[Bf-cycles] Improved sampling methods for rectangular area lights

[Matthew Heimlich]

Hi all,

Not sure how this slipped under my radar since I was keeping an eye out for
it leading up to its release, but the folks over at Solid Angle recently
released a white paper on improved sampling of area lights by allowing MIS
of the solid angle between the sample point and the light. The results are
pretty impressive in terms of noise reduction, and it seems like Cycles
could benefit. Results are especially good when sampling an area light
through heterogeneous volumes, so it could help lay some groundwork to
avoid headaches with scattering later.

Anyway, thought I'd bring it to the attention of anyone who may have missed
out.

http://lsi.ugr.es/curena/inves/egsr2013/

~Matt Heimlich
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.blender.org/pipermail/bf-cycles/attachments/20130717/173129ac/attachment.htm