Sequencing and Pseudo-Randomness
-----------------------------------

"The aim of synthesis is not to imitate nature, but to imitate nature's
way of functionning" (First law of synthesis).

Western symphonic orchestra musicians are often ill at ease
when they have to perform a graphical score requiring them to play
random notes and various noises on their instruments. They are also discontent when
the public doesn't observe complete silence during their performance.

Oriental musicians, on the contrary, consider unplanned random events
to be an intricate part of the performance. To them, it is a
reflection of nature's dynamism : aleatoric movements flow and changes
are essential renewal factors in any complex system.

So, accepting the idea of incorporating unplanned pseudo-random occurences
in a work of art is, essentially, a philosophical issue, often conditionned by
a cultural worldsview...
 

The clock's tyranny

Musical sequences often appear to be lifeless or too machinistic :
this is generally due to a misunderstanding of the master clock
potentialities.
Too often, usually due to musical archetypes reasons, musicians tend to restrict
the behavior of the master clock to a standard 4/4 rhythm.

As everyone knows, the major drawback of using a clock for timing purposes
is that all steps in the sequence have the same set value in time...
Let's not talk about the restrictions of another important parameter: silence...

For example, in the acoustical world, the shape of
envelope parameters are not static at all: their shapes are constantly changing
in time, depending on the types of "phrasing" techniques used and in what octave
range the musical piece is performed.

For those willing to innovate with new rhythms, it is sometimes interesting to
use the clock's CV input, to frequency modulate it from an external voltage,
such as row B of the sequencer or any device generating a DC voltage.

In some cases, the best way to get round is not to use a master clock to trigger
a sequencer. Instead, use the master clock pulse to trigger sophisticated up/down
counters and/or divide by n counters : they are, usually, the best alternative
for all your complex clocking needs.(ref: my previous column on "3D
Polyphonic sequencing").

Once you have made up your mind about your complex rhythmic base, you can experiment
with other timing devices to simulate the "off-tempo" typical phrasing of a solo
performance : this is when envelope shapers come in handy.
Indeed, triggering sequencers with envelope generators offers great advantages :
events in time are now interspersed with meaningful silences!

Obviously, triggering a sequencer from a clock driven envelope generator depends on
the pulse-like shape of the envelope and the gating/triggering characteristics
of your gear. In most synths and sequencers, triggering occurs either on the
rising or falling edge of the pulse or of any other waveform used.
 
 

Dynamic time sequencing

In the above patch, the sequencer's clock input is "tricked" to accept an
envelope output as its timebase : the envelope mimicks a typical A/R (Attack/
Release) envelope, by using a rising-edge sawtooth-like waveform, with a variable
"fall" slope, in lieu of a standard clock.

For those not familiar with block diagrams, this is the explanation of the patch:
The X output of an analog X/Y DC Joystick, with scalable -/+ Volt range selection for
each axis, controls the frequency of a master clock, while the Y output is patched
to one of the inputs of a CV mixer (the Y axis acts as a manual bias).
The master clock's output is then sent to the trigger input of an envelope generator
and to the logic inputs of two Sample and Hold modules.
The first one, is configured as a stepped generator (arpeggiator): i.e a slow
triangular waveform is patched to the analog input. The second, is configured
as a Sample & Hold exhibiting pseudo-random behaviors : i.e white noise
is patched to the analog input.
The two Sample & Hold outputs are then sent to two inputs of a CV mixer for attenuation
and processing.
The combined DC outputs of the CV mixer are used to voltage control
the "release" parameter of a simple A/R envelope. This variable
slope envelope is then sent to the clock input of a sequencer and to the trigger
input of a second envelope generator. Finally, the second A/R envelope, which is
set in a log configuration, is patched to the control inputs of a VCF and VCA.

A note for Serge users: I have also shown the patch using Dual Universal
Slope Generators. As a reminder, all Serge's triggering occurs on the rising-edge
of the waveform used. If you want to use the falling-edge of a VCO sawtooth output
as an envelope, you'll need to process it into a scalable inverter :
this rising-edge sawtooth waveform will then trigger the DSG.

The originality of this patch is that the time value of each step in the
sequence changes constantly and is interspersed with silences, while remaining
virtually in rhythm with the master clock's beat!

Evidently, the ratio between the master clock's frequency, the
triangular waveform's frequency and the range of randomness exhibited
by the second Sample & Hold is critical : an envelope with a sharp release will
follow the master clock's beat faithfully, while one with a
longer release time will skip a number of clock beats, thus stretching the time value
of that step momentarily!

This ratio of relationship can be tailored at will within the CV mixer, by
attenuating/emphasizing some functions, as necessary.
If needed, the joystick Y output DC voltage can be used as an additional bias
to shorten/stretch the release part of the envelope manually.
 

André Stordeur
10.27.2002