Re: [orca-list] audio and pulseaudio: was: Re: Punctuation, capital letters, exchange of characters and strings, generally error in the design of Orca

[Michael Whapples <mwhapples aim com>]

On Tue, 2008-04-15 at 13:50 +0200, Hermann wrote:
> Hi,
> now I've tested Hardy for a few hours, and I must say that I'm completely 
> fed up with that "sound improvements":
That's how I felt when using pulseaudio for one evening (I was using it
on slackware for notes).
> If you want to be happy with Hardy, stay away from SD!
> When installing SD, you have to realize, that it works only with Espeak, 
> when you change the output to "oss". But what advantage do you get from 
> using Pulseaudio? The sound mixing has gone then.
> Neither Alsa nor Pulse work as SD output modules for Espeak.
Not knowing about ubuntu specifically, but part of the problem might be
to do with pulseaudio taking exclusive control of the ALSA device. If
you want to use ALSA applications with pulseaudio, then you are meant to
set up the pulse ALSA device in the .asoundrc or asound.conf files and
then get the ALSA application to use that device. You can set the pulse
ALSA device as the default ALSA device but then make sure you change the
pulseaudio output device!
> But it came harder: To avoid this I tried to switch to Espeak-generic as 
> Output. To make a long story short, it breaks the Pulseaudio server, so 
> that 
> you can't use neither Gnome-Speech nor any media player at all. As soon as 
> Alsa is involved, everything is broken. You can use SD in the console only. 
May be the problems are as above. The only problem I have with the
above, how does one alter the output device of ALSA applications when
pulseaudio gets in between/gets in the way?
> In Gnome SD talks over Espeak-generic, but that's the only thing you hear. 
> No system sounds, no media files.
That is probably due to the pulseaudio system not gaining exclusive
control of the audio device so not starting, in which case it means your
sound system is now broken.
> So folks, what the hell have you done? If you want to bann SD, be that 
> honnest and admit it. I want to use SD with brltty in the console, and 
> setting to the OSS output forces me to use Alsa-oss, and I don't know 
> whether this will work in this setting; haven't tested that. It's enough 
> for today.
As I understand it, pulseaudio in ubuntu is being run per user in gnome,
so when you don't have gnome running (IE. working purely in the console)
pulseaudio won't be running, so anything which would try to output via
pulseaudio (or its compatability system) will have broken audio and no
output will be heard. On the other hand, use SD with ALSA directly on an
actual ALSA hardware device, and pulseaudio will now be broken.
> The only solution would be to completely deactivate Pulseaudio and restore 
> the settings known from Gutsy. They are not perfect, but compared to that 
> Hardy Pulseaudio stuff they seem to be paradise. Hacking Gnome-speech by 
> using aoss is a small issue compared to the stress you get with Hardy, when 
> you don't want to use the standard settings.
Either when you don't want the defaults, or when your distro doesn't
provide pulseaudio as default, and it is this complexity which puts me
off it. As I said I have ALSA set up quite well and it covers most
software, but for pulseaudio I would still need to have alsa configured
but also configure pulseaudio. 

The only other possibility which is occurring to me, although I haven't
tried it, is to set up a dmix device in .asoundrc or asound.conf and
have other virtual devices using that, one for pulseaudio and one for
other ALSA software. This doesn't solve the complexity problems though.

> Hermann
Michael Whapples