MediaScribe Narrate

MediaScribe Narrate generates a second, narrated audio track that describes the on-screen action for blind and low-vision viewers, so your video meets accessibility requirements without hiring a narrator or waiting weeks for production.

How it works

Upload a video in any format. Narrate transcribes the dialogue, finds the gaps of three seconds or more where a description fits without talking over anyone, and analyzes the picture for the elements that carry meaning: who is speaking, what is on a slide, a change of scene. It writes descriptions timed to those gaps, voices them with neural text-to-speech at a natural reading pace, and gives you a timeline editor with a waveform to review and adjust every line.

What you get

• Descriptions placed in the natural pauses in dialogue, so they never overlap the meeting
• A two-pass timing system that shortens a description when the gap is tight, keeping the essential information
• Audio ducking that lowers background sound for safety-critical announcements
• Speaker identification and sentence detection for accurate timing
• An interactive timeline editor with waveform display for review and refinement
• Turnaround measured in hours instead of weeks

Accessibility

Audio description supports WCAG 2.1 AA and the ADA Title II requirements taking effect April 26, 2027. It is delivered as a selectable track: SAP on the television, and a toggle in your web, app, and OTT players.