AI Vocal Clone Creation
Publicada el 2026-07-26
Descripción de la oferta
I’m producing a set of new tracks and want them to feature an AI-generated voice that is indistinguishable from my preferred singer. You will receive polished, professional studio recordings as the reference material, so there’s plenty of clean data to train on. What I need from you is a fully trained model capable of rendering original vocal lines with the same tone, timbre, pitch accuracy and expressive nuance heard in the references. The end result should drop straight into a modern music-production workflow without noticeable artefacts or “robotic” flavour. Deliverables (discrete and specific) • Trained vocal-cloning model files (or cloud instance access) • Simple usage guide covering installation, inference settings, and best practices for DAW integration (VST, AU, or standalone inference script) • A short demo stem singing lyrics I will supply, so I can quickly judge quality before final hand-off I’m happy to discuss preferred frameworks (e.g., Diff-Singer, iZotope’s new toolkit, proprietary Python pipelines) as long as the output meets commercial release quality. Please outline your approach, any GPU requirements, and the estimated turnaround time when you reply.
Skills
Fuente original: freelancer