(In reply to Takashi Iwai from comment #10) .. > boot option for applying the same quirk entry as the device 1025:0283 (Acer > TravelMate 837 in this case). That's what I mentioned -- you can do > trial-and-error for all those entries, and it's a huge list. Yes really. It looks as if this approach will have to be partially automated in order to keep throughput times short. modprobe -v snd_hda_intl displays a list of loaded modules. I have transferred the loading of the modules mentioned to a script cat << EOF > test-load-snd-module lsmod | grep snd_ | gawk '{ print $1}' | sudo xargs rmmod sleep 2 insmod ....ko ... insmod snd_hda_intel.ko. model=$1 sleep 2 speaker-test -twav -c2 -Dhw:0 EOF chmod a+x test-load-snd-module and can now do this with ./test-load-snd-module 1025:0283 to start it. Are there any disadvantages with this procedure compared to a full reboot, which would take much longer? > Maybe you'd better to figure out a similar model and try those at first. I only found the following two entries: SND_PCI_QUIRK(0x10ec, 0x118c, "Medion EE4254 MD62100", ALC256_FIXUP_MEDION_HEADSET_NO_PRESENCE), SND_PCI_QUIRK(0x161f, 0x205d, "Medion Rim 2150", ALC880_FIXUP_MEDION_RIM), ./test-load-snd-module 10ec:118c ./test-load-snd-module 161f:205d but this did not help. Then I tried: ./test-load-snd-module auto ./test-load-snd-module nofix ./test-load-snd-module generic with no luck. The current problem is mentioned on https://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound.git/tree/Documentation/sound/hd-audio/notes.rst#n230 and reference is made to an external amplifier device as a possible source of error. So I tried SND_PCI_QUIRK(0x17aa, 0x9e54, "LENOVO NB", ALC269_FIXUP_LENOVO_EAPD), ./test-load-snd-module 17aa:9e54 and ./test-load-snd-module alc285-hp-amp-init ./test-load-snd-module alc298-spk-volume but none of them worked. It looks like detailed information about the actual configuration of the chip is required to get any further. Is there such a possibility ?