Graphics Computer Science

The Z-buffer keeps the depth (Z value) of the nearest drawn object per pixel; a new fragment overwrites that pixel only if it is closer, efficiently deciding which surfaces are visible. It turns tricky geometric sorting into per-pixel depth comparison and is easy to implement in hardware, forming a basis of modern GPU rasterization. The idea was introduced by Catmull in 1974.

Raster graphics store an image as a grid of pixels (e.g., PNG, JPEG) and lose quality or show jaggies when scaled up; vector graphics describe shapes with mathematical paths (points, lines, curves), such as SVG, and stay sharp at any zoom. Raster suits continuous-tone content like photos, while vector suits icons, fonts, and graphics needing lossless scaling. The trade-off centers on fidelity versus file size and editability.

Ray tracing synthesizes images by shooting rays from the camera into a scene and simulating their interaction with objects and lights (reflection, refraction, occlusion); because it follows physical light paths, specular reflection, refraction, and exact visibility come out naturally. Whitted's 1980 improved illumination model with recursive reflection and refraction laid the foundation of modern offline rendering. Basic ray tracing, however, yields only hard shadows—soft shadows, depth of field, and global illumination require Monte Carlo methods such as distributed ray tracing (Cook et al. 1984) or path tracing (Kajiya 1986).