MASTERING ENVELOPES, Part 1
 Envelopes   Quick Reference   Tutorial 1  Tutorial 2   Timings

A Better Envelope

A Walkthrough of Time/Level Pairs versus old-school ADSRs

If you have no idea what envelopes are in any way, then this article is not for you. Go hereOr here if you need more info.

If you can already tweak ADSR sliders to get the sound you want, then you’ve come to the right place. This article introduces a better concept in envelopes, which the ESQ/SQ80 just happens to use.  Best of all, you’ll be able to wrap your brain around this concept in just a few short minutes.

Tricky Envelopes

I’ll freely admit it.  One of the most confusing and intimidating screens on the ESQ/SQ80 is the envelope page.  We’re all familiar with something like this (see photo to the right):

It is the famous ADSR envelope with simple slider controls.  You knob freaks may want to see round potentiometers, but it ADSR really makes more sense as linear sliders.  You get a nice visual representation of the current state of your envelopes.

Unfortunately, many of you who were thrilled with the ease, speed and logic of editing the rest of the ESQ/SQ80, turned to the ENVELOPE pages and were gobsmacked by the following image:

What are all those numbers?  And how do they relate to the little graph on right side?

Here's one way of looking at it: Those number represent definitive proof of one thing:

ADSR is fake!

You heard me.  I have not come to praise ADSR, but to bury it.

Remember those College entry board problems (SAT, ACT) where they trick you into using different units in ways that don’t mix ("A train leaves Chicago at 9:40 furlongs per fortnight and  travels fifteen hectares at the rate of 75 liters per metric hour.  What is the color of the polar bear?")?  ADSR is the same kind of rip-off.

Here's the proof.  Let’s step through ADSR one letter at a time:

 

 

A=Attack:
I understand this. I want my attack to be fast, sort-of-fast (half-fast?) or slow.

D=Decay:
So far so good.  I can set a similar delay to the attack, or something completely opposite.  Because they are both "slopes", at least they relate to each other.  I can set a fast attack-slope, a slow delay-slope, or any combination of the two. I trust the synth maker to make sure that if I set the "A" and "D" sliders to similar positions, they will take roughly the same amount of time to attack and decay.

Now jump ahead to...

R=Release:
I get this one, too. I can have a fast or slow release. If you think about it, "release" is like a second decay. Grizzled old-time synth curmudgeons are now yelling that the simplest envelope is nothing but "Attack" and "Release", with "Decay" being a newfangled fad among us whippersnappers (Now get off my lawn!).

Now jump back in "ADSR" one letter, and that brings us to...

S=Sustain:
Sustain is defined as “the constant volume at which the sound remains following the decay, but before the note is released.”

Here is where things go off the rails. Sustain has nothing to do with the speed (or slope) of the attack, the speed of the decay or the speed of the release! It is a volume level!

How can you have a "fast sustain"? Or a slow one? While sustain is certainly a property of acoustic instruments (and other real-world sounds), it is the odd-man-out of the envelope world.

 

One of these things is not like the other

Sustain specifies a volume level, not a time period.  But we can’t simply ignore or get rid of it.  Sustain is most certainly a vital part of the way audio envelopes work.  Imagine a bland organ sound, but with a nice Leslie effect. The Leslie magic is heard during the sustain portion of the sound. And where would rock guitar be without the sustain pedal? Cute Puppies and Synthesizers?  What is the world coming to?

I'm certainly not calling for the “repeal” of the sustain parameter. After all, Sustain is the only ADSR parameter that comes close to allowing you to set a specific volume level. Unfortunately, that volume control is achieved in a very roundabout and imprecise way.

Well, I want more control.  I want to specify the volume level of my sustain. I want to set exactly how loud the sustain level is (in some sort of volume units).  Not just a general idea, mind you, like the approximation of the ADSR sliders.

I also want to set exactly how much time my envelope takes to get to the sustain level.  I don’t want to set a general “slope”.  I want the sustain to kick in in exactly 609 milliseconds.  Now how do I do that with ADSR sliders?

By saying all of this, what I’m really asking for is the ability to set precise volumes and times for each stage of the ADSR. The simple idea of setting volumes and times for different envelope stages opens up a new world of possibilities!  And the ESQ/SQ80 gives you this ability!

Unlike ADSR, the attack of any natural sound isn’t a simple fade-in slope. It’s more accurate to say that attack is “how much time it initially takes ('Time 1', or 'T1'), to get to the first volume level (Level 1, or 'L1').”

Similarly, for the decay stage (which we could call 'Time 2', or 'T2'), we could do the same thing, specifying exactly when Time 2 will drop down to "volume level 2" (L2).

The beauty of this technique is that we can say exactly how long times 1 and 2 are (in milliseconds), and exactly how loud levels 1 and 2 are.  Instead of just guesstimating the whole thing with a single "Attack" or "Decay" slider!

You are just lucky I didnt put up a picture fro Release"I see where you're going with this," you say, "but I've found a huge flaw in your time-and-level scheme. I've already set the volume LEVEL (L2) for "Time 2" (which you say is the decay stage). In this scheme, how do I set the sustain volume level (the 'S' slider of the old ADSR)?

That's the beauty of this technique. YOU DON'T HAVE TO!!! You've already set the sustain volume level (L2)!!! No further intervention is required. It's a huge two-fer.  You get an amazing degree of time-and-level control that no simple ADSR can match.

These four parameters (T1, L1, T2 and L2) have managed to do--with precision--what three old parameters (ADS...) used to do with rough approximation.

By learning to use Time/Level envelopes, you can do anything ADSR can do, but with unprecedented control, just by mastering one more parameter (T1/L1 T2/L2 vs. ADS...).

Exactly How Long/Exactly How Loud?

Now I'm not saying this is easier than ADSR sliders.  Indeed it forces you to think about how loud and how long you want to set each envelop phase. 

But for a slight bump in complexity, your sound design jumps from abstraction to precision.  (See the following diagram comparing ADSR with T/L pairs).

Now if we want to be control freaks, let's agree that two Time/Level pairs are not enough (you may have inferred this from the diagram). I'm going to replace the old fashioned "Release" slider you normally use with shiny new "Time 3" (T3) and volume level 3 (L3) parameters. Now how much would you pay?  More, you say?  Ok, I'll throw in T4 as well!

Your mind is already racing ahead to different ways you can use the time/level pairs.  Now you can do fancy, multi-stage magic that plain vanilla ADSR envelopes never could! 

But, hey!  Where is the L4 parameter to match the T4 I just gave you?  The truth is, you don't need it.  The final level is always assumed to be zero. That's the case not just for any plain volume envelope, but also any fancy automation you might perform on a filter, or the pitch of an oscillator.  It all comes back to zero.

Sidebar:

What if you want a sound to continue after a keystroke?  There are ways, young Jedi.  Stand by for further info on Envelope Magic, to be posted shortly...

 

Now lets look at the ESQ/SQ80 envelope screen. Most of this should now look pretty familiar:

The first two envelope parameters of the ESQ and SQ80 are marked "L1" and "T1". 

And you'll see all the other Time/Level pairs up to T4.  Just remember that T4's invisible partner, the zero-value L4 is the only one that is inferred and you don't have to set it.

And a quick glance over at that funny diagram on the left shows all these Times and Levels (T's and L's) on an ADSR-like diagram. And that suits us to a "T" (sorry).

I’m going to say it again: Just by mastering the “Time/Level Pair” concept, you can do everything an ADSR envelope can, plus a whole lot more. 

Go Forth, Young Graduate!

Congratulations! You now know the most important concept toward mastering envelopes--not just on the ESQ/SQ80, but anywhere, because you understand the difference between real times/levels and ADSR simplifications.  Consider yourself a Bachelor of Envelope Science from the School of  ESQ/SQ80 Studies.

But there may be a few advanced research topics you might consider:

For example, I'll bet there are a couple things on the diagram or the “Green Screen” envelope page that trouble you.

For starters:

  • What’s all this LV, TK, and T1V stuff?
  • On the SQ80 front panel, why is the diagram "double printed" with 2 envelopes over the top of each other?  (How can you have two envelopes at the same time?)
  • What about the place where the graph dips below the zero axis?  How can you have a "negative sound"?
  • And sometimes after I enter a value, the letter “R”, “L” or “X” appears. That has to be a bug, right?

The fact that you are aware of these questions qualifies you for Part II of this series. You are ready to earn your Masters Degree in Envelope Science, but don’t worry. Compared to the Time/Level lesson you just learned, the next part is easy.

Coming soon, Part II of “A Better Envelope”, including info about the effects of keyboard tracking on envelopes, negative envelope values (how and where to use them), and funky, crazy shapes that you could never get with ADSR envelopes.