How it works
Browser to BWAV.
Four steps.
No scheduling. No studio. No engineer-on-the-clock.
01 · Producer
Upload.
Drop in the video and the reference voice. We extract cues and build the script book in ten minutes.
02 · Voice artist
Record.
A browser tab. A 3-pop streamer. Live QC catches clipping, noise floor, length variance, and misreads before the next cue.
03 · Supervisor
Review.
Approve takes between meetings. Pickup requests auto-queue back to the artist with your note attached.
04 · Delivered
Stems.
Frame-aligned BWAV per character with bext + iXML metadata. Drops straight onto the timeline.
Built around the voice artist.
Friendly tone. No red warnings. Local backup. Reference-voice fader. Three-take prompt. Block recording for fight scenes.
Our promise
100% real human voices.
Zero voice cloning. Zero synthesis.
Every cue you receive is recorded by a working voice artist on a real microphone, in their own studio. No text-to-speech. No generative voice models. No fine-tuning on your performers. We do not sample, embed, fingerprint, clone, or synthesize voices — ever. That commitment is contractual, not aspirational — see the Privacy Policy § 7.
Voice
Real artist, real microphone, real take. Performance stays the artist's property.
Training
Recordings never train, fine-tune, or evaluate a voice-cloning system or TTS model.
Future
Any synthetic / digital-replica use requires separate, written, compensated consent.
What you get
Six things every delivery includes.
- Capture
- 48 kHz · 24-bit
- Sync
- ±1 frame
- Stem format
- BWAV + bext + iXML
- Per-character
- video-length single file
- Optional split
- Netflix Dialog / Vocalizations / Narration / Group_ADR
- Delivery
- signed R2 URL · 1 hour expiry
Pilot a project.
Tell us the show, the line count, the deadline. Quote inside one business day.