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

[Richie Burnett]

The intrinsic 1-e^(-t) attack of the classic analogue ADSR circuit matches what you get if you apply constant drive energy to a linear resonant system. E.g. Constant bowing of a string, constant blowing across the top of a bottle, or constant sinewave drive into a resonant analogue or digital filter.

The amplitude of the oscillation increases until you reach an equilibrium where the energy lost from the resonator equals that being pumped into the resonator. (Energy lost can be audible sound radiation, heating due to friction, etc.)

At the initial onset of the drive the string or whatever isn't vibrating, so there's no stored energy, and therefore no energy loss from the string. Therefore all the input energy goes into pumping up the oscillation. As the oscillation builds more and more energy is lost as sound and heat, leaving a smaller portion of the drive energy left to pump up the oscillation further and the rate of rise slows. Eventually energy going in equals energy coming out and we reach the steady-state amplitude.

In theory it takes forever to reach the steady state level, so ADSR circuits typically cut the attack phase short once it reaches about 70% of the way to the steady state settled level, and switch over to the decay phase then. 

I hope this helps, 

-Richie, 


Sent from my Xperia SP on O2

---- Vinicius Brazil wrote ----

>Hi Richie,
>
>
>the fall in amplitude is auditory sense to me, but what about the attack?
>
>A linear attack on a ADSR make more sense?
>
>For example a bow on a violin?
>
>
>Vinicius
>
>
>On Sat, Feb 6, 2016 at 6:29 PM, Richie Burnett <rburnett at richieburnett.co.uk> wrote:
>
>Yes, exactly. It matches the way the amplitude falls off when the energy lost from the string (or whatever is vibrating) over one cycle is a fixed fraction of the energy stored.
>
>-Richie,
>
>Sent from my Xperia SP on O2
>
>---- Colin f wrote ----
>
>
>>
>>> -----Original Message-----
>>> From: Synth-diy [mailto:synth-diy-bounces at dropmix.xs4all.nl] On Behalf Of
>>> Tom Wiltshire
>>> Subject: [sdiy] Lin/Log VCAs and envelopes - compensating for log VCAs
>>
>>> We're usually told that the standard exponential cap-charging/discharging
>>> ADSR curve into a linear VCA produces a "natural" sound, because our ears
>>> understand volume logarithmically - e.g. we hear in dB.
>>
>>Surely an exponential decay curve into a linear VCA sounds "natural" because
>>it mimics the natural decay curve of any struck resonator (piano strings,
>>bells, etc) ?
>>
>>Cheers,
>>Colin f
>>
>>
>>
>>_______________________________________________
>>Synth-diy mailing list
>>Synth-diy at dropmix.xs4all.nl
>>http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>_______________________________________________
>Synth-diy mailing list
>Synth-diy at dropmix.xs4all.nl
>http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20160206/ddfcd700/attachment.htm>