[Bf-cycles] Realistic Nk-Data Metal BSDF
lukas.stockner at freenet.de
Thu Dec 26 17:45:45 CET 2013
Wow, adding a node was really difficult, but I got it working. Attached is a
patch that adds a "Conductive BSDF" which just calculates reflectance based on
the Fresnel equation for conductors.
Its inputs are Eta and K, the coefficients of the complex IOR, and Roughness. It
supports Sharp, Beckmann and GGX microfacets. Performance and sampling look
just as good as the Glossy BSDF. I haven't tried the OSL version yet since I
can't get it to build, but the shader itself works in standard 2.69.7 in the
script node. GPU should work, but not tested since my GPU still is Compute
Two questions remain: First, how should the ETa and K values be input?
Currently, since thay are RGB, they just show up as a color. However, their
compontents can easily reach 5 and higher. 6 Sliders, on the other hand, ane
probably overkill. Second, should there be a set of preset coefficients to
choose from (like in the LuxRender plugin), or just a list somewhere in the
I have rendered a simple demo, the image is on
http://postimg.org/image/70vyf00yp/ . From left to right, the materials are:
Silver, Gold, Chrome, Copper and Aluminium. The coefficients are simply from
multiplication of the PBRT measurements with the CIE XYZ curves and a
transform to sRGB. If someone is interested, I could post an Octave script for
> Yes, that's what I meant, although as he already mentioned in the
> description, a single node would be more useful.
> I'm currently trying to get Blender to build, once it works, I'll try to
> implement it.
> The basic formula isn't that complicated, the most difficult part will
> probably be adding a node :D
> > Their is already broadstu, who made a nice nk-data node tree for cycle.
> > Check blends was for this. It's named something like realistic metals or
> > physical based metals.
> > Lukas Stockner <lukas.stockner at freenet.de> a écrit :
> > >Hi everybody,
> > >do you think a BSDF node for Conductive Materials using measured Nk-Data
> > >and Fresnel's equation for conductors would be useful? Getting Nk-Data
> > >shouldn't be a problem, for example, the PBRT distribution has 162
> > >datasets. Of course, full spectral rendering would be ideal for these,
> > >but
> > >simple RGB should also work fine.
> > >
> > >Lukas Stockner
> > >_______________________________________________
> > >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
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 17481 bytes
Desc: not available
Url : http://lists.blender.org/pipermail/bf-cycles/attachments/20131226/7db0f47a/attachment.bin
More information about the Bf-cycles