Re: [orca-list] linux sound issues Re: Opensolaris Vs. Ubuntu
- From: Michael Whapples <mwhapples aim com>
- To: Halim Sahin <halim sahin t-online de>
- Cc: orca-list gnome org
- Subject: Re: [orca-list] linux sound issues Re: Opensolaris Vs. Ubuntu
- Date: Thu, 13 Nov 2008 12:17:22 +0000
I have to say I agree so much there.
To answer a few of the questions:
Opensolaris uses gnome-speech with espeak. This uses the standard Sun
audio system as it is the only system (as far as I can tell).
I don't think viavoice would work on solaris, but I may be wrong (having
looked for eclipse for solaris it seemed like they were unable to use
precompiled linux versions, although a recompile seemed to solve it, may
be that's all viavoice would need but as users we can't do that).
I agree that gnome-speech should provide audio output for the synths
rather than leaving it to the synth. If this was done then that is one
large problem with gnome-speech solved.
As for making speech-dispatcher default, does speech-dispatcher work on
all the platforms orca does? If not then this could be a problem. As
orca is part of the gnome desktop and so is gnome-speech then whilst it
may not work with the console, if the sound issue with gnome-speech was
seen to it wouldn't be hard to change to speech-dispatcher output in orca.
As for pulseaudio, it causes more headaches than its worth. When you
rely on audio working to use your computer (speech output) you want as
little as possible in the chain to go wrong. Whilst not having used
pulseaudio (because I have been unable to find a suitable set up with
it) I have heard comments that it can fail on people and that the
authors don't intend to improve its reliability (understandable if you
only want it to play sounds, music, etc, but a totally different matter
if you then have to recover the system without any output before you can
resume what you are doing). I suppose what I am getting at here is the
audio system to us is like the video system to a sighted user, could
they live with having a video server like pulseaudio (eg. could be
simulated by turning off the monitor when ever pulseaudio crashes and it
can only be turned back on when they have restored pulseaudio). If they
feel what I have suggested is too much, then pulseaudio is not suitable.
I feel ALSA is sufficient for a desktop use of audio. As for where a
sound server is needed (eg. for gnome's system sounds) I use ESD which
slackware provides with ALSA output compiled in. ESD mixes with all
other sounds coming through my ALSA devices.
Michael Whapples
On 23/12/42 20:59, Halim Sahin wrote:
Hello,
The mentioned audio problem in linux does only ixist because no configuration
was made to reflect accessibility needs.
I think I have written this about 20 times in this list:
Opensound drivers were not free which was the reason for creating
alsa.
OSS had no softmixing features at this time!!!
Currently a user can play and enjoy his favourite song together with speech
output.
This needs complex configuration because no distribution does a usefull
one at this point.
Currently the biggest problem is gnomespeech.
Orca uses as default gnomespeech to provide speech.
This is not a good solution because
1. It depends on gnome but we want to use the
textconsole as well
2. it doesn't handle soundoutput itself.
All synths must implement their own output methods
This is really bad for endusers.
They want use viavoice wich is (outdated).
It uses oss as default and BTW. does it run under opensolaris?
To use oss apps with orca we have two things.
1. use direct oss compat layer output e. G. through /dev/dsp (wich is the bad
solution).
2. Use aoss wrapper wich is should be prefered because of softmixing!!!
The best solution for all problems is Speech-dispatcher.
It has native alsa output and works great for me on all my machines
my drivers are:
snd-hda-intel, snd-emu10k1, snd-intel8x0 snd-via17xxx, snd-usb-audio.
speech-dispatcher plays the speechoutput and doesn't relay on synthesis output stuff.
I am voting for
1. using speech-dispatcher as default speech system under orca.
2. Distributions shouldn't try to route sound of speechsystems through soundservers like
pulseaudio!!!
Regards
Halim
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]