I wrote:
> Well, actually it doesn't. It'll happily start up, but trying to actually
> play anything outputs... nothing. With no error message or anything, it
> just produces no sound, and AFAICT never actually connects to PulseAudio
> (even if I try explicitly setting the ALSA output device to default or
> pulse). The command-line fluidsynth is broken the same way. (This shows
> another issue with the current situation: some programs claim to support
> several APIs, but only one gets actually tested.) Well, at least there's
> timidity++ which actually works...
FYI, the issue there is just with the default settings: there is no
soundfont configured by default in QSynth or fluidsynth. (Thanks to Jon
Escombe for pointing this out.) When I configure one by hand (why do I have
to do this? Should it not be the packager's job to set up a default
soundfont?), it "works", for some definition of "working". After bumping up
the buffer size (with the default, I just get noise as output), it suffers
from horrible arhythmy, probably due to ALSA being used both on the
sequencer end and on the output end, I've seen the same happen with
timidity++ back when I was still using dmix (now with PulseAudio,
timidity++ uses libao which uses the native PulseAudio protocol, so this
issue is gone). Jon Escombe also pointed out that the latest fluidsynth
release has a native PulseAudio driver (but unfortunately it's not
available in Fedora yet), that one should work better.
But this is getting way off topic for this list.
Kevin Kofler
--
fedora-devel-list mailing list
fedora-devel-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/fedora-devel-list
