How does CGI rendering work?

Every frame of an animated film starts as millions of rays fired from a camera that does not exist. A virtual camera looks through an image plane cut into pixels. For each pixel, a ray leaves the eye and runs into the 3D scene; the first thing it hits gives the pixel its colour.

Read how it works: How does CGI rendering work? · The history of CGI rendering · More boxes on Glassbox

Chapters in this interactive model

1. One ray for every pixel: A virtual camera, a grid of pixels, and a ray fired through each one.
2. Rasterising vs ray tracing: Games throw triangles at the screen. Films follow rays of light.
3. Light bounces, and noise: Path tracing follows random light paths. More samples, less noise.
4. Shaders: what a surface does with light: Matte, glossy, mirror, glass and skin, all from a few rules.
5. The render farm: Hours per frame × frames = years of computer time. So share it out.
6. Ray tracing in real time: One noisy sample a frame, then clever cleaning. The gap is closing.