TTS chipset with decent quality voicec

drill in this search on speech synthesis +chipset.

There may be some links of merit.

http://www.google.com/search?hl=en&q=%22SPEECH+synthesis%22+%2Bchipset&btnG=Search

Guest

I'm quite interested in the results of your search effort. My need is for text-to-speech technology that is capable of tracking phrasing and intonation of a voice sample but substituting the manufactured voice.

Yamaha Industries has been licensing some interesting software that will sing (in a male or female voice) text that has been set to a written melody line. Again, fairly mechanical sounding to my ear, but an interesting direction of development.

In my opinion, the way to do it is to create the artificial voice from text, identify the various vocal components therein (vowel sounds, diphthongs, etc.), then access a library of sampled voice characteristics (John Wayne, Howard Cossell, Bette Midler, for instance, or even myself) and reconstruct the voice with those library characteristics of pitch, tonality, phrasing, etc. Voice recognition is already doing most of this in hands free, voice activated devices.

Some of this is also being done fairly well with the voicing of various musical instruments--trumpets and clarinets for instance that actually sound like trumpets and clarinets, at least until you push them beyond the capabilities of their original voices.

I can imagine a multitude of applications for a chipset capable of accomplishing this task. I am able to devote some time to testing and evaluation of same if you wish to correspond with me and share ideas and information.

Lonnie