[sdiy] fast-attack EGs and trigger sync

[Donald Tillman]

   > Date: Mon, 10 Nov 2008 23:05:03 -0500 (EST)
   > From: "mcb, inc." <mcbinc at panix.com>
   > 
   > Someone on [AH] brought up the question of fast attacking
   > EGs with Cwejman's adsr-vc2 given as an example at 100uS.
   > This recalled something I've thought about but never
   > experimented with.  With fast attacks, would delaying the
   > trigger to a zero crossing of the controlled signal (or
   > even better a zero crossing of the root tone) produce
   > an audible improvement at least when used to control
   > a VCA?  Thoughts?

Another approach is to use an exponentially decaying pulse of the
right amplitude and polarity to cancel the transient before the VCA.

For example, if your audio input is a 1.0 V zero-to-peak sine wave,
and it's up at 0.75 V at the moment of the trigger, you would add in a
-0.75 V exponentally decaying pulse to effectively move the trigger
point to a zero crossing.  

Implementation is easy, put the signal through a lossy S/H, going to
lossy hold on trigger, and subtract that from the original signal.

(This is from some classic DSP Fourier windowing technique, but I
can't remember the name of it right now.)

I was going to ask if any modules do it this way, but... they would
need to integrate both the EG and VCA into the same module for that to
work.

  -- Don

-- 
Don Tillman
Palo Alto, California
don at till.com
http://www.till.com