Generate custom speech from text and voice reference
Sampling/Granular/Concatenative Synth using Neural Codec