Formant Morph

Parameters

Morph Amount — Morph position between A and B, 0–1. 0 = pure Source A formants applied to the chosen excitation; 1 = pure Source B formants. Mid values blend the two formant envelopes — at 0.5 you get a hybrid vocal tract that’s neither A nor B but interpolates between them. Animatable for vowel-shifting effects.

Envelope Order — Cepstral order for the formant extraction, 2–40. Lower values (2–10) capture only the broadest formant shape — useful for very smooth vowel morphing without picking up pitch information. Higher values (20–40) track formants more precisely but can leak pitch content into the envelope. 20 is balanced for most vocal material.

Additional controls

Excitation — Which input’s pitch/excitation content the morphed formants are applied to:

• Source A — uses A’s excitation; result has A’s pitch with morphed (A↔B) formants. The most common choice — keep A’s vocal performance, change A’s vowel to a hybrid of A and B.
• Source B — uses B’s excitation; result has B’s pitch with morphed formants.
• Blend — interpolates excitation between A and B by morph_amount as well. Both pitch and formant change together — produces full vocal-tract morph rather than just vowel shift.

Phase — How phase information is reconstructed:

• Preserve — keeps the chosen excitation’s phase (cleanest default).
• Dominant — uses the loudest input’s phase.
• Blend — interpolates phase per morph_amount.

About Formant Morph

Formant Morph splits each input into “excitation” (pitch/source) and “envelope” (formants/filter), interpolates the envelopes between A and B, then re-applies the chosen excitation. The result: vocal-tract shape can morph from A to B while keeping the pitched performance from one source intact. Use it for: vowel morphing on vocals (smooth transitions between “ah” and “ee” by setting B to a different vowel), cross-singer transformations (apply singer B’s vocal-tract shape to singer A’s performance), or instrument-vocal hybrids (apply a vocal envelope to an instrument’s excitation). Compare with the Spectral Vocoder which is simpler (carrier × modulator’s amplitude per band) and the Source-Filter approach used by the Spectral Envelope Extractor for manual decomposition.

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