Newsgroups: comp.dsp,comp.speech,comp.ai
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!toads.pgh.pa.us!news.pgh.net!news.fast.net!netaxs.com!tezcat!cam-news-hub1.bbnplanet.com!cpk-news-hub1.bbnplanet.com!newsfeed.internetmci.com!howland.erols.net!cs.utexas.edu!usc!news.isi.edu!gremlin!shomase!jbarnett
From: jbarnett@nrtc.northrop.com (Jeff Barnett)
Subject: Re: Voice Box Modelling
Message-ID: <E000HF.C52@gremlin.nrtc.northrop.com>
Sender: news@gremlin.nrtc.northrop.com (Usenet News Manager)
Reply-To: jbarnett@charming.nrtc.northrop.com
Cc: betarose.ix.netcom.com, jbarnett@nrtc.northrop.com
Organization: Northrop Automation Sciences Laboratory
References: <3270D3E9.9D0@ix.netcom.com>
Date: Mon, 28 Oct 1996 18:24:51 GMT
Lines: 28
Xref: glinda.oz.cs.cmu.edu comp.dsp:33716 comp.speech:11267 comp.ai:41721

In article <3270D3E9.9D0@ix.netcom.com>, Bruce Lamar Rosenberg <betarose@ix.netcom.com> writes:
|> Hello,
|> I am looking for a program which simulates the human articulatory 
|> apparatus and its dynamics.  Anyone have information on a source of 
|> software simulating a human voice box in the production of speech and 
|> other utterances?  Thanks.

You might try a literature search using the names Denis Klatt and
Ken Stevens (MIT) or Peter Ladefoged (sp?) (formerly at UCLA).
These folks are (were) first-class acoustic-phonetcs researchers.
There is also an IEEE journal Trans. Audio & Electroacoustics
that might be useful.  In general, the specch production system
is modeled as a bunch of organ pipes and the mathematics associated
with linear systems analyses is used.  At least this was the
state of the art a few decades ago.  Speech synthis implementations
typically model the vocal cords as a periodic impulse function
and various other parts of the anatomy as filters/resonators/organ
pipes.  The tongue separates the mouth into a few resonator
cavities and the nose, when it is used, is modeled as a side-
branch resonator that puts zeroes in the output spectrum.  Lots
of work in this area but you must be willing to read a lots of
papers written by non-computer scienc folks who think more of the
details than of the system.  As a final thought, flip through the
pages of the old Bell Labs research jpournal.

Jeff Barnett


