PULSEAUDIO Crackling / Distortion

me4tw · 2011-04-03 04:44:47

This distortion is like a crackly echoey distortion. It actually has an echo component to it, When I change the output and it goes normal for a few seonds and tehre is a song playing, then it kicks in, the sound quality goes way down, it has this crackling sound, and very noticeable echo comes on the voices and the drums (which you can barely make out over the crackling). Output is perfect if forced to use ALSA directly, or if I boot into windows Seven. This is all pulse audio (probably), the piece of crap.

When I go "pulseaudio --kill" and then "mplayer -ao alsa:device=hw=0.0 <file>" it plays perfectly without distortion. When I go alsamixer -c0 none of the sliders have any gain, they are all at 0db or -1db or so.

To stop the distortion, chanigng from 44100 to 48000 and so on in the /etc/pulse/ whatever file doesnt work. What does work is opening pavucontrol and changing the internal audio on the configuration page. I hav a Barts thing that is turned off and an Internal Audio that can be set to a huge list of things like Analog Surround 5.0 Out, Analog Stereo Duplex, Analog Surround 7.1 Output + Analog Stereo Input, so on and so forth.

Now changing from one to another will randomly make the distortion go away for 30 seconds, until I open a new program, close pavucontrol, or just wait a while, then it will come back on, and I'll have to change it to another, which won't work at first usually but randomly will. There is NO CORRELATION between what is actually chosen and what does and doesnt have distortion straight away.

Basically PulseAudio is $#!+ all the sudden, and I don't know why. I have a fully updated system, hardware is as follows:

$ lspci | grep udio
00:14.2 Audio device: ATI Technologies Inc SBx00 Azalia (Intel HDA) (rev 40)
01:00.1 Audio device: ATI Technologies Inc Barts HDMI Audio [Radeon HD 6800 Series]

Also plays it fine as testing with mplayer, so this is a pulseaudio thing (i think). WTH????

[EDIT] And here is the lsmod:

$ lsmod
Module Size Used by
uas 8170 0
usb_storage 43628 1
ipv6 280362 48
md5 4127 2
hmac 2937 1
nls_utf8 1320 5
cifs 265094 6
fscache 39883 1 cifs
it87 28123 0
hwmon_vid 2796 1 it87
usbhid 36375 0
hid 78087 1 usbhid
snd_hda_codec_hdmi 22857 1
snd_hda_codec_realtek 295717 1
snd_seq_dummy 1479 0
snd_seq_oss 29240 0
fglrx 2679751 87
snd_seq_midi_event 5516 1 snd_seq_oss
vboxdrv 1760035 0
snd_seq 50562 5 snd_seq_dummy,snd_seq_oss,snd_seq_midi_event
snd_hda_intel 21837 6
snd_hda_codec 74609 3 snd_hda_codec_hdmi,snd_hda_codec_realtek,snd_hda_intel
snd_seq_device 5281 3 snd_seq_dummy,snd_seq_oss,snd_seq
r8169 36414 0
ohci_hcd 21338 0
ppdev 5854 0
firewire_ohci 26921 0
snd_hwdep 6222 1 snd_hda_codec
snd_pcm_oss 39509 0
parport_pc 31800 1
snd_mixer_oss 17730 1 snd_pcm_oss
lp 8992 0
firewire_core 50038 1 firewire_ohci
snd_pcm 72321 6 snd_hda_codec_hdmi,snd_hda_intel,snd_hda_codec,snd_pcm_oss
mii 3842 1 r8169
evdev 9361 12
fuse 65179 5
xhci_hcd 65218 0
parport 30087 3 ppdev,parport_pc,lp
edac_core 35034 0
i2c_piix4 8176 0
ehci_hcd 38878 0
crc_itu_t 1313 1 firewire_core
snd_timer 19537 2 snd_seq,snd_pcm
button 4882 1 fglrx
i2c_core 19217 1 i2c_piix4
processor 25265 0
wmi 8061 0
pcspkr 1835 0
sg 25972 0
k10temp 2771 0
serio_raw 4566 0
usbcore 139496 7 uas,usb_storage,usbhid,ohci_hcd,xhci_hcd,ehci_hcd
edac_mce_amd 9223 0
shpchp 26725 0
snd 58906 22 snd_hda_codec_hdmi,snd_hda_codec_realtek,snd_seq_oss,snd_seq,snd_hda_intel,snd_hda_codec,snd_seq_device,snd_hwdep,snd_pcm_oss,snd_mixer_oss,snd_pcm,snd_timer
soundcore 6161 1 snd
snd_page_alloc 7361 2 snd_hda_intel,snd_pcm
pci_hotplug 24687 1 shpchp
ext4 334690 1
mbcache 5802 1 ext4
jbd2 69898 1 ext4
crc16 1313 1 ext4
sd_mod 27120 4
ahci 20465 1
pata_jmicron 2464 0
libahci 17952 1 ahci
pata_acpi 3296 0
libata 169396 4 ahci,pata_jmicron,libahci,pata_acpi
scsi_mod 125814 5 uas,usb_storage,sg,sd_mod,libata

Also my last message on dmesg is:

hda-intel: IRQ timing workaround is activated for card #0. Suggest a bigger bdl_pos_adj.

But there is only one of these lines in the output and the distortion kicks in several times. This was right after the line:

chromium-sandbo (2511): /proc/2509/oom_adj is deprecated, please use /proc/2509/oom_score_adj instead.

So it happened after I opened my web browser (I only just booted this pc up I had it turned off over the night). This is a recurring problem over the past couple of days, BTW. It was probably the last update because when I done it I didn't reboot for awhile.

Last edited by me4tw (2011-04-03 05:03:26)

Niclas · 2011-04-10 14:56:06

Hi,

I have the exact same problem. Did you find any solution yet?

Here is my onboard soundcard:

$ lspci | grep Audio
00:14.2 Audio device: ATI Technologies Inc SBx00 Azalia (Intel HDA)

Switching the audio profile randomly does help, until the problem occures again after ~30 seconds.
If I just use plain alsa (without pulse) everything works fine.

Any suggestions? I've also tried to set the default-sample-rate to other values, and I tried what is suggested here http://pulseaudio.org/ticket/697#comment:2 . Maybe we should fill in a bug report?

swanson · 2011-04-10 15:11:16

A long time ago I had this exaxt problem on Ubuntu and it was the pulseaudio module module-x11-xsmp that didn't get loaded at boot/login. I might be way out here but it might be worth checking since pulseaudio can be tricky. Try this command and see;
pactl list | grep module
or
pactl list | grep module-x11-xsmp

With pactl you can also load pulse modules, do a pactl -h and see. I used to load this; pactl load-module module-x11-xsmp when starting X.

Last edited by swanson (2011-04-10 15:12:26)

clearloon · 2011-04-10 19:42:50

I'm getting this behaviour too, but only with vlc & skype... kaffeine & mplayer are fine.

FWIW, my onboard audio is ATI Technologies Inc SBx00 Azalia (Intel HDA)

Last edited by clearloon (2011-04-10 19:44:01)

R00KIE · 2011-04-11 18:08:53

I have also stumbled upon this problem.

Hardware: Audio device [0403]: ATI Technologies Inc SBx00 Azalia (Intel HDA) [1002:4383]
Codec: ALC660-VD

I did notice that both vlc and skype were having trouble outputting to pulse directly, skype worked fine with a usb headset though.

I have edited /etc/pulse/daemon.conf and uncommented 'enable-remixing = yes' and it seems to have helped at the expense of slight increase of cpu usage.

My /etc/pulse/daemon.conf (only the uncommented parts)

resample-method = ffmpeg
enable-remixing = yes
flat-volumes = no

Technotux · 2011-04-11 21:51:18

I have this problem too, with VLC, Skype and me-tv, although I think me-tv uses VLC components. The sound is aweful, heavily distorted, and crackly until I change the profile, then a few minutes later it happens again. I happen to have the same audio hardware as the others above, and using pulseaudio, I've tried all the tricks, the only thing that did seem to help was adding 'tsched=0', to the end of the line, '#load-module module-udev-detect' in the file '/etc/pulse/default.pa', uncomment it, and restart pulseaudio. As I said it seems to help but I had pulseaudio die playing some games in WINE after adding this, so its not a fix. I think pulseaudio has a bug with these applications as they are supposed to use pulseaudio directly.

$ lspci | grep Audio
00:14.2 Audio device: ATI Technologies Inc SBx00 Azalia (Intel HDA)

Last edited by Technotux (2011-04-11 21:59:19)

clearloon · 2011-04-12 19:50:06

[EDIT: appears to be messed up again...]

Today's update has resolved this for me.

Last edited by clearloon (2011-04-15 19:04:51)

bluej774 · 2011-04-15 06:31:07

Oops. My problem isn't strictly related.

Last edited by bluej774 (2011-04-15 06:32:36)

Ichimonji10 · 2011-04-16 21:56:11

I'm affected by nearly this exact problem as well. Here's the sound cards I have in my system:

~$ lspci | grep Audio
00:14.2 Audio device: ATI Technologies Inc SBx00 Azalia (Intel HDA)
02:00.1 Audio device: ATI Technologies Inc HD48x0 audio

And only certail apps are affected by the crackling, popping, and distortion. VLC is affected, while mplayer and aplay are not.

I've seen this line pop up in dmesg:

hda-intel: IRQ timing workaround is activated for card #1. Suggest a bigger bdl_pos_adj.

(I've also seen a similar message with card #0 instead) But I did not have any line related to chromium-sandbox pop up. FWIW, I don't have chrome installed, only firefox 3.

Changing my Internal Audio profile with pavucontrol helps, but only for a while. Attempting this fix did not help, nor did editing /etc/pulse/daemon.conf help. Loading module-x11-xsmp with pactl did not help either (I unloaded it after trying the fix).

I've been using my computer for months with alsa alone, with no problems. However, I recently installed gnome3, and pulseaudio along with it. That's when this problem started.

benny_kenobi · 2011-04-17 08:52:02

I'm experiancing same symtopms as Ichimonji10 on the same hardware. So far only vlc is affected.

However I noticed 2 things: 1st when I let the video play, after 2-5 minutes the sound clears - stops crackling and echoing. I know it might not sound likly, but anyway. The other thing is that according to pavucontrol vlc is trying to output sound of 5.1 speaker system - and I only have to speakers. The configuration checks out everywhere and the outputs is set to stereo.

I've tried almost every available sound configuration in vlc to no avail. When I thought something worked it didn't next time I tried it. Seems like after that magical couple of minutes the issue goes away.

--
00:14.2 Audio device: ATI Technologies Inc SBx00 Azalia (Intel HDA)
02:00.1 Audio device: ATI Technologies Inc HD48x0 audio

heftig · 2011-04-18 10:40:21

Try adding tsched=0 to the load-module module-udev-detect line in /etc/pulse/default.pa so it reads:

load-module module-udev-detect tsched=0

benny_kenobi · 2011-04-18 17:41:32

@heftig : It seems to be working for me. I'll try it a couple times more to be 'sure'. Thanks! I have no idea what these 8 chars do, but it must be something close to magic

debijan · 2011-04-18 17:59:28

hi - i had the same problem, but only with vlc - but I love vlc *sad*.

00:14.2 Audio device: ATI Technologies Inc SBx00 Azalia (Intel HDA) (rev 40)
Subsystem: Micro-Star International Co., Ltd. Device 7599
Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=slow >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
Latency: 64, Cache Line Size: 64 bytes
Interrupt: pin ? routed to IRQ 16
Region 0: Memory at f9ff4000 (64-bit, non-prefetchable) [size=16K]
Capabilities: <access denied>
Kernel driver in use: HDA Intel
Kernel modules: snd-hda-intel

lsmod | grep -i snd:
snd_seq_dummy 1455 0
snd_seq_oss 28052 0
snd_seq_midi_event 5332 1 snd_seq_oss
snd_seq 48705 5 snd_seq_dummy,snd_seq_oss,snd_seq_midi_event
snd_seq_device 5100 3 snd_seq_dummy,snd_seq_oss,snd_seq
snd_pcm_oss 37890 0
snd_mixer_oss 14851 1 snd_pcm_oss
snd_hda_codec_realtek 294053 1
snd_hda_intel 21738 2
snd_hda_codec 73739 2 snd_hda_codec_realtek,snd_hda_intel
snd_hwdep 6134 1 snd_hda_codec
snd_pcm 71032 3 snd_pcm_oss,snd_hda_intel,snd_hda_codec
snd_timer 18992 2 snd_seq,snd_pcm
snd 55132 15 snd_seq_oss,snd_seq,snd_seq_device,snd_pcm_oss,snd_mixer_oss,snd_hda_codec_realtek,snd_hda_intel,snd_hda_codec,snd_hwdep,snd_pcm,snd_timer
soundcore 5986 1 snd
snd_page_alloc 7017 2 snd_hda_intel,snd_pcm

uname -r : 2.6.38-ARCH

the hint with pulse an tsched=0 does help too :-)

@heftig: danke ;-)

Last edited by debijan (2011-04-18 18:03:43)

heftig · 2011-04-19 10:57:50

tsched=0 disables PulseAudio's timer-based scheduling and uses the classic interrupt-driven approach. Timer-based scheduling allows for better latency management and reduced CPU usage. It puts very strict requirements on the ALSA drivers. Unfortunately, some drivers do not handle it well.

More information:
http://0pointer.de/blog/projects/pulse-glitch-free.html

tokoro · 2012-05-21 03:35:40

heftig wrote:

Try adding tsched=0 to the load-module module-udev-detect line in /etc/pulse/default.pa so it reads:
load-module module-udev-detect tsched=0

This worked amazingly well. Is this a fix or just a workaround?

maverick340 · 2012-05-21 19:39:46

Same problem for me too. It was working fine, all of a sudden (maybe one week) it started this awful crackling sound with a delay after the normal sound plays. Affected programs are Skype and VLC. It also makes some noise when switching between different system sound effects rapidly.
Audio device is

lspci | grep Audio
00:07.0 Audio device: NVIDIA Corporation MCP67 High Definition Audio (rev a1)

I am running an update as we speak, Will let you know if that fixes the problem. If not will try the workaround suggested by heftig.
Wonder why the problem started all of as sudden.

king.flasher.dave · 2012-05-22 21:38:40

I have the same issue with VLC, but not with Skype. VLC has become unusable because of this.
Is this related to pulseaudio 2.0? https://bugs.archlinux.org/task/29906

As soon as VLC is running and playing video with audio this will also effect other sound sources.
If I start a video (will crack and noise), then pause it and play music using an audio player like gmusicbrowser (using gstreamer), pause and unpause it, the audio stream will also be effected as long as VLC is opened.
As soon as VLC is closed all works fine again.

Might this just be a VLC and pulse problem? I don't have any other audio/video player acting up so far, since the update to pulse 2.0.

And: heftig's fix/workaround works for me as well. It's ugly to have a higher CPU load though

Last edited by king.flasher.dave (2012-05-22 22:14:13)

mandog · 2012-06-03 13:53:30

This is very dirty it only achieves the goal of cleaning up sound in skype/vlc at the expense of killing sound in wine

boast · 2012-09-01 14:51:24

any new info on this? or a bug report or something?

DaneM · 2012-11-19 19:48:23

Hello. I was just searching for a fix for this problem, and this was the first thread that came up on Google. The solutions mentioned herein didn't work for me, but I did find something that does, so I think it only fair to post back. (Note: I'm a Linux Mint user, but since this was the first thing on Google's results page, and since the problem isn't Mint/Ubuntu-specific, I figure it's good form to post my solution here, anyway.)

Here's what I've discovered: there's a "proper" workaround, and a very dirty hack. Both are below. (The hack was mentioned here, previously.)

"Proper" workaround:
Run:

sudo -H alsamixer

...and set PCM to something substantially below maximum. Having PCM at or near maximum causes crackling and distortion. Mine is at "-4.40, -4.80"--whatever that means. Then:

sudo -H alsactl store

The "-H" option tells sudo to use the current user's home directory instead of setting it to "/root". Not using this option will result in a "home directory /home/blah is not ours" error. This isn't necessary on all systems, presumably, and it can probably be done with "su", instead.

DIRTY hack:
To reduce the "crackling" in (some) multimedia apps at the expense of realtime performance, do this:

In /etc/pulse/default.pa, change this:

load-module module-udev-detect

...to this:

load-module module-udev-detect tsched=0

This is a very dirty workaround that bypasses PulseAudio's scheduling mechanisms in favor of the traditional interrupt approach. Use with caution.

I hope this helps you all, in case you haven't yet solved this problem. There's one other thread I found in Arch Forums that seems to need this information, so I hope a "double post" is OK just this once. :-)

beroal · 2012-11-19 20:09:50

DaneM wrote:

set PCM to something substantially below maximum. Having PCM at or near maximum causes crackling and distortion. Mine is at "-4.40, -4.80"--whatever that means

You are just changing a volume level in the analog mixer of your sound card.

Once I encountered distortion caused by high PCM volume. That occurred in a sound card embedded into a motherboard. I think that it was because of clipping audio signal.

DaneM · 2012-11-19 20:18:07

You are just changing a volume level in the analog mixer of your sound card.
Once I encountered distortion caused by high PCM volume. That occurred in a sound card embedded into a motherboard. I think that it was because of clipping audio signal.

Thanks for the reply, beroal. I'm not entirely clear on what you're saying, though; can you elaborate?

I've found that there's certainly more than one type and cause of "crackling," and I am using onboard audio (Intel HD Audio, with Realtek audio codec); is that what you mean? While the above solutions seem to help both kinds of crackling in my case, I do suspect that the remaining periodic "tick" sound is because of a signal interruption/delay/something. (This doesn't happen in Windows 7, so I suspect it's a Linux sound thing.) Thoughts?

beroal · 2012-11-20 07:35:57

DaneM wrote:

I'm not entirely clear on what you're saying, though; can you elaborate?

The operation you did in "alsamixer" sounds like a volume adjusting. I will be more sure if you supply a screenshot. Should I elaborate on "analog mixer" or "clipping audio signal"?

DaneM wrote:

I've found that there's certainly more than one type and cause of "crackling,"

I also think so.

DaneM · 2012-11-20 10:28:14

Here's a screenshot:

http://s1183.beta.photobucket.com/user/ … r.png.html

As I understand it, PCM is the software equivalent of "gain," such as one would find on an amplifier or mixer board. I think it's a sort of "pre-amp." As such, it adds amplitude to the signal at the cost of signal clarity. Keeping it down to a reasonable level doesn't significantly harm the sound, but cranking it up does the same as an "overdriven" amplifier--it distorts the sound, just like with an electric guitar. (I'm a guitarist, FYI. :-) Using no PCM at all will give you basically no sound, though, so you can't really get away with turning it down to 0.

I hope that makes more sense than my previous posts. :-)

As for "clipping audio" and "analog mixer," I think I could use some clarification on how you're using those terms.

Have a nice day.

Mixu · 2012-12-03 19:42:13

Hi people!

I've got a similar problem, but only when I play some audio, and only at the begginning.
It happens to me for Spotify, webpages like youtube, grooveshark and so.
Whe I click on play it goes doing crackling sounds and the video playing is fast, after some time (sometimes minutes, sometimes seconds) it plays fine and without speed increase.

Anyone knows anything about it?

I have tried the tsched=0 and the daemon.conf changes, but still happens the same

Thanks!

Last edited by Mixu (2012-12-03 19:43:00)

Arch Linux

#1 2011-04-03 04:44:47

PULSEAUDIO Crackling / Distortion

#2 2011-04-10 14:56:06

Re: PULSEAUDIO Crackling / Distortion

#3 2011-04-10 15:11:16

Re: PULSEAUDIO Crackling / Distortion

#4 2011-04-10 19:42:50

Re: PULSEAUDIO Crackling / Distortion

#5 2011-04-11 18:08:53

Re: PULSEAUDIO Crackling / Distortion

#6 2011-04-11 21:51:18

Re: PULSEAUDIO Crackling / Distortion

#7 2011-04-12 19:50:06

Re: PULSEAUDIO Crackling / Distortion

#8 2011-04-15 06:31:07

Re: PULSEAUDIO Crackling / Distortion

#9 2011-04-16 21:56:11

Re: PULSEAUDIO Crackling / Distortion

#10 2011-04-17 08:52:02

Re: PULSEAUDIO Crackling / Distortion

#11 2011-04-18 10:40:21

Re: PULSEAUDIO Crackling / Distortion

#12 2011-04-18 17:41:32

Re: PULSEAUDIO Crackling / Distortion

#13 2011-04-18 17:59:28

Re: PULSEAUDIO Crackling / Distortion

#14 2011-04-19 10:57:50

Re: PULSEAUDIO Crackling / Distortion

#15 2012-05-21 03:35:40

Re: PULSEAUDIO Crackling / Distortion

#16 2012-05-21 19:39:46

Re: PULSEAUDIO Crackling / Distortion

#17 2012-05-22 21:38:40

Re: PULSEAUDIO Crackling / Distortion

#18 2012-06-03 13:53:30

Re: PULSEAUDIO Crackling / Distortion

#19 2012-09-01 14:51:24

Re: PULSEAUDIO Crackling / Distortion

#20 2012-11-19 19:48:23

Re: PULSEAUDIO Crackling / Distortion

#21 2012-11-19 20:09:50

Re: PULSEAUDIO Crackling / Distortion

#22 2012-11-19 20:18:07

Re: PULSEAUDIO Crackling / Distortion

#23 2012-11-20 07:35:57

Re: PULSEAUDIO Crackling / Distortion

#24 2012-11-20 10:28:14

Re: PULSEAUDIO Crackling / Distortion

#25 2012-12-03 19:42:13

Re: PULSEAUDIO Crackling / Distortion

Board footer