Website powered by

Bubbles

A simulation of a spectral thin-film interference BSDF in Blender.
It is not entirely accurate in the aspect that the angle of refraction is not accounted for when computing the depth at which the ray travels through the soap-water film. I may make another version sometime that is more accurate, and perhaps has a prettier scene than just a boring checker plane.
This also works by multisampling over an additional wavelength component (and hence adds another parameter/dimension to integrate over in the rendering equation), so it would be good to create a method to explicitly calculate the "final" converged color of light at a point, probably by approximating the integral over every wavelength in the wavelength-intensity distribution produced after wave interference.

Ah yes, thin film interference.

Ah yes, thin film interference.

Looking at only the reflected light.

Looking at only the reflected light.

Oh, my! Quite noisy running on my laptop even for a couple of hours. Rendering this nicely and quickly would require a better PC, and a better rendering method – and perhaps would best be done outside of Blender.

Oh, my! Quite noisy running on my laptop even for a couple of hours. Rendering this nicely and quickly would require a better PC, and a better rendering method – and perhaps would best be done outside of Blender.

Ok, so the samples converge nicely when we forget about shadows and caustics.

Ok, so the samples converge nicely when we forget about shadows and caustics.