# Thread: Optimizing an off-axis reflector system?

1. ## Optimizing an off-axis reflector system?

Hi Forum.

I have a sticky situation on my hands.

I'm trying to find a reflector which would focus an image plane onto another image plane. Seems simple enough, right?

But, what I noticed is, modern softwares (Zemax, Code V) really aren't designed to do this.

The only way you can define coordinate systems is in relation to the other elements, which is fine, when the elements are reflecting or refracting on their optical axis'. But, as soon as you decenter a surface, changing the geometry of anything knocks the reflected rays out of alignment, which means that, you can't really optimize, eg, zernike polynomials using a cost function. If you try changing the surface too much, the rays may move off the detector, which would cause the optimizer to fail.

I'd think to succeed, you'd have to optimize to generate the surface in chief-ray space [from (x,y) some point on surface != (0,0)], not pariaxial coordinate space (x, y) == (0,0) :-(

A few years ago I tried to write something which would design an arbitrary reflector to bounce rays from one surface to another. I found is the system is under-constrained if you try say, gradient descent, and that, a method I tried which was something akin to drawing a mirror surfel, then calculating the next mirror surfel based on the solid angle shift of the next ray, also failed. It didn't really converge, and solutions were radically different based on the step size.

I'm a bit stumped here.

It seems like most people's solutions involve just picking some "approximate" surface, then working backward to generate optics from that surface. But, that surely doesn't lead to the best system now, doesn't it?

2. ## Re: Optimizing an off-axis reflector system?

I've no idea what you're talking about.
But I will allow the general forum to have a go at your problem.
Ive always considered that changing a plane with a curved mirror to be at best awkward at worst a waste of time

3. ## Re: Optimizing an off-axis reflector system?

I'm trying to find a reflector which would focus an image plane onto another image plane.
What does that even mean?

If you could explain what the application is, we might have a better idea of what solutions might work.

4. ## Re: Optimizing an off-axis reflector system?

Is this what you're trying to do? https://en.wikipedia.org/wiki/Reflec...Schiefspiegler

5. ## Re: Optimizing an off-axis reflector system?

Is this what you're trying?

Welcome To DGM Optics

6. ## Re: Optimizing an off-axis reflector system?

Originally Posted by KeithBC
What does that even mean?

If you could explain what the application is, we might have a better idea of what solutions might work.
This is very close!

In summary, let's say you have a plane, or aperture. For example, a photograph.

You'd like to reflect this photograph, such that its image is projected onto another surface, or a point even. What is the reflector which would preform this function?

18009972_1453235244740893_44608317_n.png

Something like this is a system which has corrective optics for a conic surface. Can the number of these optics be reduced, with an arbitrarty, non-conic surface?

18034892_1453238654740552_882302861_n.png

My simulation / attempt at this seems to be misguided, as the surface is assymetric given symmetric inputs, and, discontinuous. (notice the non-smooth upper edge).

7. ## Re: Optimizing an off-axis reflector system?

Ah, then, what you want is sufficiently complex that you should pay an optician a consulting fee to help with your design, IMO.

8. ## Re: Optimizing an off-axis reflector system?

This unfortunately doesn't seem to be something well-studied in literature, and I haven't yet met an optician who has experience in these kind of computational optics :-/

9. ## Re: Optimizing an off-axis reflector system?

Have you tried OSLO? I have installed it but never used it. Looking at the menus there are ways to optimize and define an error function. It can handle reflective surfaces and lenses.

It is easy to ask simple questions with incredibly complex solutions, it is not a slam dunk. What have you programmed top help solve this problem?

10. ## Re: Optimizing an off-axis reflector system?

It may be a problem without a solution. I can't think of any reason to assume that a single reflective surface can produce a flat image of a flat object. A parabola would be a good first approximation, but the resulting focal 'plane' will not be flat. The best solution may be a primary reflector with corrective lenses.

Page 1 of 2 12 Last