Add one or two pitched copies of a file on top of the original, at intervals you choose, to build harmony without singing or playing a second take.
Two independent voices
The file is analysed once per frame, the same way the Pitch Shifter measures each partial’s exact frequency from how far its phase turns between frames. That single analysis then feeds two separate resynthesis passes, one per voice, each moved by its own ratio. Voice 1 and voice 2 can point in different directions, a third up and an octave down, and neither affects the other.
Level and mix
Each voice has its own level, so a harmony can sit quieter than the lead it doubles. Mix scales both voices together on top of the always-present original: at 0% only the source is heard, and raising it brings the harmony in without ever removing the note it’s harmonising.
Picking intervals
A major third above and an octave below, the default, is consonant in any key because an octave is always safe and a third only mildly colours the chord. A perfect fifth, 7 semitones, is another interval that rarely clashes. Intervals chosen to match a specific chord need to match the source’s actual key, since the shift itself has no idea what key the recording is in.
Example: thickening a solo vocal line
Voice 1 at 4 semitones and 60% level, voice 2 at -12 semitones and 45% level, mix 80%.
The result reads as a small backing section behind the lead rather than a single processed voice, with the octave underneath adding weight and the third on top adding width.