[x3d-public] [x3d] X3D meeting minutes 3 FEB 2019: X3Dv4 mantis github specification changes, glTF plans, MetadataSet

Michalis Kamburelis michalis.kambi at gmail.com
Fri Feb 8 12:43:05 PST 2019


 Brutzman, Donald (Don) (CIV) <brutzman at nps.edu> wrote:
> 5. glTF.  Much excellent recent progress.  We discussed next steps for Mantis and github.  Email threads:
>
>         http://web3d.org/pipermail/x3d-public_web3d.org/2019-February/010001.html
>
>         http://web3d.org/pipermail/x3d-public_web3d.org/2019-February/010008.html
>
> Briefly discussed issue of representation
> (a) glTF as "black box" that gets rendered, or
> (b) expansion of glTF internals as X3D.
>
> Certainly understanding the correspondences between ID/labeling, and correspondences between 3D functionality, is centrally important. Ongoing implementation work by Andreas Timo Michalis Leonard and others is showing that such correspondences are indeed possible.  Not yet clear where we should "draw the line" regarding internal glTF access by X3D, along with agnostic/suggested/recommended/required conversion guidance.
>
> Future Khronos intent might have a bearing here as well, but please note our task at hand is to support glTF 2.0 in X3Dv4.0.  Internal editing of content is not apparently a supported feature; so X3D SAI access to glTF internals might not last over long term.  We must proceed carefully.
>
> It would certainly be helpful for authors to be able to reach into a glTF model and get PBR material values, which might then be DEF/USE reused in an X3D scene for model consistency.
>
> The issue of inconsistent lighting among glTF implementations appears to be a side effect of their not yet formally specifying lights.  Nevertheless the PBR lighting equations are well defined.  Therefore we expect to be able to upgrade the X3Dv4 lighting model, lights, materials and rendering to support Physically Based Rendering PBR materials consistently and unambiguously.
>
> Next-step X3D glTF actions of Mantis summaries and specification placeholder is deferred for upcoming call when Michalis can be present.
>

Sorry for not attending today -- it was exhausting week, and in effect
I just overslept a whole day :)

Thank you for the detailed meeting notes.

My personal plan for PBR in X3D is:

- Take my notes on
https://github.com/michaliskambi/x3d-tests/wiki/How-to-add-PBR-to-X3D%3F

- Write up a "rough" version of PhysicalMaterial specification, as a
3rd subpage under
https://github.com/michaliskambi/x3d-tests/wiki/Include-PBR-(PhysicalMaterial-and-related-concepts)-in-the-official-X3D-specification

- Along with it, write a specification of extensions for (Phong)
Material (to add diffuseTexture, specularTexture etc. there, to be
consistent with PhysicalMaterial) and Appearance (to provide
normalmaps regardless of Phong/PBR lighting). Read
https://github.com/michaliskambi/x3d-tests/wiki/How-to-add-PBR-to-X3D%3F
for a discussion why it seems necessary to modify these nodes, along
with adding PhysicalMaterial, to have a complete and consistent
specification.

- Afterwards, implement it in Castle Game Engine, of course.

- I suspect that new PhysicalMaterial will be largely, if not
completely, consistent with what X3DOM is doing, so test it.

I don't really have a timeline when I'll be able to do it yet -- too
much work, and a large backlog of tasks from 2018, is still hanging
over me :) First, I need to make a new Castle Game Engine 6.6 release,
hopefully very soon.

In parallel, somewhere in February, I'll add support for animations
(of transformations/camera) in glTF to Castle Game Engine.

Regards,
Michalis



More information about the x3d-public mailing list