Project: CPU Path Tracer

• Parent: PathTracer Learning
• Phase: PathTracer Learning - Phase 2 - CPU Ray Tracing

Goal

• Render a Cornell box scene with global illumination
• Implement all core path tracing concepts before moving to GPU

Milestones

• Milestone 1: Ray-sphere intersection + simple shading
  • Cast rays through pixels
  • Intersect with spheres (simpler than triangles)
  • Shade with surface normal (no lighting yet)
  • Output PPM image
• Milestone 2: Ray-triangle intersection + BVH
  • Implement Möller–Trumbore
  • Build BVH with SAH
  • Load OBJ file (Cornell box)
• Milestone 3: Lambertian path tracing
  • Cosine-weighted hemisphere sampling
  • Recursive path tracing (fixed depth)
  • Accumulate samples over multiple frames
• Milestone 4: Russian roulette + NEE
  • Add Russian roulette termination
  • Add next event estimation for area lights
  • MIS weighting
• Milestone 5: GGX specular
  • Implement GGX BRDF
  • GGX importance sampling
  • Mix diffuse and specular based on metallic/roughness

Reference Scenes

• Cornell box — classic test scene
  • Two colored walls (red left, green right)
  • White floor, ceiling, back wall
  • Two boxes (one tall, one short)
  • Area light on ceiling
• Veach MIS test scene — tests MIS correctness
  • Multiple lights of different sizes
  • Glossy floor with varying roughness

Output

• PPM or PNG image
• Resolution: 512×512 for development, 1920×1080 for final
• Samples: 64 spp for quick preview, 4096 spp for reference

Code Structure

src/
  math/     - Vec3, Ray, Transform
  geometry/ - Triangle, AABB, BVH
  material/ - Lambertian, GGX, BRDF interface
  scene/    - Scene, Camera, Light
  renderer/ - PathTracer, Accumulator
  main.cpp  - Entry point, image output