The big problem (as I see it) with designing an additive synthesizer
oscillator is controlling it. Especially when you want to integrate it into
an analog voltage-control system. You figure you need the 1v/OCT master
input, FM inputs, and so on.... And then for each harmonic you would want at
the very least a CV input for amplitude, then probably phase, and finally
possibly an FM-per-harmonic input.
In order to do anything the least bit sophisticated with additive, you
probably want to control the first 128 harmonics, or lets say 64 with the
ability to 'cascade' two oscillators as funds permit. You are still talking
about a module with 200 "CV IN" jacks on it!!! And what good are the jacks
unless you have LFO's and EGs to plug into them!!
So the real challenge in creating a VCASO ("Additive Synthesis Oscillator")
is designing a concise way of making important changes through minimal
voltage control without sacrificing the flexibility that you went to
additive for in the first place. (You don't have to worry about this stuff
in analog, because you are causing harmonic changes via the ∗waveshape∗, not
via ∗partial manipulation∗. E.g., pulse-width modulation creates a radical,
systematic alteration of harmonics, and a pleasing effect, but it's done
using simple analog means (shoving a comparator whose reference voltage is
modulated onto a triangle wave). To do that same thing in additive takes
loads of controls, math, etc.).
One thing you can do is have certain program banks on a rotary switch and
that's all you get. In that case, a Wavetable oscillator is a lot easier (or
VCDO as they tend to be called).
Another is to stick a jack on the panel to which a PC is attached, but that
makes me irritable. Maybe it's a necessary evil for such a project though.
Really what's needed is a "way of looking at" creating an additive VCO. The
circuitry would then fall into place by comparison. But how could the
important aspects of additive be harnessed in a realistic way? ("Realistic"
meaning "get real," not "imitative".)
One method would be to group the harmonics, so you would be controlling
'bands' of them like a vocoder or graphic EQ. Not real flexible, but
workable. This could be made better if you could group them by the front
panel. You know, "The amplitudes of harmonics 2, 4, 7 & 17 are controlled by
CV1, 5 & 16 by CV 2," and so on. This cuts down on jacks, but beefs up the
number of switches needed for selection, unless you went with an LCD
interface and +/- buttons to navigate a menu, which you don't want to see
too many of on an analog synthesizer (we went with knobs to ∗escape∗ those
menus, right?). I would think you'd want some control method that is
intuitive enough that you wouldn't want or need a 'memorization' scheme for
a thousand settings on the VCO in order to get the sound you want.
Maybe you want to group the harmonics into bands, but "coloristically"
rather than "consecutively." So one CV would control the Partials that land
on the Unison/Octaves (1, 2, 4, 7... &c.), the next on the 'fifths' (3, 5,
&c.), 'thirds,' 'sevenths' (speaking crudely of course). You don't need
super-precise control, in my opinion, because in our context we are creating
an additive source that will be followed with subtractive filters and so on.
Even a final VCA. You just want some new, sophisticated colors at the "front
end" to add to the simple waveforms we now use.
This will take a lot more coffee and daydreaming. I keep thinking that
through meditation or experimentation (which is harder) there might be a
"eureka!!" solution to this, some way of making nifty changes to an additive
signal through broad methods.
-----Original Message-----
From: Steve Maietta [mailto:
revtor@...]
Sent:Tuesday, 05 December, 2000 11:19 AM
To:
motm@egroups.comSubject:[motm] MOTM 910 Nice Job Paul!!
...
Ken T, Yeah I actually thought of that chip as a base for an
additive osc module.
It would be something DIFFERENT!!
Digital modules should be easy because absolute sound perfection
wouldn't be as important than with our beloved analog pieces.. anyone
for a digital osc, or digital noise or a >gasp< digital filter??
DIFFERENT!!