[radiance] is radiance glass material reflective by default?

I created a Radiance glass material with zero transmittance. However, there seems to be reflections on the glass surface based on the rendering shown below (those floating surfaces).

May I ask if Radiance glass material reflective by default? If it is the case, shall I just change the material to plastic since the transmittance is zero?

Thanks.

Replies are closed for this discussion.

Replies to This Discussion

Permalink Reply by Grasshope on August 10, 2017 at 2:54am

For those Radiance users, have any of you encountered such issue before?

Thanks.

Permalink Reply by Chris Mackey on August 12, 2017 at 4:30am

Hi Ji,

I am not as much of a radiance guru as others on this forum but I can say that the Radiance glass material is a simplified model and I would not be surprised if there was some default reflectance assumed for it. From your image, it looks like it is assuming some specular reflectance to account for the mirror-like properties of smooth glass. If you need precise control over the amount of light that is reflected vs transmitted vs absorbed, I would recommend using the Radiance trans material. Otherwise, if you just need a material with 0 reflectance, the opaque material should suit your needs.
I hope that helps,
-Chris

Permalink Reply by Sarith Subramaniam on August 12, 2017 at 6:26pm

1. The core algorithm inside Radiance essentially solves Robert Kajiya's famous Rendering Equation. One of the key principles in Kajiya's theory relates to balancing radiant energy between any two points.

2. The "glass" primitive is implemented inside Radiance as a simplification of the "dielectric" primitive. It only accounts for transmission and reflection of light and assumes a default value of 1.52 as refractive index. This simplistic assumption works for most architectural applications as we are mostly interested in just transmission and reflection. More complex phenomena such as bending of light inside thick glasses, caustics etc can be studied with "dielectric" instead (which usually increases the simulation time significantly).

If you consider #1 and #2 from above, you have your answer. With regards to what to use for your simulation, I agree with Chris. Use an opaque material instead (like plastic or metal). Keep the specularity to zero if your material is not shiny.

A few years ago, I had done some tests on this while learning Radiance. A couple of images from then are below. The values are from 0 transmittance to 100 % transmittance (left-right). The color palette for falsecolor is the default color palette usually set in Radiance.

If you set the refractive index to 1 instead of 1.52, you'd get rid of any presence of material at all (because 1 is for air). Here is an example of that with dielectric (indices decrease from 1.52 to 1 in a linear order).