[Raytrace] Optimization

[Peter John Smith]

Richard,


> Adding a sixth wavelength to the second case (0.555, weight 1e-6),
> making it the primary, regenerating the default merit function to
> include the new wavelength, and reoptimizing changes the result by
> about 0.002%, so it is not the Polychromatic Strehl reporting function.
> (I had surmised this from cross-checking the results with the OSLO
> Polychromatic Strehl reporting.

Well, I thought it was a possibility but you now confirmed its not.

What weightings are you giving to the 0.482, 0.522, 0.571, 0.628, and 0.670
?

Maybe by shifting all wavelengths up a bit the design is more tolerant which
is a possibility.  Performance is usually harder to extend into the far blue
than far red because of the more rapidly changing index at the blue end.

Its the only thing I can think of.

Peter.