Supersampling

SuperSampling, or supersampling anti-aliasing (SSAA), is a computer graphics technique that renders a scene at a higher resolution than the target display and then downscales the result. The goal is to reduce jagged edges and other aliasing artifacts on rendered images.

It achieves this by increasing sampling density during rendering. The image is shaded at the higher resolution, and the final image is produced by averaging or filtering multiple samples per output pixel during downsampling. Common configurations render 2x, 4x, or more samples per axis. This approach can improve edge fidelity across geometry and textures.

Variants of SSAA include full-scene SSAA, which renders the entire frame at the elevated resolution, and partial or adaptive SSAA, which targets specific areas such as edges or shading regions. TemporalSupersampling combines information from multiple frames to augment spatial sampling, and some implementations mix SSAA with other anti-aliasing techniques to balance quality and performance.

Advantages of SSAA include high-quality edge smoothness and broad compatibility with shaders and rendering pipelines, as it does not rely on the specifics of a shader or scene content. Disadvantages are its heavy computational and memory costs, since rendering at a higher resolution multiplies workload; it can be impractical for real-time applications, and downsampling filters must be carefully chosen to avoid blur.

Historically, SSAA was widely used in older games and professional workflows before more efficient anti-aliasing methods emerged. Modern real-time rendering often relies on alternatives such as multisample anti-aliasing (MSAA), post-process anti-aliasing techniques, or upscaling-based approaches, sometimes in combination with temporal methods to achieve similar quality at lower cost.

See also: anti-aliasing; downsampling; texture filtering; temporal anti-aliasing.