Using differential ray tracing in stray light analysis

This paper describes differential ray tracing and shows how it can be used to great advantage in stray light analysis. Tracing a differential ray from a source to a target yields a set of derivatives that contain a complete first-order description of a ray bundle surrounding the ray being traced. These derivatives provide the information needed for aiming rays and transforming a sample area on a target surface into a solid angle seen by the source. By using targeted differential rays, we eliminate the need for defining importance curves for generating scattered rays. Convergence is accelerated, and the resulting irradiance distributions end up smoother than what one usually obtains with the traditional Monte Carlo approach. This paper also shows how the derivatives from a single ray can be used to define and propagate Gaussian beams without the need for secondary rays.