Subtractive Synthesis Explained: Oscillators, Filters, and Envelopes

Subtractive synthesis is where most people begin their journey into sound design — and for good reason. It maps naturally onto how we hear and describe sound, its controls are intuitive, and its results are immediately musical. The Minimoog, the Sequential Circuits Prophet-5: virtually every synthesizer that defined the sound of popular music from the 1970s through the 1990s was subtractive at heart. Modern soft synths like Massive, Sylenth1, and U-He Repro continue the tradition.
This is part of our Synthesizers Explained guide.
The Core Concept: Start Rich, Carve Away
The name says it all. Subtractive synthesis starts with a harmonically rich sound — a waveform that contains many overtones — and subtracts frequencies using a filter until the desired timbre is achieved. It’s the sculptor’s approach: remove what doesn’t belong.
This is the opposite of additive synthesis, which builds sounds by layering simple sine waves. Subtractive synthesis has a practical advantage: filters are intuitive to control and respond musically to modulation, making it easy to create patches that feel expressive and alive.
The Signal Path
A classic subtractive synthesizer routes audio through three core stages, typically labelled VCO → VCF → VCA in analog hardware (where VCO = Voltage Controlled Oscillator, VCF = Voltage Controlled Filter, VCA = Voltage Controlled Amplifier). In software, the same stages exist but the voltage-control terminology is often dropped.
VCO — The Oscillator
The oscillator generates a continuous waveform. The classic shapes each have a distinct harmonic character:
- Sawtooth wave — contains all harmonics (both even and odd) at decreasing amplitudes. The brightest and most harmonically dense. The foundation of string synths, bass patches, and leads. Classic Moog and Prophet sound.
- Square wave — contains only odd harmonics. Hollow and woody — think woodwind and clarinet-like patches. Also produces the classic 8-bit game sound at high frequencies.
- Pulse wave — a variable-width square wave. Narrower pulse widths produce thinner, more nasal tones. Pulse Width Modulation (PWM) — animating the pulse width with an LFO — creates the characteristic chorusing shimmer of the Roland Juno and Oberheim synthesizers.
- Triangle wave — contains only odd harmonics, but they fall off faster than a square wave. Soft and flute-like. Often used for sub-bass oscillators.
- Sine wave — a pure tone with no harmonics. Not much for the filter to work on, but essential as a sub-bass layer or for FM synthesis.
- Noise — random signal containing all frequencies simultaneously. Used for percussion, wind effects, breath, and textural layers. White noise emphasizes high frequencies equally; pink noise rolls off with frequency.
Most subtractive synthesizers provide two or more oscillators that can be detuned against each other, producing a fat, chorusing sound from the beating between slightly different frequencies. This is the fundamental technique behind the “big synth” sound.
VCF — The Filter
The filter is the heart of subtractive synthesis — arguably the most characterful and musically important single component. It determines the timbre of the sound far more than the oscillator shape in most patches.
The most common type is the low-pass filter (LPF), which passes low frequencies and cuts everything above the cutoff frequency. Moving the cutoff down makes the sound progressively darker; moving it up makes it brighter. The rate at which the filter attenuates frequencies above the cutoff is described by its slope — measured in dB per octave. A 24dB/octave (4-pole) filter, as in the classic Moog ladder filter, cuts sharply and aggressively. A 12dB/octave (2-pole) filter cuts more gently with a different character.
Resonance (also called Q or emphasis) boosts frequencies near the cutoff point, creating a peak that emphasizes the filter’s position. At moderate resonance levels this adds presence and colour; at high levels the filter begins to ring or self-oscillate — producing a pure sine wave at the cutoff frequency even with no oscillator input. Self-oscillation is a characteristic sound of the Moog filter and is deliberately exploited in many bass and lead patches.
Other filter types include high-pass (cuts low frequencies), band-pass (passes a band of frequencies, cuts above and below), and notch/band-reject (cuts a specific frequency band). Many modern synthesizers offer all types with variable slope.
VCA — The Amplifier
The VCA controls the output volume of the synthesizer over time. On its own, with no modulation, it’s simply a volume control. Its musical value comes entirely from what is applied to it — specifically, an amplitude envelope.
Envelopes: Shaping Sound Over Time
An envelope generator produces a control signal that changes over time in response to a note being played. The standard shape is ADSR:
- Attack — how long it takes the signal to reach its maximum level after a note is triggered. Short attack = immediate onset (piano, plucked string). Long attack = slow fade-in (pad, swelling strings).
- Decay — how long it takes to fall from the peak to the sustain level.
- Sustain — the level at which the signal remains while the note is held (not a time value — a level).
- Release — how long the sound continues after the note is released.
A separate envelope is typically applied to the filter (VCF envelope), allowing the filter cutoff to sweep over time independently from the amplitude. This filter envelope sweep — cutoff opening up then falling back — is perhaps the most iconic sound in synthesis: the classic subtractive “wah” or “vowel” movement heard on countless bass lines and leads.
LFOs: Continuous Modulation
An LFO (Low Frequency Oscillator) is an oscillator running below the audible range — typically 0.1 to 20 Hz — whose output is used to modulate other parameters rather than produce audio. Routing an LFO to oscillator pitch produces vibrato; to filter cutoff produces a wah effect; to the VCA amplitude produces tremolo; to oscillator pulse width produces the PWM shimmer described above.
LFO rate and depth control how fast and how extreme the modulation is. Syncing the LFO rate to the host tempo creates rhythmic, beat-locked modulation — a filter slowly opening and closing in time with the music, or a tremolo pulsing on the quarter note.
Classic Subtractive Synthesizers
- Minimoog Model D (1970) — the first commercially successful portable synthesizer. Three oscillators, the legendary Moog ladder filter, and a simple, direct architecture that made synthesis accessible. Its bass and lead sounds remain the benchmark. The modern Moog Minimoog Voyager, Subsequent 37, and numerous software recreations (Arturia Mini V, Softube Model 72) carry this lineage.
- Sequential Circuits Prophet-5 (1978) — the first fully programmable polyphonic synthesizer. Five voices, two oscillators each, the Curtis CEM filter chip. The Prophet-5 defined the polyphonic synth pad sound of the late 1970s and early 1980s. Its lush, warm character comes from the interaction of its oscillators and a filter notably different in character from the Moog.
- Roland Juno-106 (1984) — a DCO-based (digitally controlled oscillator) six-voice polysynth famous for its built-in chorus effect, which is inseparable from the instrument’s character. Simpler than the Prophet-5, it was enormously popular and its chorus shimmer defines much of 1980s pop. The Roland JU-06A is a modern recreation.
- Roland TB-303 Bass Line (1981) — a failed bass accompaniment device repurposed by acid house producers in the mid-1980s. Its distinctively squelchy, resonant filter sweep (the “acid” sound) became one of the most recognisable sounds in electronic music history, despite the instrument being widely considered a commercial failure at launch.
- Oberheim OB-X / OBXa (1979–1981) — fat, creamy polyphonic synthesis with a warm, characterful sound distinct from either Moog or Sequential. The OB-X sound is synonymous with 1980s arena rock and pop.
Subtractive Synthesis in Modern Production
Modern subtractive soft synths implement the same architecture with additional flexibility: more oscillator types, multiple filter types and slopes, extensive modulation matrices, and built-in effects. U-He Repro-5 and Repro-1 model the Prophet-5 and TB-303 respectively with component-level accuracy. Arturia’s V Collection models dozens of classic instruments. U-He Diva takes virtual analog to its highest expression — multiple interchangeable oscillator and filter circuit models, each with distinct character.
DAW Spotlight: Retrologue 2 in Cubase
Cubase’s included Retrologue 2 is a genuine virtual analog subtractive synthesizer with two oscillators (eight waveforms each), a noise generator, a sub-oscillator, a multi-mode filter (LP 24/12dB, HP 24/12dB, band-pass, notch), two ADSR envelopes (amplitude and filter), two LFOs, and an arpeggiator. Its step modulator adds a third, pattern-based modulation source. For learning subtractive synthesis, it covers every fundamental concept and sounds good doing it — a complete education tool without loading any third-party plugins.
Quick Tips to Carry Forward
- Start with a sawtooth wave — it’s the most harmonically rich and gives the filter the most material to work with
- The filter cutoff and resonance are your primary tone-shaping controls — spend more time here than at the oscillator
- The filter envelope (not just the amplitude envelope) is what makes a sound move and breathe
- Detune two oscillators slightly against each other for instant width and fatness
- PWM on a pulse wave gives you a chorusing, shimmering quality without any effects — learn this early
- LFOs are for continuous animation; envelopes are for one-shot shapes — use both
Subtractive synthesis is where synthesis begins for most people — and for many professionals, it’s where they spend most of their time regardless of what other tools are available. Master its fundamentals and every other synthesis type will be easier to understand.
Related: Additive Synthesis Explained • Wavetable Synthesis Explained • Synthesizers Explained (complete guide)
