PulseAudio and Alsa

[Mitchell Smith]

Just to update the list on my progress, I have completely removed PulseAudio 
and reconfigured Ubuntu to use Alsa by default.

This seems to work for the most part, Orca is still working with the 
gnome-speech driver, and system sounds appear to be working correctly.

I then installed the speech-dispatcher speech-dispatcher-flite and 
python-speechd packages.

I first tried configuring speech-dispatcher as a user service after adding 
my user to the audio group, the test was successful and I can get spd-say to 
work correctly however when speech-dispatcher is running orca is no longer 
able to speak.

I also tried configuring speech-dispatcher as a system service with exactly 
the same result.

I am able to run python -c "import speechd" with no problem however Orca 
does not see speech-dispatcher as an option I can use as my speech system, 
only gnome-speech.

Of course I have to kill off speech-dispatcher in order to get Orca to read 
anything so maybe it can't see it because it's not running at the time.

I seem to recall that in the past if you used Alsa you are able to have 
multiple apps sharing the same sound device, it is only with OSS that you 
could only have one sound source at a time, is that not the case?

Any input on this subject would be greatly appreciated.  I seem to be pretty 
close to what I am trying to achieve, but just can't quite get it right.

Thanks,

Mitch

----- Original Message ----- 
From: "Mitchell Smith" <mjsotn at gmail.com>
To: <ubuntu-accessibility at lists.ubuntu.com>
Sent: Sunday, January 04, 2009 1:27 AM
Subject: PulseAudio and Alsa


> Hi,
>
> I raised this question on the gnome-orca list, however it is pretty Ubuntu 
> specific, so I thought I would ask here as well.  For those on both 
> mailing lists, please forgive the cross posting.
>
> I would like to use Alsa on my Ubuntu 8.10 installation, firstly because
> from reading the Orca wiki I see that speech-dispatcher is slightly more
> responsive with Alsa, and secondly I have a couple of third party apps 
> which
> require Alsa for sound output.
>
> What I wanted to ask was, can PulseAudio and Alsa co-exist on Ubuntu with
> minimal problems, or do I need to completely remove PulseAudio and 
> recompile
> any applications and libraries that depend on it to use Alsa instead?
>
> Obviously the best solution would be to leave the system as untouched as 
> possible, only having those applications that require low latency access 
> Alsa directly, and anything else less timing critical continue to use 
> PulseAudio, however I am not sure if both systems will play nicely 
> together.
>
> A quick Google turned up an article on how to disable PulseAudio in a 
> non-destructive way, but a read through some of the comments suggest this 
> does not come with out some significant side effects.
>
> So has anyone managed to get speech-dispatcher working nicely on Ubuntu 
> 8.10 with nice fast responsive speech, but keeping most of the PulseAudio 
> configuration in place, or does PulseAudio really need to go and I 
> reconfigure my system to do things the old way entirely with Alsa?  Or as 
> a final thought, would I be better off going with something a little more 
> minimalist like Debian or Gentoo where I can make those sorts of decisions 
> for myself and build my entire environment from scratch with out any 
> PulseAudio at all (obviously not the preferred approach).
>
> Any comments or suggestions in relation to this topic would be very much 
> appreciated.
>
> Cheers,
>
> Mitch
>
>