Re: [orca-list] Perhaps a bit OT: SD and punctuation



Hi Herman,

Hermann wrote:
SD's handling of punctuation cannot be set up by Orca, and it has to be
done in the SD config files.
How can I do this?

The generic module for espeak doesn't implement punctuation settings, so
it currently ignores the configured SD punctuation settings.  You can,
however, set the espeak punctuation level in
/etc/speech-dispatcher/modules/espeak-generic.conf.  Just add the
--punct option to the GenericExecuteSynth command.  See espeak --help
for the description of the --punct option.  For example the following
GenericExecuteSynth command will read a dot comma and a semicolon and
will ignore all other punctuation characters:

GenericExecuteSynth \
"echo \"$DATA\" | espeak -w /tmp/espeak.wav -v $VOICE -s $RATE -a
$VOLUME -p $PITCH --punct=".,;" --stdin && aplay /tmp/espeak.wav"

Please note, that the above has nothing to do with the punctuation
settings in Orca.  I just looked to the code and I see that Orca expects
the speech subsystem to actively add the punctuation verbalization into
the spoken text.  I didn't implement this in the SD backend, so Orca
punctuation settings are currently ignored.  It seems, however, very
simple to implement, so I will send a patch soon.  Until I do so, you
can use the solution described above.

I couldn't also manage to set a higher pitch for caps.

Speech Dispatcher can announce capital letters by a "sound icon".  It
should work with Festival, I think this is not possible with espeak.  I
will try to find out the simplest way to set it up with Orca and let you
know.

And finally: Does SD support saying text by sentence, instead of line?

As far as I know, Orca sends the text by line, so the speech system has
no notion of the original sentence structure.  Maybe Will will have some
advice on this?  Would it be possible in Orca?  This would be definitely
useful to get the text by larger chunks in the SayAll function and
probably on other places too (jumping by paragraph in Firefox etc).

Best regards, Tomas



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]