To change pace from doing some experiments (not ended up in the wave system yet) I’ve added render passes to the render setup. I’ve also created a very simple Nuke multi-pass comp.
In order to not have fresnel calculations attenuate the reflection and refraction passes, I had to do some minor rework on the shader which now multiplies the fresnel calcs with the rendered reflection/refraction passes only for the beauty pass. I added a separate output for the fresnel output itself as a separate render pass so the correct attenuation can be done in the comp, be it with a lot more control.
While doing this I also included a initial fix for a shading problem where a dark green (with the current colors) line was visible inside the wave. This was caused by the internal fresnel attenuation inside the gather loop that calculates the complete reflection/refraction color of the wave. I’m currently investigating a way to control these ‘internal’ fresnel calcs in more detail than the global multiplier I’m using to reduce the fresnel value for the refraction in the current fix.
The attached image shows the different passes, I currently have in my output EXR:
Currently, I’m having some trouble with rendering out control passes for the wave zones as they show black where the lip-spray particles are visible which is not right for these specific passes. This might mean I have to add a very fast and simple extra render that outputs these zone-masks separately. Another approach could be to render the particles separately, which will make it hard to get correct reflections and shadowing as the interaction between the particles and wave shading would be lost.