No sound from laptop speakers on Omarchy

Silent built-in speakers on Omarchy are usually one of five unrelated faults: the stream is going somewhere else, the audio DSP never loaded firmware, the session manager built sinks with no ports, a hardware mixer control is muted where PipeWire cannot see it, or a smart amplifier refused to load its tuning blob. The steps below split them apart. Everything here was checked against the v4.0.4 source tree, with notes where 3.x differs.

The fix

1. Confirm where the sound is actually going. Open the audio panel with SUPER + CTRL + A, or from a terminal:

wpctl status
pactl list sinks | grep -E "Name:|Active Port:|Mute:|Volume:"

If the default sink is an HDMI monitor or a USB dock, pick the speakers in the panel. SHIFT + XF86AudioMute cycles outputs. If the only sink listed is Dummy Output, skip to step 3.

2. Restart the audio stack. On 4.x:

omarchy restart audio

The same thing lives in the Omarchy menu under Update > Hardware > Audio. omarchy-restart-audio restarts pipewire, pipewire-pulse and wireplumber, and if wpctl still does not answer it looks for a USB audio device stuck in SETUP state and resets it. On 3.x the command was omarchy-restart-pipewire, which only restarts the services.

3. No soundcard at all, or only Dummy Output, on an Intel laptop. Check the kernel log:

sudo dmesg | grep -i sof

Lines like SOF firmware and/or topology file not found and Check if you have 'sof-firmware' package installed mean the DSP never booted. Install the firmware and reload the driver, or reboot:

sudo pacman -S sof-firmware

Omarchy 3.8.3 fixed the install hook that caused this. Before it, the hook only matched Panther Lake, so Arrow Lake, Meteor Lake and similar machines shipped without sof-firmware, as hyprcat documented in issue #6110. Since 3.8.3 the gate is omarchy-hw-intel-sof, which matches any Intel audio controller, and the same release shipped a migration that adds the package on existing installs. If pacman -Q sof-firmware still comes back empty on an updated machine, install it by hand as above.

4. ASUS ROG laptops: remove the soft-mixer config. Omarchy installs ~/.config/wireplumber/wireplumber.conf.d/alsa-soft-mixer.conf on ROG hardware. It sets api.alsa.soft-mixer = true, which tells PipeWire to do volume in software and never touch the ALSA controls. beyondeye’s issue #4821 traced the consequence on an ALC285 Strix G16: the hardware Master switch started out muted at zero, the kernel’s jack detection switched Headphone off on every boot, and with the soft mixer in place PipeWire never ran the mixer paths that would unmute Master, Headphone and Speaker. That report was about the headphone jack, but owners of the Zephyrus G14, G16, M16 and Flow X13 confirmed the same fix in the thread. The fix that report gives:

rm ~/.config/wireplumber/wireplumber.conf.d/alsa-soft-mixer.conf
systemctl --user restart wireplumber pipewire pipewire-pulse

If you would rather keep the soft mixer, unmute and raise the hardware control by hand instead:

card=$(aplay -l | grep -i ALC285 | head -1 | sed 's/card \([0-9]*\).*/\1/')
amixer -c "$card" set Master 100% unmute
sudo alsactl store

Omarchy’s own script sets that control to 80 percent, which rdoupe measured as a 12.75 dB ceiling under the soft mixer in issue #11046.

5. Framework Laptop 13 AMD: reset the card profile. Omarchy’s install script for this machine puts the card on HiFi (Mic1, Mic2, Speaker), the profile that carries both microphones and the speakers. If pactl list cards shows a different active profile, put it back:

pactl list cards short
pactl set-card-profile <card-name> "HiFi (Mic1, Mic2, Speaker)"

6. Dell XPS with CS35L56 amps: read the amp log before changing anything.

journalctl -k -b | grep -i cs35l56

FIRMWARE_MISSING, Calibration disabled due to missing firmware controls and Can't read tuning IDs mean the amplifiers bound but never downloaded their tuning. On the XPS 13 DX13260 with SKU 0E53 this is an upstream kernel bug: the amps misread their speaker ID as 0, and linux-firmware only carries spkid1, spkid2 and spkid3 for that subsystem. Gundrak confirmed the root cause in issue #11320 and closed it as a duplicate of wizaj’s #9687, which is tracked at kernel.org. Cirrus supplied a driver patch for spi-cs42l43 that three owners on #9687 built as an out-of-tree module and confirmed, with the caveat that every kernel update reverts it until the module is rebuilt. Two things to know before you copy a recipe from those threads:

• wizaj retracted the symlink workaround that aliased spkid0 to spkid1. It made sound, but it applies protection parameters that were never validated for those drivers. Do not use it.
• The dell-xps13-sidecar-amps package, merged as omarchy-pkgs PR #150 alongside PR #7032, writes options snd_soc_sof_sdw quirk=65536. The same quirk value restored bass for jonnyace in #7427, but MBvisti’s #10543 shows the package producing complete silence on another 0E53 unit. PR #7032 landed on the quattro branch on 2026-08-25, yet the hook is absent from the v4.0.0 through v4.0.4 release tags, so nothing installs the package for you.

The only reversible thing to try is routing playback back through the codec instead of the sidecar amps, which a1local reported as audible on #9687:

echo 'options snd_soc_sof_sdw quirk=0' | sudo tee /etc/modprobe.d/sof-sdw-quirk.conf
sudo limine-mkinitcpio && reboot

Use sudo mkinitcpio -P instead if limine-mkinitcpio is not present. That route gives you the tweeters only, with no bass and no amp firmware. Delete the file to undo it.

7. Sinks exist, nothing links to them, and it started with a September update. Issue #11350 opened as a kernel regression between 7.1.9 and 7.2.3 and was corrected in its own thread: with WirePlumber 0.5.17 every ALSA sink is created with zero ports, wpctl get-volume @DEFAULT_AUDIO_SINK@ answers '-1' is not a valid ID, and raw ALSA playback with speaker-test -D hw:1,2 still works. zygote55 traced it to the v4l2 create-device script stalling the event dispatcher whenever a rule disables a camera device, which the intel-ipu7-camera package Omarchy installs on Intel laptops does. The fix is merged upstream but Arch still ships 0.5.17-1. Two reporters restored audio by downgrading wireplumber and libwireplumber to 0.5.15-1 from the pacman cache or the Arch Linux Archive:

pacman -Qi wireplumber | grep Version
sudo pacman -U /var/cache/pacman/pkg/{wireplumber,libwireplumber}-0.5.15-1-x86_64.pkg.tar.zst

Add IgnorePkg = wireplumber libwireplumber to /etc/pacman.conf until a fixed release lands, then remove it. zygote55’s comment on #11350 also has a one-line override of the script under ~/.local/share/wireplumber/scripts/ if you would rather not downgrade.

Verify it worked

speaker-test -c 2 -t wav -l 1
wpctl status | head -30
journalctl -k -b | grep -ciE 'firmware_missing|sof_probe_work failed'

You want an audible left and right test tone, a real sink name in wpctl status rather than Dummy Output, and a zero from the last command. If a speaker tuning applies to your machine, omarchy audio tuning status should print an active host service and a present tuning sink.

Why it happens

Three designs meet here. PipeWire and WirePlumber own routing. ALSA still owns the hardware mixer underneath, and Omarchy’s soft-mixer config deliberately hides that layer from PipeWire, which is why a muted hardware switch looks like a working desktop. Modern laptop speakers then add a third layer: Cirrus, TI and similar smart amplifiers that refuse to make sound until they load a tuning blob keyed to the machine’s subsystem ID. When the vendor never shipped the blob variant your unit asks for, no userspace setting can help.

On top of that, Quattro moved the audio UI. v4.0.0 replaced wiremix with the shell’s own audio panel and added output switching that preserves playback. v4.0.0 also introduced per-laptop speaker tunings, which are PipeWire filter chains in front of the speaker sink, currently shipping only for the 2026 XPS 14 and XPS 16.

If that did not work

Collect a report with omarchy-debug and search the tracker before filing. Cases with no clean fix yet:

• Lenovo Legion Pro 7 16IRX8H, where the TAS2781 amplifier binding changes with a codec SSID override, in issue #12086.
• MacBook9,1 and 10,1, where karlentwistle found in #12008 that the out-of-tree Cirrus driver is not enough on its own and the NVRAM startup chime has to be unmuted before firmware powers the speaker amp.
• The linux-omarchy kernel, which 4.0.4 installs on every x86 machine while leaving the previous kernel in place. No audio report has pinned a fault on it yet, and the one that tried, #11350, turned out to be WirePlumber. Booting the stock linux entry from the Limine menu rules it in or out in one reboot.

The Lenovo Yoga Pro 7 14IAH10 bass pin quirk already ships in install/hardware/lenovo/fix-yoga-pro7-bass-speakers.sh, so that model needs nothing from you.

Related

• /hardware/audio/ for the wider audio picture
• /hardware/dell-xps-13-2026/ and /hardware/asus-rog-zephyrus/
• /fix/bluetooth-stops-after-resume/ if only your headset is silent
• /releases/v4.0.4/ for what the kernel change in 4.0.4 covers
• Omarchy manual: Troubleshooting