A vocoder imposes the spectral envelope of a modulator signal onto a carrier using a bank of matched band filters

A vocoder analyzes a modulator signal (typically voice, via Surge XT’s audio input) by splitting it into a bank of frequency bands and tracking each band’s amplitude with an envelope follower. Those amplitude envelopes are then applied to the same bands of a carrier signal (the synth’s oscillators), so the carrier ‘speaks’ with the modulator’s changing formant/spectral shape — the source of the classic robot-voice and talking-synth sounds. Surge’s implementation offers 4–20 bands spread evenly in pitch between a Min and Max frequency; more bands give higher spectral resolution and intelligibility, fewer bands give a coarser, more synthetic character. A Q control sets filter steepness, a Gate silences bands below a threshold to reduce noise, and Env Follow sets how fast the amplitude tracking responds.

Examples

Feed a spoken phrase into Surge’s audio input as modulator with a bright sawtooth carrier and 20 bands for an intelligible talking synth; drop to 8 bands for a rougher, more robotic timbre.

Assessment

In a vocoder, which signal is the carrier and which is the modulator, and what does each contribute to the output? How does increasing the band count change intelligibility versus character?