Inconsistency between conductor and dielectric plugins

[mitsuba-renderer]

Inconsistency between conductor and dielectric plugins

[wjakob]

Hi @schunkes,

apologies in the delay responding to your message. That is indeed an omission, and we should properly compute the (complex-valued!) relative OR when a real-valued external IOR is provided. Probably not too much an issue in most cases (air), but potentially problematic when metals are observed under water. A PR would be welcomed, though the important thing will be to retain compatibility with the parameter conventions of Mitsuba 1.

Now that said, we could adopt a new set of parameter names (e.g. int_eta, int_k, and ext_eta) that are documented exclusively and preferred by the plugin if available.

Best,
Wenzel

[schunkes]

Hi @wjakob,

thank you for your answer. At this moment we are not targeting under water scenes as you mentioned them, so I expect I will not have the time to create such a PR. However, should problems with the conductor type plugins block our work, I'll make sure to create a PR for our changes.

Best wishes,
Sebastian