Money can't buy happiness, but it sure makes living with misery a lot easier!
Norme 150mm MCT f/13, 31% CO Yellow Zone "People say I'm in denial. I disagree."
Quote:A Schmidt corrector has net zero optical power. It doesn't matter to any meaningful degree which way it's facing. All this plate is doing is introducing spherical aberration; this will be the same whether forward or backward.
Quote: In detail there is a difference in the emerging aberrated wavefront if the two surfaces differ, but is it of any meaningful degree? After all, the corrector is both a positive and a negative lens, where the lateral color induced in the inner and outer zones is reversed. The steepest slope, at the outer edge, is what, something like 1 wavelength per 3mm? That's a 'prismatic wedge' of angle a bit larger than 1 arcminute. Whether that sloped surface is forward or backward facing, the *difference* in the refracted angle tor light goung on to the primary is mighty small indeed.
Quote:There is no appreciable difference of any kind.
Quote:That's a 'prismatic wedge' of angle a bit larger than 1 arcminute.
Quote:Indeed for such small angles, the ray deviation is independent of the incident angle and it depends only on the refractive index of glass and the wedge angle
Quote: … It is correct too, that there is no appreciable lateral color error with the Schmidt corrector. The higher order coefficients are there to correct higher order terms of spherical aberration.
Quote: There is no appreciable difference of any kind.There is certainly a difference - just as there is in reversing any non-symmetric lens. But indeed *normally* it would not be noticeable or appreciable.But with a real lens from a real, mass produced system - there is not only a difference, but likely an appreciable one. That's why the corrector and the secondary have marks on them in the first place. This is not to say they aren't made well or won't perform well - it just says their orientations have hopefully been dialed in optimally when they were assembled to cancel the imperfections as much as possible - and should be kept that way.
Quote:Using wavefronts to understand the operation of the corrector plate is perfect, but I want to avoid any confusion here. A Schmidt corrector can, in principle, be shaped with a curve of the form z=Ar^2 + Br^4 + Cr^6…; however, most (all?) correctors have the form z=Ar^2 + Br^4. In a classic Schmidt camera with the stop at the CC, terms higher than r^4 are negligible and can be ignored. That may not necessarily be true for other configurations such as a SCT, but I’m not aware of any correctors that are made with higher order correction—it just gets too hard to manufacture in a cost effective way.
Quote:I view any realistic optical surface as having an arbitrary mixture of high order zernike's present - and any time the system is assembled and then final figuring performed to achieve an autocollimator null - any change of the component orientations - including reversal of supposedly neutral elements - could play a non-negligible role.
Quote:When you are trying to achieve a particular rms target in the overall wavefront, it makes perfect sense to allow the surface errors on the individual components to be relatively large as long as the overall wavefront is nulled in a cost and time-effective manner by retouching one of the surfaces in an autocollimator. But you have to keep the orientations the same in the final product - which normally isn't a problem unless people go willy-nilly yanking stuff out and then throwing it back in any which way.
Quote:Light refracts because of variation in glass thickness, and nothing else, and flipping the plate doesn't change that.
Quote: John, is not the Schmidt profile and resulting wave a higher order term? I'm not sure how such a profile corrects for primary SA alone, especially with a thicker edge. Both the edge and the center should be delayed relative to the 70% zone. To my eye, this does not seem to correct for lateral aberration especially for the paraxial rays which are largely masked anyway. It seems to better correct for transverse aberration from the edge. This seems to result in bringing marginal and best focus closer together but sending paraxial focus further from best focus. As I understand it, this delay from the edge is a higher order term.Anyway, it's a fascinating question and any difference is likely to be very small, I would think.
Quote:This implies polishing on a *spot*, without rotating the optic about its axis.
Quote:Vla, it's interesting how a reversed corrector normalizes the depth of the wavefront to what it would have been by giving the medial zone a bit of an advance as it refracts from the flat final surface. That's just hard to imagine. I see it the same way Frank does. The formed wavefront refracts a second time inducing some unintended curve. The glass depth perspective makes sense, but the angle of refraction is different.
Quote:the only difference is the slight difference in the ray height on the primary - resulting in the whooping 1/440 wave p-v difference in spherical correction...
Quote:It scares the heck out of me to contemplate what could happen were I to attempt to restrict work to just some isolated spot on one side of the optic.
Quote:A corrected system does not deliver a plane wavefront, but rather a spherical one. A wavefront on a converging light cone is a section of a sphere whose radius is at the focus. A plane wavefront never comes to a focus.
Quote:A Schmidt corrector has net zero optical power. It doesn't matter to any meaningful degree which way it's facing. All this plate is doing is introducing spherical aberration; this will be the same whether forward or backward. [/quoteGlenn is ABSOLUTLY correct, rotation can make a fairly big difference,