What holds it together

Voices and languages

Who speaks, in which language, lip-synced to the picture.

Anything spoken in a take can be spoken in a voice you choose, in a language your audience actually speaks, with the lips matching.

Choosing a voice

Every shoot that has someone talking offers a voice picker with two shelves, never mixed:

  • Your voices — audio passports you saved, including one cloned from a clip of your own.
  • Platform voices — the studio's own library.

Or leave it on the model's own voice, which it invents to suit the character.

Voices are described by how they sound — warm, bright, low, unhurried — never by a character's name, because the voice is yours to cast.

Languages

Lines can be spoken in Indian languages as well as English, and they are not translated word for word. Each language has its own lesson — how people actually speak it aloud, which written forms to avoid, and examples — and every translated line is checked by translating it back and comparing. A line that loses its meaning is rewritten, not shipped.

A language the speech model does not carry is refused before anything is spent, naming what it can do.

How it reaches the film

For a spoken scene the line is synthesized first, attached to the render as the performance, and the picture is filmed to match it — so the mouth follows the words rather than the words being laid over a mouth doing something else.

Keep exploring

  • Influencer UGC — where most spoken takes are made
  • Explainers — narration over b-roll

Open the studio→