Voices and victories

[Ken Norris]

Hi David,

> From: davidsyes/nac
> Subject: Re: Voices and victories
> Date: Sat, 28 Jun 2003 23:27:33 -0700
> 
> Is it possible to "impersonate" the voices? Or, are the voices tied to a SDK?
----------snip
Nope, they're tied to a TTS engine, like AT&T's NaturalVoice, which requires
licensing to use in distributable projects.
---------- 
> So, Tuviah may want use own-voices and then charge self for royalties. This
> way Tuvia can say, "If my voice is successful, and I want more royalties, I
> will sue myself and I will be victorious".
----------
:-) Well, you can synthesize voice-modeling by hyperextensive sampling and
digitizing, but you're talking about hundreds (and perhaps 1000's) of man-
(or woman-) hours in recording, analyzing, tweaking results, etc. which is
why AT&T has to recover those costs and try to make money on it, otherwise
it's monetarily a bad deal, and it's probably a safe bet they're not gonna
give away the best TTS voices in the world out of the kindness of their
hearts.

But maybe Tuviah will ;-)

Ken N.