https://bugs.winehq.org/show_bug.cgi?id=54543
Bug ID: 54543 Summary: Avantree Leaf USB audio device does not play sound in applications run via wine Product: Wine Version: 8.1 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs@winehq.org Reporter: linux@bernd-steinhauser.de Distribution: ---
I'm using an USB audio device called "Avantree Leaf" to pass audio via Bluetooth to a receiver. After this change in the kernel https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h... applications run by wine via the pulseaudio output cannot initiate a stream on the audio device and there is no sound. To test, I've used the audio player foobar2000, but it's the same for other Windows applications. Native Linux applications work without problems so far. The latest Wine version that I tested is 8.1, but I haven't found any between Wine versions in this regard. It also does not work, if I select the usb device specifically in winecfg or in foobar2000 itself. I do however get sound if I select another device, e.g. my onboard sound card or the HDMI output, thus I know it is a problem with the usb device only.
I usually use pipewire/wireplumber as an audio server with pipewire-pulse as a Pulseaudio replacement, but the problem also exists if I'm using standard Pulseaudio instead.
There are two possible workarounds that I can use to get sound from the device when running Wine application: 1. Disable the usb audio latency management by loading the module with the parameter lowlatency=0 2. Set clock.min-quantum 512 in pipewire/wireplumber, which I think also increases the latency a bit
The maintainer of the kernel usb audio driver thinks that this an application problem and that the kernel driver is working correctly. Therefore, I'm now opening this bug report. The corresponding bug report on the kernel side is, there I also have some additional information available, not sure if I should replicate all of that here: https://bugzilla.kernel.org/show_bug.cgi?id=216909
Also, see this discussion on a pipewire issue, which was the first report I did about this: https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/2739#note_1713023