Mirrors, arbitrary shape, non-paraxial

[vitalychepel]

Dear Michael,

I am interested in reconstruction of images of a series of point source by a spherical mirror (further - by a more complex surface with rotational symmetry defined, say, by a spline), which is set inside a glass cylinder, all viewed through a camera lens. I was looking to your python rayoptics page (https://ray-optics.readthedocs.io/en/latest/index.html#) but could not find any mentioning of reflecting surfaces. Can the package handle them? Another detail - the ray traces are expected to be far from the paraxial conditions. Still manageable?

Appreciate your advice.
Best,
Vitaly Chepel

[mjhoptics]

Hello @vitalychepel,

Thank you for the question. Yes, ray-optics can handle reflecting surfaces, of potentially arbitrary profile. The ray-tracing is based on Snell's Law, without the paraxial approximation. As long as geometrical optics is a satisfactory approximation, ray-optics ray tracing should be fine.

I have a separate repo of jupyter notebooks that demonstrate different capabilities. Two notebooks in particular are worth noting:

Off-axis parabola notebook v3.0 Demonstrates spot diagram and geometrical PSF with a parabolic reflector.
focus tilt notebook Also uses a parabolic reflector but demonstrates wavefront analysis and diffraction PSF

ray-optics doesn't have an implementation of a spline surface currently. ray-optics does support:

• Conic
• Even (radial) Polynomial
• Radial Polynomial
• X and Y Aspheric Toroids

These are all part of the profiles module. To add a new surface profile, one would write a class that inherits from SurfaceProfile and implements functions that evaluate the surface profile function and gradient at a point, as well as a few other support functions. Given this information, the ray-optics ray trace can calculate ray/surface intersections and refraction/reflection in a general way.

Please feel free to ask further questions. Thanks!
MIke Hayford