Posts

Update 6/5

Image
Reverted back to old code since we had partial mis working and a lot of random code everywhere. Attempt 2 at shooting rays through the object and taking 50% contribution from the object color. For some reason dragon appears to be emissive too but results look at bit better. 20% contribution here

Update 6/3/20

Image
An attempt at shooting ray straight through object if jade material intersected. After reading more on this topic, it seems like much of the fundamentals build off of being able to shoot rays into the object and processing them there. So this attempt was to try to simulate something glass-like where the ray would shoot through the object first. This would provide a better basis for being able to manipulate the ray after shooting it into the object.

Attempted addition of Diffusion Approximation(Jensen 2.1)

Image
We attempted to implement the equation in Jensen 2.1: The current issue is the calculation of dv and dr. They are defined as follows: dr - ||x - xr|| or distance between our point and the real source dv - ||x - xv|| or distance between our point and the virtual source Looking further through the paper, these seem to perhaps depend on us calculating the single-scattering term and utilizing the point from there into this calculation but we are unsure as of now.

Initial Images

Image
     This is our initial path tracer rendering the dragon. As you can see, subsurface scattering has not been implemented yet. Right now, it is rendering the Stanford dragon using brdf importance sampling, nee, ggx, and rr. This gives us the baseline so we can tell how subsurface scattering changes the dragon.