Synthesis Deep Dive: Understanding Subtractive Synthesis

The Language of Synthesis

Subtractive synthesis is the foundation of electronic music production. Nearly every synthesizer you've used — hardware or software — implements some version of it. Understanding how it works doesn't just help you program better patches; it helps you hear what's happening in other people's music and understand how to achieve similar results.

The core idea is simple: start with a harmonically rich waveform, then subtract frequencies using filters until you have the sound you want.

Oscillators: The Sound Source

An oscillator generates a repeating waveform at a specific pitch. The shape of that waveform determines its harmonic content — which overtones are present and at what relative amplitudes.

Sawtooth wave — Contains all harmonics (odd and even) at decreasing amplitudes. This is the richest waveform, giving you the most material to work with. Classic for lead and bass sounds.

Square wave — Contains only odd harmonics. Has a hollow, woody character. A 50% duty cycle square wave is what you hear in classic chip music and clarinets (approximately).

Triangle wave — Like a square, contains only odd harmonics, but they fall off more steeply. Much softer and flute-like. Good starting point for bass sounds when you want less aggression.

Sine wave — No harmonics at all. Just the fundamental frequency. Sounds pure and clean — useful for sub bass and as a modulation source.

Most synthesizers give you several oscillators that you can tune relative to each other. Detuning two sawtooth oscillators slightly against each other creates that classic "supersaw" chorus effect.

Filters: The Heart of Subtractive Synthesis

The filter is what makes subtractive synthesis subtractive. You're removing frequencies from the oscillator's output to shape the timbre.

Low-pass filter (LPF) — Passes frequencies below the cutoff, attenuates frequencies above it. This is the most common filter type. Opening the cutoff brightens the sound; closing it makes it darker and more muted.

High-pass filter (HPF) — The opposite of LPF. Passes high frequencies, attenuates lows. Useful for thinning out bass sounds and removing rumble.

Band-pass filter — Passes a band of frequencies around the cutoff, attenuates both above and below. Nasal, vocal-like quality.

Resonance — Boosts frequencies at the cutoff point. High resonance creates a sharp peak and, in analog-modeled filters, can cause the filter to self-oscillate — generating its own sine wave tone. This is the distinctive "squelch" of TB-303 basslines.

The slope of a filter (measured in dB per octave) describes how steeply it attenuates frequencies past the cutoff. A 24 dB/octave filter (4-pole) has a more aggressive, surgical character than a 12 dB/octave (2-pole) filter.

The Amplifier and Envelope

After the filter shapes the timbre, the amplifier controls the volume over time. The ADSR envelope modulates the amplifier:

• Attack — How long it takes the sound to reach full volume after a note is triggered
• Decay — How long it takes to fall from peak to the sustain level
• Sustain — The volume level held while the key is held
• Release — How long the sound fades after the key is released

Short attack and decay with zero sustain gives you a percussive pluck. Long attack and release creates pads that breathe in and out slowly.

The same ADSR envelope shape is often applied to the filter cutoff frequency — so a sound can start bright and get darker over time, or start dark and bloom open. This filter envelope modulation is what gives classic analog sounds their movement and life.

LFO: Modulation Over Time

A Low-Frequency Oscillator (LFO) generates a waveform, like a regular oscillator, but at sub-audio rates — typically 0.1 to 20 Hz. Instead of producing sound, it's used to modulate parameters over time.

LFO to pitch creates vibrato. LFO to amplitude creates tremolo. LFO to filter cutoff creates a wah-like effect. LFO to oscillator pulse width on a square wave creates the classic "pulse width modulation" sound that moves and breathes.

Putting It Together: Designing a Patch

Start with a goal: a warm bass, a cutting lead, an evolving pad.

1. Choose an oscillator waveform based on harmonic content — saw for rich, square for hollow, sine for clean
2. Set the filter cutoff and resonance to get the basic timbre
3. Shape the amplitude envelope for the articulation you want
4. Apply the filter envelope for timbral movement
5. Add LFO modulation for life and motion

Every great synth patch starts with these basics. Once you understand subtractive synthesis deeply, the more complex techniques — FM synthesis, wavetable synthesis, granular synthesis — all become easier to approach because you have a mental model for how synthesis works.