So path tracing is only used for the secondary rays and beyond?
Re:Path Tracing 3D Fractals
by Dave_Rosenthal on Monday January 19, 2015 @11:48AM
(#8912669)
I think think that's kind of a philosophical question. Wikipedia, for example, explains path tracing as integrating over incident light on a surface. You could either think of that as incident light at a point on a virtual film/sensor plane behind a lens or incident light on the first object you hit in the scene.
Such techniques are often generally known as "Global Illumination".
Often in physics based rendering you are attempting to estimate the BSDF (bi-directional scattering distribution function, often broken into T, transmission and R reflection) parts. This gets complicated by interactive media and subsurface scattering (i.e the photons go into a material and bounce around) but in the simple case of a surface interaction you are estimating an integral over solid angle for all light interactions. Monte Carlo path tracing is one way to do this estimation.
- Ray-casting is the general technique of using ray/object intersections (usually to solve a graphics problem). Could be via analytical solutions or not.
- Ray-marching refers to using an iterative technique for computing ray/object intersections. This type of strategy is used for these 3D fractals.
- Ray-tracing creates an image by casting rays for each pixel in the image (from the camera).
- Path-tracing is one of many random ray-tracing techniques. It averages across many random deep bounces of rays following light paths to accurately solve for lighting.
In this case, the ray-tracing casts rays against the 3D fractal via ray-marching and uses path-tracing to compute accurate lighting.