This question comes up from time to time. There are two issues to get past if you want to perform this analysis for a different system. One is that the Gaussian to top-hat analysis has to be done for different beam shapes. The other is that the analysis in the KBA article makes use of the radial symmetry of the system, so everything can be done with only radial sample rays. Using a line would require a fully 2-D analysis and subsequent optimization.

We don’t have any further examples like this, but certainly encourage any users who work one out to share with the larger community.

Thanks a lot for your reply Kevin.

For me it is not clear in which way the KBA article makes use of the radial symmetry in a way we can not use it for optimizing a line.

Let me try to explain my point of view. In the KBA article a “rotational symmetric” Gaussian beam is formed into a rotational symmetric top-hat. For analysis they use the Geometric Image Analysis (Cross X setting - or Cross Y, which leads to a identical result) where in case of a good optimization a rectangular profile is plotted. Now we can make the assumption, that only rays exactly in the Y-Z plane of the entrance pupil have an effect on the GIA Cross Y plot, since any other rays are refracted away from the Y-Z plane. According this assumption for the analysis it doesn´t makes a difference anymore whether a cylindrical or a rotation symmetric lens is uses, since the lens curvature in the Y-Z plane is the same for both.