Feeds fresh audio to Justin's voice clone so the accent stays honest — no creeping mid-Atlantic drift. Drop in a recording, the diarizer fishes out the bits where Justin's actually talking, you approve the keepers, ElevenLabs learns from them.
Upload audio or video
Drop one or more files here or
MP4, MP3, WAV, M4A, MOV, etc. Up to 2 GB per file. Drop several at once — they're queued and processed automatically.
Review extracted clips
Untick anything where Justin isn't alone (background music, other voices, mid-word cuts). The remaining clips from each interview will be stitched into a single sample, padded with a brief silence between segments.
Preview stitched samples
One file per interview — just the parts where Justin's speaking, joined with ~0.3 s of silence between segments. Listen through; if anything sounds off, untick clips above and click Stitch & preview again.
The source label appears in each uploaded sample's filename so you can find them later.
Retrain voice
Triggers a 6–24 hour retraining run using all samples currently on the voice. Click only when you're done adding samples for this session.