ferraus vocoder

Vocoder: make a carrier speak or sing with the input's voice

Synopsis

ferraus vocoder <input> [options] [-o <output>]

The input is the modulator (usually a voice). --carrier is an audio file, or a built-in oscillator (saw, square, pulse, noise) at --note, following the voice's detected pitch (--follow), or playing a MIDI file (--midi part.mid). Bands are log-spaced over --range. With --unvoiced noise, sibilants (detected by their high-band energy ratio) are re-synthesized from a noise carrier so consonants stay intelligible. --mode classic is a filter-bank channel vocoder; --mode spectral is STFT cross-synthesis (smoother, suits high band counts). The output level matches the modulator.

This command is also an fx stage and a pipeline op: ferraus fx in.wav -o out.wav --fx "vocoder:key=value".

Examples

Classic robot voice on a C3 saw

ferraus vocoder vox.wav --carrier saw --note C3 -o robot.wav

Saw that follows the singer's pitch

ferraus vocoder vox.wav --carrier saw --follow -o sung.wav

Vocoded chords from a MIDI part

ferraus vocoder vox.wav --carrier square --midi chords.mid --bands 32 --mode spectral -o choir.wav

Make a pad talk, formants up 3 semitones

ferraus vocoder vox.wav --carrier synth_pad.wav --formant-shift +3 -o pad_talk.wav

Options

Flag	Type	Default	Meaning
`--carrier`	text	`saw`	Audio file, or saw, square, pulse, noise or plugin:<name> (an installed instrument plugin playing --midi, --follow, or --note)
`--note`	frequency	`C3`	Oscillator pitch when not following or playing MIDI
`--follow`	flag	—	Oscillator follows the modulator's detected pitch
`--midi`	path	—	MIDI file whose notes play the oscillator carrier (polyphonic)
`--bands`	integer	`24`	Number of bands Range: 8..64.
`--range`	text	`80Hz..8kHz`	Frequency range of the bands (log spaced)
`--attack`	time	`5ms`	Band envelope attack
`--release`	time	`30ms`	Band envelope release
`--formant-shift`	cents\|semitones	`0`	Shift the vocal formants in semitones (+ brighter/smaller, - darker/bigger)
`--unvoiced`	noise \| off	`noise`	Re-synthesize sibilants from noise
`--mode`	classic \| spectral	`classic`	Filter-bank or STFT vocoder
`--mix`	percent	`100%`	Dry/wet balance: 0% returns the input unchanged, 100% is fully processed

Global options

Every command also accepts the global options (-o, --json, --dry-run, --overwrite, --progress, ...). See Global options.

Exit codes

Code	Meaning
0	Success
1	Processing failed
2	Usage or parameter error
3	Input missing or unreadable
4	Unsupported format or codec
7	Plugin crashed, timed out, or was not found
8	Cancelled
70	Internal error (bug)

Common mistakes

Forgetting that outputs are never overwritten: pass --overwrite or a new -o path when re-running.
Beat-based times (4b, 1/4) need a tempo: pass --bpm or write 4b@120.

For models

MCP tool: ferraus_vocoder
Schema: ferraus schema vocoder
Pipeline: { op: vocoder, args: { ... } } (flag names with _ instead of -)
fx stage: yes

{"command": "vocoder", "defaults": {"carrier": "saw", "note": "C3", "bands": "24", "range": "80Hz..8kHz", "attack": "5ms", "release": "30ms", "formant_shift": "0", "unvoiced": "noise", "mode": "classic", "mix": "100%"}}