48000
-p256 -n3 -P
In which way are the plughw:* different than the hw:*? Which one should be
used for jack?
I am always using the hw:* and I don't get such errors. At least not since
upgrading to 2.6.11-gentoo...
BTW: Is this the right place, or should I post it to
LAD?
You are a user asking for help with his problems? So LAU is the right place to
ask. LAD is for developers with problems with their programming.
What should LAU be for if we outsource the music _and_ the user-problems?
Arnold
--
There is a theory which states that if ever anyone discovers exactly what the
Universe is for and why it is here, it will instantly disappear and be
replaced by something even more bizarre and inexplicable.
There is another theory which states that this has already happened.
-- Douglas Adams, The Restaurant at the End of the Universe
9:22 a.m.
Hi Arnold,
In which way are the plughw:* different than the hw:*?
AFAIK plughw adds an extra layer between the application and
the card. Plughw can do resampling and so on.
Which one should be used for jack?
I guess that hw: is better for realtime issues.
I am always using the hw:* and I don't get such
errors. At
least not since upgrading to 2.6.11-gentoo...
The 2.6.11 did cause some problems with Dell notebooks, and
guess what I'm using, so I wanted to wait for 2.6.12 ;-) .
But perhaps it's a good idea to check a 2.6.11 if the problems
are gone or not.
Thanks & best regards
ce
30 Mar
30 Mar
10:28 p.m.
On Sat, 2005-03-26 at 23:17 +0100, Christoph Eckert wrote:
Hi,
it's easter time, so I play a bit with aeolus ;-) .
Anyway, I have a Dell Inspiron 8200 with a vanilla 2.6.10
kernel. The only patch applied is the LSM-RT-patch.
The Inspiron offers a USB 1.1 port. A USB 2.0 hub is connected
to it.
According to the kernel developers this is indeed the kernel's fault.
On Wed, 2005-03-30 at 14:57 -0800, David Brownell wrote:
"Well, I'd not yet expect USB audio to work over EHCI quite yet,
though one of the patches Greg just posted should help some of
the issues with full speed iso through USB 2.0 hubs. (At least
for OUT transfers as to speakers.)"
Please try this patch against 2.6.12-rc1:
http://cache.gmane.org//gmane/linux/usb/devel/33380-001.bin
Lee
31 Mar
31 Mar
9:16 a.m.
According to the kernel developers this is indeed the
kernel's fault.
On Wed, 2005-03-30 at 14:57 -0800, David Brownell wrote:
"Well, I'd not yet expect USB audio to work over EHCI quite
yet, though one of the patches Greg just posted should help
some of the issues with full speed iso through USB 2.0
hubs. (At least for OUT transfers as to speakers.)"
Please try this patch against 2.6.12-rc1:
http://cache.gmane.org//gmane/linux/usb/devel/33380-001.bin
Thanks, Lee, for your investigation. I'll try 2.6.12 with the
patch mentioned, but it will last some days (due to the fact
that I'm not too familiar with building kernels, even I know
how to do it :) .
Thanks & best regards
ce
2 Apr
2 Apr
6:27 p.m.
Lee,
According to the kernel developers this is indeed the
kernel's fault.
On Wed, 2005-03-30 at 14:57 -0800, David Brownell wrote:
"Well, I'd not yet expect USB audio to work over EHCI quite
yet, though one of the patches Greg just posted should help
some of the issues with full speed iso through USB 2.0
hubs. (At least for OUT transfers as to speakers.)"
Thanks a lot for these infos.
Please try this patch against 2.6.12-rc1:
http://cache.gmane.org//gmane/linux/usb/devel/33380-001.bin
well, did last a while. I'm not that familiar with kernel
building. But a friend was helpful, so I have now
Linux Grandevitesse 2.6.12-rc1-ce #1 SMP Sat Apr 2 22:59:58
CEST 2005 i686 Intel(R) Pentium(R) 4 Mobile CPU 1.60GHz
GenuineIntel GNU/Linux
installed. Unfortunately, it doesn't change anything
concerning the usb_audio behaviour. I get notches every 3 to
7 seconds again caused by xruns.
Except, OTOH, I did something wrong when patching (sorry for
my ignorance). Can I somehow check if the patch has been
applied successfully by probing the current running kernel? A
location in /proc probably?
Thanks a lot & best regards
ce
8:59 p.m.
On Sat, 2005-04-02 at 23:27 +0200, Christoph Eckert wrote:
installed. Unfortunately, it doesn't change
anything
concerning the usb_audio behaviour. I get notches every 3 to
7 seconds again caused by xruns.
OK. The patch isn't expected to solve the problem completely.
The Linux USB developers would like to be cc'ed on any future bug
reports.
Except, OTOH, I did something wrong when patching
(sorry for
my ignorance). Can I somehow check if the patch has been
applied successfully by probing the current running kernel? A
location in /proc probably?
You can edit the Makefile and append something to the EXTRAVERSION.
Then, after applying the patch, check for rejects like this:
$ find -name \*rej
If you get no output, go ahead and compile the kernel, install it, and
reboot. Then run:
$ uname -r
If it shows the string you appended to EXTRAVERSION, you are indeed
running the patched kernel.
Lee
9:55 p.m.
Hi,
Looks promising:
./2.6.12-EHCItest/linux-2.6.12-rc1/drivers/usb/host/ehci-sched.c.rej
./2.6.12-EHCItest/linux-2.6.12-rc1/drivers/usb/host/ehci.h.rej
OK. The patch isn't expected to solve the
problem
completely. The Linux USB developers would like to be
cc'ed on any future bug reports.
Huh. How can I do so without being a subscriber of the kernel
mailing list?
Except, OTOH,
I did something wrong when patching (sorry
for my ignorance). Can I somehow check if the patch has
been applied successfully by probing the current running
kernel? A location in /proc probably?
You can edit the Makefile and append something to the
EXTRAVERSION. Then, after applying the patch, check for
rejects like this:
$ find -name \*rej If you get no output, go ahead and compile the
kernel,
install it, and reboot. Then run:
$ uname -r
vi /usr/src/linux/Makefile
VERSION = 2
PATCHLEVEL = 6
SUBLEVEL = 12
EXTRAVERSION =-rc1-ce
uname -r
2.6.12-rc1-ce
Seems I've been successful compiling and installing the
kernel. Thanks for the help!
If it shows the string you appended to EXTRAVERSION,
you
are indeed running the patched kernel.
Yes, seems to be running, and there are surely xruns!
Well, the xruns are bad, but I'm happy to have the patched
kernel running, regardless that the bugs are still
there ;-) .
Thanks a lot & best regards
ce
7176
days inactive
7183
days old
linux-audio-user@lists.linuxaudio.org
15 comments
6 participants
participants (6)
-
Arnold Krille -
Christoph Eckert -
Dubphil -
Frank Barknecht -
Lee Revell -
Ryan Gallagher