[SOUND] speechd lets your Computer speak

Message

kavau · Post by **kavau** » Sat Sep 11, 2004 5:14 pm

nastassja wrote:*edit* finally i fixed it, it works very fine.. till i start kde (2 arts) :/ *edit*

try the following:

echo "Hello World" | artsdsp festival --tts -

this should work within kde as well. Don't know how to do the same thing for speechd, though...

Viza · Post by **Viza** » Sat Sep 18, 2004 9:43 pm

For anyone having the "socket: bind failed" error while trying to run speechd (festival --server):

You have to run "killall festival" to kill previous processes that sometimes don't die on their own. I had this problem before and whenever I get that error, I do killall festival and everything works perfectly fine.

Centove · Post by **Centove** » Wed Sep 22, 2004 11:02 pm

I saw a couple of people wanted speechd and festival to play 'nice' with alsa but no real answer.

First you'll want to have the dmix stuff working:
http://forums.gentoo.org/viewtopic.php? ... mix+plugin

Once that is working now get festival in the happy family.

/etc/festival/server.scm

Code: Select all

(Parameter.set 'Audio_Method 'Audio_Command)
(Parameter.set 'Audio_Command "sox -t raw -sw -r $SR $FILE -c2 -t wav - | aplay")

With that I can be playing music and listening to my computer babble at me.

Have fun.

Cinder6 · Post by **Cinder6** » Sun Sep 26, 2004 6:22 am

bassvandijk wrote:This is really cool!!!

Also nice:

forune > /dev/speech
or
dmesg > /dev/speech

or emerge kde > /dev/speech

Blood_Roses75 · Post by **Blood_Roses75** » Sat Oct 02, 2004 5:44 am

Another really cool thing you can do with festival is have it say a message to you whenever you startup fluxbox.

In ~/.fluxbox/startup:

# Applications you want to run with fluxbox.
echo "Welcome to fluxbox" | festival --tts -

joseh · Post by **joseh** » Fri Oct 08, 2004 4:42 pm

Turns out you can use any capable player you want to play the speech data. Example, festival had trouble with aRts and artsdsp etc, so I did this to my ~/.festivalrc

(Parameter.set 'Audio_Command "/usr/local/bin/mplayer $FILE > /dev/null 2>&1")
(Parameter.set 'Audio_Required_Format 'wav)
(Parameter.set 'Audio_Method 'Audio_Command)

and viola, it sends the data output as wav and plays with mplayer, which I have set up to work just fine with aRts.

undrwater · Post by **undrwater** » Wed Oct 13, 2004 3:17 pm

joseh wrote:Turns out you can use any capable player you want to play the speech data. Example, festival had trouble with aRts and artsdsp etc, so I did this to my ~/.festivalrc

(Parameter.set 'Audio_Command "/usr/local/bin/mplayer $FILE > /dev/null 2>&1")
(Parameter.set 'Audio_Required_Format 'wav)
(Parameter.set 'Audio_Method 'Audio_Command)

and viola, it sends the data output as wav and plays with mplayer, which I have set up to work just fine with aRts.

Do you notice any change in sound quality doing it this way?

joseh · Post by **joseh** » Wed Oct 13, 2004 10:00 pm

no. It simply works much better with my system because I run KDE and aRts all the time and mplayer works very nicely with aRts

riggagoogoo · Post by **riggagoogoo** » Sun Oct 31, 2004 1:05 pm

is there a new ebuild available for Festival as the current one in portage does not work.....

Hydraulix · Post by **Hydraulix** » Wed Dec 15, 2004 1:07 pm

I have to give this a shot.

tecknojunky · Post by **tecknojunky** » Sun Dec 26, 2004 6:37 am

Perl!?!

I have a whole instance of the slow and mommoth Perl interpreter in memory just to pipe text to a voice synthetiser?!?

Some explaining is in order.