acknowledgments.txt

FreeTTS is based upon both the Festival Speech Synthesis System from
the University of Edinburgh and the Flite Speech Synthesizer from
Carnegie Mellon University.

Festival (http://www.cstr.ed.ac.uk/projects/festival/) was primarily
written by Alan W. Black, Paul Taylor and Richard Caley. Festival sits
on top of the Edinburgh Speech Tools Library, and uses much of its
functionality.  The duration cart tree and intonation (accent and F0)
models were derived from the models in the Festival distribution,
which in turn were trained from the Boston University FM Radio Data
Corpus.

The initial development of Flite was primarily done by Alan W. Black,
and Kevin A. Lenzo was involved in the conversion techniques and
representions for the voice distributed with Flite (as well as being
the actual voice itself).

The included lexicon is derived from Carnegie Mellon University's
CMULEX (http://www.speech.cs.cmu.edu/cgi-bin/cmudict) and the letter 
to sound rules are constructed using the Lenzo and Black techniques 
for building LTS decision graphs.

Support for MBROLA voice output was contributed by Marc Schröder,
text-to-speech Researcher in the Language Technology Lab at DFKI, Saarbrücken,
Germany.

The speech team at Sun graciously thanks all involved, especially Alan
W. Black and Kevin A. Lenzo for their expertise and for making both
Festival and Flite available via open source.