Auto cue extraction
WhisperX forced alignment turns a script and reference voice into timed cues in around 10 minutes — about 99% accurate when reference matches script verbatim.
Features
We obsess over the little things — gain headroom, frame alignment, scene navigation, metadata embedding — so the takes that come back are the ones you would have approved in a studio anyway.
Features
WhisperX forced alignment turns a script and reference voice into timed cues in around 10 minutes — about 99% accurate when reference matches script verbatim.
Actors open a link and record. No app to download, no driver to install, no permissions to wrangle.
Broadcast-quality capture in the browser, delivered as BWAV with embedded bext, iXML, and LIST/INFO metadata.
A five-tier indicator catches mic distance, room tone changes, and likely clipping before the actor hits record.
One BWAV per character, spanning the full episode, with silence between cues. Drop directly onto a Pro Tools timeline.
Scene navigation, line context, character notes, and supervisor comments in a single view alongside the recording widget.
Pay for the lines that get recorded, not for studio time you did not need. Volume pricing for series and slates available from sales.
Pin notes to a specific video frame. Threads, mentions, resolutions, and full history exported as JSON with the deliverables.
Single line, cluster (consecutive cues), full scene, and Effort/Wild ADR. Pick the flow that matches the moment.
Optional split into Dialog, Vocalizations, Narration, and Group_ADR groups, named to match the Netflix Sound Mix Specifications.
Performer agreements signed in-platform, with a public hash-verification page and a tamper-evident audit trail. No third-party signature vendor.
Cloudflare for edge and storage; Supabase, Fly.io, Modal, and Stripe handle their own slices. Listed on the security page so security teams can review.
Walk through the upload, recording, and review steps with live screenshots and timing.
How it works