Hi Terry, Le 7/11/24 à 06:00, Terry D. Cudney a écrit : > --> The problem will them be concurrent access to the audio card between > --> espeakup and speech-dispatcher. If in the desktop you use alsa and the > --> dmix module, that should work fine. > > Ah! That seems to be where the problem lies! The desktop is using pipewire/pulseaudio. > > > **--**> *** Key question: > What is the best way to disable pulseaudio/pipewire/wireplumber/whatever and force the desktop to use alsa, dmix? You do not have to. You can set in /etc/speech-dispatcher/speechd.conf alsa as AudioOutputMethod or set it as libao and set default_driver=alsa on /etc/libao.conf. Then you can redirect the audio output deop puse to dmix. For instance in /etc/pulse/default.pa: ### In Slint, we want to share audio resources between speech apps that ### rely on alsa and other apps that rely on pulseaudio. load-module module-alsa-sink device=dmix load-module module-alsa-source device=dsnoop With this setting pulse is not disabled (but started only in demand, by applications that require it), I have speech during boot up, both if the system stats in console mode or in a graphical environment, I can hear the music started in a graphical environment while in a tty. Caveat: I remember that Samuel had objections about this setup, I do not remember which ones, maybe he does? All I can say is I do not receive complaints by Slint users about these default settings. Another solution is proposed in the Arch wiki: https://wiki.archlinux.org/title/PulseAudio/Examples#PulseAudio_as_a_minimal_unintrusive_dumb_pipe_to_ALSA > -->If you use pulseaudio or pipewire > --> you'll get a conflict. This is a concern that I've never found time > --> to tackle, and I have hoped for years that somebody else would have > --> a look at it since it won't involve coding, but discussing properly > --> with pulseaudio/pipewire people to work out a proper solution. In the > --> meanwhile people have rather looked at workarounds and stay half-happy > --> with that... > --> > --> Samuel > > I'm happy to lose pulseaudio/pipewire et al and use alsa/dmix. > > Question out of curiosity: What does pulseaudio add; why are distributions seeming to abandon alsa in favour of pulseaudio? Well, to know everything about pulseaudio, you can read PulseAudio under the hood from Victor Gaydov: https://gavv.net/articles/pulseaudio-under-the-hood/ > My naive observation is that pulseaudio just adds a lot of complexity... Amazingly the introduction of pulseaudio in Slackware triggered the same question some years ago and even flame wars, leading Patrick Volkerding to during some time propose a variant "alsa only" ;) All this being said I know nothing about pipewire and systemd. Cheers, Didier
