Abstract

In this talk I will discuss Raymarching, which is a technique used to render scenes. It involves taking steps along a ray until we reach an object. A basic implementation of this will be inefficient, so I will go over how we can optimise the renderer by using a distance estimator function. This will allow us to get a high level of precision with a small number of steps.

Raymarching excels when scenes can be built up using iterated reflection, scaling, and translation. Applying a sequence of these transformations repeatedly can be used to build up highly detailed fractals, and then render them in an efficient manner. This technique also allows effects such as soft shadows and ambient occlusion to be approximated very cheaply.

I will then cover Raymarching’s limitations and compare it to other rendering techniques.