Go look in conf/lang/en/en.xml. Near the top of that
file is a place to specify the default TTS engine and TTS voice
for your language. Try setting those to "flite" and "kal" or
whatever voice you are using.
-MC
On Fri, Jun 24, 2011 at 8:57 AM, Wes <wes-fs@499x.com>
wrote:
Thanks!... I took a
look, and got one working like this:
<include> <!--This line will be ignored it's here
to validate the xml and is optional -->
<macro name="press_x_to_continue">
<input pattern="^(\d+)$">
<match>
<action
function="play-file" data="voicemail/vm-press.wav"/>
<action function="say"
data="$1" method="pronounced" type="items"/>
<action
function="play-file" data="voicemail/vm-continue.wav"/>
<action
function="speak-text" data="yo, you must press $1 to
continue, do it now"/>
</match>
</input>
</macro>
</include>
but when it got to the speak-text function, it gave this
error:
I have the flite engine active, so how do I configure
speak-text to try using the flite engine instead of
cepstral?
On 6/22/2011 6:42 PM, Michael Collins wrote:
Have you researched the subject
of "phrase macros"? That may be your golden ticket. If
you have the FS book then look at the latter half of
ch 6. Otherwise check out these resources:
The voicemail system uses phrase macros a ton,
and effectively, I might add. In short, phrase
macros let you piece together sound prompts, calls
to the say app, calls to the speak app, calls to the
sleep app, etc. I think you would be most interested
in the "speak-text" action.
-MC
On Wed, Jun 22, 2011 at
3:09 PM, Wes <wes-fs@499x.com>
wrote:
or maybe
we just need a new version of this method
called "speakAndGetDigits" that specifically
takes a couple of string parameters for the
words to speak, instead of taking the location
of the wav files...
On 6/22/2011 4:13 PM, Wes wrote:
Sorry, I was
thrown off by the dialplan example...
I'm trying to use speak... I changed my
example to try the speak command and I
still have the same issue with the
quotes.
Is it possible to send in a "speak"
phrase as the phrase to use in the PAGD
command? Or do I have to record a wav
file and pass the name of the wave file
in for that parameter?
On 6/22/2011 3:51 PM, Michael Collins
wrote:
Can you confirm:
are you trying to use TTS or the say
engine? These are two completely
different subjects. If you want to use
TTS then you need the "speak" app, not
the "say" app: