Catalog Series

Extended Vocal Techniques Spectrum (Music)

Rights-cleared, articulation-level recordings of musical extended vocal techniques— captured as controllable phonation and resonance behaviors for expressive singing, character performance, and generative music systems (not speech datasets).

Why this series matters

Musical voice is an instrument—capable of harmonic complexity, nonlinear behaviors, and technique-driven transitions that don’t appear in typical speech corpora. This series isolates extended techniques so models can learn them as explicit, controllable performance behaviors rather than “style” artifacts.

Performative vocal color

Nonlinear phonation

Harmonic control

Technique transitions

What makes it hard to capture

Technique boundaries are subtle and require performer-grade labeling
Many sounds involve bifurcations and subharmonics that confuse generic feature extractors
Clean isolation must preserve intent while avoiding fatigue-driven artifacts
Metadata must encode technique and control parameters—not vague descriptors

Typical use cases

Expressive singing and character-vocal generation for music, games, and film
Technique-aware controllability (e.g., phonation mode, harmonic emphasis, transitions)
Musical technique classification and boundary detection
Robustness testing beyond clean singing / “standard” vocal tones

Proteus + Orpheus

All datasets in this series are protected under the Proteus Standard™. Orpheus metadata may be applied at the Enterprise tier to support instruction-tuned control vocabularies for technique intent, phonation mode, and performance dynamics.