Problems with Software Speech

Problems with Software Speech

[ace]

Hi all

I just spent the last two weeks building a LFS system then installing 
necessary packages from the BLFS book.  Now, I am to the point of installing 
software speech.  Speech-Dispatcher is installed, as is Speechd-Up.  Espeak 
is my synth of choice.  I also setup my kernel to compile sftsyn as a 
module.  In the default synth field in the kernel, though, I have ltlk as 
that is my primary synth.  I have verified that Speech-Dispatcher and 
Speechd-up are working because I performed the speechd-up probe test and it 
was spoken.  The problem happens when I do modprobe speakup_sftsyn.  The 
system simply freezes; no speech, no beeps, nothing.  In fact, I'm forced to 
restart forcefully with the ower button.  Where should I look to find 
whatever error may be occurring?  If I need to get sighted assistance, 
that's not a problem.

Thanks,
Robby 

Re: Problems with Software Speech

[Gregory Nowak]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

It seems you may be running into a similar problem which I reported
happens when switching synths with the doubletalk pc driver built into
the kernel, and set as the default. Try starting the system with none
as the default synth, and then loading software speech, starting with
the sftsyn module. If that works like it's supposed to, then it looks
like the ltlk driver has issues as well.

Greg


On Mon, Jun 18, 2007 at 03:21:19PM -0400, ace wrote:
> Hi all
> 
> I just spent the last two weeks building a LFS system then installing 
> necessary packages from the BLFS book.  Now, I am to the point of installing 
> software speech.  Speech-Dispatcher is installed, as is Speechd-Up.  Espeak 
> is my synth of choice.  I also setup my kernel to compile sftsyn as a 
> module.  In the default synth field in the kernel, though, I have ltlk as 
> that is my primary synth.  I have verified that Speech-Dispatcher and 
> Speechd-up are working because I performed the speechd-up probe test and it 
> was spoken.  The problem happens when I do modprobe speakup_sftsyn.  The 
> system simply freezes; no speech, no beeps, nothing.  In fact, I'm forced to 
> restart forcefully with the ower button.  Where should I look to find 
> whatever error may be occurring?  If I need to get sighted assistance, 
> that's not a problem.
> 
> Thanks,
> Robby 
> 
> 
> _______________________________________________
> Speakup mailing list
> Speakup@braille.uwo.ca
> http://speech.braille.uwo.ca/mailman/listinfo/speakup

- -- 
web site: http://www.romuald.net.eu.org
gpg public key: http://www.romuald.net.eu.org/pubkey.asc
skype: gregn1
(authorization required, add me to your contacts list first)

- --
Free domains: http://www.eu.org/ or mail dns-manager@EU.org
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFGdtza7s9z/XlyUyARAowVAJsG+gI++/HpsVLhbDvQDfHsLexjngCfUbHz
f42LsQPVqBrB0CYymgbaDxc=
=Bcwz
-----END PGP SIGNATURE-----

Re: Problems with Software Speech

[Jan Buchal]

>>>>> "a" == ace  <ace@talkingirc.net> writes:

    a> Hi all I just spent the last two weeks building a LFS system then
    a> installing necessary packages from the BLFS book. Now, I am to
    a> the point of installing software speech. Speech-Dispatcher is
    a> installed, as is Speechd-Up. Espeak is my synth of choice. I also
    a> setup my kernel to compile sftsyn as a module. In the default
    a> synth field in the kernel, though, I have ltlk as that is my
    a> primary synth. I have verified that Speech-Dispatcher and
    a> Speechd-up are working because I performed the speechd-up probe
    a> test and it was spoken. The problem happens when I do modprobe
    a> speakup_sftsyn. The system simply freezes; no speech, no beeps,
    a> nothing. In fact, I'm forced to restart forcefully with the ower
    a> button. Where should I look to find whatever error may be
    a> occurring? If I need to get sighted assistance, that's not a
    a> problem.

It seems that modprobe freezes. You can test it if you will have the
other console and there you do killall modprobe, then do modprobe
speakup_softsyn again, I think that will works. If yes, you can use the
script which I sent here in one of previous email.

Have a nice day

-- 

Jan Buchal
Tel: (00420) 24 24 86 008
Mob: (00420) 608023021

Re: Problems with Software Speech

[ace]

Greg,

That was indeed the problem.  I carried out your steps and it works.  I 
still don't think that software speech under Linux is all that impressive. 
According to my experiences, it is somewhat slow and often dies completely 
if a lot of text is sent to it too quickly.  Could be just my hardware but 
others have had similar experiences.

In the espeak-generic.conf file, should I have left it creating wav files in 
the command line parameter?  I took that out so that text was sent to the 
espeak application directly; seems to me that making wav files would just 
slow the process down even more.

Thanks,
Robby
----- Original Message ----- 
From: "Gregory Nowak" <greg@romuald.net.eu.org>
To: "Speakup is a screen review system for Linux." <speakup@braille.uwo.ca>
Sent: Monday, June 18, 2007 3:28 PM
Subject: Re: Problems with Software Speech


> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> It seems you may be running into a similar problem which I reported
> happens when switching synths with the doubletalk pc driver built into
> the kernel, and set as the default. Try starting the system with none
> as the default synth, and then loading software speech, starting with
> the sftsyn module. If that works like it's supposed to, then it looks
> like the ltlk driver has issues as well.
>
> Greg
>
>
> On Mon, Jun 18, 2007 at 03:21:19PM -0400, ace wrote:
>> Hi all
>>
>> I just spent the last two weeks building a LFS system then installing
>> necessary packages from the BLFS book.  Now, I am to the point of 
>> installing
>> software speech.  Speech-Dispatcher is installed, as is Speechd-Up. 
>> Espeak
>> is my synth of choice.  I also setup my kernel to compile sftsyn as a
>> module.  In the default synth field in the kernel, though, I have ltlk as
>> that is my primary synth.  I have verified that Speech-Dispatcher and
>> Speechd-up are working because I performed the speechd-up probe test and 
>> it
>> was spoken.  The problem happens when I do modprobe speakup_sftsyn.  The
>> system simply freezes; no speech, no beeps, nothing.  In fact, I'm forced 
>> to
>> restart forcefully with the ower button.  Where should I look to find
>> whatever error may be occurring?  If I need to get sighted assistance,
>> that's not a problem.
>>
>> Thanks,
>> Robby
>>
>>
>> _______________________________________________
>> Speakup mailing list
>> Speakup@braille.uwo.ca
>> http://speech.braille.uwo.ca/mailman/listinfo/speakup
>
> - -- 
> web site: http://www.romuald.net.eu.org
> gpg public key: http://www.romuald.net.eu.org/pubkey.asc
> skype: gregn1
> (authorization required, add me to your contacts list first)
>
> - --
> Free domains: http://www.eu.org/ or mail dns-manager@EU.org
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.6 (GNU/Linux)
>
> iD8DBQFGdtza7s9z/XlyUyARAowVAJsG+gI++/HpsVLhbDvQDfHsLexjngCfUbHz
> f42LsQPVqBrB0CYymgbaDxc=
> =Bcwz
> -----END PGP SIGNATURE-----
>
> _______________________________________________
> Speakup mailing list
> Speakup@braille.uwo.ca
> http://speech.braille.uwo.ca/mailman/listinfo/speakup
>
>
> __________ NOD32 2336 (20070618) Information __________
>
> This message was checked by NOD32 antivirus system.
> http://www.eset.com
>
> 

Re: Problems with Software Speech

[Michael Whapples]

The reason for the wave file being generated was for two reasons.
Firstly by using aplay I found better compatability with my sound cards,
some were taking a long time to start the speech when letting espeak use
portaudio for output. Secondly portaudio v18 didn't support alsa, and
while portaudio v19 can use alsa, at the time not all distros, and see
first comment as portaudio v19 seemed just as bad or possibly worse when
using the alsa driver, so as other speech-dispatcher generic modules use
creating a temp wave file I decided to go with that (for me the results
weren't bad, still much better than flite or festival for
responsiveness).

Further to the above, some users of orca have found stuttering with the
gnome-speech espeak driver (which produces audio through portaudio), but
found by using the speech-dispatcher module (when creating a wave file)
the stuttering went away.

Admittedly it probably isn't the most efficient way to create speech,
but it seems the most reliable to me, but if you find you don't need
this then why not customise it for yourself. 

From
Michael Whapples
On Tue, 2007-06-19 at 11:16 -0400, ace wrote:
> Greg,
> 
> That was indeed the problem.  I carried out your steps and it works.  I 
> still don't think that software speech under Linux is all that impressive. 
> According to my experiences, it is somewhat slow and often dies completely 
> if a lot of text is sent to it too quickly.  Could be just my hardware but 
> others have had similar experiences.
> 
> In the espeak-generic.conf file, should I have left it creating wav files in 
> the command line parameter?  I took that out so that text was sent to the 
> espeak application directly; seems to me that making wav files would just 
> slow the process down even more.
> 
> Thanks,
> Robby
> ----- Original Message ----- 
> From: "Gregory Nowak" <greg@romuald.net.eu.org>
> To: "Speakup is a screen review system for Linux." <speakup@braille.uwo.ca>
> Sent: Monday, June 18, 2007 3:28 PM
> Subject: Re: Problems with Software Speech
> 
> 
> 
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA1
> >
> > It seems you may be running into a similar problem which I reported
> > happens when switching synths with the doubletalk pc driver built into
> > the kernel, and set as the default. Try starting the system with none
> > as the default synth, and then loading software speech, starting with
> > the sftsyn module. If that works like it's supposed to, then it looks
> > like the ltlk driver has issues as well.
> >
> > Greg
> >
> >
> > On Mon, Jun 18, 2007 at 03:21:19PM -0400, ace wrote:
> >> Hi all
> >>
> >> I just spent the last two weeks building a LFS system then installing
> >> necessary packages from the BLFS book.  Now, I am to the point of 
> >> installing
> >> software speech.  Speech-Dispatcher is installed, as is Speechd-Up. 
> >> Espeak
> >> is my synth of choice.  I also setup my kernel to compile sftsyn as a
> >> module.  In the default synth field in the kernel, though, I have ltlk as
> >> that is my primary synth.  I have verified that Speech-Dispatcher and
> >> Speechd-up are working because I performed the speechd-up probe test and 
> >> it
> >> was spoken.  The problem happens when I do modprobe speakup_sftsyn.  The
> >> system simply freezes; no speech, no beeps, nothing.  In fact, I'm forced 
> >> to
> >> restart forcefully with the ower button.  Where should I look to find
> >> whatever error may be occurring?  If I need to get sighted assistance,
> >> that's not a problem.
> >>
> >> Thanks,
> >> Robby
> >>
> >>
> >> _______________________________________________
> >> Speakup mailing list
> >> Speakup@braille.uwo.ca
> >> http://speech.braille.uwo.ca/mailman/listinfo/speakup
> >
> > - -- 
> > web site: http://www.romuald.net.eu.org
> > gpg public key: http://www.romuald.net.eu.org/pubkey.asc
> > skype: gregn1
> > (authorization required, add me to your contacts list first)
> >
> > - --
> > Free domains: http://www.eu.org/ or mail dns-manager@EU.org
> > -----BEGIN PGP SIGNATURE-----
> > Version: GnuPG v1.4.6 (GNU/Linux)
> >
> > iD8DBQFGdtza7s9z/XlyUyARAowVAJsG+gI++/HpsVLhbDvQDfHsLexjngCfUbHz
> > f42LsQPVqBrB0CYymgbaDxc=
> > =Bcwz
> > -----END PGP SIGNATURE-----
> >
> > _______________________________________________
> > Speakup mailing list
> > Speakup@braille.uwo.ca
> > http://speech.braille.uwo.ca/mailman/listinfo/speakup
> >
> >
> > __________ NOD32 2336 (20070618) Information __________
> >
> > This message was checked by NOD32 antivirus system.
> > http://www.eset.com
> >
> > 
> 
> 
> 
> 

Re: Problems with software speech

[Michael Whapples]

I think the alsa configuration still applies. They are all called things
like asound.rc or .asoundrc.conf. Doing a quick google brings up
www.thepenguin.org.uk/alsa/ which seems to discuss a very similar case
for software mixing and contains a link to the alsa dmixer page.

If you are struggling with using alsamixer to get a list of the
controls, then you may find it useful to define a window and use speakup
in read window mode. Use speakup-key and f2 for setting the window (eg.
on my system in alsamixer I would use the review keys to go to the start
of the 5th row and then press speakup-key and f2 and then use the review
keys to go to the end of the row and press speakup-key and f2 again,
then switch to read window mode. You may find the item row is a
different row on your system), now when you use the cursor keys the item
will be spoken.

From
Michael Whapples

Re: Problems with software speech

[Alex Snow]

I've seen lots of standard PCs with this built-in speaker, back when I 
worked in a primarily Dell shop, we had the option of ordering 
machines for clients with this option.  Also the ancient deskpro 4000 
I run as a server has one of these, and I've seen many more machines 
that do also.
On 
Sun, Mar 11, 2007 at 04:57:34PM +0000, Michael Whapples wrote:
> On Sun, 2007-03-11 at 09:08 -0400, Erik Heil wrote:
> > Hi their.  This isn't related to Intel hardware, but this is actually 
> > something in my opion they should implement.  It wouldn't require much 
> > cost up-front at all.  On my Sun workstation, their is actually a 
> > dedicated headphone output on the back of the machine.  Also, to make 
> > things better, their is actually a real speaker embedded in the front.  So 
> > this means that even if you didn't have a set of speakers, flite, 
> > festival, etc. will by default use /dev/audio which in this case by 
> > default accesses the system's speaker.  No of course if you want to use 
> > speakers, go right head, the capability is there.  No sorround sound, but 
> > certainly CD-quality audio.  Along with this, you have your standard 
> > line-in and MIC inputs.
> > That's interesting about the in built speaker, probably is useful.
> I question though the advantage of a headphone socket on the back,
> surely the front would be better. Some PC cases have front panel audio
> connectors for headphones, but I don't know if they automatically will
> turn off the speaker connection, and it does also require the soundcard
> has front panel connectors to accept the wires for these sockets.
> > On Sun, 11 Mar 2007, Michael Whapples wrote:
> > 
> > > Date: Sun, 11 Mar 2007 11:23:17 +0000
> > > From: Michael Whapples <mikster4@msn.com>
> > > Reply-To: Speakup is a screen review system for Linux.
> > >     <speakup@braille.uwo.ca>
> > > To: Speakup is a screen review system for Linux. <speakup@braille.uwo.ca>
> > > Subject: Re: Problems with software speech
> > > 
> > > On Sat, 2007-03-10 at 23:08 -0600, Farhan wrote:
> > >> Hello, i'm not sure about your second question about the headphone issues, but most laptop soundcards now a days are cingle channel only, they don't do multichannel, sinse Windows xp has some weird kernel implementation to do that all for you, I think the only true multichannel soundcards you can get for laptops are the soundblaster external pcmcia cards.
> > >>
> > > Let's be helpful here, now the reason may be known, but how to solve it.
> > > Well as far as I know, alsa can do software mixing as well, you will
> > > need to use a dmix device in the alsa configuration files. All the
> > > possibilities with alsa configuration is quite wide, so it would be best
> > > to read about it in the alsa documentation or search the internet for a
> > > suitable solution. Also make sure that the software itself is really
> > > using alsa, some programs still use OSS (trplayer, realplayer, flite,
> > > freetts (so firevox plugin if you are using freetts), etc). If you need
> > > to use something which uses OSS then it is still possible to use
> > > software mixing, get it to use the alsa-OSS compatibility system by
> > > putting the aoss command at the beginning of your command for the app
> > > which uses OSS (eg. "aoss realplayer" will run realplayer with OSS
> > > output going through alsa-OSS compatibility layer).
> > >
> > > I don't actually know about your headphone issue, but it sounds like
> > > alsa doesn't know about configuring (or it isn't configuring) a setting
> > > for muting speakers when using headphones (it sounds like the switch is
> > > software based rather than being a simple switch system. I think due to
> > > the comment about the behaviour when using windows). Guesses at a
> > > possible fix:
> > > Check that alsa doesn't have a control for it somewhere.
> > > See if the problem is solved in updated versions of alsa (if you are not
> > > using the latest).
> > > Last resort might be to find out if headphone and speaker volume can be
> > > set separately (I know my laptop soundcard alsa gives a headphone
> > > volume), and then you create two scripts (or one which either takes
> > > options or knows how to toggle) that you run to mute speakers but not
> > > headphones and to unmute speakers, and then you run this/these when
> > > connecting and disconnecting headphones.
> > >
> > > Hope some (if not all) of this is useful.
> > >
> > > From
> > > Michael Whapples
> > >
> > >
> > >
> > >
> > > _______________________________________________
> > > Speakup mailing list
> > > Speakup@braille.uwo.ca
> > > http://speech.braille.uwo.ca/mailman/listinfo/speakup
> > >
> > 
> > eheil@sdf.lonestar.org
> > SDF Public Access UNIX System - http://sdf.lonestar.org
> > 
> > 
> 
> 
> 
> _______________________________________________
> Speakup mailing list
> Speakup@braille.uwo.ca
> http://speech.braille.uwo.ca/mailman/listinfo/speakup

-- 
I once witnessed a long-winded, month-long flamewar over the use of
mice vs. trackballs... It was very silly.
	-- Matt Welsh

Re: Problems with software speech

[Zachary Kline]

Hello,
    Thank you very much for this information.  I am using the in-kernel Alsa 
support, so am a little unsure if configuration files apply here.  Might 
anybody be able to enlighten me on this?
I've checked my alsamixer setup, and there doesn't seem to be any kind of 
headphone line volume.  It's funny: when I boot into Windows first, I have 
master and PCM, and when I don't, I have...Something else.  I am not very 
good at diagnosing Alsa problems, and not good either at navigating 
Alsamixer, so...
Thanks for all your help, and I hope to hear from you soon.
Have a good day,
Zack.

----- Original Message ----- 
From: "Michael Whapples" <mikster4@msn.com>
To: "Speakup is a screen review system for Linux." <speakup@braille.uwo.ca>
Sent: Sunday, March 11, 2007 4:23 AM
Subject: Re: Problems with software speech
> Let's be helpful here, now the reason may be known, but how to solve it.
> Well as far as I know, alsa can do software mixing as well, you will
> need to use a dmix device in the alsa configuration files. All the
> possibilities with alsa configuration is quite wide, so it would be best
> to read about it in the alsa documentation or search the internet for a
> suitable solution. Also make sure that the software itself is really
> using alsa, some programs still use OSS (trplayer, realplayer, flite,
> freetts (so firevox plugin if you are using freetts), etc). If you need
> to use something which uses OSS then it is still possible to use
> software mixing, get it to use the alsa-OSS compatibility system by
> putting the aoss command at the beginning of your command for the app
> which uses OSS (eg. "aoss realplayer" will run realplayer with OSS
> output going through alsa-OSS compatibility layer).
>
> I don't actually know about your headphone issue, but it sounds like
> alsa doesn't know about configuring (or it isn't configuring) a setting
> for muting speakers when using headphones (it sounds like the switch is
> software based rather than being a simple switch system. I think due to
> the comment about the behaviour when using windows). Guesses at a
> possible fix:
> Check that alsa doesn't have a control for it somewhere.
> See if the problem is solved in updated versions of alsa (if you are not
> using the latest).
> Last resort might be to find out if headphone and speaker volume can be
> set separately (I know my laptop soundcard alsa gives a headphone
> volume), and then you create two scripts (or one which either takes
> options or knows how to toggle) that you run to mute speakers but not
> headphones and to unmute speakers, and then you run this/these when
> connecting and disconnecting headphones.
>
> Hope some (if not all) of this is useful.
>
> From
> Michael Whapples
>
>
>
>
> _______________________________________________
> Speakup mailing list
> Speakup@braille.uwo.ca
> http://speech.braille.uwo.ca/mailman/listinfo/speakup
> 

Re: Problems with software speech

[Michael Whapples]

On Sun, 2007-03-11 at 09:08 -0400, Erik Heil wrote:
> Hi their.  This isn't related to Intel hardware, but this is actually 
> something in my opion they should implement.  It wouldn't require much 
> cost up-front at all.  On my Sun workstation, their is actually a 
> dedicated headphone output on the back of the machine.  Also, to make 
> things better, their is actually a real speaker embedded in the front.  So 
> this means that even if you didn't have a set of speakers, flite, 
> festival, etc. will by default use /dev/audio which in this case by 
> default accesses the system's speaker.  No of course if you want to use 
> speakers, go right head, the capability is there.  No sorround sound, but 
> certainly CD-quality audio.  Along with this, you have your standard 
> line-in and MIC inputs.
> That's interesting about the in built speaker, probably is useful.
I question though the advantage of a headphone socket on the back,
surely the front would be better. Some PC cases have front panel audio
connectors for headphones, but I don't know if they automatically will
turn off the speaker connection, and it does also require the soundcard
has front panel connectors to accept the wires for these sockets.
> On Sun, 11 Mar 2007, Michael Whapples wrote:
> 
> > Date: Sun, 11 Mar 2007 11:23:17 +0000
> > From: Michael Whapples <mikster4@msn.com>
> > Reply-To: Speakup is a screen review system for Linux.
> >     <speakup@braille.uwo.ca>
> > To: Speakup is a screen review system for Linux. <speakup@braille.uwo.ca>
> > Subject: Re: Problems with software speech
> > 
> > On Sat, 2007-03-10 at 23:08 -0600, Farhan wrote:
> >> Hello, i'm not sure about your second question about the headphone issues, but most laptop soundcards now a days are cingle channel only, they don't do multichannel, sinse Windows xp has some weird kernel implementation to do that all for you, I think the only true multichannel soundcards you can get for laptops are the soundblaster external pcmcia cards.
> >>
> > Let's be helpful here, now the reason may be known, but how to solve it.
> > Well as far as I know, alsa can do software mixing as well, you will
> > need to use a dmix device in the alsa configuration files. All the
> > possibilities with alsa configuration is quite wide, so it would be best
> > to read about it in the alsa documentation or search the internet for a
> > suitable solution. Also make sure that the software itself is really
> > using alsa, some programs still use OSS (trplayer, realplayer, flite,
> > freetts (so firevox plugin if you are using freetts), etc). If you need
> > to use something which uses OSS then it is still possible to use
> > software mixing, get it to use the alsa-OSS compatibility system by
> > putting the aoss command at the beginning of your command for the app
> > which uses OSS (eg. "aoss realplayer" will run realplayer with OSS
> > output going through alsa-OSS compatibility layer).
> >
> > I don't actually know about your headphone issue, but it sounds like
> > alsa doesn't know about configuring (or it isn't configuring) a setting
> > for muting speakers when using headphones (it sounds like the switch is
> > software based rather than being a simple switch system. I think due to
> > the comment about the behaviour when using windows). Guesses at a
> > possible fix:
> > Check that alsa doesn't have a control for it somewhere.
> > See if the problem is solved in updated versions of alsa (if you are not
> > using the latest).
> > Last resort might be to find out if headphone and speaker volume can be
> > set separately (I know my laptop soundcard alsa gives a headphone
> > volume), and then you create two scripts (or one which either takes
> > options or knows how to toggle) that you run to mute speakers but not
> > headphones and to unmute speakers, and then you run this/these when
> > connecting and disconnecting headphones.
> >
> > Hope some (if not all) of this is useful.
> >
> > From
> > Michael Whapples
> >
> >
> >
> >
> > _______________________________________________
> > Speakup mailing list
> > Speakup@braille.uwo.ca
> > http://speech.braille.uwo.ca/mailman/listinfo/speakup
> >
> 
> eheil@sdf.lonestar.org
> SDF Public Access UNIX System - http://sdf.lonestar.org
> 
> 

Re: Problems with software speech

[Erik Heil]

Hi their.  This isn't related to Intel hardware, but this is actually 
something in my opion they should implement.  It wouldn't require much 
cost up-front at all.  On my Sun workstation, their is actually a 
dedicated headphone output on the back of the machine.  Also, to make 
things better, their is actually a real speaker embedded in the front.  So 
this means that even if you didn't have a set of speakers, flite, 
festival, etc. will by default use /dev/audio which in this case by 
default accesses the system's speaker.  No of course if you want to use 
speakers, go right head, the capability is there.  No sorround sound, but 
certainly CD-quality audio.  Along with this, you have your standard 
line-in and MIC inputs.

On Sun, 11 Mar 2007, Michael Whapples wrote:

> Date: Sun, 11 Mar 2007 11:23:17 +0000
> From: Michael Whapples <mikster4@msn.com>
> Reply-To: Speakup is a screen review system for Linux.
>     <speakup@braille.uwo.ca>
> To: Speakup is a screen review system for Linux. <speakup@braille.uwo.ca>
> Subject: Re: Problems with software speech
> 
> On Sat, 2007-03-10 at 23:08 -0600, Farhan wrote:
>> Hello, i'm not sure about your second question about the headphone issues, but most laptop soundcards now a days are cingle channel only, they don't do multichannel, sinse Windows xp has some weird kernel implementation to do that all for you, I think the only true multichannel soundcards you can get for laptops are the soundblaster external pcmcia cards.
>>
> Let's be helpful here, now the reason may be known, but how to solve it.
> Well as far as I know, alsa can do software mixing as well, you will
> need to use a dmix device in the alsa configuration files. All the
> possibilities with alsa configuration is quite wide, so it would be best
> to read about it in the alsa documentation or search the internet for a
> suitable solution. Also make sure that the software itself is really
> using alsa, some programs still use OSS (trplayer, realplayer, flite,
> freetts (so firevox plugin if you are using freetts), etc). If you need
> to use something which uses OSS then it is still possible to use
> software mixing, get it to use the alsa-OSS compatibility system by
> putting the aoss command at the beginning of your command for the app
> which uses OSS (eg. "aoss realplayer" will run realplayer with OSS
> output going through alsa-OSS compatibility layer).
>
> I don't actually know about your headphone issue, but it sounds like
> alsa doesn't know about configuring (or it isn't configuring) a setting
> for muting speakers when using headphones (it sounds like the switch is
> software based rather than being a simple switch system. I think due to
> the comment about the behaviour when using windows). Guesses at a
> possible fix:
> Check that alsa doesn't have a control for it somewhere.
> See if the problem is solved in updated versions of alsa (if you are not
> using the latest).
> Last resort might be to find out if headphone and speaker volume can be
> set separately (I know my laptop soundcard alsa gives a headphone
> volume), and then you create two scripts (or one which either takes
> options or knows how to toggle) that you run to mute speakers but not
> headphones and to unmute speakers, and then you run this/these when
> connecting and disconnecting headphones.
>
> Hope some (if not all) of this is useful.
>
> From
> Michael Whapples
>
>
>
>
> _______________________________________________
> Speakup mailing list
> Speakup@braille.uwo.ca
> http://speech.braille.uwo.ca/mailman/listinfo/speakup
>

eheil@sdf.lonestar.org
SDF Public Access UNIX System - http://sdf.lonestar.org

Re: Problems with software speech

[Michael Whapples]

On Sat, 2007-03-10 at 23:08 -0600, Farhan wrote:
> Hello, i'm not sure about your second question about the headphone issues, but most laptop soundcards now a days are cingle channel only, they don't do multichannel, sinse Windows xp has some weird kernel implementation to do that all for you, I think the only true multichannel soundcards you can get for laptops are the soundblaster external pcmcia cards.
> 
Let's be helpful here, now the reason may be known, but how to solve it.
Well as far as I know, alsa can do software mixing as well, you will
need to use a dmix device in the alsa configuration files. All the
possibilities with alsa configuration is quite wide, so it would be best
to read about it in the alsa documentation or search the internet for a
suitable solution. Also make sure that the software itself is really
using alsa, some programs still use OSS (trplayer, realplayer, flite,
freetts (so firevox plugin if you are using freetts), etc). If you need
to use something which uses OSS then it is still possible to use
software mixing, get it to use the alsa-OSS compatibility system by
putting the aoss command at the beginning of your command for the app
which uses OSS (eg. "aoss realplayer" will run realplayer with OSS
output going through alsa-OSS compatibility layer).

I don't actually know about your headphone issue, but it sounds like
alsa doesn't know about configuring (or it isn't configuring) a setting
for muting speakers when using headphones (it sounds like the switch is
software based rather than being a simple switch system. I think due to
the comment about the behaviour when using windows). Guesses at a
possible fix:
Check that alsa doesn't have a control for it somewhere.
See if the problem is solved in updated versions of alsa (if you are not
using the latest).
Last resort might be to find out if headphone and speaker volume can be
set separately (I know my laptop soundcard alsa gives a headphone
volume), and then you create two scripts (or one which either takes
options or knows how to toggle) that you run to mute speakers but not
headphones and to unmute speakers, and then you run this/these when
connecting and disconnecting headphones.

Hope some (if not all) of this is useful.

From
Michael Whapples 

Re: Problems with software speech

[Farhan]

Hello, i'm not sure about your second question about the headphone issues, but most laptop soundcards now a days are cingle channel only, they don't do multichannel, sinse Windows xp has some weird kernel implementation to do that all for you, I think the only true multichannel soundcards you can get for laptops are the soundblaster external pcmcia cards.

On 3/10/2007 at 23:05 Zachary Kline said
Hi,
    I admit there are probably better forums to adk this question.  However, I've tried looking for answers and come up with a blank.
I use one of those Intel Centrino mobile laptops, an HP DV2050 to be exact.  This comes with an 'High Definition Intel' sound card, which Alsa supports...Up to a point.
For whatever reason, I can't mix sounds with the provided drivers.  At all.  I can't play music, or listen to streaming radio, or anything along those lines.
But the far stranger problem is what happens when I boot my machine into Linux: it has a duel boot setup.  I use headphones for software speech.  The problem is that when I plug them in, they work, but the sound also comes out of internal laptop speakers.  
Funny thing is, when I boot into Windows first, and without shutting down the machine fully boot into Linux, I get a similar problem: only this time, speech comes out of headphones only, and no speakers.  So...
I guess I was wondering if anyone, anyone at all, has had this problem, and what they've done about it?
Thanks,
Zack.
_______________________________________________
Speakup mailing list
Speakup@braille.uwo.ca
http://speech.braille.uwo.ca/mailman/listinfo/speakup

Problems with software speech

[Zachary Kline]

Hi,
    I admit there are probably better forums to adk this question.  However, I've tried looking for answers and come up with a blank.
I use one of those Intel Centrino mobile laptops, an HP DV2050 to be exact.  This comes with an 'High Definition Intel' sound card, which Alsa supports...Up to a point.
For whatever reason, I can't mix sounds with the provided drivers.  At all.  I can't play music, or listen to streaming radio, or anything along those lines.
But the far stranger problem is what happens when I boot my machine into Linux: it has a duel boot setup.  I use headphones for software speech.  The problem is that when I plug them in, they work, but the sound also comes out of internal laptop speakers.  
Funny thing is, when I boot into Windows first, and without shutting down the machine fully boot into Linux, I get a similar problem: only this time, speech comes out of headphones only, and no speakers.  So...
I guess I was wondering if anyone, anyone at all, has had this problem, and what they've done about it?
Thanks,
Zack.