[school-discuss] etc ... releases new audio books, new technology

[Mike Eschman <meschman@etc-edu.com>]

April 12, 2003.

Engima Technologies Corp., 
etc ... 
"Not just an afterthought ...
http://www.etc-edu.com
meschman@etc-edu.com

A new production release of our audio book creation software is ready for 
distribution.  This release evaluates paragraph, sentence, and word lengths 
and ratios in determining inflection and pacing.  Additionally, over three 
hundred common word prefixes, suffixes and second syllables are evalutaed to 
futrther refine speech inflection.
These are new features.

The end product book that will be used to calibrate future audio books is "The 
Time Machine" by H. G. Wells.
It is available on the download page of http://www.etc-edu.com.
The download is 50 meg., as it is in stereo on .mp3 sampled at 22050 khz.
A mono version would be a little more than half this size.

"The Time Machine" is also being broadcast from http://www.etc-edu.com in 
stereo.

Up to twenty books will be produced using this release of the software.
During this production run, we will be administering the following speech 
therapy 
tests to the books in the production run :

Bankson-Bernthal Test of Phonology.
Compton Phonological Assessment of Foreign Accent.
Compton-Hutton Phonological Assessment.
Arizona Articulation Proficiency Scale.

The results of this testing will be the next upcoming production release, 
which must pass these tests with 
a perfect score, where attainable.

The release being issued today has a far more natural presentation than 
anything i have heard anywhere else.
The phonological and articulation testing will perfect that presentation.  
When that point is reached,
we are going to attempt a waveform smoothing of a Michael Rennie reading from 
"The Day the Earth Stood Still"
into the computer voice output.  We suspect this will result in a completely 
natural sounding product.

In accomplishing these goals, we have eschewed the use of a dictionary.  The 
processing is reliant on a 
three word pipe, the prior word, the current word and the next word.  This 
insures our ability to produce
audio books on demand, using a server equipped with multiple soundboards.  It 
also opens the door to a 
real time embedded solution.

As part of the on-going effort, we are revising the Linux kernel to optimize 
for real time audio processing 
with multiple boards.

I am asking everyone who reads this to listen to "The Time Machine" and let me 
know what you think.

Specifically, I want exact complaints, as in "the word foo is slurred or 
hurried, why is that?

or "the phrase 'up the down staircase' is choppy.

thank you for your time, interest and patience.

mike eschman, etc ...
 

-- 
gutenberg! yum!