software synths

[David Bruzos]

Hi:
Actually the software speech for speakup does not come with Fedora.  You will have to configure and install a few things 
before you have it running.  What comes with Fedora is Festival which is just a speech synthesizer.  Festival is part of 
what you need, but it is not all of it.  I am writing a document explaining how to do this, but it is taking forever.
I am putting below a message I posted to someone else about this, so you can take a look at it.  If you have questions, you 
can drop me a line directly.

Here is the message:

Hi there:
I think you should use "flite", because it is written in C and it is faster and more responsive than "festival".  However, 
if you like "festival" better, there is nothing wrong with that...

To test the synths do:

	1. $ flite -f /path/to/file
	2. $ festival --tts /path/to/file

Where /path/to/file is the path to some text file that you want flite/festival to speek.

To get speakup working with software speech you will need some other applications.  They are:
	1. speech-dispatcher
	2. speechd_up

I think you can get speech-dispatcher from your apt-get repos.  Speechd_up you will have to download from its website.  I 
don't remember what that is.  Someone else can tell you that or google it...
Remember to look at the speech-dispatcher configuration if things are not working...

To test speech-dispatcher do:
	$ speech-dispatcher
	$ spd-say "some text"
	
Where "some text" is just some text you want it to speek.

Now, you have to create the device /dev/softsynth with numbers 10/26.  Use the command:
	$ mknod /dev/softsynth c 10 26

After you have done all of this, run speech-dispatcher:
	$ speech-dispatcher
Load the "sftsyn" speakup module into memory:
	$ modprobe speakup_sftsyn
and run speechd_up:
	$ speechd_up

Note: you must remove the current speakup_xxxx module from the kernel before loading the speakup_sftsyn module.  Do 
something like:
	$ rmmod speakup_xxxx
Trying to load both modules at the same time, could crash your box.

You should have a software speech enabled system at this point.  I am writing a howto on how to do this, but it has taken 
me rediculously long to finish it!  I guess life/baby/school/etc has gotten in the way.

David B.