Ubuntu 8.04

I am trying to find a good quality text to speech software that you can use to give you a real sounding voice, that can be adapted to use for animation character sound. All that I have found so far, just give a robotic sounding voice, and you can not adapt the sound to different characters.

Like when they put John Wayne in the tv ad and made him say things by computer.

"Real voice" is a bit subjective. You'll need to determine that by your own ear. I find that some text to speech software is acceptable and others are just plain terrible. I suggest you check out http://www.cstr.ed.ac.uk/projects/fe.../download.html.

You might try Cepstral's (http://cepstral.com) voice engine.

The AT&T Natural Voices (http://www.naturalvoices.att.com) sound the least robotic of all of the above that I have listened to. They have a demo page where you can listen to the voices at http://www.research.att.com/~ttsweb/tts/demo.php. The restriction is that you can not use the generated samples in any work that will be widely distributed (e.g., the internet) unless you buy a license.

Cepstral was the better I've hear on linux. But machine often get intonation wrong, even if voice and speech quality is very good. So I don't think you'll be able, say, to use machine as voice actor.

Nice find!
Hehehe, I could entertain myself for days making the voices say inappropriate messages to friends, family, and colleagues. :-)

thanks for the links to some text to speech options.

What I am looking for is something that will allow you to change the voice so that you can make it sound like different characters in a movie. This would allow you to change settings to make it fit the various characters, and even to simulate other voices. For this is would have to be a real sounding voice, and not a robotic voice. The use is not for text to speech, but for movies, so it needs to sound good, like a voice over.

Your best bet is to look for a program that alters the sound of your voice. In the near term, you are not going to find any computer generated voices that will "pass" as a human voice. By frequency shifting you can make a male voice sound female(or female to male) and yet still keep most of the intonation that is so critical to understand what is being said beyond the words spoken. By altering the rate of the voice you can also dramatically change the way the voice is perceived. A google search on "linux voice editing" provides a lot of hits.