Lecture 13: Global Illumination & Path Tracing (106)
dhruvchowdhary
In P.T., splitting light into direct and indirect parts helps us focus on what really affects how things look. By choosing where to put our effort, like which lights or reflections matter more, we get better images without wasting time on less important details. This smart split shows how breaking down a big problem into smaller pieces can make solving it both easier and give us nicer results.
rcorona
Are there heuristics one can use to dynamically vary the number of bounces one allows from certain objects?
For example, I think it was mentioned in lecture that objects made out of glass generally gain more information from a bounce than matte/diffuse objects.
Therefore, instead of globally performing n-bounces through a scene, could one have a max number of bounces allowed, but terminate certain paths earlier depending on the material properties of the objects the ray has bounced through?
In P.T., splitting light into direct and indirect parts helps us focus on what really affects how things look. By choosing where to put our effort, like which lights or reflections matter more, we get better images without wasting time on less important details. This smart split shows how breaking down a big problem into smaller pieces can make solving it both easier and give us nicer results.
Are there heuristics one can use to dynamically vary the number of bounces one allows from certain objects?
For example, I think it was mentioned in lecture that objects made out of glass generally gain more information from a bounce than matte/diffuse objects.
Therefore, instead of globally performing n-bounces through a scene, could one have a max number of bounces allowed, but terminate certain paths earlier depending on the material properties of the objects the ray has bounced through?