← Previous

Synth Education

 Next →
Part 64

Beyond Subtractive: Wavefolding

For 63 parts we filtered complex waveforms down to shape timbre. Wavefolding turns that idea inside out: start with a simple sine wave and fold it back on itself to add harmonics. This is the foundational technique of West Coast synthesis, pioneered by Don Buchla in the 1960s.

East Coast vs West Coast

In the 1960s, two radically different approaches to electronic music emerged on opposite sides of the United States. On the East Coast, Bob Moog built synthesizers around subtractive synthesis: start with harmonically rich waveforms (sawtooth, square) and sculpt the sound by filtering harmonics away. This is the paradigm we have explored throughout this series. Oscillators, filters, envelopes, and amplifiers form a signal chain designed to remove energy.

On the West Coast, Don Buchla took the opposite approach. His instruments began with simple, harmonically pure waveforms, typically sine waves, and made them complex through waveshaping, wavefolding, and nonlinear processing. Waveshaping is the general category: any process that reshapes a waveform by running it through a transfer function. Wavefolding is a specific type of waveshaping where the signal folds back on itself. Nonlinear means the output is not a simple scaled copy of the input; the relationship between input and output changes depending on the signal level. Rather than carving a sculpture from a block of marble, West Coast synthesis builds one up from clay. The timbral palette is different: where subtractive synthesis excels at brass, strings, and pads, wavefolding produces metallic, bell-like, vocal, and otherworldly tones that are difficult or impossible to achieve with filters alone.

Neither approach is better; they are complementary. Many modern synthesizers combine both. But understanding wavefolding opens a door to sounds that subtractive synthesis simply cannot reach.

How Wavefolding Works

A wavefolder applies a transfer function to an audio signal. When the signal stays within the threshold range (roughly -1 to +1), it passes through unchanged. But when it exceeds the threshold, instead of clipping flat (as a distortion would), it folds back, reflecting the signal in the opposite direction. Push the signal further and it folds again, and again, each fold adding new harmonics to the output.

The transfer function is a zigzag pattern. Increasing the number of folds compresses the zigzag, creating more reflections per unit of input amplitude. A sine wave through 2 folds gains a few overtones; through 8 folds it becomes a dense, metallic wash of harmonics. The preview below shows how the transfer function changes. The horizontal axis is the input signal level, and the vertical axis is what comes out.

Asymmetric Folding

When the fold pattern is perfectly symmetric around zero, the output contains only odd harmonics (like a square wave). Adding a DC offset shifts the signal so that the positive and negative halves fold differently, breaking the symmetry and introducing even harmonics. This is the "Symmetry" control in the demo below. At zero the folding is symmetric; pushing it positive or negative creates an asymmetric fold pattern that produces richer, more complex spectra including woodwind-like and vocal-like tones.

The demo below also includes an LFO (low-frequency oscillator) that automatically sweeps the fold amount up and down. The LFO Rate controls how fast it sweeps, and LFO Depth controls how far. This lets you hear how the timbre evolves when the fold count changes continuously, a preview of the modulation techniques we will explore in later lessons.

Now try it: play the keyboard and drag the Folds slider to hear harmonics appear.
Wave
Preset
Wavefolder
LFO
Transfer Function
Waveform
Spectrum

Further Reading