Ray tracing's main advantage over rasterization is computing secondary effects: reflections, refractions, and shadows

While both rasterization and ray tracing can handle primary visibility, ray tracing is uniquely powerful for secondary effects. By sending secondary rays from a surface hit point — reflection rays, refraction rays, shadow rays — the renderer can compute physically accurate reflections of arbitrary objects on arbitrary surfaces, correct refractions through transparent materials, and soft shadows from area lights. Rasterization approximates these effects (cube map reflections, shadow maps) but the approximations break down: cube maps are only accurate at one point, shadow maps have resolution limits and can’t handle area lights, and global illumination has no good rasterization solution at all. Ray tracing handles these by sending rays to the appropriate regions of the scene and finding actual hit points.

Examples

Shadow rays from a surface point to an area light: sample multiple points on the light, trace a ray to each, count what fraction is unoccluded — yields accurate soft shadows. Reflection rays: trace a ray in the mirror direction from the hit point to compute what is actually reflected.

Assessment

For each effect — hard shadows, soft shadows, mirror reflections, blurry reflections, global illumination — state whether rasterization can handle it acceptably and explain why or why not.