- [✨ feature request] - [❔ other question] In polarized mode, reflection on plastic and thindielectric should be polarized

[MaxDevas]

Summary

It seems plastic and thindielectric materials don't polarize reflected light

• Compiled variants:
  • scalar_spectral_polarized

Description

It is written in the docs : "Mitsuba 2 includes pBSDF implementation for conductors and dielectrics (following the polarized Fresnel equations), as well as standard optical elements such as linear polarizers and retarders."

Yet according to the BSDF overview schema, plastic and thindielectric materials should act like dielectric material regarding the polarization.

When light reflects on a thin dielectric, it should have the same polarization properties as if it reflects on a dielectric, shoudn't it?

In this case, plastic also should act so, due to its thindielectric coating. is this model (diffuse + clear coating) accurate for polarization?

[tizian]

Hi, You are correct that, from a physical perspective, the two materials you mention should in general reflect some polarized light.

The interesting part of the two models is that they offer closed form expression including all multiple-scattering of light between the top and bottom layer. As far as I know, Fresnel equations for multi-layered dielectric plates are well understood, so generalization of the thin dielectric should be possible. I'm less certain how a (analytical) formula for the diffuse + specular layered case of the plastic material should work though.

Due to time constraints, looking into these extensions is currently not on our roadmap unfortunately. The next release of Mitsuba 2 will however include a simplified form of a diffuse+specular pBRDF from this article by Baek et al. 2018.

Best, Tizian