Optimizing the Rendering of Hair: Volume and Geometry Shaders
(Author's Note: this is not really a tutorial, just my notes from a presentation I did on the topic at Siggraph.) Explanation:
In a scene involving hair it is often useful to divide up the rendering into separate passes based on the scene contents. Often backgrounds, set elements, characteers and special effects should be rendered seperately and then composited later for the final shot. This method allows creativity at composite time, and segments problems into manageable sets. It also allows for more efficient rendering of large scenes. Hair rendering may present special efficiency and optimization considerations. In this sort presentation we will examine the impact of various Hair rendering settings in XSI Orleans.
This paper will not discuss hair creation and grooming Ð these aspects have been covered in the Hair seminars and are better known.

The scene we will use is a cartoon scene using a hybrid render style: lambert shaded materials, cartoon inked characters, illuminated primarily with Final Gathering. Francis the bad clown is riding his pogo stick past the clown wagon, against the backdrop of the circus. He already has dynamic squash and stretch cartoon style effects on his body, we need now to add some good cartoon hair, with dynamiccs to finish the effect. The scene is collected into layers to speed interactivity.