Summary of Available Synthesis Engines that Work with Orca

The following table provides information on which speech synthesis engines are available for various languages on various platforms, and which also work with Orca. Engines marked with "*" are open source. Please feel free to update this with your experiences.

Ubuntu users: see also https://wiki.ubuntu.com/Accessibility/doc/LanguagesAndSpeechSynthesis for a very thorough page on synthesis engines available on Ubuntu.

Language

Platform

Engine(s)

NOTES

English (en_US)

Linux

Festival*, eSpeak*, FreeTTS*, Cesptral Swift, Fonix DECtalk, IBMTTS

English (en_US)

Solaris

Festival*, FreeTTS*, Cepstral Swift, Fonix DECtalk

FreeTTS ships with Solaris

German

Linux

eSpeak*

...tbs...

Notes on Various Engines

The following sections provide more detail about specific speech synthesis engines. Engines marked with a "(*)" are open source.

eSpeak*

  • Web page: http://espeak.sourceforge.net/

  • NOTES: Formant synthesis engine. Small and fast. Adopted by Ubuntu Feisty as the default synthesis engine (replacing Festival). Supports a large number of languages.

Festival*

  • Web page: http://www.cstr.ed.ac.uk/projects/festival/

  • NOTES: Concatenative synthesis engine. A very popular platform for speech synthesis research. Open source, and written in C with a Scheme configuration/customization interface.

FreeTTS*

  • Web page: http://freetts.sourceforge.net

  • NOTES: Concatenative synthesis engine. Primarily English (en_US) with German support obtainable with a bit of hacking. Written entirely in the Java programming language.

Cepstral Swift

  • Web page: http://www.cepstral.com/

  • NOTES: Concatenative synthesis engine. Very human sounding at normal rates of speech.

Fonix DECtalk

IBMTTS

Projects/Orca/SpeechSynthesisEngines (last edited 2013-11-22 19:22:28 by WilliamJonMcCann)