The Festival Speech Synthesis

Disclaimer

The official web pages for the Festival Speech Synthesis system are:

• http://www.cstr.ed.ac.uk/projects/speech_tools/
• http://festvox.org/festival/index.html

This github repository distributes a modified version of festival. Hopefully, these modifications will be merged in the next festival version. If you find a bug using a downloaded festival from this repository, do not contact upstream authors asking for help and open an issue here instead. Bug/patch will be forwarded upstream if required.

Some of the modifications are taken from the Debian TTS group, from the Debian festival package maintainers, from Fedora package maintainers, etc. Other changes are fixes or features implemented by myself. If you are an author of any of these modifications and you notice that your name is missing, please open an issue.

Available documentation:

• Documentation
• Static Analyzer results (generated with clang) (pull requests to fix them are welcome. There may be false positives)
• Code (browse)

Changes with respect to festival-2.1-release:

• Dropped ESD Audio
• Added PulseAudio support [Original implementation by Matthias Clasen, Michal Schmidt]
• Tenths of bugs fixed [Peter Drysdale, Samuel Thibault...]
• Add alaw support [Debian maintainers]
• Honor CFLAGS, CXXFLAGS, LDFLAGS [Sergio Oller]
• Huge performance improvements in text2wave. [Sergio Oller]

This modification of Festival is maintained by zeehio

Hosted on GitHub Pages