software synths

[Michael Whapples]

I don't know if you have worked out how to install all you need, but as you 
said you are using debian, here is a guide. Use apt to get and install the 
stuff for you (use a command such as:
apt-get install speechd-up speech-dispatcher espeak
in fact you probably could miss out speech-dispatcher as it should be a 
dependency of speechd-up (not sure, so make sure that speech-dispatcher gets 
installed if you don't specify it). Also the above assumes you wish to use 
espeak. I don't know what versions of the software, so you may wish to get 
the packages from their original sources and compile them yourself (not too 
hard, just make sure you have everything required installed). It may be only 
a few packages which need compiling, eg. speech-dispatcher may be version 
0.6.4, but if you wish to use espeak with it, then you need above a certain 
version number (to use the espeak specific module), so it might be best to 
compile the latest espeak).

Once the above are installed, you may wish to edit speech-dispatcher's 
configuration files to get espeak as the default module. If you installed 
speech-dispatcher through apt, the configuration file is probably 
/etc/speech-dispatcher/speechd.conf, and if from source it probably is 
/usr/local/etc/speech-dispatcher/speechd.conf unless you specified otherwise 
in which case you should know where it is. This file is fairly well 
commented, a # symbol indicates a commented out line, some options may need 
un-commenting. to use espeak as default synth, make sure the default synth 
line is uncommented and set to espeak and make sure the espeak module line 
is uncommented in the add module section (you may wish to comment out unused 
modules). Now save and make sure the espeak module is set as you want. This 
is in the modules/espeak.conf file in the speech-dispatcher configuration 
directory (ie. if speech-dispatcher was installed through apt, 
/etc/speech-dispatcher/modules/espeak.conf). The main thing I would suggest 
is to make sure that output is set to alsa (you will need to make sure when 
compiling (if you compile speech-dispatcher) to have alsa support). Use alsa 
as this is the current system for audio output, and offers better features 
(particularly when using software speech, eg. software sound mixing (have 
speech and music at the same time). save this and restart or start if you 
previously didn't start speech-dispatcher to load the new settings, and 
change speakup output to sftsyn (I would use echo sftsyn 
 >/proc/speakup/synth_name, but I am not sure how well this works with the 
modular structure used in the debian kernel). Now load speechd-up and you 
should now have software speech from speakup. You can use scripts on startup 
to load speech-dispatcher, speechd-up and change speakup output to sftsyn, 
but I won't list it here, as I am more familiar with slackware start scripts 
than debian ones.

From
Michael Whapples