API v1 Sceneloom

A lesson script in.
A narrated, animated video out.

Send plain text to one endpoint. The service cuts it into scenes, draws them, speaks them in a voice of the lesson's language, and answers with a link that plays.

  1. 1Scriptyour own words
  2. 2Scenespictures, placed on words
  3. 3Voiceone per language
  4. 4Render1080p, with captions
  5. 5Linkplays from Amazon S3

Themes

A theme sets the colours of the frame, the cards, and every drawing. Name it in the order: "theme": "it". Without one, the default look is used. These pictures are frames of the same lesson, drawn by the service itself.

How to use it

  1. Order a video

    Send the script with your key. The answer comes at once, with the identifier of a job. sample.json is a complete order: a script and its directions.

  2. Ask how far it is

    One video is made at a time, and making one takes minutes, not seconds. Ask every few seconds until the status is SUCCEEDED or FAILED.

  3. Play it

    video_url is a link to Amazon S3. It plays in a browser, in a <video> element, or in any player, and it can be downloaded. Captions come beside it as WebVTT and SRT.

The fields of an order
script The lesson as plain text, in the form shown below. Or send request, a request that is already directed.
directions Optional. Your own choice of scenes and pictures.
theme Optional. The key of a theme.
voice Optional. The name of a voice. Without it, the voice of the lesson's language speaks.
narration "voice", or "silent" for a free trial without sound.

During maintenance, an order is answered with 503 and the code MAINTENANCE, and Retry-After says when to try again. GET /v1/capabilities shows the same state as "status". A silent trial still works.

The script

A few lines about the lesson, a line of three dashes, then the explanation in your own words. Every sentence is spoken exactly as written. A paragraph usually becomes one scene.


          
title
Shown on the opening card.
discipline
Chooses the drawings the lesson may use.
language
The language of the text, such as en-GB, sk-SK, or ar-SA. Arabic is laid out from right to left.
org_line, series_code, series_line, episode_line, author_line, meta_line
Optional lines of the opening card and the frame.
The text
Write numbers and symbols the way they should be spoken.

Directions

Directions say which sentences form a scene, what the scene shows, and the phrase at which each element appears. The phrases must be in the script, word for word. This is the first scene of the sample.

Scene types

What the service can draw. Directions name one of these for each scene and fill in its data.

Voices

Each language has a voice that speaks when an order names none. Name another with "voice".

Try it

This form sends the sample lesson to the API, exactly as the commands above do. A silent trial costs nothing.