FM and Additive Synthesis

Core Idea

FM (Frequency Modulation) synthesis and additive synthesis represent two fundamentally different approaches to generating complex timbres — both contrast sharply with subtractive synthesis's start-and-remove approach.

In FM synthesis, one oscillator (the modulator) modulates the frequency of another oscillator (the carrier) at audio rates. When modulation occurs in the audio frequency range (above ~20 Hz), it generates new sidebands — frequency components that do not exist in either oscillator alone. The sidebands appear at frequencies of carrier ± (n × modulator), where n is an integer. This means a small change in the modulation index (depth) creates dramatic, nonlinear timbral shifts. The classic FM sound — bells, electric pianos, clangorous metallic tones, glassy organs — comes from these harmonic and inharmonic sidebands. The Yamaha DX7 (1983), built on John Chowning's FM research at Stanford, defined an era of pop music. FM synthesis is CPU-efficient, generating complex timbres from simple sine waves, but parameter-to-sound relationships are nonintuitive and require experience to navigate.

Additive synthesis works from first principles: the Fourier theorem states that any complex waveform can be expressed as a sum of sine waves at different frequencies, amplitudes, and phases. Additive synthesis constructs sounds by explicitly controlling each partial — ideally hundreds of them — with individual amplitude envelopes. This offers complete control over every harmonic, allowing precise reproduction of acoustic instruments or the creation of sounds physically impossible in nature. The computational cost is high (many oscillators per voice), and real-time control of hundreds of partials requires specialized interfaces. Spectral modeling tools like SPEAR and iZotope RX decompose recorded audio into additive components, enabling hybrid resynthesis and transformation.

The operator concept from FM bridges into FM-additive hybrids: operators can be configured in various algorithms (parallel or series chains of modulators and carriers), with each configuration producing distinct spectral character.

Explainer

FM synthesis revolutionized electronic music production in the 1980s by enabling complex, bright, metallic timbres from inexpensive hardware — a Yamaha DX7 cost far less than a Moog but could produce electric piano, vibraphone, and percussive sounds with remarkable realism for the era. The limitation — opaque parameter relationships — became an aesthetic signature: the DX7's "wrong" patches became iconic sounds in their own right.

Additive synthesis offers a complementary perspective: where FM achieves complexity through modulation interactions, additive achieves it through explicit harmonic specification. Understanding additive synthesis provides deep insight into the structure of timbre itself — the fact that all steady-state sounds can be decomposed into sinusoidal components. This connects directly to Fourier analysis, spectrograms, and spectral audio processing.

Modern synthesizers often combine synthesis approaches. Serum combines wavetable oscillators with FM modulation. Ableton Operator is a four-operator FM synthesizer. Spitfire's granular engine resynthesizes recordings as additive partial tracks. Understanding FM and additive synthesis unlocks these hybrid instruments and provides the theoretical framework to understand spectral transformations applied in mastering and restoration processing.