Refactor text according to physical ground truth policy

[portsmouth]

This modifies the text and stated model to try to be consistent about having the model define an as-specific-as-possible concrete physical material model with known BSDFS, which determines a well-defined ground truth.

This is stated in the introduction section as:

Unlike previous models such as Autodesk Standard Surface [Georgiev2019], we define the model by describing the physical structure as unambiguously as possible, rather than by specifying a particular form of implementation. At the level of computer graphics, the model is completely physically specified if the BSDF (Bidirectional Scattering Distribution Function) of each slab interface, and the volumetric properties of each slab medium, are specified. The ground truth appearance of the model is then defined to be given by the BSDF obtained by physical light transport through the whole structure of component slabs. Implementations thus have an unambiguous goal appearance which they can reason about and approximate logically, to a greater or lesser degree of fidelity, on the basis of physics and light transport. The closer an implementation is to reproducing this ground truth physical appearance, the better it conforms to the specification.

Specific changes are:

• some rewording/shuffling in the intro and formalism section to make it read more clearly. Particularly, keeping the introduction as short/terse as possible with focused paragraphs of the structure:

  • we need a standard material model
  • general frameworks exist, but we need a standard uber-shader as well
  • history of uber-shaders
  • but all those shaders were proprietary
  • this model aims to be practical, supports most common use-cases, and physically-based/logical
  • here is the specific, informal structure and the types of common materials we can support
  • we define the physical structure in detail using a simple formalism
  • we define the ground truth appearance to be the physical BSDF of this structure
  • the model is controlled by artistically intuitive parameters driving the physical structure

• modify layering formalism section to make it clear that the ground truth is the complete light transport in the layer, e.g. .

Complete conformance to the spec is defined as reproducing all the physical inter-layer light transport effects, though this is not typically practical. In practice, each implementation must decide what level of approximation to use for the light transport within layers, trading off accuracy for efficiency according to its own particular use case.

• moved some discussion around for more sensible structure, e.g. the discussion about "more accurate layering models" and the mixture model approximation of the shader, is moved into the "Model" section after we introduced the full model.

• the Glossy-diffuse slab was previously not presented as a physical model or a concrete BSDF. It's not obvious how to write down a specific BSDF though, since the Arnold one is not reciprocal and the ASM one does not support Oren-Nayar roughness. So I refactored to instead define it as a physical layer of dielectric "gloss" on top of a diffuse base. This at least is completely well-defined, and reasonably approximated by what we do in Arnold/ASM/MaterialX. It does not change anything at the level of our implementations.

  • the Coat layer is stated to have transmittance defined by the square root of the coat_color. This defines it physically, then we state that all the darkening/roughening effects of the bounces should be taken into account in the ground truth (but explicitly not the more obscure effects due to coat "dissolving" into the base..). The existing implementations are then still reasonable approximations of this.

[virtualzavie]

Let's try to merge it, lower priority than some other changes.