Attempted to linearise the curve moreso tonight.
Learned a few things.
The first experiment, initialised a patch, and set the Decay knob to zero, Sustain Level to maximum, and used the (Sustain)Time knob to simulate decay, instead of the Decay knob itself. Believe it or not, the slope of the Sustain(Time) is
also a logarithmic curve, and is not a linear slope (although it appears a teensy bit more linear than the actual proper Decay curve)!
So I thought to myself, if the Decay curve isn't linear, the SustainTime curve isn't linear, what IS linear?...
... A falling sawtooth waveform.
Re-initialised a patch again. Then turned Decay to zero, SustainLevel also to zero.
Then chose an LFO, selected a sawtooth LFO waveform for it, put the LFO into EnvMode (one shot), made sure the LFO is in poly mode (so a new LFO is used per voice, as needed), and set the LFO Assign destination as AmpSustain with the amount set to maximum (+63).
The LFO 'Rate' effectively controls the length of the decay hereon.
Worked well, until I played several notes in quick succession, the LFO glitches up and doesn't retrigger new notes correctly, even though the LFO is in Poly mode, and even with TrigPhase set to 1 (to make sure the LFO is firing from the start [the top] of the sawtooth).
Did a bit more testing and I found it only farks up when you hold down one note and then play another note elsewhere.
So, it appears I've found my first bug on the Virus (Indigo)!
However! Using the
ModMatrix to assign LFO > AmpSustain, instead of selecting AmpSustain from within the
LFO section itself, works properly, as it should do, with no LFO poly issues or glitches.
It's still not a perfectly linear slope, but you lose the rapid fall off at the beginning and the half-life of the waveform is more towards where you want it to be, along with a more consistent linear fade-out at the very end.
And it's yet another different curve to add to your synthesis toolbox (compared to modulating FiltSustain by FltEnv, or AmpSustain by AmpEnv, earlier in the thread).
BTW, for the above screengrab from Wavelab, I modulated the Amp envelope -
not the Filter envelope - as it's easier to get a visual representation of an envelope when merely recording signal levels as opposed to a filter sweep (furthermore, audio frequencies themselves aren't linear).