Harmonizer

Parameters

Mix — Equal-power blend between dry input (0) and harmonized voices (1). Default 0.5 — equal mix lets you hear original + harmony. Push to 1 to hear only the shifted voices (drop the original from the texture); pull toward 0 to keep harmonies as a subtle bed under the lead.

Output Gain — Output trim, ±12 dB. Multi-voice stacking can pile up a lot of energy quickly — use this to recover headroom after summing 3 or 4 voices.

Additional controls

Fft Size — FFT window size for the underlying phase vocoders (one per voice): 1024 / 2048 / 4096.

• 1024 — lower CPU per voice, smears tonal content slightly. Acceptable for percussive sources.
• 2048 — default. Good balance for vocals and instruments.
• 4096 — best frequency resolution, more CPU. Best for sustained material; stay within ±12 semitones to avoid formant artifacts regardless of FFT size.

Larger FFTs improve pitch precision but smear transients — same trade-off as Pitch Shift, multiplied by voice count.

Voice Widget — Multi-point widget for the harmony voices — 2 to 4 voices, drag points to dial each voice’s interval, scroll-wheel (or tertiary axis) to set per-voice level. Double-click empty area to add a voice (up to 4); right-click a point to remove (down to 2). Y axis = interval in semitones (−12 to +12); X axis is locked (one column per voice). Default ships a 4-voice chord: M3 / P5 / Octave-up / Octave-down. Per-voice intervals and levels are not animatable — they drive the offline phase vocoder, which can’t vary mid-buffer.

About Harmonizer

Harmonizer stacks 2–4 pitch-shifted copies of the input on top of itself for harmony generation — set intervals via the multi-point widget, balance each voice with its level. Internally it runs one phase vocoder per voice, so CPU scales linearly with voice count and roughly with fft_size. For best results, keep intervals within ±12 semitones — beyond that, formants shift audibly (chipmunk / monster character) on every voice. Use for: vocal harmonies (M3 + P5 = chord), instrument doubling (octave down for body, octave up for shimmer), pad thickening (small detunes around 0). Unlike Pitch Shift (single-voice transposer) or Pitch (lightweight per-slice shifter), Harmonizer’s whole purpose is the simultaneous stack. For complex pitch automations, use a separate Pitch Shift on each branch and sum manually — Harmonizer’s per-voice intervals are static by design.

Generated 2026-05-05 from K2K_Dev@96730bdc by scripts/gen_lexique.py. Edit _intros/ or _overrides/, not this file.