[sdiy] Taking a Step towards - - --((FUTURE-PREDICTIONS))-- - -

[Richard Wentk]

At 23:17 09/01/2004 -0500, harrybissell wrote:
>Yes... but you get to stay up all night programming each of the ADSRs for
>all the oscillators...  :^P

No.

>The reason why Additive Synthesis seldon is practical has nothing to do with
>the
>oscillators... but in the man-machine interface.  How do you handle it ?

Analysis and resynthesis, the latter under algorithmic macro control.

>Resynthesis... heck why not just sample then...

Because you can use the analysis as a starting point for creative 
resynthesis. E.g.

Smooth out the envelopes for timbral warping
Change the speed at which you scan through the resynthesis data for live 
time-stretching under real-time control
Make the resynthesis morph between analysis tables under real-time control 
(using velocity, an LFO, whatever)
Quantise the pitch information towards various overtone grids (with 
variable quantisation under real time control)
Add controlled variations to pitch and/or envelope data

...And so on.

>Draw the waveforms... doesn't work. looks do not equal sounds.
>Figure out a set of parameters that have the same ball-wrenching effects
>as twisting a resonant filter knob and you will be getting there...

Only if you're stuck in a 1970s synthesis time warp. ;)

What *really* pisses me off about the MT scene at the moment is that a lot 
of this stuff isn't even new. You could do useful resynthesis and make some 
*very* cool additive sounds on a Synclavier in the mid-80s.

A proper additive synth project sounds like a very fun thing. Where do I sign?

Richard