Text to speech on Linux, possibly with Festival
 

Sorry if I'm reaching my question quota but this is the first day I've tried to run Linux for acctual use; really makes things I haven't thought about before standout.

I saw an artical on a program called festival and tried my best to install it. It looks like it is built in 2 steps (requires like 5 archives) and I got both parts to successfully build but when I try to run the server scripts it made they throw errors, kinda seems like it was supposed to register as a command but didn't. I was also to get festival run to it's prompt without language file errors:

Couldn't get it to do anything from there. The readme says to use (SayText "hello world"), that doesn't work.

I saw a guide on this site about it but it used a weird method using rpm's, doesn't seem like what I need.

Any advice on getting this thing up and running? All the tutorials I found were out of date and even the readme file that came with it is refering to wrong versions of things.

Thanks guy, this ones hard, even for linux.

That's as far as I got as well. Look forward for success stories.

It seems the settings files have to be altered after all that compile. Also, i am under impression, the binaries were not copied into /usr/* after make in my case. It was too stressful to move further. TTS on linux is my last frontier... Too bad. KDE seems to have good support for TTS.

suslik/tibberous

This is what I did on Redhat 9 and the basic speech function works

Download from http://www.cstr.inf.ed.ac.uk/downloads/festival/1.95/

festival-1.95-beta.tar.gz
festlex_OALD.tar.gz
festlex_POSLEX.tar.gz
festvox_rablpc16k.tar.gz
speech_tools-1.2.95-beta.tar.gz

I unpacked them into a directory called home/tts, this gave me two directories festival and speech_tools

Then as root I cd'd into the speech_tools directory, in my case /home/tts/speech_tools and ran ./configure then make I also ran make test which gave no errors

Next again as root I went to the festival directory cd /home/tts/festival and ran make

Then as a normal user from the festival directory I ran bin/festival and got

Festival Speech Synthesis System 1.95:beta July 2004
Copyright (C) University of Edinburgh, 1996-2004. All rights reserved.
For details type `(festival_warranty)'
festival>

From here all the following work perfectly

(SayText "hello world")
(intro)

Then enter (quit) and script mode works using the following

examples/saytime

Hope this helps, I'm no expert just lucky I guess

Pete

I love you!

Followed your steps and it works, not sure what I was doing wrong, maybe didn't have the right files?

This is great, it sounds like an old British guy!

I'll look into it more later (reading clipboard speech, speeding it up, possibly changing the british guy) but I'm just glad to have it working. Nothing is worse than having to reboot in XP to read a book about Linux :)

Thanks again,

Trent

Trent

Here's the tutorial I used, it's hard work but hopefully after a while it will make sense

http://www.festvox.org/docs/manual-1...tival_toc.html


Glad I could help

Pete

I also used your instructions to try the "Kmouth" program included in the Yoper OS that I installed on this laptop, and I must say your instructions were excellent. It worked the first time. Are there more voice packages on the download page where I got festival and speech tools? and if so which ones are they? I also read the festival tutorial, and it spoke of "rab_diphone something files. I found that file in festival/lib where it should be,but is it possible to add more voices to that file? thanks for the help and the great easy to follow instructions. Next project is " speech to text" any suggestions?