Page 1 of 1

build my own tts voice using recordings

PostPosted: Mon Jun 22, 2009 3:30 pm
by jkenn337
Hello,
I want to be able to build my own tts voices by using recordings in .wav format and then be sent an executable of my sapi5 voice through email. I also want to be able to make voices by uploading an mp3 of a book, all phonemes are read from the book, extracted, then sapi5 voice is created and the download link is in my email. I also would like to be able to make my own voices with effects using a vocoder to make a singing pipe organ voice and voices that take breaths, or whatever voice I may have the power to create and turn it into a sapi5 voice that I can share with others for free, or sell on my own website if I choose. check out my blog for a soon to come petition to cepstral so this will be doable very soon. ModelTalker lets you do this but its interface for a blind person using jaws is a bit unstable and cumbersome at times. As far as I know, cepstral only lets commercial companies, not individuals to create their own voices. This has to change. Please let us as home users, students, and individual consumers make our own sapi5 voices!
Read my blog at: http://jkenn337.klango.net .

Josh Kennedy