Menu

Tree [61be2e] master v0.1.0 /
 History

HTTPS access


File Date Author Commit
 examples 2010-06-02 Daniel van Niekerk Daniel van Niekerk [85c76e] Ported halfphone script to new Speect...
 modules 2010-06-04 Daniel van Niekerk Daniel van Niekerk [d2e379] Cleaned up script metadata...
 scripts 2010-06-04 Daniel van Niekerk Daniel van Niekerk [d2e379] Cleaned up script metadata...
 .gitignore 2010-03-08 Daniel van Niekerk Daniel van Niekerk [f5ac57] small changes...
 HOWTO 2010-06-02 Daniel van Niekerk Daniel van Niekerk [85c76e] Ported halfphone script to new Speect...
 INSTALL 2010-05-27 Daniel van Niekerk Daniel van Niekerk [a14910] added INSTALL and README files...
 LICENSE 2010-06-04 Aby Louw Aby Louw [3496fe] License file
 README 2010-06-04 Daniel van Niekerk Daniel van Niekerk [61be2e] Added version number to README

Read Me

========================
 VOICETOOLS Version 0.1
========================

1. Introduction
---------------

This package is a collection of scripts to assist with the process of
constructing voices for the Speect Speech Synthesis System. These
utilities are designed to run on standard GNU/Linux systems. The
following are included in this package:

 - A standalone script to perform phonetic alignment using the Hidden
   Markov Model Toolkit (HTK).

 - Scripts to compile resources for the construction of run-time
   voices for Speect (unit-selection and HMM-based techniques).

 - Examples of various configuration files necessary to construct a
   working run-time voice for Speect.



2. Package layout
-----------------

The top-level structure of this package should be as follows::

 .
 |-- README
 |-- INSTALL
 |-- examples
 |-- modules
 `-- scripts

where the directories "examples", "modules" and "scripts" contain
configuration examples, Python modules required by these scripts and
executable scripts respectively.



3. Software dependencies
------------------------

The following software is required:

 - Python 2.6 (http://python.org/)
 - HTK 3.4 (http://htk.eng.cam.ac.uk/)
 - Praat 5.0.32 (http://www.fon.hum.uva.nl/praat/)
 - Edinburgh Speech Tools 2.0.95 (http://www.cstr.ed.ac.uk/projects/speech_tools/)
 - SoX 14.3.0 (http://sox.sourceforge.net/)
 - normalize-audio 0.7.7 (http://normalize.nongnu.org/)
 - Speect 0.9.0 (http://speect.sourceforge.net/)

Python modules:

 - Numpy 1.3.0 (http://numpy.scipy.org/)
 - Scipy 0.7.0 (http://www.scipy.org/)

Optional Python modules:

 - PyParsing 1.5.2 (http://pyparsing.wikispaces.com/)
 - Psyco 1.6 (http://psyco.sourceforge.net/)



4. Installation
---------------

See the ``INSTALL`` file.



5. Usage
--------

Usage information for each script can be obtained through the
``--help`` command line parameter. For more information on the process
of constructing voices for Speect using these utilities, refer to the
``HOWTO`` file.

Want the latest updates on software, tech news, and AI?
Get latest updates about software, tech news, and AI from SourceForge directly in your inbox once a month.