Rendering technique that converts 3D polygons directly into screen pixels. Fast enough for real time, making it the standard in game engines and virtual production, unlike slower ray tracing.
Rasterization answers a simple question for every triangle in the 3D scene: Which pixels of the image does it cover? The geometry is projected into image space, and for each pixel, color, texture, and depth are determined; a Z-buffer handles what is in front. This method originates from the early days of computer graphics and is the reason why graphics cards look the way they do — their hardware is built around precisely this pipeline.
Why It Matters On Set
The strength of rasterization is speed. An LED wall in virtual production must keep up with the camera without stuttering — leaving only a few milliseconds per frame. Game engines like Unreal therefore calculate based on rasterization and achieve lighting and reflections through tricks: pre-calculated lightmaps, screen-space reflections, probes. Anyone planning a volume stage should know that the image on the wall comes from this pipeline — with its strengths and its artifacts.
Distinction from Ray Tracing
Ray tracing physically follows light rays through the scene and delivers correct shadows, reflections, and indirect lighting — but at a multiple of the computation time. Rasterization only approximates these effects. In practice, the boundary blurs: modern real-time renderers work hybridly, rasterizing the basic geometry while individual effects like reflections or soft shadows are added via ray tracing. Final frame rendering for VFX shots, on the other hand, mostly sticks to full path tracing.
- For LED wall shoots, test engine limitations early: reflective surfaces and soft shadows are the typical weak points of rasterized images.
- Not every preview is the final image — a raster viewport in previs says little about the later path-traced final.
Related terms
Quiz
1. Was beschreibt „Rasterisierung" am besten?
2. Zu welchem Department gehört „Rasterisierung"?