Project 3: Griffin Evans

[grievans]

Repo Link

https://github.com/grievans/Project3-CUDA-Path-Tracer/tree/main

Core Features Completed

• Shading kernel with BSDF evaluation (diffuse, perfect specular surfaces)
• Stream compacted path termination
• Sorting by material type
• Stochastic sampled antialiasing

Extended Features Completed

• Visual Improvements
  • [2] Refraction with Fresnel effects
  • [2] Physically-based depth-of-field
  • [3] Motion blur by sampling at different times, with interpolated keyframes for object movements
• Mesh Improvements
  • [4] glTF model loading, in conjunction with BVH mentioned below
• Performance Improvements
  • [1] Russian roulette path termination
  • [6] Hierarchical spatial data structure using BVH
• Other Improvements
  • [2] Environment mapping

Other Features and Details

• I used the tinygltf library to load data from glTF files. In doing so I also needed to change stb_image.h and stb_image_write.h from those originally included in the project to the newer versions used by tinygltf, for compatibility.
• I also modified the top-level CMakeLists.txt in order to get Visual Studio to properly recognize and provide tooltips for some of the already-used libraries' classes/functions (particularly Thrust).
• I made various scene files for testing, mostly variations of the Cornell Box with relevant components added (e.g. adding a mirror and glass material to test reflection and refraction, placing objects at multiple distances to test depth-of-field, setting an object to have keyframed changes in translation to test motion blur, loading in various glTF models). I also made scenes to both test loading of multiple & more complex models and act as more of a showcase of the path tracer's features—currently I've settled on a scene of ducks about a chessboard, which includes refractive and reflective materials, loading of multiple glTF models, and environment mapping, and well-showcases the effect of depth-of-field through objects being at a variety of distances from the camera. I plan though to potentially tweak the scenes shown further while finishing my readme.

README Completion Checklist

• Cover image in readme does not use Cornell Box
• Descriptions, screenshots, debug images, side-by-side comparisons of features implemented
• Analysis
• Scenes and meshes included or linked
• Third-party library changes or compilation changes documented
• Bloopers (optional)

Late Days Used

I am using 3 late days. I am using an additional, 4th late day for the readme.

Project feedback

[grievans]

Sorry for my impunctuality; I hope this has not caused too much of an issue.