Rendered Multispectral 3D Scenes Using Physical Illumination Models
This project demonstrates physically-based spectral rendering using spectral reflectance, spectral power distributions (SPDs), and camera spectral sensitivity models. The goal is to explore how varying illumination conditions impact the appearance of materials and the accuracy of color reproduction.
---- Spectral Rendering Pipeline built with PBRT, Blender, and Python.
- Custom Materials designed using spectral BRDFs and real-world reflectance data.
- Illuminants modeled via real spectral power distributions (D65, Tungsten, etc.).
- Camera Simulation based on spectral sensitivity curves for realistic capture.
- Experiments conducted to evaluate material appearance shifts under different lighting.
Spectral-Rendering/
│
├── materials/ # Material definitions with spectral BRDFs
├── textures/ # Texture maps used in scenes
├── reflectance/ # Measured reflectance spectra
├── spds/ # Spectral Power Distributions of illuminants
├── geometry/ # Geometry and models (e.g., BMW M6)
├── experiments/ # Scene files with different material and light settings
├── images/ # Output renders and visualizations
├── .gitignore
├── BLENDSWAP_LICENSE.txt
├── README.md
└── *.pbrt # PBRT scene configuration files- PBRT-v3/v4: Physically based rendering toolkit
- Blender: Geometry modeling and export
- Python: For data preprocessing and automation
- NumPy, Matplotlib: Spectral data manipulation and visualization
- Install PBRT (v3 or v4) from official repo.
- Clone this repository:
git clone https://github.com/huynhtruc0309/Spectral-Rendering.git cd Spectral-Rendering - Modify or create your own
.pbrtscene files using provided materials and SPDs. - Run PBRT:
pbrt bmw-m6-dragon_250.pbrt
- Visualize output images and compare spectral effects.
| Illuminant | Material: Mirror Coating |
|---|---|
| D65 | ![]() |
| D50 | ![]() |
Geometry and assets from BlendSwap licensed under Creative Commons.
Other code and assets are for educational and non-commercial use only.
Jennie Truc Huynh
Master's in Computational Colour and Spectral Imaging
GitHub Profile





