Well, you did ask for it.
At the moment I am trying to keep my head focused on building up my
first (and just ordered second) set of MOTM for the core of my
modular. I think that it will be pretty good sized when it is done,
mostly because I have the money, time and desire to throw at it right
now.
But I am also thinking about trying to build some stuff myself some
later time down the road. I am not analog expert, but I can hack DSP
stuff with the best of them.
Back in the ancient days when I was in collage, computers and music
were a hot ground for weird ideas and mad scientist types. This was
my sorta place! If you got back and look at issues of "Computer Music
Journal" of the time, you will see that folks at places like UCSD,
CERN, IRCAM, and UofI were all up to some strange tricks to get CPUs
to produce odd farting noises. Most of these guys were good little
scientists and published their works for the world. Many of them got
swept under the rug of time as the masses consumed powerful synthesis
techniques in mass produced keyboards. In that day and age, there was
not a small single chip computer with the power that could do the
stuff that they described. Well, today there are.
Here is what I have as my pie in the sky type thoughts
ADO - Additive Digital Osc
This module would be a DSP based additive VCO. That means it would
read control voltages through some A/D process to come up with the
fundamental frequency to do the additive calculation. Now this could
be fun in itself, but hardly worth losing your mind over. So my
thoughts are that you need to have a nice selection of control voltage
inputs to make the ADO's world a lot more chaotic. Here are some of
the
parameters I think need to be CV controlled
· Wave - This would be some raw additive spectra to serve as a
starting point for our fun. I think these could be like 1-99, I would
love to had a nice little knob next to a 2 digit LED for this.
(comment, I think the MOTM stuff is amazing, but I am an LED freak....
so if you see my MOTM some day, and you think I have put too many LEDs
on stuff.... sorry).
· Pitch - Just like the VCO that exists today. There would be CV
inputs for this, as well as a coarse and fine adjustment on the front
panel
Control Modifiers
· Low Harmonics - This would modify the level of the lower 16
harmonics en masse based on a CV amount
· High Harmonics - This would modify the level of the upper 16
harmonics en masse based on a CV amount
· Even / Odd - would modify the level of the even (or odd) harmonics
en masse based on a CV amount
· Brightness - would apply a "brightness curve" to the harmonics based
on a CV amount
· Darkness - would apply a "darkness curve" ..
· Octave - Would adjust the level of the octave harmonics
· 5th - Would adjust the level of the 5th harmonics
· Spread +/- - Crazy little function that would spread the harmonics
out or contract them based on CV
WTO - Wave Table Osc
I dearly love my PPG. It is a true marvel of design and sound. I
liked it so much I even bought the dodgey Waldorf knockoffs. But I
have always wondered what you could do with a nice wavetable VCO in a
modular environment. Wiard's Miniwave is a good start, sure but I
think one could get a lot more wacked out . Once again there would be
some core control voltages that would determine the pitch of the WTO,
and of course you could do audio frequency FM type poo into the CV
that controls the pitch. But past that is where we get really weird,
and this is a list of the parameters that should be CV controlled and
what they do:
· Pitch - Just like the VCO that exists today. There would be CV
inputs for this, as well as a coarse and fine adjustment on the front
panel
· WaveTable - Little knob and LED to choose which of the wave tables
you will use.
· StartWave - Little knob and LED to choose the start wave within the
wavetable
· Phase - The phase of the resulting waveform offset from zero
Control Modifiers
· WaveTable - CV input to adjust which wavetable we are using for the
playback
· WaveNumber - CV to adjust the wave number that we are playing back.
This lets us sweep wavetables like a PPG can.
· Wave Phase - The phase of the wave relative to zero. Now, put a
lovely audio signal in this and you can play casio tricks... whooo...
Phase distortion synth...
There are a few other gags that can be done with Wave Tables, but I
will not bore you with them here...
SSO - Spiral Synthesis Osc
Once upon a time, some wacky guy was looking at the physics and
calculus behind spring motion. Those of you who took too much collage
calculus or physics know that the Fourier functions that lie at the
guts of DSP also seem to have this grip over springs, slinkys and
blocks of Jello. While to most this is the subject of annoyance or at
worst difficult omputations, I say we put this to our use! So imagine
if you would that we have this really wonderful spring, and we can
control some of its motion through control voltages. Now shine a
light from above the spring onto a piece of paper below it. It
produces a shadow that looks a lot like a wave form! Ok, lets output
that wave form. Gee
sounds kinda
weird
But wait! Theres more.
Lets make this spring out of some stuff that no one has ever seen
before, and allow it to operation in more than just 3 dimensions, lets
be able to warp and reflect and twist it based on control voltages.
Now you are talking weird stuff. (ok, I coded this one in school
. It
is one fruity sound generator). Once again we have control voltages
that determine the fundamental frequency of the output spiral
waveform. Here are some of the control inputs that I think would be
fun, and what they do
· Pitch - Just like the VCO that exists today. There would be CV
inputs for this, as well as a coarse and fine adjustment on the front
panel
· Drive Waveform - a CV input from an audio VCO. More about the
horror this could cause in a bit. If you don't supply this it uses
its internal sawtooth
Control Modifiers (some of these are weird, ok - But remember the
shadow metaphor)
· X Reflect +/- - Think if it as using a CV to folding a single cycle
of the waveform about the X axis.
· Y Refect +/- - Think if it as using a CV to do the same to the Y
Axis
· X Sheer +/- Think of it as using a CV to move the light source along
the X Axis
· Y Sheer +/- Same trick to the Y Axis
· Twist + / - - How much torque you put on the spring. Oh yeah, audio
CV in this one and the world comes unglued
· Dither - This would be placing a diffuse sheet between the spring
and the paper that shows our wave form.
There are others, but this should be enough to cause trouble for the
next while.
One last note on this one. There is an implied driver waveform that
is a sawtooth. I did not have the CPU power back in the early 80's to
try a different driver. Today that is not a problem
The sawtooth
simulates the back -and- forth motion of the spring excitation, so if
you feed it something other than a sawtooth it becomes a very
complicated waveshaper.
Beyond that, I have always been in love with tempo, and the ability to
mathematically relate temp within a sound and across a musical piece.
Analog is so lovely for the rich sound you can produce with it, and
the absolute flexibility you can have to cross pollinate between the
signal and the control domain. But I want to introduce tempo into my
modular. So it would be fun to have a series of modules that do some
odd tempo style things.
(these thoughts are a little less well formed)
CLK - Clock Driver
This would read an timing input (midi or DIN) to perform some basic
tempo calculations. It could also generate an internal tempo clock.
It is really sort of the master interface to the other weird and goofy
things that follow.
PCV - Patterned Control Voltage
This would allow the user to select a pattern that would be tempo
driven from a sync signal from the clock drive. Those of you who have
goofed around with Propellerhead's Rebirth tool know about the pattern
controlled filter they have. I like this a lot. But why just a
filter? Pattern control my PWM baby! Or my VCA's that are pumping
audio FM source between VCO's. Modular is about breaking the rules.
You could probably put a couple of these in one panel, and have 2 or 3
different patters coming out of the same 2U gizmo. (Ideas here?) It
would be able to do these step patterns on any note basis (1/32nd to
multiple bars per step).
TLO - Tempo lock LFO
Ok, I have an Emu Audity 2000. Great little sample who-ha. But the
magical thing is that all umpteen gazillion LFOs can synch to the midi
clock. Oh my, and they can cross modulate each other, and they have
like 16 different waveform. So as I am a quasi digital goof, why not
unleash this concept on the modular world. I figure you could get 2
of these in a 2U panel. Sure it won't be analog, but it will be weird.
It would be able to produce a low frequency wave form on any note
basis (1/32nd to multiple bars per cycle). Ok, want to make your head
explode? Imagine hooking a few of these up to the spiral VCO I
described above, temp lock them to different beat divisions.
Each of these could probably run off a mid range embedded processor. I
think all of them could be built off of a single "base module" that
handled the D/A and A/D functions and reading the knob inputs. I am
not sure I could pull of the design myself, but at some point I will
probably give it a go. At that point each of these is a different
front panel and different ROMs. There is no telling if I will ever
get around to building any of these, but what the heck, it makes for
some fun discussion.
Now aren't you sorry you asked?
Bruce
PS - You don't want to know what would happen if you applied modular
concepts to something called a "Convolution Matrix"