Melchior FRANZ wrote:
* John Wojnaroski -- Monday 06 February 2006 17:25:
The original code I sent on to Dave Luft had a socket interface in the
AI module for the ATIS and AI controller.
It still has. With hardcoded IP address, which I find a bit suboptimal
for usage by anyone but you:
#define ATC_SERVER_ADDRESS "192.168.2.15" // adddress of machine running
festival server
We shouldn't have to advise users to run a festival server on machine
192.168.2.15, port 7100, when the *vast* majority of users would
prefer localhost:1314, which is the festival default. :-)
Don't have the code in front of me, IIRC that address is where FG sends
the text string to a small wrapper program that runs the festival client
which in turn parses the data packet and sends it to the festival server
on the same machine at the default. The intent was to provide a means to
create a data packet that would contain additional info regards the text
string, such as a repeating string for ATIS, voice ID to identify
speaker, etc, etc. The hard coding was merely a convience to get
something to run. I agree it does need to be configurable for general use.
My approach is to run festival on a seperate machine so as not to reduce
the frame rate.
Somewhere on a disc that code and a "howto" I started on setting up the
festival server/client interface to FG exists.
I would be interested in that. I think that the festival interface needs
some rework, but I don't want to remove current capabilities. I'd like
to merge that into my approach. I'm now supporting multiple voices that
can speak at the same time, each with different pitch & speed. Individual
volume doesn't work yet, and I'm not sure if it can be done at all. Better
voices don't work yet, either. (If anyone can tell me how to do that
over telnet, please tell me.) I have two such groups in my preferences.xml:
<voice>
<desc>Copilot/Instructor Voice</desc>
<text type="string"></text>
<volume type="double">1.0</volume>
<pitch type="double">150.0</pitch>
<speed type="double">0.9</speed>
<preamble type="string">(SayText "FlightGear Flight Simulator")</preamble>
</voice>
The preamble is really for selecting a different voice or testing
purposes. <text> is the property that one sets to make the voice
subsystem speak with this voice.
I'll rummage through some old drives I have on the shelf and find where
I saved it and send it over.
There was also a speech recognition program using SphinxII under [...]
I had looked at that a long time ago and wasn't thrilled. I don't
remember if it was complicated to set up or didn't compile. It's certainly
not in wide use yet, which festival is. I'm sure that several Linux
distributions come with festival already. Mine does.
It was moderately difficult. The biggest item was creating the word
dictionary or corpus. One advantage since ATC speech is somewhat
codified, recognition is quite good
Regards
John W.
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
_______________________________________________
Flightgear-devel mailing list
Flightgear-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/flightgear-devel
