On Sun, 2007-08-05 at 20:32 -0500,
asterisk-users-request at lists.digium.com wrote:> Date: Sun, 5 Aug 2007 19:08:25 -0300
> From: Jo?o Paulo Vanzuita <joaopaulo at riseup.net>
> Subject: Re: [asterisk-users] text2wave Voices Improvements?
> To: asterisk-users at lists.digium.com
> Message-ID: <20070805190825.1a7e7443.joaopaulo at riseup.net>
> Content-Type: text/plain; charset=US-ASCII
>
> On Sat, 04 Aug 2007 19:52:21 -0400
> Matthew Rubenstein <email at mattruby.com> wrote:
>
> > I currently have an AGI that calls the Festival text2wave app
> to write
> > a wav file that my dialplan plays into a call with the Background()
> > command. But the voice sounds terrible: like SAM, the 1980s 6502
> voice
> > synthesizer. I tried to slow it down by calling (text2wav -eval
> > "(Parameter.set 'Duration_Stretch 1.4)" -scale 2.0
[...]), but it
> still
> > sounds like it's talking while sucking down a strawful of
spaghetti.
> How
> > do I install a different voice, to speak basically simple emails?
> I'm
> > (APT) installing on Debian 3.1/Sarge, Asterisk 1.4.x .
>
> works fine to me installing "festvox-kallpc16k" for speaker
> apt-get install festvox-kallpc16k festival
I apt-get install'ed that package, and rablpc16k and kdlpc16k . But
though it rab and kal voices work, kdl does not:
#text2wave -o say.wav say.txt -eval "(voice_kdl_diphone) (Parameter.set
'Duration_Stretch 1.4)" -scale 2.0
SIOD ERROR: unbound variable : voice_kdl_diphone
They all sound like a 1980s synth (which wasn't so bad, if you were
familiar with the burbly voice). I'm using the first 2 sentences of (man
man) as my test string: "man is the system?s manual pager. Each page
argument given to man is normally the name of a program, utility or
function." Maybe there's some kind of hint characters I can insert to
make it sound better?
--
(C) Matthew Rubenstein