Re: Tragic State of FreeBSD Audio/Sound

On Wed Jul 22, 2026 at 2:05 AM EEST, Shawn Webb wrote:
> On Wed, Jul 22, 2026 at 01:26:17AM +0300, Christos Margiolis wrote:
>> On Mon Jul 20, 2026 at 4:59 PM EEST, Shawn Webb wrote:
>> > I've been experiencing audio issues as well since the recent work
>> > started a little while ago (a year? two years? my chronological memory
>> > isn't the best.)
>> >
>> > It's broken bad enough that I now don't use audio on my systems at
>> > all. I haven't filed any bug reports, though, because it just hasn't
>> > been a priority for me. I've got tons of other systems where audio
>> > works fine and it's not really a bother to use those instead of my
>> > fbsd/hbsd systems.
>>=20
>> Can you mention one example of "bad enough"? I've been lurking this
>> thread to find concrete problems to work on and I only see stuff like
>> this.
>
> LibreWolf just does not play audio at all now. When I open
> pavucontrol with a video/audio tab open and playing, LibreWolf isn't
> even listed in the "Playback" tab. It used to prior to the sound
> system overhaul.

I daily-drive LibreWolf without any issues. I can listen play sound and
video call normally. But I use native OSS as the audio backend, not
pulseaudio or anything else.

You can try setting media.cubeb.backend =3D oss in about:config.

> Audio recording device issues as well. No matter what, I cannot get my
> wireless (via non-BT USB wireless dongle) headphone's microphone to
> work with video conferencing websites (in LibreWolf). I haven't tried
> anything with Chromium, though I suppose I could add that to the list
> of things to try.

Please feel free to open a PR about this and we can look into it. I
don't do BT work, but I suppose yours is a USB audio device, which
handles BT on its own.

> With me having zero experience in the audio stack, I'm not entirely
> sure where to start debugging--especially given the number of moving
> components. Is this a PulseAudio problem? Is this a kernel problem? Is
> this because we've hardened certain interfaces in HardenedBSD, and the
> audio stack could be impacted (unintentionally, of course)?
>
> I don't feel comfortable asking others to do that initial debugging
> work (since I don't feel entitled to dictate other's priorities and
> time). I don't know how well my issue can be reproduced. The zfs
> datasets for this laptop date back to 2013-ish, so my issues could
> also be due to mysterious old cruft surviving 13 years across multiple
> laptops.
>
> If I can find some spare time, I'll try to dig in. What would be
> helpful is some sort of "get started debugging the audio stack(s)"
> sort of guide. I remember seeing the audio stack receiving some
> coverage in the FreeBSD Journal, perhaps that could be my starting
> point.

Yes, the Journal article did mention how to file reports with sufficient
(as a first step) information.

> Thanks,

Christos