Open Sound System

From ArchWiki
Revision as of 12:12, 1 November 2012 by Alexfikl (Talk | contribs) (Comparisons with ALSA: of course OSS has better support for OSS apps)

Jump to: navigation, search

zh-CN:Open Sound System The Open Sound System (or OSS) is an alternative sound architecture for Unix-like and POSIX-compatible systems. OSS version 3 was the original sound system for Linux, but was superseded by the Advanced Linux Sound Architecture (or ALSA) in 2002 when OSS version 4 became proprietary software. OSSv4 became free software again in 2007 when 4Front Technologies released its source code and provided it under the GPL license.

Comparisons with ALSA

Some advantages and disadvantages compared to using the Advanced Linux Sound Architecture.

OSS Advantages (users)

  • Per-application volume control.
  • Some legacy cards have better support.
  • Initial response time in audio applications is usually better.

OSS Advantages (developers)

  • Support for drivers in userspace.
  • Cross-platform (OSS runs on BSDs and Solaris).

ALSA advantages over OSS

  • Better support for USB audio devices.
  • Support for Bluetooth audio devices.
  • Support for AC'97 and HD Audio dial-up soft-modems such as Si3055.
  • Better support for MIDI devices.
  • Support for suspend.
  • Better support for jack detection.
  • OSS has experimental output support for USB audio devices, but no input.
  • OSS supports MIDI devices with the help of a software synthesizer such as Timidity or FluidSynth.


Install oss from the official repositories.

This will install the OSS files and run the OSS install script that will temporarily disable the ALSA modules, and install the OSS kernel modules. Since ALSA is enabled by default in the boot scripts, you need to disable it so it does not conflict with OSS when booting. You can do this blacklisting the module:

install soundcore /bin/false

Start the oss daemon and add oss to your DAEMONS array so that it starts automatically on boot.

Add yourself to the audio group:

Note: Re-log in for the change to take effect.
# gpasswd -a $USER audio

In case OSS is not able to detect your card when starting it, run:

# ossdetect -v
# soundoff && soundon


Warning: The default volume is very loud, avoid using earphones and physically lower the volume of your speakers (if possible) before running the test.

Test OSS by running:

$ osstest

You should be able to hear music during the test process. If there is no audio, try to adjust the volume or refer to the troubleshooting section.

If you want to hear sounds from more than one application simultaneously, you need vmix, OSS's software mixer.

Check that vmix is enabled by running:

$ ossmix -a | grep -i vmix

You should see a line like vmix0-enable ON. If you do not see any lines beginning with vmix, it probably means that vmix has not been attached to your sound device. To attach vmix, issue the command

$ vmixctl attach device

where device is your sound device, eg., /dev/oss/oss_envy240/pcm0.

To avoid having to issue this command manually in the future, you can add it to /usr/lib/oss/soundon.user, as suggested here.

If you get a "Device or resource busy" error, you need to add
Template error: are you trying to use the = sign? Visit Help:Template#Escape template-breaking characters for workarounds.
to /usr/lib/oss/conf/osscore.conf, and then reboot.

See which devices are detected by running:

$ ossinfo

You should be able to see your devices listed under Device objects or Audio Devices. If the device that you want to use is not at the top on Audio devices or Device objects sections, /usr/lib/oss/etc/installed_drivers needs to be edited. The driver for the device that needs to be used should be at the very top. It may be required to do a:

$ soundoff && soundon

If this does not work, comment all drivers listed that are not your preferred device.

Volume Control Mixer

To control the volume of various devices, mixers levels will need to be set. There are two mixers. One is a command line mixer called ossmix. This one is very like the BSD audio mixer (called mixerctl).

The second mixer is a graphical mixer called ossxmix and will require gtk2 to be installed to run it.

The basic ossxmix controls look like:

 / High Definition Audio ALC262 \    --------------------------------> 1
|                                                                 \
| [x] vmix0-enable [vmix0-rate: 48.000kHz]      vmix0-channels    |--> 2
|                                               [ Stereo [v] ]    |
|                                                                 |
|  __codec1______________________________________________________ |
| |  _jack______________________________________________________ ||--> 3
| | |  _int-speaker_________________   _green_________________  |||
| | | |                             | |                       | |||
| | | |  _mode_____ | |             | |  _mode_____   | |     | |||
| | | | [ mix [v] ] o o [x] [ ]mute | | [ mix  [v] ]  o o [x] | |||
| | | |             | |             | |               | |     | |||
| | | |_____________________________| |_______________________| |||
| | |___________________________________________________________|||
| |______________________________________________________________||
| ___vmix0______________________________________________________  |
| |  __mocp___  O O   _firefox_  O O  __pcm7___  O O            | |--> 4
| | |         | O O  |         | x x |         | O O            | |
| | | | |     | x O  | | |     | x x | | |     | O O            | |
| | | o o [x] | x x  | o o [x] | x x | o o [x] | O O            | |
| | | | |     | x x  | | |     | x x | | |     | O O            | |
| | |_________| x x  |_________| x x |_________| O O            | |
| |_____________________________________________________________| |
  1. One tab for each sound card
  2. The vmix (virtual mixer) special configurations appear at the top. These include sampling rate and mixer priority.
  3. These are your sound card jack configurations (input and output). Every mixer control that is shown here is provided by your sound card.
  4. Application vmix mixer controls and sound meters. If the application is not actively playing a sound it will be labeled as pcm08, pcm09... and when the application is playing the application name will be shown.

Color Definitions

For high definition (HD) audio, ossxmix will color jack configurations by their pre-defined jack colors:

Color Type Connector
green front channels (stereo output) 3.5mm TRS
black rear channels (stereo output) 3.5mm TRS
grey side channels (stereo output) 3.5mm TRS
gold center and subwoofer (dual output) 3.5mm TRS
blue line level (stereo input) 3.5mm TRS
pink microphone (mono input) 3.5mm TS

Saving Mixer Levels

Mixer levels are saved when you shut off your computer. If you want to save the mixer level immediately, as root:

# savemixer

savemixer can be used to write mixer levels to a file with the -f switch and restore by the -L switch.

Other Mixers

Other mixers that have support for OSS:

  • GNOME — Gnome Volume Control || gnome
  • KDE — Kmix || kdemultimedia-kmix
  • LXDE — VolWheel || volwheel

After installing VolWheel, you will need to do the following to enable OSS support:

  • add it to the LXDE autostart file:
 echo "volwheel" >> ~/.config/lxsession/LXDE/autostart
  • set the correct preferences:
    • Driver: OSS
    • Default Channel: vmix0-outvol (find out what channel to use from ossmix)
    • Default Mixer: ossxmix
    • Tab MiniMixer (optional): vmix0-outvol and optionally others.

Configuring Applications for OSS


The skype package only includes support for ALSA. To get an OSS-capable Skype, install the bin32-skype-oss package from the AUR.


To set OSS support in Wine:

  • Run winecfg.
$ winecfg
  • Go to the Audio tab.
  • Select OSS Driver.


By default Gajim uses aplay -q to play a sound. To change this go in Advanced Settings and search for the soundplayer variable. The ossplay program included in the oss package is a good replacement: ossplay -qq.


To use MOC with OSS v4.1 you must change OSSMixerDevice to /dev/ossmix in your config (located in ~/.moc).

Alternatively, you can use the moc-svnAUR package from the AUR that has OSS support. For issues with the interface try changing the OSSMixerChannel: press 'w' in mocp (to change to sofware mixer).

Applications that use Gstreamer

Remove pulseaudio and install the gstreamer0.10-good-plugins package which is needed for oss4sink and oss4src.

To change the GStreamer setting to output the sound to OSS instead of the default ALSA, run:

$ gstreamer-properties

Change the Default Output plugin to custom and the change the pipeline to:


For the input:

Note: It is not certain that the input will sound better with oss4src compared to osssrc, so change this only if it improves your input sound. < confirmation on this please >
Note: For some applications (e.g. Rhythmbox, Totem) the gstreamer-properties have no effect, as they rely on "musicaudiosink" instead of "audiosink" (which is modified by gstreamer-properties). Workaround: Set audiosink with gstreamer-properties and use gconf-editor to copy the value of "/system/gstreamer/0.10/default/audiosink" to "musicaudiosink" (at the same location)

If you are using phonon with the gstreamer backend you will need to set the environmental variable. To add to your current user:

export PHONON_GST_AUDIOSINK=oss4sink

Add this to your ~/.bashrc to be loaded on login.


If you are using gui (smplayer etc.) you will find the oss output at the audio settings. Using on cli you should specify the sound output: mplayer -ao oss /some/file/to/play.mkv If you do not want to bother typing it over and over again add "ao=oss" to your config file. ("$HOME"/.mplayer/config)

Music Player Daemon

MPD is configured through /etc/mpd.conf or ~/.mpdconf. Check both of these files, looking for something that looks like:

 audio_output {
       type           "alsa"
       name           "Some Device Name"

If you find an uncommented (the lines do not begin with #'s) ALSA configuration like the one above, comment all of it out, or delete it, and add the following:

 audio_output {
       type           "oss"
       name           "My OSS Device"
Note: I had to put this configuration in my ~/.mpdconf for it to work properly, but it ought to work in /etc/mpd.conf as well.

Further configuration might not be necessary for all users. However, if you experience issues (in that MPD does not work properly after it has been restarted), or if you like having specific (i.e. more user-configured, less auto-configured) config files, the audio output for OSS can be more specifically configured as follows: First, run:

 ossinfo | grep /dev/dsp

Look for the line that says something similar to /dev/dsp -> /dev/oss/<SOME_CARD_IDENTIFIER>/pcm0. Take note of what your <SOME_CARD_IDENTIFIER> is, and add bolded lines to your OSS audio output in your mpd config file:

 audio_output {
       type            "oss"
       name            "My OSS Device"
       device          "/dev/oss/<SOME_CARD_IDENTIFIER>/pcm0"
       mixer_device    "/dev/oss/<SOME_CARD_IDENTIFIER>/mix0"


If audacity will start but either complains that it cannot open the device or simply doesn't play anything, then you may be using vmix. Audacity expects exclusive access to your sound device. To fix this, before running audacity, run:

 ossmix vmix0-enable OFF

You can restore vmix after ending audacity with

 ossmix vmix0-enable ON

Other applications


Troubleshooting HDAudio devices

Understanding why problems arise

If you have a HDAudio sound device, it is very likely that you will have to adjust some mixer settings before your sound works.

HDAudio devices are very powerful in the sense that they can contain a lot of small circuits (called widgets) that can be adjusted by software at any time. These controls are exposed to the mixer, and they can be used, for example, to turn the earphone jack into a sound input jack instead of a sound output jack.

However, there is a side effect, mainly because the HDAudio standard is more flexible than it perhaps should be, and because the vendors often only care to get their official drivers working.

Then, when using HDAudio devices, you often find disorganized mixer controls, that does not work at all by default, and you are forced to try every mixer control combination, until it works.

How to solve

Open ossxmix and try to change every mixer control in the middle area, that contains the sound card specific controls, as explained in the previous "Volume Control" section.

You will probably want to setup a program to record/play continously in the background (e.g. ossrecord - | ossplay - for recording or osstest -lV for playing), while changing mixer settings in ossxmix in the foreground.

  • Raise every volume control slider.
  • In each option box, try to change the selected option, trying all the possible combinations.
  • If you get noise, try to lower and/or mute some volume controls, until you find the source of the noise.

Please note again that you do not need to change any controls in the top area nor in the bottom area, as they are virtual vmix-related mixer controls.

  • Editing /usr/lib/oss/conf/oss_hdaudio.conf uncommenting and changing hdaudio_noskip=0 to a value from 0-7 can give you more jack options in ossxmix

I had to edit mine to hdaudio_noskip=7 for my sub/rear speaker to work on my laptop, restart oss for the changes to take effect /etc/rc.d/oss restart

MMS sound cracking in totem

If your stream sounds with cracks or strange noise in totem like it did with me then you could try to play it with another backend like ffmpeg (mplayer). That "fixed" the issue for me. This will not fix the issue that somehow pops up in gstreamer when playing MMS streams but it will give you the option to play it with good sound quality. Playing it in mplayer is simple:

# mplayer mmsh://yourstreamurl

Microphone playing through output channels

OSS by default plays back the microphone through the speakers. To disable this in ossxmix find the misc section. Check off every "input-mix-mute" to disable this.

Troubleshooting other issues

  • If you get distorted sound, try lowering some volume control sliders.
  • If you need to change the default sound card, look at here.
  • If you have another issues, try searching or asking for help at the 4front forums.

Tips and tricks

Using multimedia keys with OSS

An easy way to mute/unmute and increase/decrease the volume is to use the ossvol script available in AUR.

Once you installed it try to toggle the sound:

$ ossvol -t

Type ossvol -h for the other commands.

If you do not know how to assign commands to your multimedia keys, see Extra Keyboard Keys.

ossvol troubleshooting

If you get an error like:

Bad mixer control name(987) 'vol'

you need to edit the script (/usr/bin/ossvol) and change the value of the CHANNEL variable which is at the beginning of the script. For example mine is CHANNEL="vmix0-outvol".

  • Note if you are using xbindkeys for your multimedia keys adding this
"if ossmix misc.speaker-mute|grep OFF;then ossmix misc.speaker-mute on;else ossmix misc.speaker-mute off;fi;"

toggle the sound

"ossmix vmix0-outvol -- +1"

raise volume

"ossmix vmix0-outvol -- -1"

lower volume

to the raise/lower volume section of your .xbindkeysrc file is an easy way to adjust the volume

Changing the Sample Rate

Changing the output sample rate is not obvious at first. Sample rates can only be changed by the superuser and vmix must be unused by any programs when a change is requested. Before you follow any of these steps, ensure you are going through a receiver/amplifier and using quality speakers and not simply computer speakers. If you are only using computer speakers, do not bother changing anything here as you will not notice a difference.

By default the sample rate is 48000hz. There are several conditions in which you may want to change this. This all depends on your usage patterns. You want the sample rate you are using to match the media you use the most. If your computer has to change the sampling rate of the media to suit the hardware it is likely, though not guaranteed that you will have a loss in audio quality. This is most noticable in downsampling (ie. 96000hz → 48000hz). There is an article about this issue in "Stereophile" which was discussed on Apple's "CoreAudio API" mailing list if you wish to learn more about this issue.

Some example sample rates:

  • 44100hz - Sample rate of standard Red Book audio cds.
  • 88000hz - Sample rate of SACD high definition audio discs/downloads. It is rare that your motherboard will support this sample rate.
  • 96000hz - Sample rate of most high definition audio downloads. If your motherboard is an AC'97 motherboard, this is likely to be your highest bitrate.
  • 192000hz - Sample rate of BluRay, and some (very few) high definition downloads. Support for external audio reciever equipment is limited to high end audio. Not all motherboards support this. An example of a motherboard chipset that would support this includes Intel HDA audio.

To check what your sample rate is currently set to:

  1. Run "ossmix | grep rate".

You are likely to see "vmix0-rate <decimal value> (currently 48000) (Read-only)".

If you do not see a "vmix0-rate" (or "vmix1-rate", etc.) being outputted, than it probably means that vmix is disabled. In that case, OSS will use the rate requested by the program which uses the device, so this section does not apply. Exception: envy24(ht) cards have a setting envy24.rate which has a similiar function (see "oss_envy24" manpage). You can follow these steps, but at step 2, change with ossmix the value of "envy24.rate" as well.

Steps to affect the change:

  1. First, make sure your card is able to use the new rate. Run "ossinfo -v2" and see if the wanted rate is in the "Native sample rates" output.
  2. As root, run "/usr/lib/oss/scripts/". Be aware, this will close any program that currently has an open sound channel (examples being media players, Firefox as of 3.5 if you have xulrunner-oss installed, and the gnome volume control).
  3. After all programs occupying vmix are terminated, run as root: "vmixctl rate /dev/dsp 96000" replacing the rate with your desired sample rate.
  4. Run "ossmix | grep rate" and check for "vmix0-rate <decimal value> (currently 96000) (Read-only)" to see if you were successful.
  5. Make changes permanent use the soundon.user file to set the rate for every soundon
write "vmixctl rate /dev/dsp 96000" in the file /usr/lib/oss/soundon.user and make it executable.

Changing the Default Sound Output

When running osstest, the first test passes for the first channel, but not for the stereo or right channel, it sounds distorted/hisses. If this is what your sound is like, then it is set to the wrong output.

      *** Scanning sound adapter #-1 ***
      /dev/oss/oss_hdaudio0/pcm0 (audio engine 0): HD Audio play front
      - Performing audio playback test... 
      <left> OK <right> OK <stereo> OK <measured srate 47991.00 Hz (-0.02%)> 

The left sounded good, the right and stereo were the distorted ones.

Let the test continue until you get a working output:

      /dev/oss/oss_hdaudio0/spdout0 (audio engine 5): HD Audio play spdif-out 
      - Performing audio playback test... 
      <left> OK <right> OK <stereo> OK <measured srate 47991.00 Hz (-0.02%)> 

If this passed the test on all left, right and stereo, proceed to next step.

So from here: Changing_the_default_sound_output you get the command to change the default output; change according to what works for you

      sudo ln -sf /dev/oss/oss_hdaudio0/spdout0 /dev/dsp_multich

With 5.1 surround, chose dsp_multichannel; with 2 channel, dsp should work.

Creative Sound Blaster X-Fi Surround 5.1 SB1090 USB

This information is completely from ; courtesy of kristian and Maxa. Thanks!!

It is surprising to learn that the external card does not work just because of a missing true return value in the function write_control_value(...) in ossusb_audio.c.

To fix this, a recompile of oss is nessasary, for now.

  • Grab the latest oss from the Arch Repo
  • Extract it
  • cd to the folder, I renamed the folder to oss
  • run makepkg --nobuild
  • cd to src/kernel/drv/oss_usb/ ; edit ossusb_audio.c ; add a return 1
    • should look like so:
    static int
write_control_value (ossusb_devc * devc, udi_endpoint_handle_t * endpoint,
                     int ctl, int l, unsigned int v)
    return 1;
  • cd to src/kernel/setup and edit, search for -Werror and remove it, otherwise OSS will not compile.
  • do makepkg --noextract

Now you must install the package with pacman -U ; remove oss first if already installed (pacman -Rd oss)

A simple systray applet

Want a applet to control volume like in GNOME? From here I got a usable one.

Download this script and rename whatever you want, e.g.: ossvolctl. run the following command:

$ chmod +x ossvolctl
# cp ossvolctl /usr/bin/ossvolctl


# install -Dm755 ossvolctl /usr/bin/ossvolctl

There's also the volwheel package.

Start ossxmix docked to systray on startup


Create an application launcher file named ossxmix.desktop in you local application launchers directory (~/.local/share/applications/ then enter:

[Desktop Entry]
Name=Open Sound System Mixer
GenericName=Audio Mixer
Exec=ossxmix -b

To add it to autostart when loading the desktop environment:

System Settings > Advanced tab > Autostart. Then click add program and choose it from the 'Multimedia' list.


  • As Root create a file /usr/local/bin/ossxmix_bg with the following content:
exec /usr/bin/ossxmix -b

Goto System > Preferences > Start Up Applications

  • Click Add, Type OSSMIX in Name field and /usr/local/bin/ossxmix_bg in Command field then click Add button.
  • Login and Logout to see the changes.

Record sound output from a program

Suspend and Hibernation

OSS does not automatically support suspend meaning that OSS must be manually stopped prior to suspending or hibernating.

OSS provides soundon and soundoff to enable and disable OSS, although any processes that use sound must be terminated first.

The following script is a rather basic method of automatically unloading OSS prior to suspending and reloading afterwards.




case "$1" in
 *) exit $NA

Save the contents of the script (as root) into /etc/pm/sleep.d/50ossound and make it executable. chmod a+x /etc/pm/sleep.d/50ossound

Note: This script is rather basic and will terminate any application directly accessing OSS, save your work prior to suspending/hibernating.

OSS does not support suspending but we do not care or better s2ram works fine without stopping OSS. Just create a nice suspend script to /sbin/suspend and make it executable.


 ## Checking if you are a root or not
 if ! [ -w / ]; then
   echo >&2 "This script must be run as root"
   exit 1

 s2ram -f

 sleep 2

 /etc/rc.d/oss restart 2>/tmp/oss.txt ||
 echo "OSS restart failed, check /tmp/oss.txt for advice"

That is all your apps are fine and suspend works. \o/

Note: If you are using Opera you must kill operapluginwrapper before suspend. To do this add pid=$(pidof operapluginwrapper) && kill $pid before s2ram -f.

ALSA emulation

You can instruct alsa-lib to use OSS as its audio output system. This works as a sort of ALSA emulation.

Note, however, that this method may introduce additional latency in your sound output, and that the emulation is not complete and does not work with all applications. It does not work, for example, with programs that try to detect devices using ALSA.

So, as most applications support OSS directly, use this method only as a last resort.

In the future, more complete methods may be available for emulating ALSA, such as libsalsa and cuckoo.


  • Install the alsa-plugins package.
# pacman -S alsa-plugins
  • Edit /etc/asound.conf as follows.
pcm.oss {
   type oss
    device /dev/dsp

pcm.!default {
    type oss
    device /dev/dsp

ctl.oss {
    type oss
    device /dev/mixer

ctl.!default {
    type oss
    device /dev/mixer
Note: If you do not want to use OSS anymore, do not forget to revert changes that you do here in /etc/asound.conf.

Settings for specific driver

If something is not working, there is a possibility, that there are specific settings for specific driver (this way I have enabled jack-sense on my laptop)

  • Find out which driver is used
$ lspci -vnn | grep -i -A 15 audio
00:1e.2 Multimedia audio controller [0401]: Intel Corporation 82801FB/FBM/FR/FW/FRW (ICH6 Family) AC'97 Audio Controller [8086:266e] (rev 03)
	Subsystem: Hewlett-Packard Company NX6110/NC6120 [103c:099c]
	Flags: bus master, medium devsel, latency 0, IRQ 21
	I/O ports at 2100 [size=256]
	I/O ports at 2200 [size=64]
	Memory at d0581000 (32-bit, non-prefetchable) [size=512]
	Memory at d0582000 (32-bit, non-prefetchable) [size=256]
	Capabilities: <access denied>
	Kernel driver in use: *oss_ich*
	Kernel modules: snd-intel8x0
  • Locate configuration file for device in:
# cd /usr/lib/oss/conf/
  • Try changing defaults. There are only few settings, and they are self explanatory


ich_jacksense = 1 

in oss_ich.conf turned on jack-sense on my laptop (now plugged headphones are recognized, and speaker muted).

  • Restart oss for changes take effects.
# /etc/rc.d/oss restart
  • oss_hdaudio.conf has hdaudio_jacksens too. Maybe it will work for you. Unfortunately not for everyone.

Experimental packages

Mercurial repository version

There is a oss-hg package in AUR. This package compiles and installs the latest OSS development version direcly from the Mercurial repository.

You can try this package if you want to contribute code to OSS or if only a very recent change in OSS code introduced support to your sound device.

Note: This package installs the daemon start script to /etc/rc.d/oss4 (not oss) so update your rc.conf