Core concepts

These are the building blocks every JSON2Video request shares. Understanding them is enough to read and write any movie JSON.

Movie. The final video output. A movie is a JSON document describing the video's resolution, scenes, and movie-level elements that overlay every scene.
Scene. A segment of the movie. A movie contains one or more scenes, played back-to-back. Scenes cannot overlap; each renders independently and is then chained into the final output.
Element. The atomic content unit inside a scene or movie. Element types include image, video, text, audio, voice, subtitles, component, html, and audiogram.
Duration and timing. Every element has a start (when it appears) and a duration (how long it plays). duration accepts a number in seconds, or -1 for "natural duration of the asset", or -2 for "match the container".
Coordinates. Positioning works like HTML/CSS: the canvas origin (0, 0) is the top-left corner. x and y move right and down. Width/height match the movie's resolution.
Layering. Elements stack in the order they appear in the JSON array. Later elements paint on top of earlier ones, the same way HTML siblings stack along the z-axis.
Caching. The renderer caches downloaded assets and intermediate scene renders to speed up repeat requests and reduce cost. Set cache: false on an element to bypass the cache for that asset.

Core concepts#