31 May
2010
31 May
'10
7:16 p.m.
This morning when I turned on my computer it started playing
stuff on Pandora, but more slowly and at a lower pitch than
normal. I wondered if this was just a problem with Pandora, so I
checked my own song files with Amarok. Again, that music plays
more slowly and at a lower pitch than I'd expect. 1-2 days ago
this very same computer played at the right pitch and speed.
What could have gone wrong? The built-in (only) sound card's
clock is now running more slowly? Some system wide configuration
changed for the sound system, and it now slows things down?
If it's a configuration thing, then I didn't intend to do that,
nor did I even touch the configuration. This computer runs
PulseAudio on Mandriva 2009.1. The driver is snd_intel8x0.
Any ideas folks?
Thanks....
--
Kevin
31 May
31 May
7:48 p.m.
On 31.05.2010 19:16, Kevin Cosgrove wrote:
This morning when I turned on my computer it started playing
stuff on Pandora, but more slowly and at a lower pitch than
normal. I wondered if this was just a problem with Pandora, so I
checked my own song files with Amarok. Again, that music plays
more slowly and at a lower pitch than I'd expect. 1-2 days ago
this very same computer played at the right pitch and speed.
What could have gone wrong? The built-in (only) sound card's
clock is now running more slowly? Some system wide configuration
changed for the sound system, and it now slows things down?
If it's a configuration thing, then I didn't intend to do that,
nor did I even touch the configuration. This computer runs
PulseAudio on Mandriva 2009.1. The driver is snd_intel8x0.
Any ideas folks?
I had the same problem sometimes on my last computer. It had Gentoo on
it and onboard soundcard. Sometimes it just did that (usually in the
middle of some high demanding task, like compilation). Solution was
sometimes to just restart alsa, but that didn't work more times that it
did. Restarting whole system usually helped.
Unfortunately, I never found out what caused the problem so this is not
very helpful (I am not using that machine any more, but now it runs just
fine with Linux Mint on it).
Cheers!
Igor
Thanks....
--
Kevin
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
8:30 p.m.
On 31 May 2010 at 19:48, Igor Brkic <igor(a)hyperglitch.com> wrote:
On 31.05.2010 19:16, Kevin Cosgrove wrote:
Your reply IS helpful. Now I know that someone else had the
issue. While away from email for a bit, I wondered if maybe
the temperature of my system was high enough to cause a clock
rate to lower. I can imagine that a badly designed system might
even allow a lower CPU speed to alter the audio clock rate too.
Your comment about the sound speed/pitch changing under high
load would be consistent with this theory, in that high load can
increase the system's internal temperature.
Thanks....
--
Kevin
This morning when I turned on my computer it started playing
stuff on Pandora, but more slowly and at a lower pitch than
normal. I wondered if this was just a problem with Pandora,
so I checked my own song files with Amarok. Again, that
music plays more slowly and at a lower pitch than I'd expect.
1-2 days ago this very same computer played at the right
pitch and speed.
What could have gone wrong? The built-in (only) sound
card's clock is now running more slowly? Some system wide
configuration changed for the sound system, and it now slows
things down?
If it's a configuration thing, then I didn't intend to
do that, nor did I even touch the configuration. This
computer runs PulseAudio on Mandriva 2009.1. The driver is
snd_intel8x0.
Any ideas folks?
I had the same problem sometimes on my last computer. It had
Gentoo on it and onboard soundcard. Sometimes it just did
that (usually in the middle of some high demanding task, like
compilation). Solution was sometimes to just restart alsa,
but that didn't work more times that it did. Restarting whole
system usually helped.
Unfortunately, I never found out what caused the problem so
this is not very helpful (I am not using that machine any more,
but now it runs just fine with Linux Mint on it).
9:24 p.m.
On 31.05.2010 20:30, Kevin Cosgrove wrote:
Your reply IS helpful. Now I know that someone else
had the
issue. While away from email for a bit, I wondered if maybe
the temperature of my system was high enough to cause a clock
rate to lower. I can imagine that a badly designed system might
even allow a lower CPU speed to alter the audio clock rate too.
Your comment about the sound speed/pitch changing under high
load would be consistent with this theory, in that high load can
increase the system's internal temperature.
I don't know if temperature can cause a clock rate to lower, but that
explanation would certainly make sense. My old machine was (at that
time) five years old laptop which sometimes had problems with heating (I
was cleaning it from dust pretty often, but fan was almost always on).
Cheers!
Igor
9:39 p.m.
On Mon, May 31, 2010 at 09:24:21PM +0200, Igor Brkic wrote:
On 31.05.2010 20:30, Kevin Cosgrove wrote:
Incorrect sample rate?
I've had this problem when having my jackd set to a different sample rate than, say,
fluidsynth.
Perhaps your media player does not resample, or is getting confused based on what ALSA is
reporting as a sample rate?
Are you using ALSA, or JACK, or Pulse, or what?
Also, do other players have this problem (i.e. mplayer, aplay, etc.) or just the one you
normally use?
How much is the pitch off, exactly? If you have a guitar tuner, playing a sine wave and
then placing the guitar tuner in front of the speakers might tell you how off exactly it
is. If it's off by, say, the same ratio as 44.1/48, then there's your answer.
I really doubt it's any kind of hardware problem. Sounds to me like a sample rate
glitch.
-ken
Your reply IS helpful. Now I know that someone
else had the
issue. While away from email for a bit, I wondered if maybe
the temperature of my system was high enough to cause a clock
rate to lower. I can imagine that a badly designed system might
even allow a lower CPU speed to alter the audio clock rate too.
Your comment about the sound speed/pitch changing under high
load would be consistent with this theory, in that high load can
increase the system's internal temperature.
I don't know if temperature can cause a clock rate to lower, but that
explanation would certainly make sense. My old machine was (at that
time) five years old laptop which sometimes had problems with heating (I
was cleaning it from dust pretty often, but fan was almost always on).
9:52 p.m.
On 31 May 2010 at 12:39, Ken Restivo <ken(a)restivo.org> wrote:
Incorrect sample rate?
One would thing so. But, I'm using only consumer targeted
software, where nobody would even know what sample rate was, let
alone how to change it. The two progs that had trouble were
Firefox and Amarok.
I've had this problem when having my jackd set to
a different
sample rate than, say, fluidsynth.
Perhaps your media player does not resample, or is getting
confused based on what ALSA is reporting as a sample rate?
Are you using ALSA, or JACK, or Pulse, or what?
Unlike my recording system, this one doesn't use JACK, only ALSA
with PulseAudio
Also, do other players have this problem (i.e.
mplayer, aplay,
etc.) or just the one you normally use?
I noticed the trouble when using firefox, then tried Amarok and
found the same symptom.
How much is the pitch off, exactly? If you have a
guitar tuner,
playing a sine wave and then placing the guitar tuner in front
of the speakers might tell you how off exactly it is. If it's
off by, say, the same ratio as 44.1/48, then there's your
answer.
I didn't check, but it seemed like maybe a 4th or 5th down?
I really doubt it's any kind of hardware problem.
Sounds to me
like a sample rate glitch.
Like Igor, a reboot fixed my symptom. I'll be taking the cover
off soon to look for large balls of dust, and I'll clean it out
just to be safe.
I'll also go find the lm_sensors howto and get that config'd so
that I can see what's going on inside this machine. The fan has
been getting louder, in a way that tells me that it's working
harder, not that the bearings are failing.
Being a hardware engineer I'm the first to blame software! ;-)
But, I still gotta at least look at the hardware.
Thanks....
--
Kevin
1 Jun
1 Jun
12:08 a.m.
Just a random thought:
in ALSAmixer, if they exist, check the values of 'Multi Track Rate
Locking' and 'Multi Track Rate Reset' .
Both should be off. 'M' for mute will turn them off and on. These
settings seem "sticky" so you'll need to unload/reload the drivers,
or reboot, in order for then to take effect. [[ if any alsa expert has
opinions on what these settings are for -- other than causing massive,
inexplicable hard-to-debug problems with soundcards and/or causing
Jackd to lockup when they're set... ]]
Much more rarely, I've also seen problems w/ "ondemand" scheduling
(CPU clock changes) versus "performance" (steady clock at top speed).
for example, on a dual processor machine, as root, try:
/usr/bin/cpufreq-selector -c 0 -g performance ;
/usr/bin/cpufreq-selector -c 1 -g performance
For example, if you use the program 'mixxx' and output to ALSA
directly, running in "ondemand" mode causes it to lock up. Running it
through jack allows it to work in either ondemand or performance mode.
Niels
http://nielsmayer.com
2 Jun
2 Jun
11:05 a.m.
On Mon, 31 May 2010 15:08:09 -0700
Niels Mayer <nielsmayer(a)gmail.com> wrote:
Just a random thought:
in ALSAmixer, if they exist, check the values of 'Multi Track Rate
Locking' and 'Multi Track Rate Reset' .
Both should be off. 'M' for mute will turn them off and on. These
settings seem "sticky" so you'll need to unload/reload the drivers,
or reboot, in order for then to take effect. [[ if any alsa expert has
opinions on what these settings are for -- other than causing massive,
inexplicable hard-to-debug problems with soundcards and/or causing
Jackd to lockup when they're set... ]]
Much more rarely, I've also seen problems w/ "ondemand" scheduling
(CPU clock changes) versus "performance" (steady clock at top speed).
for example, on a dual processor machine, as root, try:
/usr/bin/cpufreq-selector -c 0 -g performance ;
/usr/bin/cpufreq-selector -c 1 -g performance
For example, if you use the program 'mixxx' and output to ALSA
directly, running in "ondemand" mode causes it to lock up. Running it
through jack allows it to work in either ondemand or performance mode.
Is this anything to do with one of the multi-core processors, sorry I
forget which one, where changing processor clock rate for power saving
reasons could cause the TSC registers on each core to drift appart?
Steve.
6:17 p.m.
On 2 June 2010 at 10:05, Steve Fosdick <sjflists(a)btinternet.com> wrote:
Is this anything to do with one of the multi-core
processors, sorry I
forget which one, where changing processor clock rate for power saving
reasons could cause the TSC registers on each core to drift appart?
The computer in question is a hyperthreaded Intel something. It
shows up as having two CPUs, even though it's really only got one.
The problem has not reappeared after that one reboot.
Last night I opened the case and cleaned out a lot of dust. The
processor heat sink was maybe half blocked from dust. The machine is
quieter now that it's clean.
TTFN....
--
Kevin
5088
days inactive
5090
days old
linux-audio-user@lists.linuxaudio.org
8 comments
5 participants
participants (5)
-
Igor Brkic -
Ken Restivo -
Kevin Cosgrove -
Niels Mayer -
Steve Fosdick