Lightworks engineer Manuel Gamito will be presenting a paper entitled ‘Accurate Multi-Dimensional Poisson-Disk Sampling’ at SIGGRAPH in LA, discussing how the technique can be used to improve performance of Global Illumination and Anti-Aliasing processes.  His paper synopsis:

Poisson-disk distributions have been found to be optimal for computer graphics because they are random and yet all the samples are reasonably equidistant to their nearest neighbours. This has significant applications for graphics techniques such as anti-aliasing, final gathering or texturing. Previous techniques were either very expensive or had to introduce simplifications, due to speed concerns, that compromised the quality of the sample distributions. The proposed method is statistically correct, works on 2D, 3D or even higher dimensions, and is made efficient through the use of a spatial subdivision tree.

Manual says that the work is already integrated into the upcoming version the Lightworks Rendering software.  You can check out his presentation on Tuesday 27th July, as part of the ‘Geometry Algorithms and Sampling’ session, between 15.30 and 17.30 in Room 502B.  Full release after the break.

Lightworks to present paper at Siggraph Conference

SHEFFIELD, UK (13 July 2010) – Lightworks, the world’s leading supplier of rendering solutions for developers of advanced 3D computer graphics software, today announced that Manuel Gamito – one of the Company’s development team – will be presenting a paper at this year’s Siggraph Conference in Los Angeles.

The Siggraph Computer Graphics Conference is renowned for being the leading conference in its field and Lightworks is delighted that Manuel will be taking an active part in this important conference. Manuel has been working for Lightworks since 2008 and his work is focused on efficient parallel photorealistic image generation. His paper, entitled ‘Accurate Multi-Dimensional Poisson-Disk Sampling’ is co-authored with Steve Maddock from The University of Sheffield and presents research on techniques for increasing the performance of Global Illumination and Anti-Aliasing processes in real-world situations.

Manuel commented, “The technique that I will be presenting has already been implemented in the new rendering software that Lightworks is developing. It addresses the efficient generation of Poisson-disk sample distributions in a space, with any number of dimensions.”

Siggraph Conference delegates can attend the presentation of this paper on Tuesday 27th July, as part of the ‘Geometry Algorithms and Sampling’ session, between 15.30 and 17.30 in Room 502B

Synopsis

Poisson-disk distributions have been found to be optimal for computer graphics because they are random and yet all the samples are reasonably equidistant to their nearest neighbours. This has significant applications for graphics techniques such as anti-aliasing, final gathering or texturing. Previous techniques were either very expensive or had to introduce simplifications, due to speed concerns, that compromised the quality of the sample distributions. The proposed method is statistically correct, works on 2D, 3D or even higher dimensions, and is made efficient through the use of a spatial subdivision tree.