Niklas Harrysson

Thanks for the great feedback @spiffmon and @meshula @spiffmon You are precisely right that the `thin_walled` input is a bit redundant in this sense. The main reason it was added...

I sense we may have been talking about different kinds of sidedness above, so first I want to clarify that the _material sidedness_ we want to describe in MaterialX was...

@proog128 You could flip between two dielectric_bsdf's, using a mixer with the thin_walled parameter as mixing weight, so it's either one with refractions or the other without refractions. This is...

This is a great discussion. But I think there are two parallell questions being discussed: A) _Should_ thin-walled be a feature added to this twosided surfacematerial node? B) _How_ should...

I mean the flag could be added only to the nodedef's of shading models that requires this, and then used as needed in the corresponding BSDF nodegraphs. There is already...

Yes for the transmission part I would use a dielectric_bsdf with ior = 1.0. But for the reflection part we would still have the user-controlled ior. `BSDF thin_walled_dielectric = layer(top...

Thanks for the additional details @proog128, these are some great points. To my mind they are strong enough arguments that we may consider **A** to be resolved, with the conclusion...

Glad to hear that @proog128. @spiffmon, @meshula just let us know if there are any concerns from USD and Pixar side. Thanks for all feedback!

Excellent! Thanks @meshula

Thanks @spiffmon for the additional feedback. I think there are two questions here that need further discussion: 1. Could thin-walled be inferred from the presence of a backsurface connection on...