New Czech/Slovak TTS module for speech dispatcher

[Daniel Tihelka]

Dear speech-dispatcher comunity,

we deal with the research of TTS synthesis for several years at our Department 
of Cybernetics at the University of West Bohemia (see www.artic.cz). We focus 
on Czech, but we have voices for Slovak (and one older for German) as well. 
The result of our research are every year presented and discussed on 
significant speech-related international conferences such as Interspeech, 
LREC, TSD and others.

The commercial version of our TTS is distributes by spin-off firm SpeechTech 
s.r.o (see www.speechtech.cz, where various voices and engine versions can be 
tried in web demo). The licence for a TTS module can be bought at a 
fairly accepted price, and as far as I know, the users are largely satisfied 
with the quality of voices. However, although the module is platform 
independent (and it is developed on Linux, actually), the voices are 
distributed for Windows users only at this time, mainly due to 
the existence of standardised SAPI interface. 


Just recently, our student Michal Siroky developed GPL module 
(speechtech_tts.c and speechtech_tts.conf.dummy in the attachment) which 
allows to use the Linux version of our TTS together with the 
speech-dispatcher (the module is built and tested against 
speech-dispatcher-0.6.7). Now, I would like to give you this module, and we 
will be happy if it become the part of speech-dispatcher.

The module is GPL, of course, and it calls functions from the non-GPL 
SpeechTech TTS library (I suppose that this does not violate the GPL licence). 
For the purposes of development and testing, we have also created very 
simple dummy TTS (attached in SpeechTechTTSDummy.tar,bz2), just replaying the 
given file.

We will appreciate your revision of the module code - you have much deeper 
knowledge about SpeechD internals and the requirements for the module 
behaviour than we do. If you have any comments, improvement suggestions or 
questions, do not hesitate and send them to us (to me and to the Michal 
<msiroky at students.zcu.cz>). We will try to improve or fix the issues 
reported.


We wish you many satisfied users of speech-dispatcher. Best regards on behalf 
of the whole TTS group,
D. Tihelka


-- 
Daniel Tihelka, dtihelka at kky.zcu.cz ? ? ? ? ?
http://www.kky.zcu.cz/cs/people/tihelka-daniel
http://www.kky.zcu.cz

Phone : +420 377 63 2531 ? ? ? ? ? 

Department of Cybernetics - Laboratory of Artificial Intelligence
University of West Bohemia
Univerzitni 8, 
306 14 Plzen
Czech Republic
-------------- next part --------------

# -- SOUND ICONS --

# Debug turns debugging on or off
# See speechd.conf for information where debugging information is stored
# Debug 0

# DebugFile specifies the file where the debugging information
# should be stored (note that the log is overwritten each time
# the module starts)
# DebugFile "/tmp/debug-flite"


# Configuration for SpeechTechTTS - it is this file
ConfigurationFilePath 
"/home/dtihelka/Pokusy/Siroky_SpeechDisp/speech-dispatcher/etc/speech-dispatcher/modules/speech_tech_tts.conf"


#
# SpeechTechTTS config part. Luckily SpeechD can handle it (although it 
complains ;-))
# If '=' should rather not be used, we must modify the SpeechTech config reader 
...
#


# This is the configuraton file for the dummy implementation of SpeechTechTTS 
engine
# The correct config will be installed together with the real TTS engine

[SpeechTechTTS]

# The name and path to sound file which will be played by speechd.
SndFile               = 
"/home/dtihelka/Pokusy/Siroky_SpeechDisp/speech-dispatcher/share/sounds/speech-dispatcher/sentence.wav"
# The minimum number of phrases generated by debug SpeechTechTTS module.
MinPhrases            = 1
# The maximum number of phrases generated by debug SpeechTechTTS module.
MaxPhrases            = 6
-------------- next part --------------
A non-text attachment was scrubbed...
Name: speechtech_tts.c
Type: text/x-csrc
Size: 15650 bytes
Desc: not available
URL: 
<http://lists.freebsoft.org/pipermail/speechd/attachments/20090803/d4f9c9f5/attachment-0001.c>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: SpeechTechDummy.tar.bz2
Type: application/x-tbz
Size: 199004 bytes
Desc: not available
URL: 
<http://lists.freebsoft.org/pipermail/speechd/attachments/20090803/d4f9c9f5/attachment-0001.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part.
URL: 
<http://lists.freebsoft.org/pipermail/speechd/attachments/20090803/d4f9c9f5/attachment-0001.pgp>