Optimizing reflective systems using aspheric and freeform surfaces

The large number of surface types that are manufacturable today provide many available paths to the optical designer. It is not always clear which path provides the optimal balance of merit function complexity, convergence speed, and image quality after optimization. This paper shows that Q-type polynomials provide a strong benefit in both convergence speed and image quality after optimization when the footprint comprises a large portion of the area inside the normalization radius, but for off-axis systems with large decenters this benefit is weakened. Also, the sensitivity of several unobscured reflective designs is compared to show that although surface types with more degrees of freedom can reduce some sensitivities, the design form is the primary driver and a system that occupies a smaller volume will tend to have higher sensitivities to misalignments regardless of surface type.