A path tracer begins with a camera ray, but that ray is only useful if every coordinate-space conversion agrees with geometry traversal and hit reconstruction. Small convention errors here appear as broken silhouettes, shading discontinuities, or transforms that fail only in complex scenes.

This chapter follows a pixel into world space and through the first BVH, covering camera mapping, intersections, instance transforms, and the data needed to reconstruct a stable surface hit.

OpenChessSet reference render used to validate geometry and camera behavior

Read the full camera and geometry article →