Photon Parameterisation for Robust Relaxation Constraints has been selected as a notable article in computing in 2013. Computing Reviews’ Best of 2013 list consists of book and article nominations from reviewers, CR category editors, the editors in chief of journals, and others in the computing community. The complete list is here. The paper also won best paper at Eurographics 2013.
We are pleased to announce our award of Best Paper at Eurographics 2013 for our paper Photon Parameterisation for Robust Relaxation Constraints, Ben Spencer and Mark W. Jones. [Link to paper]
The paper introduces a technique that augments each photon with information about its origin trajectory. Using this, lighting is separable during density estimation queries. Additionally, we spot fine edge detail using PCA allowing us to employ photon relaxation without detrimental effects. This results in high qualify photon maps that reduce variance and can be rendered with very low bandwidth kernels reducing bias. [Link to news on EG.org]
This paper presents a novel approach to detecting and preserving fine illumination structure within photon maps. Data derived from each photon’s primal trajectory is encoded and used to build a high-dimensional kd-tree. Incorporation of these new parameters allows for precise differentiation between intersecting ray envelopes, thus minimizing detail degradation when combined with photon relaxation. We demonstrate how parameter-aware querying is beneficial in both detecting and removing noise. We also propose a more robust structure descriptor based on principal components analysis that better identifies anisotropic detail at the sub-kernel level.We illustrate the effectiveness of our approach in several example scenes and show significant improvements when rendering complex caustics compared to previous methods.
Ben Spencer and Mark W. Jones
Computer Graphics Forum, Volume 32, Issue 2pt1, pages 83–92, May 2013. [doi]
Best paper, Eurographics 2013.
Ben Spencer and Mark W. Jones have again won the Computer Graphics Forum cover competition. The winning image will be used throughout 2013 as the front cover image of the journal Computer Graphics Forum. The image shows a visualisation of the parameter space of photon trajectory from the emitting light source. When encoded into the photon map, this can be used to distinguish overlapping light envelopes typically associated with caustics. This enables more accurate density estimation along overlapping illumination, thus producing more accurate renders. The technique is described in an upcoming Eurographics 2013 paper. See their previous prize winning entry for 2009.
We introduce a novel algorithm for progressively removing noise from view-independent photon maps while simultaneously minimizing residual bias. Our method refines a primal set of photons using data from multiple successive passes to estimate the incident flux local to each photon. We show how this information can be used to guide a relaxation step with the goal of enforcing a constant, per-photon flux. Using a reformulation of the radiance estimate, we demonstrate how the resulting blue noise photon distribution yields a radiance reconstruction in which error is significantly reduced. Our approach has an open-ended runtime of the same order as unbiased and asymptotically consistent rendering methods, converging over time to a stable result. We demonstrate its effectiveness at storing caustic illumination within a view-independent framework and at a fidelity visually comparable to reference images rendered using progressive photon mapping.
Ben Spencer gives part of a SIGGRRAPH 2012 tutorial on relaxed photon mapping.
State of the Art in Photon-Density Estimation
Thursday, 9 August 2:00 pm – 5:15 pm | Los Angeles Convention Center – Room 408B
In this paper we seek to eliminate the noise caused by caustic paths during progressive Monte Carlo path tracing. We employ a filtering strategy over path space, handling each subspace using specialised derivations of path tracing and progressive photon mapping. Evaluating diffuse paths with path tracing allows the use of sample strati cation over both pixels and the image as a whole, whilst sharp detailed caustics are produced using progressive photon mapping. This is an efficient, low noise progressive algorithm with vanishing bias combining the advantages of both Monte Carlo methods, and particle tracing.
The photon mapping method is one of the most popular algorithms employed in computer graphics today. However, obtaining good results is dependent on several variables including kernel shape and bandwidth, as well as the properties of the initial photon distribution. While the photon density estimation problem has been the target of extensive research, most algorithms focus on new methods of optimising the kernel to minimise noise and bias. In this paper we break from convention and propose a new approach that directly redistributes the underlying photons. We show that by relaxing the initial distribution into one with a blue noise spectral signature we can dramatically reduce background noise, particularly in areas of uniform illumination. In addition, we propose an efficient heuristic to detect and preserve features and discontinuities. We then go on to demonstrate how reconfiguration also permits the use of very low bandwidth kernels, greatly improving render times whilst reducing bias.
Ben Spencer and Mark W. Jones won the 2009 Computer Graphics Forum cover competition with this image. The caustics underneath the sphere and leaf are generated using an enhanced photon mapping algorithm described in Into the Blue: Better Caustics through Photon Relaxation, Ben Spencer and Mark W. Jones, Eurographics 2009. The advantage of the approach is that low-noise radiance estimates may be achieved using very low bandwidth kernels. The caustic photon map in this scene contains 120,000 photons and only 50 are used in each radiance estimate. We can achieve good-quality results with as few as 20 photons. This results in the reduction of proximity, topology and boundary bias and also reduces the time required to render caustic illumination. The scene was created using a scan of a real leaf, post-processed and overlaid onto a translucent scattering dielectric film perturbed using a fractal noise function. Clip, gloss and bump maps were then created and the whole scene rendered using our own global illumination rendering platform.
Photon mapping is an efficient method for producing high-quality photorealistic images with full global illumination. In this paper, we present a more accurate and efficient approach to final gathering using the photon map based upon the hierarchical evaluation of the photons over each surface. We use the footprint of each gather ray to calculate the irradiance estimate area rather than deriving it from the local photon density. We then describe an efficient method for computing the irradiance from the photon map given an arbitrary estimate area. Finally, we demonstrate how the technique may be used to reduce variance and increase efficiency when sampling diffuse and glossy-specular BRDFs.