Good envelopes

[realmarco]

i like Absynth's envelopes you can even have them "click" no hardware analogue synth could have faster envelopes

This is not true though. If you carefully listen and start to actually measure the "click", for instance using an oscilloscope for the analogue signal, you'll quickly notice that there is way more to a simple click than meets the eye.

Fast in theory (amount of samples in a digital output) versus how fast it sounds to our ear/brain, are two different things. I know, it's hard to believe but that's how it is. Shape is everything.

For instance, in a digital system, the actual sample positions in a 8 sample (0.2ms basically at 44.1kHz) mini-click will decide very much how the sound actually sounds when it is filtered at the DA converter of your sound card. In fact, especially on these "clicks" the very type of your DA converter will change the perceived sound of the click.

I've done a lot of tests with this during various product developments and sound bank creation.

There's no such thing as a "simple click". Even an extremely short sound can have a lot of different flavors. People who are into any type of glitch music will know this very well.

Cheers!
bManic

it does, try it yourself instantiate an envelope for the amp of a oscillator, then click and drag the "a" of the envelope towards the upper left and you will hear a click

You kind of missed my point. I don't doubt that Absynth can do a "click", most synths can do that. My point was that a simple "click" can be equally fast in the analogue domain, even faster, and it has all to do with the actual shape of the impulse and how the Digital to Analogue converter treats it.

So simply saying "omg! this synth envelope is soooo fast! it can go all the way down to a "click!" doesn't really mean anything" unless you are way more specific. What kind of click? A light one? One with some bottom end to it? A "smooth" one (more like a zap!) or a chunky one (more like a "tock!"). There are so many details to the initial onset of a transient and incredibly small differences in sample position of whatever makes up the click can translate to something completely different once it's in the analogue domain (after DA conversion).

Cheers!
bManic

well hardware analog synths use transistors so its basically just electricity

but a plugin like Absynth can do faster envelopes than any analogue hardware synth and with any waveform i want

[pdxindy]

well hardware analog synths use transistors so its basically just electricity

but a plugin like Absynth can do faster envelopes than any analogue hardware synth and with any waveform i want

but it wont sound like it... I think the relatively low control rate of the soft synth plays a part cause the analogue synth has an essentially unlimited sample rate which is increasingly significant for such short bursts of sound...

[Ichad.c]

but it wont sound like it... I think the relatively low control rate of the soft synth plays a part cause the analogue synth has an essentially unlimited sample rate which is increasingly significant for such short bursts of sound...

Isn't that a 90's issue? Even SynthEdit does it's envelopes at audio-rate.
On another point - the rate of the attack - if it's very fast - is equal to an oscillator in a sense, and if you think about it as an osc - even seemingly small differences in time and shape - could have a huge impact. Also - if you have - drawable envelopes - a small 50ms< hold or clip just after the attack - will increase the punchiness.

[stonestreet]

but it wont sound like it... I think the relatively low control rate of the soft synth plays a part cause the analogue synth has an essentially unlimited sample rate which is increasingly significant for such short bursts of sound...

Isn't that a 90's issue? Even SynthEdit does it's envelopes at audio-rate.
On another point - the rate of the attack - if it's very fast - is equal to an oscillator in a sense, and if you think about it as an osc - even seemingly small differences in time and shape - could have a huge impact. Also - if you have - drawable envelopes - a small 50ms< hold or clip just after the attack - will increase the punchiness.

DSI Tempest has a peak hold parameter for this reason. The Tempest has five six stage envelopes DAPDSR.

Have fun.

[SODDI]

Just a few theoretical thoughts on envelopes.

The old school analog-style ADSR envelopes are perhaps not suited for all current time domain modulation purposes.

For instance, an envelope that modulates a filter should handle that time domain modulation differently than one that modulates amplitude - all levels should be mapped to cutoff frequency (0-18000) as compared to 0-1. In addition, it may not be desirable for a filter envelope's starting level to be zero and very desirable to be able to control levels throughout the envelope's time. (LALHDLR)

[Ichad.c]

...it may not be desirable for a filter envelope's starting level to be zero and very desirable to be able to control levels throughout the envelope's time. (LALHDLR)

Well, the first L (which I assume means level) is handeled usually by the cutoff knob. Having a different level - for the attack peak sounds interisting and useful - especially if it can go negative. The third L ? Isn't that a sustain?

[bmanic]

well hardware analog synths use transistors so its basically just electricity

but a plugin like Absynth can do faster envelopes than any analogue hardware synth and with any waveform i want

You are STILL not getting it. Please look into how a digital to analogue converter works. Sigh..

[SODDI]

...it may not be desirable for a filter envelope's starting level to be zero and very desirable to be able to control levels throughout the envelope's time. (LALHDLR)

Well, the first L (which I assume means level) is handeled usually by the cutoff knob. Having a different level - for the attack peak sounds interisting and useful - especially if it can go negative. The third L ? Isn't that a sustain?

I am not sure - I know that many envelopes modulating filters APPEAR to start at a zero level (giving you that old familiar analog MWOWM when programming a slow attack.)

When constructing simple filter set ups in Reaktor, it is usually that the filter cutoff is modulated by the envelope plus any subsequent modulations - so yeah, you're correct. It just doesn't always SOUND like that to my ears. I spend much time trying to subdue the MWOWM. Scaled curved time-domain values would really change this.

As to my theoretical filter envelope, I thought that the 3rd level could be the sustain or final release level. Maybe it would be desirable for the release endpoint level NOT to be zero.

And this would all be much easier if I could draw it.

SynCModular has these n-level ramp generators that can be used as envelopes. Maybe if I played with it there - (Reaktor only has a 6-level.

Thank you for talking to me about this - this kind of stuff just rattles around in my head and I have no one to talk to. My wife just nods and smiles.