IPOX Home Page
IPOX is an experimental, all-prosodic speech synthesizer, developed some
years ago by Arthur Dirksen and John Coleman. A preliminary version is
available for downloading. In order to install and use IPOX, you need a
PC (preferably a fast 486 or better) running Windows 3.1 or higher. Sound
output requires a 16-bit Windows-compatible sound card, such as the Soundblaster
Please note that IPOX is a rule compiler, not a text-to-speech
system. With the codes bundled in the package, it can be used to generate
short words in English, using a phonemic input (not normal spelling). It
is also possible to adjust the relative durations of syllables (e.g. to
alter the rhythm), and using a parameter editor, to tinker with the Klatt
parameter files produced by IPOX in order to improve the quality of the
sound files that it generates. It has been successfully used by researchers
at many institutions to generate synthetic speech files for use in speech
Please take note that the documentation (in the form of Windows Help
files) is incomplete. Because neither of us has any continuing motivation
to develop IPOX further, we have no plans for upgrades etc. It is what
it is, an experimental/prototype. Note that if you are looking for a text-to-speech
system, this is not one!
Arthur Dirksen's 1995 lecture
notes about IPOX are a useful source of additional, unpublished information
about the system.
IPOX is freely available for evaluation and non-profit research purposes.
However, you are required to register your copy by filling in the form
To obtain a copy of IPOX, download the following files:
To install IPOX, follow the instructions in the file readme.txt.
Please, fill in the form below, then press the Submit button.
Send any comments, questions or bug reports to firstname.lastname@example.org.
Although we are grateful for feedback, we do not offer any kind of support,
and we do not undertake to reply. Please accept our apologies for that
On-line reference material
On-line information about IPOX and related systems is available from the
Audio demonstrations complementing our 1997 paper `All-Prosodic Speech
Synthesis' (In J. P. H. van Santen, R. W. Sproat, J. P. Olive and J. Hirschberg,
eds.) Progress in Speech Synthesis. New York: Springer-Verlag. 91-108.
This page describes YorkTalk, the system which inspired the development
Arthur Dirksen / Fluency Speech Technology
John Coleman / Oxford University Phonetics Laboratory / email@example.com
Last updated: 23/2/2001