"muddy" Sound

I am using the 64 bit 10.3 version.

Sound volume is set about 30%. No problems with the volume itself or the settings.

After boot and logging into KDE, the initial startup sound is distorted and sounds like it were being stretched slower then back to normal going down in range then back up.

Using Amarok, it will occur from time to time as well. It also causes slowdowns and freezes to occur in UT2004.

This is on a 4 core AMD Opteron system using the AMD 6111 chipset for sound using the AC97 Intel810 emulation.
A list of loaded modules.

snd_pcm_oss 51776 0
snd_mixer_oss 18304 1 snd_pcm_oss
snd_seq 59376 0
snd_seq_device 9684 1 snd_seq
snd_intel8x0 37928 1
snd_ac97_codec 114376 1 snd_intel8x0
ac97_bus 3456 1 snd_ac97_codec
snd_pcm 93000 3 snd_pcm_oss,snd_intel8x0,snd_ac97_codec
snd_timer 26696 2 snd_seq,snd_pcm
snd 69368 10 snd_pcm_oss,snd_mixer_oss,snd_seq,snd_seq_device,s nd_intel8x0,snd_ac97_codec,snd_pcm,snd_timer
soundcore 9424 1 snd
snd_page_alloc 12112 2 snd_intel8x0,snd_pcm

I am not sure how to get rid of this issue. It does not appear to be an IRQ conflict, it is using IRQ 7 and not sharing it with anything else I could find.

Thoughts?

I know nothing about the 4 core type of hardware you are using, so I can only hope that this is not related to that hardware.

Thanks for the information you provided, but typically I like to get additional information, to get a better overview as to one's sound configuration. So rather than go through a post question and answer for a few times to extract all requested information, can you run a script that will obtain this information?

The script suggested by user wishie (from IRC channel #alsa) should provide what I am looking for. To download:
<span style="font-family:Courier New">bash tsalsa</span>
Then if you also ran this script, also post the URL here. ...

Url for the alsa-info script is: http://pastebin.ca/785715


talsa script info here: http://nopaste.com/p/aBJFgfwecb

It does seem from time to time it may be the disk I/O, but this problem did not occur in OpenSuse 10.2, and the drivers are SATA 1.5gb (v1.0) rated operating at UDMA5, write cache enabled.

Any assistance is greatly appreciated. I'm almost tempted to go out and buy another sound card to see if it will alleviate the problem.

Url for the alsa-info script is: http://pastebin.ca/785715
talsa script info here: http://nopaste.com/p/aBJFgfwecb

It does seem from time to time it may be the disk I/O, but this problem did not occur in OpenSuse 10.2, and the drivers are SATA 1.5gb (v1.0) rated operating at UDMA5, write cache enabled.[/b]

Thanks for this. ... I'll try and take a look at this tonight.

In the mean time, you could try increasing the buffer size in your media player settings, and see if that helps.

Thanks for this. ... I'll try and take a look at this tonight.

In the mean time, you could try increasing the buffer size in your media player settings, and see if that helps.
[/b]

I appreciate your time. I have upped the buffer size in the KDE control panel and it does help, but it is maxed out. This resolved it in the 10.2 version.

I have SATA drives in a raid configuration and I have turned off the access time stamp feature in fstab to try and optimize r/w operations. HDParm shows them in a UDMA5 configuration, but I can't toggle the 32 bit flags on with it. I've no real idea how to tune them with sdparm either although I took a look at its man page.

This really has me puzzled and I don't see an IRQ conflict in there either. I updated the BIOS on the MB last week to the latest revision as well to fix this and another issue where it is not recognizing all the memory on the system.

I really like the improvements to this version or I would drop back to 10.2. I just may if my vmware crashes continue without helping to fix it. That is something I have to have running and is not optional if I want to get paid

Url for the alsa-info script is: http://pastebin.ca/785715
talsa script info here: http://nopaste.com/p/aBJFgfwecb[/b]

Sorry, I did not get much chance to look at this. My wife got called away on business, and so I ended up with less time than normal to myself. Plus this was an opportunity to install 10.3 on her PC, so I spent my time doing that, with a result that I did not have time to research this much.

I note you have an AMD AMD8111 with AD1981B at irq 17.

# Soundcards recognised by ALSA
# -----------------------------
# 0 [AMD8111********]: ICH - AMD AMD8111
#**********************AMD AMD8111 with AD1981B at irq 17

Researching AMD8111 gave me nothing on Alsa. Researching AD1981B gave this:
http://www.alsa-project.org/main/index.php...1981B&go=Go
indicating support has been there (to a certain extent) since alsa-1.0.13.

Checking here:
http://www.mjmwired.net/kernel/Documentati...nfiguration.txt
and I found no guidance for the AD1981B, but I did for the AD1981. That "fooled" me for a while, until I noticed this:
rcalsasound restart</span></span>

Still the above is a long shot.

To investigate if this is an interrupt problem, you could look here for guidance:
http://www.linuxjournal.com/node/1000262

I can't look at this any more, as I need to run off to work ... I recommend you surf the web for "linux AMD8111" and "linux AMD1981B" to see if you come up with any hits.

I note you have an AMD AMD8111 with AD1981B at irq 17.
[code]# Soundcards recognised by ALSA
[/b]

This looks like it was it. I was looking at the bit of info you posted for the snd-intel8x0 and the following note caught my eye:
Note: the latest driver supports auto-detection of chip clock.
1186 if you still encounter too fast playback, specify the clock
1187 explicitly via the module option "ac97_clock=41194".

Bright boy I am, I pulled up the information in /etc/modeprobe.d/sound and added it like so ---
options snd-intel8x0 enable=1 index=0 ac97_clock=41194
# hlTL.jRMIPaC1_eA:AMD-8111 AC97 Audio
alias snd-card-0 snd-intel8x0

restarted alsa ( /etc/rc.d/alsasound restart ) without any errors displayed and then started Amarok.....

That wasn't the fix, but I got to listen to Alanis Morisette, Areosmiths's Pink and BTO ..... at about 7/9's normal speed. It was so unusual and funny that I left it like that for a few minutes just to catch a couple gut busters of Alanis and Aerosmith. Alanis sounds so... well, so... well, you just had to be there. So I removed the clock setting and let it default back to the 44000 that appears to be the default.

BUT, I had not noticed it was residing on IRQ 17. I assumed it was IRQ 7. Parsing the boot log quickly showed me the culprit... that is the same IRQ that the SATA drives are using. No wonder it was so odd and did not occur as often as you would think. I hae 6 gigs of memory so it wouldn't bog down much on disk access except when it was initially loading and when I was doing a lot of disk I/0 (copying files or burning a CD) because the disk caching was has so much memory to stash in that once it was cached it generally didn't need to read the drive again.

I appreciate your time and now I have a much better understanding of how the sound system works!

BUT, I had not noticed it was residing on IRQ 17. I assumed it was IRQ 7. Parsing the boot log quickly showed me the culprit... that is the same IRQ that the SATA drives are using. No wonder it was so odd and did not occur as often as you would think.[/b]

Great stuff!

Dependant on your motherboard, its possible you can go into your BIOS, and re-assign the interrupts.