About Services Tools Blog Contact us

Voice Narrated Social Video

Makes a square social video narrated in a cloned house voice. Each recording is checked against the script, and captions and sound bars are animated in time with the words.

  • Voice Cloning
  • Marketing
  • Tool
  • Productised
  • Local/private
  • 2026

The problem

Recording a fresh voiceover for every post is slow. And a caption that drifts from the spoken line looks careless when the film is watched on mute.

What it does

A voice cloning model speaks the whole script in one go. A speech recognition tool listens back, and recordings that are too slow or do not match the script are set aside. The best recording is trimmed, then set with typed captions, sound bars and a music bed under the voice. A take report shows the pace and match score of each recording.

Outcomes

  • A script becomes a narrated square film.
  • A take report lets a producer check the film before it is published.
  • Captions follow the spoken line, so the post still reads on mute.

Simplified mockup

Interactive preview with sample data. Nothing is sent to a server.

Narrate a square film

Script

38 words. Spoken in one pass so the first word of each sentence stays intact.

Voice and render
Ready to render

Take verification

Words
38
Minimum pace
1.9 w/s
Minimum match
0.85
Read-back
TakeWords/secMatch scoreDecision
13.121.00Kept

How it works

  1. Speak the script in one go with the voice cloning model, so the first word of a sentence is not clipped.
  2. Listen back to each recording with a speech recognition tool and reject any that are too slow or do not match the script closely enough.
  3. Keep the best recording, shorten long pauses and start the audio after the branded intro.
  4. Mix a music bed under the voice.
  5. Type captions on as each word is spoken, with cyan sound bars under Dash, the Web Order mascot.
  6. Export a square, high-definition file with the house logo on a dark background and no scrolling text.

Pipeline

flowchart LR
  Script[Script] --> Clone[Qwen3-TTS]
  Clone --> Read[Whisper]
  Read --> Score{Pace and match?}
  Score -->|Keep| Tight[Tighten pauses]
  Score -->|Reject| Clone
  Tight --> Mix[Music bed under the voice]
  Mix --> Caps[Typed captions and level bars]
  Caps --> Film[Square MP4 and poster]

Output

Sample output

Narrated square film
Kept voiceover take