AW: EG design

[Haible Juergen]

	>If you have a linear VCA then you need an exponential envelope
to get it to
	>sound right. This is what Moog did. But the problem is if you
want to use
	>the envelope for something else. Take panning for instance -
you don't want
	>an exponential pan envelope, you want a linear one. Buchla and
Serges
	>response was to use a linear envelope and make the VCAs
exponential. 

Point taken on panning etc.

But what about the linear envelope + exponential VCA approach?
What will it do to the *Attack* phase ?!

The common (moog and many more) method gives almost linear
ramps for attack, and expo ramps for decay / release.
(almost linear, because the asymptode is usually higher than the
peak voltage)

The lin EG + expo VCA method gives exponential *rising* attack,
which is rather unnatural, IMO.

JH.