This is very interessting as it can solve the need for a cache to compute SDS caustics.
Also it can speed up interior with lot of reflective surfaces.
And carpaint glittering could be achieveable now.
Hum I like it. Now it depend on how well the paper is wrote and how hard the method is to implement....
Support LuxCoreRender project with salts and bounties
Yup, bye-bye caches & excessive pre-computing.
In come animation w/o glittery flickering.
GGX Integral is provided, but LuxCore is lacking features.
So, it's more of a question when Cycles (and/or others - RPR, redshift, octane...) get this, what's there left for LuXcore?
Dade wrote: Sun Jun 20, 2021 11:21 am
Actually, LuxCore has GGX in Disney material.
Right, but since even 1. iteration of Principled isn't fully implemented... ... LuxCore future doesn't seem bright
Sharlybg wrote: Sun Jun 20, 2021 10:30 am
Now it depend on how well the paper is wrote and how hard the method is to implement....
With Wenzel Jakob co-authoring paper, I assume examples were most likely produced using Mitsuba so implementation could be and will probably be studied.
Dade wrote: Sun Jun 20, 2021 12:30 pm
Emm, you should read the paper ... the idea has couple of limits ... really
That one , do they mean specular to specular path (if that make any sense ) ?
The main limitation of our method is its restriction to
specular chains with a single vertex, hence other strategies remain
necessary to handle specular transport with multiple interactions.
Addressing this significantly more challenging case is an important
topic for future work.
Support LuxCoreRender project with salts and bounties
Dade wrote: Sun Jun 20, 2021 12:30 pm
Emm, you should read the paper ... the idea has couple of limits ... really
That one , do they mean specular to specular path (if that make any sense ) ?
The main limitation of our method is its restriction to
specular chains with a single vertex, hence other strategies remain
necessary to handle specular transport with multiple interactions.
Addressing this significantly more challenging case is an important
topic for future work.
Yes, among other things, it is not able to render the caustic of a sphere (2x specular intersections) over a plane
Support LuxCoreRender project with salts and bounties