[sdiy] harmonics & vibrato

[Richard Wentk]

At 18:41 18/12/2002 +0100, you wrote:
>Very good points.
>
>Randomizing Vibrato is a good start (The Yamaha VL7 has this), but I think
>it's not enough.
>
>"Random" functions are much overused / misused in music, IMO.

Yep. It's a question of how you randomise things. A normal distribution 
sounds different to a gaussian distribution. And so on.

>Randomize the pitch of individual notes on a (DCO or VA) Polysynth?
>As if the voice allocation of a polysynth would work like random!

Actually I *transformed* a patch on my Virus just by adding a touch of 
random pitch offset to each voice oscillator. Suddenly it came alive. So it 
definitely does have its uses.

Likewise for a great string sound, add small amounts of random pitch wobble 
for each oscillator separately. With three or four oscs per voice, you get 
something that sounds really huge without any need for further chorusing.

The big problem with most synth hardware is you just don't have the 
facilities to do relatively simple stuff like this. Even something like 
Reaktor - I tried prototyping a random oscillator patch, and it turns out 
all the noise sources are correlated in a single voice. So instead of four 
independent random sources you get a single source shared between four 
oscillators.

Bah.

Anyway - coincidentally I was listening to Trans Europe Express in the car 
yesterday and checking out the sound of the Mellotron. It struck me that a 
big part of the Mello sound comes from the way that it's so random it's 
practically out of tune. It's open to debate whether this is a good thing 
or not. :)

>Randomize the quantisation of drum machines to get a human touch ?
>As if a good drummer would play around the beat at random!

Agree with you on that one. That approach has always been really - uh - 
random. :)

But be fair - it is a great way to simulate a really bad drummer. :)

>Back to vibrato, I think what we need is a good control of the player
>on both, vibrato depth and vibrato rate. Routing both, vibrato depth
>and a tiny amount of vibrato speed to aftertouch is helpful. But
>vibrato depth and speed being linked too much isn't perfect either.
>
>Initial vibrato amount on a mod wheel, and then additional depth and
>rate on Aftertouch might be better.

Controllability *is* what it comes down to. A keyboard with individual 
aftertouch is about a million times more musical to play than one with 
channel aftertouch. And of course if you're playing a cello everything you 
need is right under your fingers. The closer you get to that kind of 
control on a synth, the better the result.

>Another thing to experiment: LFO with DC offset, depth (including the
>offset part) thru a VCA, controlled by Aftertouch, and then AC-coupled
>before it goes to the VCO. Thus a rather slowly varying force from
>your fingertips will be superposed to the LFO's wave.

Wouldn't that just distort the LFO waveshape? I'm not sure I follow this one.

Richard