Sound Design Basics: Synthesis Types, Techniques & How to Start

The Universal Building Blocks of Synthesis

Before diving into individual synthesis types, every synthesizer shares a set of core concepts. Master these and the transition between different synths and synthesis methods becomes straightforward.

Oscillators — The Sound Source

Oscillators generate the raw waveform. The four basic waveforms are: sine (pure tone, no harmonics), triangle (soft, flute-like, odd harmonics only), square (hollow, woody, odd harmonics), and sawtooth (bright, buzzy, all harmonics). The harmonic content of each waveform determines its timbre before any filtering happens. Most synths allow multiple oscillators running simultaneously, tuned in unison, octaves, or fifths.

Filters — Tone Shaping

Filters remove or boost specific frequency ranges. The low-pass filter (LPF) is the most common — it lets low frequencies through and cuts highs above the cutoff frequency. Resonance boosts frequencies just at the cutoff point, creating a characteristic peak. At extreme resonance settings the filter self-oscillates, producing a pure tone of its own. Filters are the primary tonal sculpting tool in subtractive synthesis.

Envelopes — Shape Over Time (ADSR)

Every sound needs a shape in time. The ADSR envelope defines this: Attack (how long to reach full level), Decay (how long to fall to the sustain level), Sustain (the held level while the key is pressed), Release (how long to fade after the key is released). Envelopes can control amplitude, filter cutoff, pitch, or any other parameter. The same ADSR principle applies across every synthesis type.

LFOs — Modulation and Movement

An LFO (Low Frequency Oscillator) oscillates below the audible range — typically 0.1Hz to 20Hz. Rather than producing audio, it modulates parameters over time. LFO on pitch = vibrato. LFO on amplitude = tremolo. LFO on filter cutoff = a rhythmic filter sweep. LFOs can be synced to the project tempo for rhythmic effects, or left free-running for organic, flowing movement.

Subtractive Synthesis — Start Here

Subtractive synthesis is the foundation of modern synthesizer design and the best place to begin learning sound design. The principle is simple: start with a harmonically rich oscillator (sawtooth or square), then use a filter to subtract harmonics and sculpt the tone.

Classic hardware synths built on this architecture — the Moog Minimoog, Roland Juno series, Sequential Circuits Prophet-5 — define the sound of recorded music from the 1970s to today. Virtually every modern software synth includes a subtractive synthesis mode.

How Subtractive Works in Practice

A typical subtractive patch starts with one or two sawtooth oscillators. One is detuned slightly above pitch and the other slightly below — this creates the familiar "fat" unison tone. The combined oscillators pass through a low-pass filter. Lowering the cutoff rolls off the high harmonics, making the sound warmer and darker. Adding resonance at the cutoff creates an emphasis that sounds nasal or aggressive. The filter cutoff is then modulated by an envelope — opening up on attack (bright) and closing on release (darker, fading) — giving the sound its characteristic feel.

Best Synths for Subtractive Synthesis

Native Instruments Massive is the reference for subtractive bass and lead design. Xfer Records Serum 2 covers subtractive alongside wavetable. For free options, Vital handles subtractive synthesis excellently and has a clear visual interface. Hardware: any vintage Moog clone or Roland SH-101 reissue.

FM Synthesis — Complex Timbres From Simple Operators

FM (Frequency Modulation) synthesis was developed by John Chowning at Stanford in the 1960s and commercialised by Yamaha in the DX7 (1983) — one of the best-selling synthesizers in history. The electric piano, bass, marimba, and bell sounds of the 1980s are almost all FM. Modern FM synthesis is experiencing a renaissance in plugins like Native Instruments FM8, Arturia DX7 V, and the Dexed freeware clone.

How FM Works

FM synthesis uses one oscillator — called the modulator — to modulate the frequency of another oscillator — the carrier. When the modulator's frequency is within the audible range and at high enough amplitude, it creates sidebands — additional frequency components around the carrier frequency. These sidebands are what give FM its distinctive harmonic complexity.

The key insight: unlike subtractive synthesis (which removes harmonics from a rich source), FM synthesis generates harmonics through modulation. The relationship between modulator and carrier frequency determines whether the resulting sidebands are harmonic (musically related to the carrier) or inharmonic (dissonant, metallic, bell-like).

The Operator System

In FM synthesis, oscillators are called operators. The DX7 had six operators. Each operator can act as a carrier (producing audible sound), a modulator (modulating another operator's frequency), or both. The routing of operators — which modulates which — is called an algorithm. The DX7 had 32 algorithms. Different algorithms produce fundamentally different timbral possibilities.

Simple algorithm: one modulator feeding one carrier. Result: a tone with sidebands above and below the carrier frequency. Complex algorithm: a chain of modulators feeding a carrier, or multiple carriers running in parallel. Result: extremely dense harmonic structures capable of anything from realistic piano to sci-fi sound effects.

The FM Ratio — Harmonic vs Inharmonic

The ratio between modulator frequency and carrier frequency is the most important parameter in FM synthesis. When this ratio is a whole number (1:1, 2:1, 3:1), the sidebands land on harmonic frequencies — the sound is tonally rich but still musical. When the ratio is fractional (1.41:1, 2.73:1), the sidebands are inharmonic — the sound becomes metallic, glassy, bell-like, or clangorous. This is how FM synthesis achieves the bell and electric piano timbres that are impossible to replicate with subtractive methods.

FM in Practice

The challenge of FM synthesis is that the interaction between ratio, modulation depth, and algorithm is non-linear — small parameter changes can produce dramatic timbral shifts. The approach is: set a ratio first (whole number for harmonic, fractional for metallic), then increase modulation depth slowly from zero while listening. At low depth, the sound is close to a sine wave. As depth increases, harmonics multiply. At extreme depth, the sound becomes noise-like or breaks into complex textures. The envelope on the modulator's amplitude is critical — it controls how the harmonic complexity changes over time.

Best Tools for FM Synthesis

Native Instruments FM8 remains the most accessible FM synthesizer with a visual algorithm editor. Dexed is free and models the DX7 exactly — excellent for learning classic FM sounds. Arturia DX7 V adds modern features to the classic architecture. In Ableton Live, Operator is a four-operator FM synth built directly into the DAW.

Wavetable Synthesis — Evolution and Movement

Wavetable synthesis was pioneered by Wolfgang Palm (PPG Wave, 1981) and refined into its modern form by Dave Hilowitz and the team behind Xfer Records Serum. It is now the dominant synthesis type in electronic music production — Serum's wavetable engine defines the sound of modern EDM, bass music, and commercial pop.

How Wavetable Works

A wavetable is a collection of single-cycle waveforms — individual snapshots of a complete waveform cycle. Each snapshot represents a different timbre. By sweeping the playback position through the wavetable (either statically or over time), the timbre of the sound morphs continuously. At position 0, you hear one waveform. At position 64, the timbre has transformed entirely. At position 127, it is a completely different sound.

The power of wavetable synthesis is in this morphing — an evolving, animated quality that no static oscillator can achieve. A wavetable pad can shift from a soft sine-like tone at the start of a note to a harmonically dense, buzzing texture as the note sustains, creating a living, breathing sound in a way that subtractive synthesis cannot match.

Wavetable Position Modulation

The wavetable position can be modulated by anything: an envelope, an LFO, velocity, MIDI CC, macro knobs — this modulation is where wavetable synthesis becomes expressive. Common approaches:

• Envelope on position: sweeps from one timbre at the attack through to a different timbre as the note decays. Classic evolving pad technique.
• LFO on position: creates a rhythmic or organic timbral wobble — the signature movement in many modern bass patches.
• Velocity on position: maps playing velocity to wavetable position — soft playing stays in the warm part of the table; hard playing drives into the bright, harmonically complex part. Extremely expressive for melodic leads.

Serum's Wavetable Editor

Serum 2 includes a built-in wavetable editor that allows importing any audio file and converting it to a wavetable, drawing custom waveforms frame-by-frame, applying spectral processing to individual frames, and morphing between frames. This means any sound — a human voice, a guitar chord, a field recording — can become a wavetable and be played as a synthesizer patch. This is how the distinctive "vocal chop" leads and "talking bass" sounds in modern bass music are created.

Unison and Detuning in Wavetable Synthesis

Wavetable synthesizers allow multiple instances of the same oscillator running simultaneously — unison — with slight pitch offsets between them. In Serum, stacking eight voices in unison with subtle detuning and panning creates the enormous, wide supersaw sounds that dominate commercial EDM. The width and density of the unison stack is one of the defining characteristics of modern synthesizer sound design.

Best Synths for Wavetable Synthesis

Xfer Records Serum 2 is the industry standard — if you learn wavetable synthesis in Serum, the knowledge transfers directly to any other wavetable synth. Vital (free) covers nearly all the same ground and is the recommended learning tool. Native Instruments Massive X, Arturia Pigments, and Roland Zenology all include excellent wavetable engines.

Additive Synthesis — Building Sounds From Sine Waves

Additive synthesis is the theoretical foundation of all synthesis types. Fourier's theorem states that any complex waveform can be represented as the sum of sine waves at different frequencies and amplitudes. Additive synthesis applies this directly — it builds sounds by combining individual sine wave partials at specific frequencies and amplitudes, each with its own envelope.

This gives enormous control — you can shape each individual harmonic independently. But this control comes at the cost of complexity. To recreate a realistic piano sound additively, you might need hundreds of partials each with their own individual envelope. This is computationally expensive and practically complex.

Modern additive synthesis tools — Camel Audio Alchemy (now built into Logic Pro), Kilohearts Phase Plant, and the spectral analysis + resynthesis modes in iZotope RX — have made additive synthesis accessible. These tools can analyse an audio file, decompose it into additive partials, and allow you to edit and transform those partials. The result is a powerful form of sound design where you can manipulate the spectral content of any sound at a granular level.

Granular Synthesis — Texture and Transformation

Granular synthesis fragments audio into tiny pieces — grains — typically 1–100 milliseconds long, and reconstructs or transforms sound from those grains. The technique was pioneered by Iannis Xenakis and Curtis Roads and has become a central tool in ambient, experimental, and cinematic sound design.

How Granular Works

A granular synthesizer takes a source audio file (or a live input) and scans through it, extracting grains at a defined scan position. Multiple overlapping grains play simultaneously — their density (grains per second), pitch, duration, and pan position are independently controllable. By freezing the scan position and playing the same grain repeatedly with slight randomisation, a granular synth can produce an infinite, evolving sustain from any audio moment — the characteristic "frozen time" texture of granular synthesis.

Key Granular Parameters

Position — where in the source audio the grains are extracted from. Modulating position creates the characteristic granular sweep effect. Grain size — smaller grains produce more abstract textures; larger grains preserve more recognisable audio content. Density — the number of grains per second. Low density creates a stuttering, rhythmic quality. High density creates smooth, continuous textures. Pitch randomisation — small pitch variations between grains add a cloud-like, diffuse quality to the texture.

Granular in Practice

Granular synthesis excels at creating pads from any audio source (including speech and field recordings), producing atmospheric textures, time-stretching audio without pitch shifting, creating glitchy or abstract sound design for film and game audio, and transforming simple sounds into complex, unrecognisable textures. Leading granular synthesis plugins include Native Instruments Kontakt's grain engine, Reason Studios Grain, Ableton Live's Granulator III, and Output Portal.

Choosing the Right Synthesis Type