<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix"><br>
Bernard Arthur Hutchins, Jr wrote:<br>
</div>
<blockquote
cite="mid:DM6PR04MB60433C1658514427B4CBEB59BCBF0@DM6PR04MB6043.namprd04.prod.outlook.com"
type="cite">
<meta http-equiv="Content-Type" content="text/html;
charset=windows-1252">
<style type="text/css" style="display:none;"><!-- P {margin-top:0;margin-bottom:0;} --></style>
<div id="divtagdefaultwrapper"
style="font-size:10pt;color:#000000;font-family:Arial,Helvetica,sans-serif;"
dir="ltr"><a moz-do-not-send="true"
href="http://electronotes.netfirms.com/AES4.PDF"
id="LPlnk294295" class="OWAAutoLink" previewremoved="true">http://electronotes.netfirms.com/AES4.PDF</a><o:p></o:p>
<p class="MsoNormal"><br>
</p>
<p class="MsoNormal">The “animator” worked through a series of
(LFO controlled – see Fig. 7) of POSSIBLE formant frequencies
(phonemes both used and unused in a particular language).<span
style="mso-spacerun:yes">
</span>It worked - it sounded human!<span
style="mso-spacerun:yes"> </span>BUT basically like a
human suffering from food poisoning!<span
style="mso-spacerun:yes">
</span>Lesson learned.</p>
<p class="MsoNormal"><span style="mso-spacerun:yes"><br>
</span></p>
<p class="MsoNormal"><span style="mso-spacerun:yes">-Bernie </span><span
style="mso-spacerun:yes"> </span><o:p></o:p></p>
</div>
</blockquote>
<br>
Heh! This reminds me of my own forays into this sort of stuff a year
or so ago.<br>
<br>
While experimenting also with analog resonator circuitry for an
analog synth chain, it dawned on me that you could synthesize speech
also.<br>
But then I heeded Gordon's advice even before he gave it (not
really, yesterday was not the first time, haha) and did that
particular thing digitally.<br>
<br>
Starting with barely any knowledge of phonetics and stuff ... I
played with primitive ways of vocal synthesis, as it was meant to
run on a cheap MCU.<br>
<br>
With just a bunch of digital chamberlin SVFs, and a bit of dynamic
parametrizing... (eh end noise source and crude vocal folds
waveform)<br>
Note that there wasn't even an attempt yet to morph between
phonemes, let alone halfway realistically - it's all just stupidly
glued to each other.<br>
The fricatives are also crap, and when I started to work on that, I
was inconveniently lured out of unemployment and had no time
anymore, lol!<br>
So it sounds like some steampunk robotic plunging going on (I mean
in the 2nd example, which was, unlike the 1st, not meant to sound
robotic ;)), but I swear, it's all "electric".<br>
<br>
Excuse very much the horrid fundamental-only melody playing - hey, I
only had a bunch of resonating filters to play (poly) melodies with
in my experimental program setup, it was a quick hack, meant as a
gag to hint at what the vocals are aimed at replicating (more or
less) ... perhaps recognized by people who listened to the readio in
the 1990's ;)<br>
The bleep stuff is over at 32secs.<br>
<a class="moz-txt-link-freetext"
href="http://www.mediafire.com/file/hb9x78fdezxcs53/u96.ogg/file">http://www.mediafire.com/file/hb9x78fdezxcs53/u96.ogg/file</a><br>
<br>
Italians - don't complain about pronunciation ^^ (although the
intellibility isn't that high in general and I didn't have a set of
specifically italian vowels at hand, I kinda winged it)<br>
<a class="moz-txt-link-freetext"
href="http://www.mediafire.com/file/pxx9iez89ziyfg8/vissidarte-lessbreathy.ogg/file">http://www.mediafire.com/file/pxx9iez89ziyfg8/vissidarte-lessbreathy.ogg/file</a><br>
<br>
<br>
- Steve<br>
<br>
</body>
</html>