Understanding ADSR

jorge · 17 Jun 2005

Hi SyrupPudding, welcome to the forum.

The ADSR will probably eventually have a Gate input as it seems to confuse a lot of new users. I'll try to explain how it works at the moment.

The MidiToPoly works in combination with the PolyToMono components to create a Poly schematic section that gets replictaded for each note you're playing. Whenever you play a new note, the ADSR starts immediatelly.

The ADSR doesnt have to be used for controlling amplitude, you can use it for controlling frequency too. The ADSR doesnt know and doesnt need to know what you're using it for. If you multiply the output of a sine osc by the ADSR you're controlling amplitude. If you multiply the frequency input of the sine osc by the ADSR you're controlling the frequency etc. as you probably know ADSR means Attack, Decay, Sustain and Release not necessarily related with amplitude.

Please feel free to ask as many question as you want, we'll do our best to help.

SyrupPudding · 17 Jun 2005

Hi Jorge, thanks for the reply.

Your response clears things up for me. Now I know the poly section is replicated for each note, I can now see why ADSR does not currently have a trigger input. Although if you added one I'm sure the people on this forum would make creative use of it.

I think I now understand the action of ADSR on the amplitude. The output from osc is simply a stream of wave amplitude values. The frequency is inferred from the rate of change of this data, but is not part of the data itself.

Thank you very much.

Time to wade in deeper (ankle deep, that is!)

RobR · 17 Jun 2005

SyrupPudding wrote: Although if you added one I'm sure the people on this forum would make creative use of it.

Indeed. Re-triggerable ADSR's can give some cool modulation options.

bones · 17 Jun 2005

jorge wrote: The ADSR will probably eventually have a Gate input as it seems to confuse a lot of new users.

Its probably confusing because it seems so limiting. It doesn't allow you to replicate the monosynth behaviour of not retriggering the ADSR when you play legato [or if it does, it doesn't allow you to have it as an option].

infuzion · 17 Jun 2005

bones wrote: jorge wrote: The ADSR will probably eventually have a Gate input as it seems to confuse a lot of new users.

Its probably confusing because it seems so limiting. It doesn't allow you to replicate the monosynth behaviour of not retriggering the ADSR when you play legato [or if it does, it doesn't allow you to have it as an option].

Quite true, since the ADSR right now is hard-coded to NOTE-ON.

On the other hand, since the output is [0,1] (zero to one Float), you can really put the ADSR anywhere, & make it do all sorts of fun stuff. You can enen invert the ouput, since it's just numbers, not "amplitute".

bones · 18 Jun 2005

Invertible modulation envelopes are the key to good pads.

platinumears · 18 Jun 2005

bones wrote: Invertible modulation envelopes are the key to good pads.

You can do that already.. 😉

I would also like them to be freely triggerable though.. my SE dynamics processors use an envelope to ramp down the gain modulation when the input level drops below a certain threshold, otherwise upwards compression or downwards expansion modes can end up with extreme values in track pauses..

I had a go at making some simple mono AD and AR envelopes.. here.

I'm sure they could be improved though..

Dan 🙂

Sk · 18 Jun 2005

bones wrote: Invertible modulation envelopes are the key to good pads.

Explain it better. I want to know more.

platinumears · 18 Jun 2005

Sk wrote: bones wrote: Invertible modulation envelopes are the key to good pads.

Explain it better. I want to know more.

say the envelope is modulating filter cutoff: an inverted enelope will start high, get gradually lower through the attack stage, then open up again through the decay stage until it reaches the sustain level. The release stage will then decay up instead of down. It is simply a way to make a sound evolve over time in an interesting way.. but its not the only way to program a pad!

Dan 🙂

Sk · 18 Jun 2005

platinumears wrote: Sk wrote: bones wrote: Invertible modulation envelopes are the key to good pads.

Explain it better. I want to know more.

say the envelope is modulating filter cutoff: an inverted enelope will start high, get gradually lower through the attack stage, then open up again through the decay stage until it reaches the sustain level. The release stage will then decay up instead of down. It is simply a way to make a sound evolve over time in an interesting way.. but its not the only way to program a pad!

Dan 🙂

Thanks!

bones · 19 Jun 2005

Sk wrote: bones wrote: Invertible modulation envelopes are the key to good pads.

Explain it better. I want to know more.

Wot 'e said. You invert the modulation envelope, modulating filter cutoff or something, and the reason it is so effective is that it tends to work against the VCA envelope which gives you many more opportunities for the sound to swirl in, out and around. OTOH, when you are programming bass sounds you want the envelopes to work together to emphasise the attack as much as possible.

platinumears · 19 Jun 2005

bones wrote: OTOH, when you are programming bass sounds you want the envelopes to work together to emphasise the attack as much as possible.

Not always: an inverted ADSR can provide a fast downward filter sweep, which then opens up again if the note is held.. often good for drum'n'bass type bass sounds.

bones · 19 Jun 2005

I got 6-point envelopes for that, although you could do it that way with ADSR. But the point was that for basslines you generally want the envelopes working together, for pads its often very cool to have them working against one-another [just in case you missed it like you missed the point about invertible envelopes already being possible and felt the need to repeat it two posts later].

platinumears · 19 Jun 2005

<sigh> bones, I'm doing my best to stay civil on this forum, out of respect for SM's creators.. could you not try to do the same?

unkargherth · 19 Jun 2005

platinumears wrote: I had a go at making some simple mono AD and AR envelopes.. here.

I'm sure they could be improved though..

Dan 🙂

Mmmmm. Shouldn't envelopes be Poly by definition? I think so, as they should affect every individual note...

SyrupPudding · 19 Jun 2005

An earlier post talked about legato, which I think means (for ADSR purposes) consecutive notes 'tapping in' to a currently triggered envelope.

One of my first lessons from jorge was that the Poly section is replicated for each note. To me this implies that one note knows nothing about the existence of any other note. So I was wondering last night (hot and humid, difficult to sleep 🙂 ) how legato could be implemented under the current architecture. Does SM need a new type of connector (e.g. shared Poly)?.

I'd be interested to hear anyone's thoughts on this. For me, thinking about how to achieve something is often more rewarding than actually doing it 😀

platinumears · 19 Jun 2005

unkargherth wrote: platinumears wrote: I had a go at making some simple mono AD and AR envelopes.. here.

I'm sure they could be improved though..

Dan 🙂

Mmmmm. Shouldn't envelopes be Poly by definition? I think so, as they should affect every individual note...

I was trying to make a monosynth.. 😉 You could make them poly by changing the monoin and monouts to polyin and polyout.. but you can make a regular ADSR act like AD or AR envelopes anyway.

infuzion · 19 Jun 2005

platinumears wrote: I had a go at making some simple mono AD and AR envelopes.. here.

Interesting... what are the Float SR inputs for please?

platinumears wrote: <sigh> bones, I'm doing my best to stay civil on this forum, out of respect for SM's creators..

I don't think bones was trying to be rude there; just stating facts.

Shifrin · 19 Jun 2005

infuzion wrote: Interesting... what are the Float SR inputs for please?

I've not looked at it myself but I'd have a guess that SR would refer to 'sample-rate' i.e. 44100/88200/48000/96000 etc...

Shif.

bones · 20 Jun 2005

SyrupPudding wrote: An earlier post talked about legato, which I think means (for ADSR purposes) consecutive notes 'tapping in' to a currently triggered envelope.

One of my first lessons from jorge was that the Poly section is replicated for each note. To me this implies that one note knows nothing about the existence of any other note. So I was wondering last night (hot and humid, difficult to sleep 🙂 ) how legato could be implemented under the current architecture. Does SM need a new type of connector (e.g. shared Poly)?.

I believe this will be addressed by the "voice management" tools that ZMalc and Jorge have hinted at. It is certainly an area that is murky at the moment and needs to be looked at.

platinumears · 20 Jun 2005

Shifrin wrote: infuzion wrote: Interesting... what are the Float SR inputs for please?

I've not looked at it myself but I'd have a guess that SR would refer to 'sample-rate' i.e. 44100/88200/48000/96000 etc...

Shif.

Good guess! 🙂

infuzion · 20 Jun 2005

platinumears wrote: Shifrin wrote: infuzion wrote: Interesting... what are the Float SR inputs for please?

I've not looked at it myself but I'd have a guess that SR would refer to 'sample-rate' i.e. 44100/88200/48000/96000 etc... Shif.

Good guess! 🙂

ok, cheers 😀