[Kde-accessibility] Another speech engine for KTTS?

Jonathan Duddington jsd at clara.co.uk
Wed Feb 8 18:22:04 CET 2006


In article <200602081210.59945.ojschmidt at kde.org>,
   Olaf Jan Schmidt <ojschmidt at kde.org> wrote:

> I can bring you in contact with Klaus Knopper, who plans to work on
> a free software solution for German speech synthesis.

> On February 25/26, we will do a usability testing of KDE
> Accessibility in German. Having the source beforehand would be great.

That sounds good.  February 25 isn't much time to do much though.
Yes, I'll be happy to discuss with Klaus whether this is useful for his
purpose, what needs to be done, and how to do it.

> > BTW I've just noticed that it's not recognising accented characters
> > in German.  Perhaps a signed/unsigned character problem.  I'll try
> > and fix that too (not that the German should be taken seriously
> > though).

> Sound like an encoding issue.

Yes.  setlocale(LC_CTYPE) was not correct, so that isalpha() was giving
the wrong result.

Is there a way I can explicitly tell the locale system to assume a
specified character set (such as ISO-8859-3) for the purpose of using
ctype functions such as tolower() and isalpha()?

> Can your speech engine accept UTF-8 input? This would really be
> useful, especially for languages with a non-Latin alphabet. Festival
> has problems in this regard as well.

The dictionary rules and lookup assume 8-bit characters, but adding a
"UTF-8 to a specified 8 bit character set" translation sounds
straightforward.  Is that sufficient, or do you mean re-writing the
rules and lookup to use wide characters?

How is the speech engine told what is the character set of the input
text?

Any ideas for a more distinctive name than "speak"?



More information about the kde-accessibility mailing list