[DECtalk] DECtalk TTS licensing

[Don]

On 8/30/2021 9:44 AM, Chime Hart wrote:
> I certainly agree that when quotes are found, the synthesizer should be able to
> switch among mail-and-female voices. But not only that, so-far I know of no
> solution which will display an emotion of an author. As an emotional Aquarian,
> I want to know if some1 sending mail is really happy, not just hearing a word
> "smile" although thats a start. There was an Israeli company called Vivotext
> which was developing expressive speech, but while you can still find their
> youtube videos, I think they may be gone? Obviously I did write-and-suggest
> them making a screen-reader.

There has been a bit of research into imparting emotion to speech.
Typically, all you get is a bit of inflection control to break up
the monotony as well as indicate questions (rising intonation
at the end of the sentence).

But, all of these things are premised on converting text that
corresponds with spoken word.  How should I present...
(rest of sentence left absent)?

For "question words" (who/how/what/why/where/when) introducing
a statement, you can infer that it will be a question.

But, what if you did not encounter such a formulaic expression?

"Mary did that?"

"Mary did that."

"Mary did..."