Under construction:
History of MarsRadio and newer N-Synth technologies.
1981:The 26-byte audiovisual version typical of all 1980's N-synths
This diagram may be in error because it was more a CA than an LFSR.


Throughout the 1980's, following this, a lot of speech synthesis,
primitive singing of typed phonetics, and distorted single bit
music and sound experiments were developed for games and
for uses of extraterrestrial fantasy.

In part, much of this is owed to the subsequent discovery of the useful
functions of Viroboxes, of which the portable entropy synthesizers
are distantly related, most Viroboxes being simple pure analog
devices with fewer than 10 transistors. Viroboxes are fantastic
alien technology with no original but later discovered practical
uses. They typically make impulse buzzing when you touch them,
respond to music, blink forever, and have a very high resistance.
The ambiguous way in which they respond to music helped the
invention of the deliberate non-automatic N-synth technologies.

In the 1990's, the program was embedded in portable devices
with a few variety controls added...

These devices were called MarsRadios and often had antennas.
They used the original algorithm with 4096 or 8192 byte loops,
but could not display video audiovisualization.
The only controls of variety here were changing the loopsize,
and adding truly random noise by using the ADC rule as long
as the ADD button was pressed. Eventually the ADC rule would
unrandomize and alternate between 2 tones, but if released
before then would add rhythm to the sound being generated.

During this time, the 1990s, more speech sampling of higher
quality was being experimented with, and early in the decade
we had a stereo BCD champernowne string player,
called CHIPMAN. It is forgotten what CHIPMAN originally
played, but it was invented in 1991, after which stereoBCD
was shelved, and in 1995 whatever it used to play was
replaced with "Great Horny Toads!". This player used a
6502 processor WITHOUT SOFTWARE and 32K of eprom.
(Actually there was one byte of code: NOP!).

One original use of CHIPMAN was for ambient sound loops.
Since then, ambient sound loop devices have become common
for helping people to relax and sleep in urban areas.

Since the introduction of Windows 95, no one has succeeded
in emulating an N-Synth, especially with realtime audiovisualization
via the screen buffer, such as was done originally in 1981.

At the turn of the century, attempts at building MP3 players
were abandoned when portable ones became available,
and there was interest in finding more automatic sound
generators.

First it was discovered that, though the original MarsRadio
was a 1-dimensional CA, it displayed as 2 dimensional patterns,
and was found to be making Sierpinski Triangles on the
edge of subsequent frames when stacked.

[insert emulation image]

And then the Sierpinski triangle was found to be
"virtually phonographically" equivalent to MarsRadio
seeded with only one "dot".
And that, arranged in 3D, would generate a Sierpinski Sponge.

Backing up a bit, in the early mid 1990's I envisioned a symbol
called VALERA/Cybernetic Apex/Fuller's Eye:
[insert image]
And this to me represented a method of deviceless communication,
suddenly seeming practical whereas beforehand was
merely the core of the extraterrestrial fantasy since I was 6
years old.
But more recently I saw a painting of the Sierpinski Triangle
with watercolor squares enclosing the lone top triangles of
each size, and saw in an instant ALL POSSIBLE DATA.

And I realized that making it into a right triangle, those
top triangles square-enclosed, when plotted floor(log2y),
makes a binary plot of the numbers...
1234567891011121314151617181920212223..oo

And thus there was a map of numbers.
I once imagined having a computer count from all zeros
to all ones, generating all possible data and programs.
But this graph shortened the ultracosmic time it would take
to a mere instant.

The Champernowne Constants provide instant infinities
to explore for raw data. Teach monkeys to count, and
tell them where to start, and they will right away
type Shakespeare!

Thus I researched #Ch.

The fractals discovered in the N-Synth/Marsradio,
were infinitely redundant, surely they contained an
infinite number of redundant everything.
Marsradio has a distinctive "essence of music" melody.

ENTROPY SYNTHESIZER, Why was it called that?
1.It was discovered while trying to generate random numbers.
2.The test for random number was to see and hear
audiovisual static on the monitor.
As the audiovisual sample of the discovery shows, it
was musical and visually patternful.
3.There was increasing complexity in the audiovisualization
as time passed after starting it.
4.At the time, I thought that Synthesizers were
intelligent music boxes.
Intelligent enough to create a symphony perhaps.
5.The advanced technology of my imaginary extraterrestrial
friends included the power of machina ex deus, and included
Music Synthesis (as well as Flying Superhero Powers).

MEMORYLESS ENTROPY SYNTHESIZER
Basically an arithmetic/logic function capable of accessing
any part of a graph at any time. First came the method of
reading the coordinates to see if it was part of
Sierpinski's triangle or not:
This is the formula for doing that.
NOT(x AND y)=0 ; AND is wordwise boolean arithmetic
And here is a MEMORYLESS circuit which uses that
to generate Marsradio Dot-seed music:

And when the formula is graphed, here are two of many
possibilities generated, the switches represent ANDing
with a presettable constant.

Green: Sierpinski Gasket , Red: Sierpinski Carpet

The binary 012345678910111213 fractal...

can be seen as the triangle rotated 180 and stretched out like silly putty.

This is the spectrum of the characteristic melody of Sierpinski's triangle:
(Rendered from the tape of MarsRadio Dot Mode)
NOTICE THE DC BIAS!


This research changed the direction of what was formerly Googolodeon.com
which attempted to calculate sounds using LFSRs.

With all this information so far,
we are still in the "realm" of ENTROPY synthesizers,
because although any digit of any number can be found at once,
it takes more information to specify that digit than is contained
in the whole part of the number.
Wait! What number? The Champernowne number.

There are formulas for doing these things:
1.Finding data (like Shakespeare) in the Number.
2.Finding what the nth digit of the Number is.
3.Finding the beginning and end of an Apparent Number.
4.Extracting specific data from the number.
This is still an Entropy synthesizer for an unmentioned reason:
THE NUMBER has infinite copies of any particular information in it!
This makes playing this number, starting at any xth digit,
a data predictor. Like the Marsradio, the number is musical,
and infinitely redundant.

Now for the future of this research project:
CONSTRAINED DATA SETS.
The result of this discovery is REVERSE ENTROPY SYNTHESIZERS!

The number of possible audible sounds of length N
is computable and less than 2^2^N, and is N!.
(N factorial)

Even without further constraints, the result is a discovery
of an "all music set player" which gets simpler the more
music is found in it. In other words, the number of the
sound is less than the number of the data in the sound,
and the more songs found, the less relative distance between
them, and thus the smaller the numbers used to find them.

The redundancy in this set is all kinds of variations of
volume and phase, and this set is like stars in a noisy
space. But the space is collapsing as more music is
found in it. When music was found in Champernowne's
number, there was a problem with expanding space, but
an expectation that since the space takes up no memory,
and can be read instantly, it didn't matter how large the...
Secondary Finding Function's data set would be, since
the whole virtual universe of data was redundantly full of
data, including the software that makes the ...
Champernowne~Turing Machine SYNTHESIZE!

But, it got much simpler when the set of all sound,
the Constrained Music Fractal,
when it's finders got progressively smaller than it's members.

This is the constraint:
Although the Champernowne number has an average digit
of 4.5, "virtually all" of the apparent numbers in it do not.
The triangle has an average bit of 0, and that is a DC bias.
All physical sounds which can be found in the Number must
be biased at average digit=4.5
By removing an infinite amount of non-4.5 numbers, the
remainder is finite and countable in an increasingly
convenient way.

Using memoryless logic, and a rotary "tuner", one can
roll through the set of "gray" numbers
(not gray code, not yet at least, but that does have relevance)
listen for good sounds in the same way you find a good
radio station on a transistor radio.

Gray code is one of several orthogonal transforms that
may make it even more convenient to explore this
wholly intangible, as intangible as any other fractal,
or irrational number, intangible universe of sound.
Why?
Because Gray code moves in a large number of "dimensions".

Orthogonal... perpendicular... moving along a different dimension.
Other Dimensions besides counting "gray" numbers are:
Gray Code
Bit Reversed Counting
Sequency Domain Counting*
Intangible Phonography (simulation of a needle following a graph)

(*) Thanks to Steve Bjork's Audio Spectrum Analyzer,
I knew in the early 1980's that Fourier Transforms do
simplify sounds as timeless chords of finite tones.
HOWEVER, Virobox buzz approaches infinite tones.

The counting which is done by Entropy Synthesizers
and by Reverse Entropy Synthesizers is not truly the
sound we perceive it to be, but the fact that it does
sound like music, makes it a suitable replacement
for music, to anyone who is pleased to listen to it.

The "gray" numbers, which are perfectly biased,
can indeed sound like songs. The Champernowne
string wrapped up in a WAV file is certainly not
music, it is even DC biased. It sounds bad. It should
not be played on a computer. It should not be stored
anywhere. It should be calculated one digit at a time
in a Reverse Entropy Synthesizer with no memory.

Not only will it sound better, but there is, somewhere,
in the all music set fractal, an excellent stereo "groove".

BEYOND:
This research includes recording dreams and imaginary images.
This research includes a unique method of voice synthesis
for singing, since finding good lyrics with the described
synthesizers would be as unlikely as finding an unpublished,
correct dictionary in the Number.
This research includes 3D VR world creation technology.

The purpose of this research is to enable myself and others
to discover beautiful sounds, drawing them out of intangible
mathspace, recognizing them as beautiful sounds, not static,
in much the same way we tune a radio. And to realize the
imaginary alien technology for making music.

The birds make free music, the wind chimes make free music,
the universe, physical and intangible makes music,
Pythagoras, in 300BC, knew this: Music is made of numbers.

The device that is being developed to play calculated music
(which is called a Reverse Entropy Synthesizer) is also known
as the "Pi-Pod" prototype.

One more thing: The CD-formatted memoryless-synthesized
file in the sound gallery, called SR01A.WAV was inspired
by and is phonographically similar to the shape in this image:


Incomplete Public Domain/(cc) disclosure of
Entropy Synthesizer
page under construction, updated 09.16.2005

ENTROPY SYNTHESIZER SOUND GALLERY