In essence, the amplitude of a band present in the modulator controls the amplitude of the same band in the carrier. In this instance, they control a VCA (voltage controlled amplifier) of the band. Envelopes create rises and falls in voltage that can control parameters such as amplitude, pitch and filter cutoff frequency etc. The amplitude of the modulator signal generates an envelope which is mapped to the same band in the carrier signal. SoundOnSound has created this handy diagram demonstrating a 10-band vocoder. This is how the characteristics of the modulator are translated to the carrier’s timbre. You can imagine a vocoder to be a little bit like a Graphic Equaliser but instead of manually amplifying or attenuating each band, it’s actually controlled by the corresponding band in the modulator signal. Older vocoders typically had fewer bands while while modern ones can have up to 32, and are therefore more detailed. The number of bands is normally somewhere between 6 and 20.

A band-pass filter isolates a frequency ㅡ you can imagine it like a high-pass filter and a low-pass filter stuck together. The carrier signal is analyzed by the vocoder and divided into a number of bands using a band-pass filter. Let’s take a brief look at what’s happening under the hood. This is one of the most common uses for vocoding, and can be more clearly heard in Giorgio Moroder’s E=MC² (1979), Devo’s Beautiful World (1981) and Kleeer’s Tonight (1984). Most commonly this is heard with a synthesiser as a carrier and a voice as the modulator, in effect making the synth ‘talk’ and mapping the harmonic characteristics to the voice.

Tal vocoder config