< Advanced Linux Sound Architecture
Advanced Linux Sound Architecture/Configuration examples
The following should serve as a guide for more advanced ALSA setups. The configuration takes place in /etc/asound.conf
as mentioned in the main article. None of the following configurations are guaranteed to work.
Note: Most things discussed here are much easier to accomplish using alsa plugins like upmix which are explained in the main article.
Upmixing of stereo sources to 7.1 using dmix while saturated sources do not get upmixed
# 2008-11-15
#
# This .asoundrc will allow the following:
#
# - upmix stereo files to 7.1 speakers.
# - playback real 7.1 sounds, on 7.1 speakers,
# - allow the playback of both stereo (upmixed) and surround(7.1) sources at the same time.
# - use the 6th and 7th channel (side speakers) as a separate soundcard, i.e. for headphones
# (This is called the "alternate" output throughout the file, device names prefixed with 'a')
# - play mono sources in stereo (like skype & ekiga) on the alterate output
#
# Make sure you have "8 Channels" and NOT "6 Channels" selected in alsamixer!
#
# Please try the following commands, to make sure everything is working as it should.
#
# To test stereo upmix : speaker-test -c2 -Ddefault -twav
# To test surround(5.1): speaker-test -c6 -Dplug:dmix6 -twav
# To test surround(7.1): speaker-test -c6 -Dplug:dmix8 -twav
# To test alternative output: speaker-test -c2 -Daduplex -twav
# To test mono upmix: speaker-test -c1 -Dmonoduplex -twav
#
#
# It may not work out of the box for all cards. If it doesnt work for you, read the comments throughout the file.
# The basis of this file was written by wishie of #alsa, and then modified with info from various sources by
# squisher. Svenstaro modified it for 7.1 output support.
#Define the soundcard to use
pcm.snd_card {
type hw
card 0
device 0
}
# 8 channel dmix - output whatever audio, to all 8 speakers
pcm.dmix8 {
type dmix
ipc_key 1024
ipc_key_add_uid false
ipc_perm 0660
slave {
pcm "snd_card"
rate 48000
channels 8
period_time 0
period_size 1024
buffer_time 0
buffer_size 5120
}
# Some cards, like the "nforce" variants require the following to be uncommented.
# It routes the audio to the correct speakers.
# bindings {
# 0 0
# 1 1
# 2 4
# 3 5
# 4 2
# 5 3
# 6 6
# 7 7
# }
}
# upmixing - duplicate stereo data to all 8 channels
pcm.ch71dup {
type route
slave.pcm dmix8
slave.channels 8
ttable.0.0 1
ttable.1.1 1
ttable.0.2 1
ttable.1.3 1
ttable.0.4 0.5
ttable.1.4 0.5
ttable.0.5 0.5
ttable.1.5 0.5
ttable.0.6 1
ttable.1.7 1
}
# this creates a six channel soundcard
# and outputs to the eight channel one
# i.e. for usage in mplayer I had to define in ~/.mplayer/config:
# ao=alsa:device=dmix6
# channels=6
pcm.dmix6 {
type route
slave.pcm dmix8
slave.channels 8
ttable.0.0 1
ttable.1.1 1
ttable.2.2 1
ttable.3.3 1
ttable.4.4 1
ttable.5.5 1
ttable.6.6 1
ttable.7.7 1
}
# share the microphone, i.e. because virtualbox grabs it by default
pcm.microphone {
type dsnoop
ipc_key 1027
slave {
pcm "snd_card"
}
}
# rate conversion, needed i.e. for wine
pcm.2chplug {
type plug
slave.pcm "ch71dup"
}
pcm.a2chplug {
type plug
slave.pcm "dmix8"
}
# routes the channel for the alternative
# 2 channel output, which becomes the 7th and 8th channel
# on the real soundcard
#pcm.alt2ch {
# type route
# slave.pcm "a2chplug"
# slave.channels 8
# ttable.0.6 1
# ttable.1.7 1
#}
# skype and ekiga are only mono, so route left channel to the right channel
# note: this gets routed to the alternative 2 channels
pcm.mono_playback {
type route
slave.pcm "a2chplug"
slave.channels 8
# Send Skype channel 0 to the L and R speakers at full volume
#ttable.0.6 1
#ttable.0.7 1
}
# 'full-duplex' device for use with aoss
pcm.duplex {
type asym
playback.pcm "2chplug"
capture.pcm "microphone"
}
#pcm.aduplex {
# type asym
# playback.pcm "alt2ch"
# capture.pcm "microphone"
#}
pcm.monoduplex {
type asym
playback.pcm "mono_playback"
capture.pcm "microphone"
}
# for aoss
pcm.dsp0 "duplex"
ctl.mixer0 "duplex"
# softvol manages volume in alsa
# i.e. wine likes this
pcm.mainvol {
type softvol
slave.pcm "duplex"
control {
name "2ch-Upmix Master"
card 0
}
}
#pcm.!default "mainvol"
# set the default device according to the environment
# variable ALSA_DEFAULT_PCM and default to mainvol
pcm.!default {
@func refer
name { @func concat
strings [ "pcm."
{ @func getenv
vars [ ALSA_DEFAULT_PCM ]
default "mainvol"
}
]
}
}
# uncomment the following if you want to be able to control
# the mixer device through environment variables as well
#ctl.!default {
# @func refer
# name { @func concat
# strings [ "ctl."
# { @func getenv
# vars [ ALSA_DEFAULT_CTL
# ALSA_DEFAULT_PCM
# ]
# default "duplex"
# }
# ]
# }
#}
Surround51 incl. upmix stereo & dmix, swap L/R, bad speaker position in room
Bad practice but works fine for almost everything without additional per-program/file customization:
pcm.!default { type route ## forwards to the mixer pcm defined below slave.pcm dmix51 slave.channels 6 ## "Native Channels" stereo, swap left/right ttable.0.1 1 ttable.1.0 1 ## original normal left/right commented out # ttable.0.0 1 # ttable.1.1 1 ## route "native surround" so it still works but weaken signal (+ RL/RF swap) ## because my rear speakers are more like random than really behind me ttable.2.3 0.7 ttable.3.2 0.7 ttable.4.4 0.7 ttable.5.5 0.7 ## stereo => quad speaker "upmix" for "rear" speakers + swap L/R ttable.0.3 1 ttable.1.2 1 ## stereo L+R => join to Center & Subwoofer 50%/50% ttable.0.4 0.5 ttable.1.4 0.5 ttable.0.5 0.5 ttable.1.5 0.5 ## to test: "$ speaker-test -c6 -twav" and: "$ speaker-test -c2 -twav" } pcm.dmix51 { type dmix ipc_key 1024 # let multiple users share ipc_key_add_uid false # IPC permissions (octal, default 0600) # I think changing this fixed something - but I'm not sure what. ipc_perm 0660 # slave { ## this is specific to my hda_intel. Often hd:0 is just allready it; To find: $ aplay -L pcm surround51 # this rate makes my soundcard crackle # rate 44100 # this rate stops flash in firefox from playing audio, but I do not need that rate 48000 channels 6 ## Any other values in the 4 lines below seem to make my soundcard crackle, too period_time 0 period_size 1024 buffer_time 0 buffer_size 4096 } }
Loopback interface with dmix external interface
Used to control which output goes to external, loopback, or both. Others have reported working setups without specifying format
# Use this to output to external pcm.dmixerout { type dmix ipc_key 1024 ipc_key_add_uid false slave { pcm "hw:CARDNAME,0" channels 2 period_time 0 period_size 1024 buffer_size 4096 rate 44100 } bindings { 0 0 1 1 } } # Use this to output to loopback pcm.dmixerloop { type dmix ipc_key 2048 ipc_key_add_uid false slave { pcm "hw:Loopback,0,0" channels 2 period_time 0 period_size 1024 buffer_size 4096 # If format is absent ALSA gives me slave PCM not usable, but it works w/o it for others format S32_LE rate 44100 } bindings { 0 0 1 1 } } # Sends to the two dmix interfaces pcm.quad { type multi # Necessary to have both slaves be dmix; both as hw doesn't give errors, but wouldn't slaves.a.pcm "dmixerout" slaves.a.channels 2 slaves.b.pcm "dmixerloop" slaves.b.channels 2 bindings { 0 { slave a; channel 0; } 1 { slave a; channel 1; } 2 { slave b; channel 0; } 3 { slave b; channel 1; } } } # Duplicates to quad, use this to output to loopback & external pcm.stereo2quad { type route slave.pcm "quad" # ttable.A.B G # where A - input channel # B - output channel # G - volume gain (1.0 = original) ttable.0.0 1 ttable.1.1 1 ttable.0.2 1 ttable.1.3 1 } # Listens to loopback # trying to play to stereo2quad when something is already listening gives me slave PCM not usable # but listening when something is already playing on stereo2quad works # and so does starting to listen, then playing to dmixerloop pcm.loopin { type dsnoop ipc_key 1111 ipc_key_add_uid false slave.pcm "hw:Loopback,1" } pcm.!default { type asym playback.pcm "plug:stereo2quad" capture.pcm "plug:loopin" }
gollark: So my traffic light control software just `pcall`ed an energy-consuming function in a tight loop.
gollark: See, 1249174 years ago, in the early days of Plethora, the energy system was more visible and caused an error if there wasn't enough instead of just waiting.
gollark: Not the recipe thing.
gollark: I think I actually caused changes to *Plethora* itself!
gollark: It was an *accident*.
This article is issued from Archlinux. The text is licensed under Creative Commons - Attribution - Sharealike. Additional terms may apply for the media files.