[sdiy] Waveforms (was dx, chorus and Spock)

[Gene Stopp]

This is exactly the kind of response I was fishin' for! I knew I would open
a can 'o worms if I used the word "better"!

In my minimoog exploits a few days ago searching for oscillator bleedthru I
did notice the shape of the waveforms and they were distinctly "warped" from
ideal. The minimoog is what it is, and it is certainly what many could
consider "musical". It has almost a "tube" warmth to it. Now whether or not
the "musicality" of the minimoog is a contextual thing, meaning I like it
because I'm so used to it, or I've heard it in so many recordings that when
I play one I think "yeah, that's the sound!", is an interesting thought
experiment in cognitive processes. But there is no right or wrong in these
things - we're talking about art.

I guess you could say that textbook waveforms could sound "different" rather
than "better". I remember working on the Big Moog (that was 10 years ago -
wow!) and I was quite surprised at how crystal clear the sounds were. Both
brittle and ballsey at the same time. It was just so different than I was
used to - it shimmered and sparkled and shook the walls. My words at the
time are in Mark Vail's book. I think I said that anything else sounds like
there are pillows in front of the speakers. I wonder if the 921 waveforms
are just more "ideal"? Or was it some kind of "warping"? Or was it that in
combination with the 904A? And then the 902's? Or the fact that the output
went straight to the PA? I'm sure it was a combination of many things.

But on to engineering. It seems that the circuits we use for wide-sweep
subaudio/audio VCO's may not have the same waveshaping capability of an
11-Mhz-capable function generator like the HP3325. We're more interested in
0.01 to 25Khz range with good exponential response than invisible vertical
lines and straight sawtooth slopes. Does a squarewave with perfectly flat
low levels and high levels and infinite slopes between sound more "hollow"?
When you detune two or more together, do they make a better "Lucky Man"
sound? ("better" in quotes, of course). I know that the 4151-based charge
pump VCO in the Rhodes Chroma has a large flat spot in the sawtooth - how
"good" does the Chroma sound? Should we attempt to idealize waveforms when
designing, or just take what we get and delight in the occasional accidental
gem?

This makes me want to put the scope on everything I own. Or maybe I should
just play and enjoy...

- Gene




From: Scott Gravenhorst

If things like *perfect* sawtooth waves sound "better" (highly
subjective statement I think), then why in God's name would we EVER
want to filter?  Using a filter changes the harmonic structure such
that in most cases, the resulting wave is not a perfect anything.  But
we do use filters.  Why?  Because doing so gives us a sound that is
correct for our purpose at the moment.

Waves are waves.  Geometric ones aren't any "better" than ones that
don't look perfect on a scope, they just have names that describe their
shape on a scope screen.  Some waves with warts sound good, others may
not.  I think that what a wave looks like on a screen has nothing to do
with whether it sounds "better" or not.  Harmonic content is what makes
a sound, and only harmonic content.  How many of us sit there and tweak
up a patch while watching a scope?  Special purposes notwithstanding,
I'd say most of us just use our ears and noodle 'til we smile.  I've
looked at waveforms after doing that and honestly, they can look pretty
nasty.  

Perfect waveforms do have their purpose, such as sweep generators, or
slope generation for accurate measuring circuits.  But this is more in
the field of instrumentation, not instruments.

I'd say Gene had a damn nice filter that was doing a great job of
isolating the harmonics.  Other repeating waveforms whether a "perfect"
something or not would still be made up of those same harmonics, just
different levels and phase relationships.  I'd bet that Gene's filter
could have isolated those harmonics in a lumpy weird looking wave too,
as long as it was frequency and phase stable and noise free.