AI Voice Data
Studio-grade voice data, director supervised
From recording to license documentation — delivered in one package.
Service Lines
Four voice-data service lines
TTS corpora, voice cloning, conversational data, annotation & cleaning — all under one roof.
TTS Training Corpus
Voice actors reading scripts, long-form recording — delivered to the client, who trains their own TTS model. Multi-style, multi-scenario.
Voice-Cloning Source Material
Single-talent directed session + license documents — client uses the material to train their own voice-cloning model. For customer-service agents and branded voice deployment.
Conversational / Emotion Corpus
Improvised conversational, multi-turn dialogue, full emotional range — client uses this to train their own conversational AI / emotion-aware models.
Data Annotation & Cleaning
Timestamp annotation, transcript proofreading, metadata tagging on existing audio — client hands over recordings and we post-process. Available as a turnkey add-on.
Workflow
Inquire
Brief + language/spec confirmation. Direction + quote within 24h.
Roster
We supply the studio-managed roster — you pick from it.
Produce
Director-led session, spec-compliant recording supervised end-to-end.
Deliver
Audio files + consent + license documents — delivered in one package.
Capabilities
Studio-Managed Workflow
One point of contact. We handle the roster, payment, and documentation — clients never onboard individual talents.
Asian-Language Depth
Mandarin (Taiwan / Mainland / HK Cantonese), Hokkien, Southeast Asian, Indian languages — depth other platforms can't match.
Compliant Licensing
Consent process, explicit license scope, renewal structure — standardized documentation, delivered together.
About Licensing & Contracts
Onyx licenses talent voices — we don't buy them out. Clients receive usage rights, not ownership. Every project documents scope (use case), term (duration), and customer range explicitly — all three parties' rights and obligations are written into the contract.
Send your voice-data brief
Tell us languages, scale, use case, and timeline. Direction + quote within 24 hours.
Submit voice-data brief