Now we’re ready to start with subtractive synthesis, which is both one of the simpler approaches to synthesis and one of the most popular. As we discussed in the Types of Synthesis lesson, subtractive synthesis starts with harmonically rich waveforms, then uses filters to carve those waveforms into musically pleasing shapes. Despite the relative simplicity of this approach, even a basic subtractive synth can seem a bit confusing and intimidating the first time you’re looking at it. Luckily, it’s actually quite simple. To master the basics, we only have to learn about five basic components – three of them make or alter audio, two of them control and shape the sound. Let’s get to it!
Audio Components
First up are oscillators. These make the initial sound. Traditionally, oscillators make a number of basic waveshapes: sawtooth and square/pulse waves, at a minimum; sine, triangle and a few other variations are also common options. Oscillators will generally have controls to select the waveshape and to set the frequency/range/tuning of the oscillators, at a minimum. More capable oscillators will offer options like sync, pulse width modulation, and more (we’ll get into those in later lessons).
Next in the audio chain typically comes the filter (altho it’s worth noting that you can put the amp before the filter, and there are occasionally reasons to do so). As we covered before, filters remove frequencies from our signal – it’s the filter that puts the subtraction in subtractive synthesis. For now, we’re going to focus on lowpass filters, which are by far the most commonly used filter type. These filters roll off the upper frequencies of a signal and would typically offer control of its cutoff frequency (the frequency at which it starts affecting the signal) and resonance, which boosts or emphasizes the frequency at the cutoff point. (Not all filters are resonant, but resonant filters are definitely preferable to non resonant filters almost always.)
Finally, the amp determines how loud the signal is – its amplitude, in other words. It’s the simplest part of the signal chain, and typically the most transparent (i.e. not prone to “color” the sound), altho some synthesizer amp stages are designed to purposely add distortion. The relevant control, if any, is simply control of the final level.
That’s it for the basics of the audio path. Now, let’s talk about the control elements.
Control Components
Besides the keyboard/sequencer that you use to trigger notes and determine pitches, there are two basic control signals that are nearly universal in the synthesizer world: envelopes and LFOs. A basic synth might offer only one of each; a more complex synth might offer as many as a half a dozen of each – or even more! Both envelopes and LFOs are control signals – they make no sound on their own. Instead, they act as automated extensions of you – instead of using your hand to twist the control knob that turns the volume of the sound (for example) up and down as it plays, you use an envelope or LFO to automate it.
The envelope is your basic sonic sculpting tool. The “shape” of a sound – whether it’s percussive or not, how quickly it decays away to nothing or if it sustains, etc – is largely determined by the use of one or more envelopes, particularly when controlling the amp. The most common type of envelope is the classic ADSR design – as in Attack, Decay, Sustain, Release.
The attack stage is typically the time the envelope takes to go from zero to max level. Decay sets the fall time from that max level to whatever the sustain level is set to (or to zero, if the sustain is set to zero). Sustain is the odd man out – it’s not a time setting, but a level setting. When you set the sustain, you are setting the level that the envelope will hold at as long as you hold the key. Finally, the release – another time setting, this one determining how long the envelope takes to go from the current level (typically the sustain level, but actually wherever the envelope is in its cycle when the key is released) to zero. Simpler envelopes typically omit the decay stage (an ASR envelope) or combine the decay and release into a single value. More complex envelopes also exist, but that’s something we’ll get into later.
Envelopes can be used to control anything and everything – pitch envelopes can help represent the kind of pitch bends certain instruments have at the onset of their sound, for example. The most common uses are the amp and filter, but envelopes are used literally everywhere.
The final piece we’ll look at in this lesson is the LFO, or Low Frequency Oscillator. Much like the oscillators we use to make our initial sound, these usually produce basic waveshapes such as triangle, saw or square. The difference is that LFOs typically run at subsonic speeds – below 20 Hz, or 20 oscillations per second. And instead of hearing them, we use them to modulate other signals. A 5 Hz LFO sent to pitch makes a pretty convincing vibrato. Send the same LFO to the amp and you have tremolo. Use it to control the filter and it’s an auto wah. And so on, and so on – anywhere you need a periodic control signal, LFOs are what you use. Typical controls for LFOs include selection of the waveshape and its speed. More advanced models might add BPM sync, control of starting phase, a delay or fade in time, and more.
Believe it or not, if you have a decent conceptual understanding of those five components, you have a handle on 85% of subtractive synthesis. And, as barebones as this configuration seems, some of the most famous synths of all time offer exactly these five components with just a couple of additional options (e.g. the Junlo-60 pictured above…). Beyond that, the vast majority of more full featured synths simply add more of the same (more oscillators, multiple envelopes, etc), variations on these parts – different types of filters, for example – or features added to these core elements to extend their capability, like pulse width modulation on the oscillators. We’ll cover all of that in detail in later lessons.
Exercises
For now, it’s time to put that knowledge into practice! Here are five exercises to get you started.
1) On whatever synth you are using, identify all of the parts named above (oscillator(s), filter(s), amp(s), envelope(s), LFO(s)). You may have multiples of some, or all, of these!
2) While holding a key and with the filter wide open and set to minimum resonance settings, switch between the sawtooth and square waveforms. Listen to each and try to imagine what kind of sounds they might be used for, and try to familiarize yourself with each waveform until you can fairly confidently identify each one without looking.
3) Select the saw wave and hold the key down again, but this time open and close the filter (i.e. turn the cutoff knob up and down) to get a sense of what the filter does to the waveform. Adjust the resonance up to a about a third and do the same open/close exercise, noting how different that makes the sound. Repeat at resonance setting of roughly ⅔ and full resonance. Now, switch to the square and do it all again, noting throughout how the tone changes and what kind of settings and combinations of waveform, filter cutoff and resonance make pleasing timbres.
4) Pick a combination of waveform, filter cutoff and resonance that you liked, then use the envelope and/or LFO to vary the sound over time. An envelope on the amp controls the general shape of the sound; the same envelope on the filter will control its brightness and timbre. Try the LFO on each of its available destinations (note: not all synths allow the LFO to go to all components, just do what your synth allows) and note how that affects the sound. When you find something that is somewhat pleasing, save it (if your synth allows saving patches, of course).
5) Finally, using just a single envelope, LFO, oscillator, filter and amp (even if your synth has multiples of all of those things!) try and create a half dozen “obvious” synth patches – the kind of stuff you’ve heard a thousand variations on over the years. Some examples would be a filter sweep (slow sound that focuses on the filter opening and then closing slowly and dramatically), a synth bass (short envelope, tuned down an octave or two) and lead (similar sound, probably more sustain, not tuned down), a “string” sound (moderately filtered saw or square, slow attack and release), and/or a bright pluck (short attack/decay, no sustain, bright n attack but fades quickly). Creating these with just those minimal components, especially if you resist using additional features your synth has even within those options, is an excellent exercise that will illustrate just how much you can accomplish with even a minimal set of synthesis tools – as well as what you might need some fancier options for.
And don’t worry, we’ll get into those fancier options in our next lesson.
Next: Intermediate Subtractive Synthesis – Coming Soon!