The role of a synthesizer's envelopes, gates, and triggers in its process

Envelopes are just ADSRs, right? Attack determines how percussive the sound is, and Decay determines how long it takes for the sound to decay to the sustain level after the initial attack phase. Finally, Release controls how long it takes for the sound to decay to zero when you release the key. Pretty simple, right? Well, yes, but that's too superficial. In fact, there are many different types of envelopes besides ADSR, some more complex than ADSR, but others simpler than ADSR. In addition, many envelopes themselves are controlled by parameters such as Amount and Invert. We'll leave the intricacies of these envelopes for another article, but today let's take a look at how common envelopes in the synthesizer you use can have unexpected effects on the sounds you make and the way you play.

Think of a common waveform, such as a sine wave or a square wave. These sound waves have simple, symmetrical shapes that repeat over time. In fact, if the shape didn't repeat, you wouldn't be able to perceive the pitch of the sound; at best you'd hear a click. Common sawtooth and pulse waves (which aren't strictly symmetrical waveforms, but have some symmetry) also repeat in this way. With this property of common waveforms in mind, look at Figure 1 again. Figure 1 shows a simple waveform that is neither symmetrical nor repetitive.

Figure 1: An irregular, non-repeating waveform

Remember what happens if you apply this waveform to another component of a synthesizer, such as a voltage-controlled amplifier (see Figure 2)?

Figure 2: Applying the waveform in Figure 1 to the VCA

If the gain value of the amplifier at any time is proportional to the wave in Figure 1, then Figure 1 will become a curve describing the loudness contour of the timbre. If the envelope in Figure 1 is applied to a low-pass filter (VCF), then the curve in Figure 1 will become the brightness curve of the timbre. Obviously, the curve in Figure 1 can be used as the output of the "envelope generator". Well, I have already introduced the above content in the third article, but this knowledge is worth repeating here again.

Figure 3: Using multiple CVs to create a composite VCA envelope

A good definition of an envelope is a graphical representation of the variation of a particular parameter over time. But in a typical synthesizer, you can use any number of modulators to vary a parameter at any time (see Figure 3). Therefore, the actual envelope of the sound may be the sum of the contributions of multiple components of the synthesizer. What we usually call an "envelope generator" is likely to be just one of many factors that determine the envelope of a parameter.

Perhaps for this reason, some early synthesizer manufacturers called what we now call an "Envelope Generator" a "Transient Generator." Since "transient generator" is more accurate, we'll use that term from now on.

Let's go back and study Figure 1. The first stage of its envelope is the attack stage of the transient, and the second stage is the decay stage. Therefore, Figure 1 shows the envelope of an "AD" transient generator. Although AD looks simple, don't underestimate their power. Figure 4 shows the effect when two different AD envelopes are applied to a device such as a VCA or VCF at the same time. You can see that the superposition of two AD envelopes can produce a complex four-stage envelope. And you can't get the same result with the common ADSR four-stage envelope.

Figure 4: Results of applying two simple AD transients to the same device

Figure 4 shows the result of applying multiple transient envelopes to a single parameter at any time. But this article is about more than that, we're also going to look at what happens when you play multiple notes at the same time.

As we all know, most analog monophonic synthesizers are controlled by a keyboard. But what you may not know is that in most synthesizers, three different signals are generated every time you press a key. The first signal is the Pitch CV, which helps determine the pitch of the sound produced by the synthesizer. The second signal is the Trigger. The Trigger is a short pulse that is generated the moment you press the key, and it triggers the synthesizer's transient generator and other components to start working. The third signal is the Gate. Similar to the Trigger, the Gate tells the rest of the synthesizer that you pressed a key. However, unlike the Trigger, the Gate remains "open" for the entire time you hold down the key. This means that the Gate also has the function of telling the rest of the synthesizer when you release the key. Figure 5 shows these three signals. Understanding the Gate and Trigger will extend our understanding of transients.

Figure 5: Trigger, Gate and Pitch CV Example

While the trigger starts the transient generator, the gate has the ability to tell the synthesizer to continue to expand the envelope contour until we release the key. Without the gate, you would only hear the beep at the beginning of the sound, but not the entire sound.

Figure 6: A trapezoidal envelope

Figure 6 shows a three-stage transient generator that uses the timing information of the gate to complete the attack and then maintains full volume for the duration of the key being pressed. We call this a trapezoidal transient generator. The EMS VCS3 is one of the few synthesizers that has a trapezoidal transient generator.

Figure 7: A typical ADSR transient curve

Slightly more complex than the three-band transient generator is the ADSR. ADSR was standard on most synthesizers in the 70s, so much so that many keyboard players called all transient generators ADSRs, even though some of them did not generate ADSR signals. Again, look at Figures 1, 6, and 7, which all assume that each transient is a separate entity and that each has enough time to run its entire cycle before the next one is triggered. But even for monophonic synthesizers, this is not the case in most cases.

We call the trigger and gate signals timing signals for obvious reasons. Every synth needs a timing signal, but why do we need both a trigger and a gate? Obviously a gate can do the same thing as a trigger, so why is a trigger necessary? This question puzzled me for many years until I was lucky enough to own both a Minimoog and an ARP Odyssey. I usually use these two synths to play well-defined (attack = 0) sounds, with a percussive attack and vivid brightness at the beginning of each note. This effect is achieved by applying envelopes to the VCF and VCA, of course.

Oddly, despite the Minimoog's acclaimed envelope generator, I've always felt that the Odyssey is better for playing fast solos, though I don't know why. I just know that the sound is "punchier" when I play the Odyssey, and I can play faster than the Minimoog. The reason for this has nothing to do with my playing skills (I'm equally bad at both), or the build quality of the keyboards. The real reason is the hardware design of the two synths: the Odyssey uses triggers, while the Minimoog doesn't.

Figure 8: Retriggering behavior on ARP Odyssey. The thick arrows show the trigger information for each key press.

Take a look at Figure 8. This shows what happens when I play a fast melody on the Odyssey. You can see that the notes overlap because I don't release my finger from the previous key in time to play the next note. But you can also see that the series of notes still has its own dynamics, with each note having a clear attack phase. This is because the Odyssey re-triggers its transient generator every time I press a key, regardless of whether I release the previous key or not, and regardless of whether the gate is open or not. This means that my notes have clear outlines, even though I don't release the keys uniformly.

Figure 9: Minimoog envelope without retrigger behavior

Let's look at the Minimoog again. The Minimoog does not have the ability to generate trigger signals. Its only timing signal is an unconventional signal called S-Trigger, which, despite the name, is actually a gate signal. Because of this, the same note played on the Minimoog will produce the filter and amplifier envelopes in Figure 9. As you can see, this waveform is far less interesting than the Odyssey. Although you can hear the second and subsequent notes playing at a sustained volume, the initial impact of the note is completely lost. As a result, fast solos on the Minimoog become sluggish and boring. To achieve the dynamics and impact of the Odyssey on the Minimoog, you must release the previous key before pressing the next key, but I was unable to do this when playing fast solos. For keyboard players of average performance level, the ARP Odyssey is a better choice than the Minimoog when playing fast solo melodies (in addition, the Odyssey keyboard is a dual-voice keyboard, which also helps the above performance, but we will leave the topic of polyphony for another day).

You may be wondering what would happen if the Minimoog and Odyssey used AD envelopes or trapezoidal envelopes. The answer to this question may surprise you. If both synthesizers use AD envelopes, the Odyssey's solo will hardly change much, while the Minimoog's sound will disappear completely after the first one or two notes (see Figures 10 and 11).

Figure 10: Retriggered AD transient

Figure 11: AD transient without trigger

Oddly enough, if both synthesizers use a trapezoidal envelope, there is almost no difference in the sound between the two, provided that the Odyssey's envelope does not decay to zero at the beginning of each note (see Figures 12 and 13).

Figure 12: Trapezoidal envelope of retrigger

Figure 13: Gate-controlled trapezoidal envelope

Wait, what does decay to zero mean? Take a look at Figure 14 below, which shows a series of “decay to zero” trapezoidal transients. Some synthesizers reset their envelope generators in this way, and this behavior can result in a series of disjointed sounds. This is particularly noticeable on synth string sounds, which have slow attacks and releases. Some early polyphonic synths also had this problem, and it sounded terrible, like the synth biting its tongue.