Re: Text 2 Speech
> I am a bit familiar with scheme. I'm having difficulty thinking of situations
> when I'd need any versatality though. Shouldn't text2speech programs simply
> take text as their input and produce speech that sounds as human as possible?
> Besides changing speaker age, accent, gender and the rate of speech, why
> would one want the UI to be versatile?
Well, I should have mentioned in my earlier post that I'm working as a
graduate student in a speech-processing lab (see the links in my .sig). The
festival package does have "quick and dirty" scripts that you can run, so it
would probably fill the bill for whatever you're trying to do.
As far as the versatility, you can do things like insert your own prosody
model or algorithm for choosing concatenation units. It's all pretty
technical stuff, and to be honest I probably couldn't explain it all in great
detail.
The main point is that festival is a serious piece of software, not like some
of the synthesis "toys" that people have to read the date and time when they
log on, not that there's anything wrong with those toys.
So to clarify: You asked for a text2speech package. I told you about
festival, with the caveat that it's not a little text2speech "toy", which may
or may not be what you're looking for.
HTH
--------------------------------------------------------------------------
Stephen W. Juranich sjuranic@ee.washington.edu
Electrical Engineering http://students.washington.edu/sjuranic
University of Washington http://ssli.ee.washington.edu/ssli
--
To UNSUBSCRIBE, email to debian-user-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: