Skip to content

geometrian/simple-spectral

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

47 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Simple Spectral

This is a simple multithreaded spectral pathtracer implementing the algorithm (linear combination of bases) described in our EGSR 2019 paper "Spectral Primary Decomposition for Rendering with sRGB Reflectance" [Mallett and Yuksel 2019]. Also implemented are [Meng et al. 2015], [Jakob and Hanika 2019], and a classic RGB renderer.

The spectral rendering uses hero-wavelength sampling on a simple pathtracer. Our implementation supports both the CIE 1931 and 2006 standard observers.

Setup

Dependencies are intended to be minimal and easy (with the exception of GLM, all required dependencies are included):

  • GLM (required)
  • GLFW (optional, adds windowing support)
  • lodepng (required, included)
  • [Meng et al. 2015] (required, included)
  • [Jakob and Hanika 2019] (required, included)

Setup is basic CMake:

  1. cd <path/to/>simple-spectral/
  2. mkdir build
  3. cd build/
  4. cmake ..
  5. cmake --build .

There are some handy options in "simple-spectral/src/stdafx.hpp" near the top of the file. Some parameters are only exposed this way for simplicity or performance.

Program Invocation

Usage, including command-line options, can be found by simply running the binary with no arguments.

However, for convenience, basic usage looks something like:

<binary> --scene=cornell-srgb -w=512 -h=512 -spp=64 --output=output.png --window

The available scenes are "cornell" (original Cornell box), "cornell-srgb" (adjusted materials; Figure 5.a), and "plane-srgb" (Figure 1).

Acknowledgments

We would like to thank [Meng et al. 2015] and [Jakob and Hanika 2019], both of which make their code easy to use. Also lodepng, for existing. More acknowledgments can be found within the paper.