Festival

Festival provides all the general tools needed to develop complete speech synthesis systems. An overview and examples can be found at Festvox.org.

=How to install= You should install by means of your distribution as described at installing software. If that is not possible, ./configure && make && make install
 * download festival and speech_tools from http://www.cstr.ed.ac.uk/projects/festival/
 * apply the patch from http://lists.kde.org/?l=kde-accessibility&m=112757271022463&w=2
 * build speech_tools:
 * build festival

=How to run= Say something: $ echo "Linux Questions" | festival --tts Read a book: $ festival --tts waroftheworlds.txt Use text2wave to convert text to an audio file, to place on media player or burn to CD. (the wave file in the example below can quickly become a large file; You can convert to MP3 using the Lame MP3 encoder or other compressed audio formats using Sox): $ text2wave -o waroftheworlds.wav waroftheworlds.txt Festival's programming language, a variant of Lisp, can be a powerful tool to develop new voices and add speech to your existing project. A complete Scheme reference is available on festvox.org. The below example simply shows how you can make your computer say "hello": $ festival Festival Speech Synthesis System 1.96:beta July 2004 Copyright (C) University of Edinburgh, 1996-2004. All rights reserved. For details type `(festival_warranty)' festival> (SayText "hello"); festival> (quit)
 * 1) 

=Improve voice quality= There are many free voices available (e.g. MBROLA, CMU Arctic, Nitech HTS). One possible starting point for a high-quality and low-cost commercial voice is Cepstral.

=See also=
 * Accessibility
 * Swift
 * text2wave

=External Links=
 * University of Edinberg - Centre for Speech Technology Research
 * MBROLA Project
 * Non-free Swift and Voices