Is it possible to add this one to Extension List in main repo?

@Reon90 — if you mean this list in the extension registry, note that it has several sections, including Khronos extensions that are (a) already ratified, or (b) in progress and actively moving toward ratification. At the moment, this PR is just a PR and nothing more.

I'm guessing that your comment is meant as a +1 for supporting SSS in glTF in the future? If so, noted! If not, please let us know if you have any other preferences or concerns here. 🙂

(Related: I think something like #1970 may help to communicate the status of PRs more clearly)

Looking at the Khronos Sample Viewer implementation for transmission and volume, I was wondering what the shader code would look like just for the SSS portion of Volume. Our team wants to implement SSS and I wanted to setup support for this extension so they could see the results, but I'm not sure how to detangle the SSS code from the volume code in 'ibl.glsl' located here: https://github.com/KhronosGroup/glTF-Sample-Viewer/blob/master/source/Renderer/shaders/ibl.glsl

Could someone help? For these type of extensions have the sample shader source is really, really useful for us devs.

Apologies in advance if this goes into bike-shedding territory, but to me the KHR_materials_sss name feels inconsistent with other extensions... perhaps KHR_materials_subsurface? Blender's UI just uses the term "Subsurface" for example, omitting the "scattering" term outside of documentation.

Apologies in advance if this goes into bike-shedding territory, but to me the KHR_materials_sss name feels inconsistent with other extensions... perhaps KHR_materials_subsurface? Blender's UI just uses the term "Subsurface" for example, omitting the "scattering" term outside of documentation.

I've no strong opinion here. Agreed that KHR_materials_sss naming feels somewhat inconsistent. On the other hand, the term SSS is almost being used like a name in itself, it's short and certainly triggers the right associations for many people.
Consistency to blender sounds useful. Not sure if I like that idea of dropping the scattering part from the name though, as that's what the extensions technically defines. But KHR_materials_subsurface_scattering is too long. :)

No strong opinion from me either, but I notice the name "scatter" is already in the parameters (scatterColor, scatterDistance) so maybe it's fine to have the name of the extension say only subsurface. Both names will be seen together in a finished glTF file, so it will be clear this is subsurface scattering.

"KHR_materials_subsurface" : {
    "scatterColor": [ 0.5, 0.0, 0.0 ]
}

Hello all! Just to chip in if I may 😄

In Unreal, we also use the shortcut subsurface to denote our SSS shading model.

Additionally, we also allow the scatterColor and scatterDistance (although we call it "density" and it goes from 0 to 1) to have texture inputs. Would that be feasible to add in the extension? i.e.: scatterColorTexture and scatterDistanceTexture?

@aidinabedi, thanks for the input. We explicitely defined the current extension proposal without texturable parameters. With the new glTF PBR extensions we decided to strictly separate surface from volume effects. Texturing volumetric parameters with a surface texture will change the results of the volume evaluation based on the volume entry point. This may be perfectly fine for a rasterizer that uses a surface-based approximation to volume integration, e.g. diffusion approximation. But for path-tracers with bi-directional integration methods such behavior will destroy reciprocity. So you end up with inconsitent results for different rendering techniques, which is something we were trying to prevent by design.

We recently added a diffuseTransmissionColorFactor to the KHR_materials_diffuse_transmission PR. Is this something that may help with the requirement for a textureable scatterColor?

Not saying that the parameterization is set in stone though, we're still discussing alternatives back and forth. If you have example scenarios and you're allowed to share, we' re more than happy to look at them!

Further discussion should take place in #2453, not here. Closing.