🚀 go-pugleaf

In article <58...@eds.ericsson.se> lme...@eds.ericsson.se (Bengt G{llmo, L M Ericsson, Stockholm, Sweden) writes:
>the random number generator, it happened to be on. The random
>number generator made the radio sound like a (US) police car! It
>whined up and down, with a period of a second or more. Now, the

This is interesting to me -- Does anyone know of any other applications
using the auditory modality for 'research' like this?  There was much
discussion about Feigenbaum, et. al.'s use of computer graphics imaging
to see patterns in chaotic systems... it seems to me that the sonic domain
would be an excellent place to check for periodicities which might otherwise
go unnoticed.

Brad Garton
Music Department
br...@woof.columbia.edu

In electronics everything is shrinking, what was musical computers is now
musical chips.

I was in my lab a couple of years ago adjusting the bit rate clock on a 
prototype LAN to see how fast it could go.  The limit was when flip-flops
inside to the protocol chip went meta-stable.  Detecting this condition was a
little difficult until I noticed that the FM radio in the lab would lock on to
the second harmonic of the bit rate.  I could track the bit rate by tuning the
radio as I increased the clock speed.  If all was going well, the radio had a
nice quiet null at the second harmonic.  When the meta-stability point was
reached, the null disappeared because the flip-flops were not changing in
unison and not enough signal was generated for the radio to capture.

An older story:
You don't need a radio or peripherals to get sound out of a computer.  All you
need is a large core that has bank set and reset capability.  Set and reset
half of the core at 100KHz.  Set and reset the other half of the core at
101KHz.  It is very difficult to locate the source of that 1KHz beat.
Of course, varying one of the frequencies varies the beat pitch.  We never
completed the task of modulating the pitch into speech.  Our goal was to have a
eerie voice come from within the memory saying "Help, I trapped in central
core".  Timing loops within timing loops to produce phonemes is difficult
(consonants are *very* difficult to make out of beat tones).

Using sound to detect periodicity is not so odd as it may seem.
Charles Dodge, a composer at Columbia in the '70's used data from
the fluctuations of the earth's magnetosphere as recorded by NASA
probes as the basis for a piece of digital electronic music called
(I believe) "The Earth's Magnetic Field" and it appeared on a
Nonesuch recording.
-- 
 /s/ rchrd <=> Richard Friedman <=>  rchrd@well
 rc...@well.sf.ca.us | {apple,pacbell,hplabs,ucbvax}!well!rchrd
 [Pacific-Sierra Research / Berkeley CA] (415) 540-5216
 (The usual disclaimers apply - I speak only for myself!)

In article <26...@cunixc.cc.columbia.edu> gar...@cunixa.cc.columbia.edu (Bradford Garton) writes:
>This is interesting to me -- Does anyone know of any other applications
>using the auditory modality for 'research' like this?  There was much
>discussion about Feigenbaum, et. al.'s use of computer graphics imaging
>to see patterns in chaotic systems... it seems to me that the sonic domain
>would be an excellent place to check for periodicities which might otherwise
>go unnoticed.

A few years ago I wrote a program to display the lexical scoping of Pascal
programs.  In doing so I had to write a parser that could read and understand
Pascal syntax.  While debugging the parser I got tired of sitting and watching
the display as it spit out token after token, waiting for it to hang up - its
usual way of crashing.  I first tried just beeping the pc's speaker as each
token was parsed but at 20 or so tokens a second, that was too monontanous.
I ended up summing the ascii value of the characters of each token and using
that for the frequency of each beep.  The result was an amazing cacophony
of sound.  At first hearing it seemed totaly random, but if you paid 
attention you could hear patterns as the same tokens were used repeatedly
within a procedure or function.  This became so much fun that I added an
option to slow down the parse when generating the audio signal so that each
tone could be heard.

A larger suprise came when I began analyzing other people's code.  There was
a program I had written with two other people, about 10,000 line of code.  I
got so that I could listen to the audio signature of the code being parsed and
identify which of the three of us had written that piece of code.  What is
worse is that one of the programmers was not very good and his code sounded
very chaotic.  My code (average quality) sounded ok, but the third person
wrote very elegant code and the audio signiture of her code reflected that.

What it suggested is that good code not only runs well, is easily maintained,
but also sounds good!

					Andrew MacRae

In article <15661@well.UUCP> rchrd@well.UUCP (Richard Friedman) writes:
>Using sound to detect periodicity is not so odd as it may seem.
>Charles Dodge, a composer at Columbia in the '70's used data from
>the fluctuations of the earth's magnetosphere as recorded by NASA
>probes as the basis for a piece of digital electronic music called
>(I believe) "The Earth's Magnetic Field" and it appeared on a
>Nonesuch recording.

I *have* this record.  I got it for $0.25 at a library sale a month
ago.  It didn't look like it had been played very much. :-)  I
suppose I'll have to actually give it a listen now ...

	Howard A. Landman
	lan...@eng.sun.com -or- sun!landman