[x3d-public] Comment of X3D V4, glTF, and Lighting

Fri Jan 15 16:36:12 PST 2021

As with all X3D lights, you can define them in the same model, or in
the "outside" model that includes another scene by "Inline".

This is the case with existing PointLight, SpotLight,
DirectionalLight, and it will equally apply to the new
EnvironmentLight. These lights also affect an inlined children content
(in X3D and glTF).

So, it seems to me that the use-case "the model may be needed in an
environment not originally planned" is covered by X3D already, and was
in X3D 3 too.

As for your statement "Requiring the lighting to be embedded with the
model severely limits the usefulness of the model. " -- this is not
the case, lights can be defined in the outside file (that merely
inlines the X3D or glTF). If you mean that "browsers should be able to
supply additional lighting, not present in the scene" -- this is
certainly an option for browsers. Various glTF viewers indeed contain
some "built-in" light sources, and nothing stops X3D browsers from
doing the same. I do not believe we need (or even should) address this
in the spec (neither does glTF), as this is browser-specific behavior
by definition.

Regards,
Michalis

pt., 15 sty 2021 o 22:14 Leonard Daly <Leonard.Daly at realism.com> napisał(a):
>
> Hi Michalis,
>
> It is almost always the case that the environment lighting needs to be external to the model. The model may be needed in an environment not originally planned -- dark vs. light room; indoor vs. outdoors, etc. Requiring the lighting to be embedded with the model severely limits the usefulness of the model. There is no issue if the provided lighting is optional (e.g., studio for product sample).
>
> Also note that the glTF lighting extension is a vendor-specific offering and may not be supported by DCC tools or other viewers of glTF models.
>
> Leonard Daly
>
>
>
> I believe Don Brutzman created a related Mantis issue about it
> recently, he will likely post a link.
>
> Image-based lighting (in the form of "EnvironmentLight" node) is
> indeed something I planned, but ultimately didn't find the time to
> finalize for X3D 4.0. But I still very much want to have it. The only
> question is "when" (for X3D 4.0 or postpone to X3D 4.1) and this
> practically depends on me too (to find time to finish the work I
> describe below).
>
> My plan about it is on
> https://github.com/michaliskambi/x3d-tests/wiki/Image-Based-Lighting-(EnvironmentLight-node)
> . There is an initial node spec, and Castle Game Engine even
> implements it and we have testcases in
> https://github.com/michaliskambi/x3d-tests/tree/master/pbr/environment_light
> .
>
> That said, there are still important questions for the design, and
> important holes in my current implementation.
>
> One of the goals is of course to make it compatible with related glTF
> extension, https://github.com/KhronosGroup/glTF/tree/master/extensions/2.0/Vendor/EXT_lights_image_based
> .
>
> I do agree it is an important feature, we definitely want to have it
> at one point in X3D. I do not think it is a blocker though -- punctual
> light sources (point, spot, directional) work with physical (and
> non-physical) materials as well, in both X3D and glTF.
>
> Regards,
> Michalis
>
>
> pt., 15 sty 2021 o 20:47 Leonard Daly <Leonard.Daly at realism.com> napisał(a):
>
> Posted on behalf of Khronos' 3D Formats Working Group - owners of the glTF specification.
>
> This message addresses the proposed use of glTF in X3D V4.0, specifically the means for lighting glTF V2.0 models.
>
> glTF V2.0 primarily uses materials that are rendered using physically-based rendering (https://github.com/KhronosGroup/glTF/blob/master/specification/2.0/README.md#materials) with the metallic-roughness model. Because of the potential for highly reflective materials, it is beneficial to simulate real-world lighting for the rendering to approximate what the eye would see with a real-world object. This simulation is typically done with image-based lighting (IBL). While IBL does not cover all situations (e.g., internal lighting in a model of a lamp), it does provide an excellent approximation of ambient and environment light source when there is a single model and the model/light interaction is not time dependent.
>
> To describe and illustrate lighting issues I wrote a blog post "Importance of Environment Lights" (https://realism.com/blog/importance-environment-lights) with an accompanying interactive demonstration "PBR Model & Lighting Inspector" (https://realism.com/Examples/pbr-lighting/). The tool allows the user to pick a model, image lighting, and environment image (Background node in X3D) and shows the real-time lighting effects. The tool uses Google's <model-viewer> custom HTML tag. Note that the lighting available in the demo is not indicative real-world lighting, but is simpler to illustrate the point. Another 3D Formats WG member (Eric Chadwick from Wayfair) rendered a different model (Fight Helmet) using closer to studio lighting and 3 Point lights (separate renders). The comparison image is at http://realism.com/Examples/pbr-lighting/FlightHelmet-3Points-IBL.png
>
> In the review copy of the X3D V4 specification, section 17 (https://www.web3d.org/specifications/X3Dv4Draft/ISO-IEC19775-1v4-WD3/Part01/components/lighting.html) describes the lighting component. There is no discussion of image based lighting. Section 9.4.2 Inline (https://www.web3d.org/specifications/X3Dv4Draft/ISO-IEC19775-1v4-WD3/Part01/components/networking.html#Inline) discusses the importing of glTF models into the X3D scene. There is no discussion of lighting there.
>
> Not including IBL with glTF model rendering creates appearances that are inaccurate and do not show off the capabilities of glTF materials and appearance. The 3D Formats Working Group highly recommends including IBL capability when rendering glTF models.
>
> --
> Leonard Daly
> 3D Systems & Cloud Consultant
> LA ACM SIGGRAPH Past Chair
> President, Daly Realism - Creating the Future
> _______________________________________________
> x3d-public mailing list
> x3d-public at web3d.org
> http://web3d.org/mailman/listinfo/x3d-public_web3d.org
>
>
> --
> Leonard Daly
> 3D Systems & Cloud Consultant
> LA ACM SIGGRAPH Past Chair
> President, Daly Realism - Creating the Future
> _______________________________________________
> x3d-public mailing list
> x3d-public at web3d.org
> http://web3d.org/mailman/listinfo/x3d-public_web3d.org