[sdiy] Lin/Log VCAs and envelopes - compensating for log VCAs

[Tom Wiltshire]

Richie,

> Have a look at this video if you haven't seen it already...
> 
> http://www.youtube.com/watch?v=0oreYmOWgYE
> 
> Nigel Redmon does a good discussion of the whole expo vs linear ADSR thing.

Yep, nice and visual. Excellent stuff. Pity about the mess-up that Neil spotted, but we'll let him off - he's a software guy, right?!
 
> The intrinsic 1-e^(-t) exponential rise in the attack phase of the ADSR intuitively feels back-to-front in terms of linear volume perception. However, it does actually resemble how energy builds up to a steady-state equilibrium in real bowed or blown instruments. A real e^t exponential function gives perceptually linear rising volume in the attack phase but in practice it lingers too long at the low volumes to be useful on a synthesiser, and just plain sounds wrong. 

That was my motivation for wanting to copy the sound of the classic ADSR into a linear VCA, rather than just using a linear envelope for a log VCA, which then gives you a weird attack. A linear decay and release is ok (although the release can have a nosedive at the far end if you like) but the Attack should compensate the Log effect and then some - "Uber-exponential", if you like.

Tom