[Bf-cycles] Raytraced subsurface scattering

[Aurel W.]

Hi,

this might be not pragmatic, but could volume rendering in cycles be
kept flexible enough to support anisotropic radiative transfer at some
point? The classical underlying theory of radiative transfer only
supports isotropic materials (which shouldn't be confused with plain
isotropic and anisotropic phase functions). Since older BSSRDF models
are derived from that equation, they all have this basic limitation.

There is this siggraph paper from 2010, which describes these
techniques and also a new BSSRDF model:

http://www.cs.cornell.edu/projects/diffusion-sg10/

Tho it might be not of priority at this point, I guess it would be
good to keep it in mind, when making design decisions with the current
implementation. I guess in a couple of years we have enough
computational power to consider such volume rendering techniques also
for animation and alike.

aurel

On 11 March 2013 13:16, Brecht Van Lommel <brechtvanlommel at pandora.be> wrote:
> I'll consider it when I plan the features for 2.68, currently don't
> want to make any promises about that.
>
> Brecht.
>
> On Sun, Mar 10, 2013 at 7:09 PM, Bryan Nearing <bryannearing at gmail.com> wrote:
>> I don't want to lock you into a fixed release date or anything but would the
>> standard skin, hair, and volume be something for 2.68 while you are doing UI
>> and small features or is it something to do after you have completed volume
>> in 2.69?
>>
>>
>> On Sun, Mar 10, 2013 at 10:53 AM, Brecht Van Lommel
>> <brechtvanlommel at pandora.be> wrote:
>>>
>>> The papers are these:
>>>
>>> Bidirectional Lightcuts - Appendix B
>>> Implementing a skin BSSRDF (or several...)
>>>
>>> Anyway, I'm finding out that these also still have some issues and I'm
>>> trying to solve them. The main one being how to normalize the amount
>>> of light being contributed for arbitrary geometry shapes. Eventually a
>>> skin shader will be there yes, but not for this release. In the end we
>>> should have a Standard Skin, Standard Hair, Standard Volume, ... maybe
>>> a few others, for quick setup.
>>>
>>> Brecht.
>>>
>>> On Sat, Mar 9, 2013 at 1:07 AM, Matthew Heimlich
>>> <matt.heimlich at gmail.com> wrote:
>>> > Brecht,
>>> >
>>> > I was listening to the podcast and heard you mention that you had
>>> > already chosen and begun to implement a SSS shader using brute force
>>> > raytraced scattering. I'd appreciate if you could link me to the paper
>>> > in question, as I haven't been able to find a good paper on raytraced
>>> > SSS anywhere.
>>> >
>>> > Additionally, I'm wondering if there are plans to add a dedicated
>>> > "skin" shader allowing for in-built support for three-layer scattering
>>> > with separate inputs for surface, dermis, and epidermis
>>> > colors/scales/weights/etc. These settings are incredibly useful for
>>> > character lookdev and save a lot of time in other raytraced renderers
>>> > with SSS implemented.
>>> >
>>> > Thanks,
>>> >
>>> > Matt Heimlich
>>> > _______________________________________________
>>> > Bf-cycles mailing list
>>> > Bf-cycles at blender.org
>>> > http://lists.blender.org/mailman/listinfo/bf-cycles
>>> _______________________________________________
>>> Bf-cycles mailing list
>>> Bf-cycles at blender.org
>>> http://lists.blender.org/mailman/listinfo/bf-cycles
>>
>>
>>
>> _______________________________________________
>> Bf-cycles mailing list
>> Bf-cycles at blender.org
>> http://lists.blender.org/mailman/listinfo/bf-cycles
>>
> _______________________________________________
> Bf-cycles mailing list
> Bf-cycles at blender.org
> http://lists.blender.org/mailman/listinfo/bf-cycles