More Statistical Rendering Papers

Statistical Error Reduction for Monte Carlo Rendering

We present a statistics-based denoising framework for Monte Carlo rendering using multiple Box–Cox transformations and a linear explained-variance correction to improve statistical robustness. We also propose variance denoising, with applications to cascaded denoising and adaptive sampling. Our fast GPU implementation provides denoising quality competitive to state-of-the-art neural denoisers.

SIGGRAPH Asia 2025 Conference Paper
A Statistical Approach to Monte Carlo Denoising

We propose a novel statistical approach to denoising Monte Carlo renderings. Building on a closed-form optimal solution, our approach matches the performance of state-of-the-art neural denoising without any hallucinations and computation-intensive pretraining.

SIGGRAPH Asia 2024 Conference Paper

This work will be presented at SIGGRAPH Asia 2026 in Kuala Lumpur, Malaysia.

Metalog
Materials

TU Wien
ACM Transactions on Graphics 45(6)
SIGGRAPH Asia 2026
(To appear)
*Joint first authors, †Independent researcher
Paper Supplementary Document Code (Coming Soon) Copy Link
A rendered scene showing multiple objects using metalog materials, illustrating their ability to mimic analytic material models, reproduce measured reflectance data, interpolate between different materials, smoothly vary materials across a surface, and support differentiable rendering for appearance optimization.

We showcase metalog materials exhibiting a diverse range of reflectance behaviors, including an analytic Trowbridge–Reitz (TR) material, measured materials from the RGL material database, and materials acquired via differentiable rendering. Our representation naturally supports interpolation between materials, enabling applications such as spatially varying bidirectional reflectance distribution functions (SVBRDFs). The plots on the left and right show bivariate metalogs for selected incident directions of the indicated materials.

TL;DR

We present a unified BRDF representation based on metalog distributions. Through quantile-based interpolation and mixture models, it captures diverse analytical and measured reflectance behaviors while remaining interpretable and efficient. Our approach supports interpolation between materials and material acquisition from real-world images via differentiable rendering.

Abstract

A central aspect of modern physically based rendering is the practical yet realistic modeling of material appearance, which requires efficient and accurate representations of bidirectional reflectance distribution functions (BRDFs). In this paper, we present a unified material reflectance representation based on metalog distributions. Introducing quantile-based metalog interpolation and metalog mixture models, our representation captures a wide range of complex reflectance behaviors of both analytic models and real-world materials, while remaining interpretable and efficient. Our approach achieves competitive runtime performance and a small memory footprint, while enabling continuous interpolation between fundamentally different reflectance behaviors. Furthermore, our model is fully differentiable with respect to the coefficients defining the distributions, thus easily integrates into existing differentiable rendering pipelines. In summary, metalog materials can effectively represent real-world reflectance data, interpolate between multiple materials within a unified representation, and be acquired from real-world images via differentiable rendering.

BibTeX

@article{sakaiHasbay2026metalog,
  author       = {Hiroyuki Sakai and
                  Saip-Can Hasbay and
                  Christian Freude and
                  Michael Wimmer and
                  David Hahn},
  title        = {Metalog Materials},
  journal      = {{ACM} Trans. Graph.},
  volume       = {45},
  number       = {6},
  pages        = {182:1--182:13},
  year         = {2026},
  url          = {https://doi.org/10.1145/3842563},
  doi          = {10.1145/3842563}
}