Making speech synthesisers speak a particular way

Jonathan Duddington jsd at clara.co.uk
Wed Jan 2 11:44:53 EST 2008


On 02 Jan, Michael Whapples <mwhapples at aim.com> wrote:

> I am looking at adding speech output to a program I am writing, and
> am wondering about the best way to achieve the style of speech (eg.
> pauses in the correct places, correct punctuation level spoken, etc)
> for this. I know of (heard of, but don't know much about) things like
> SSML, would this be what I need to look at?

Yes, SSML will give a better chance of using different synthesizers.
See:  http://www.w3.org/TR/speech-synthesis

eSpeak supports many of the SSML tags.
Although others such as <prosody contour=> make no sense to me.

The recently released eSpeak version 1.30 has improvements to its
implementation of <break> and <emphasis> tags.

<emphasis> is useful to put the emphasis on other than the final word
of a sentence.

<break time=""> was broken in eSpeak 1.30 but fixed now in 1.30.01.

I don't know whether you can use SSML through SAPI5.  It seems that, as
usual, Microsoft has its own, non-standard standard.





More information about the Speakup mailing list