Re: State of the art?

Bill Thomas <bthomas32000@...>

U of Arizona and Perkin Elmer have closed loops systems - UofA interferometer (straight fringes - Reference, computer generated hologram) - Infamous Perkin Elmer interferometer (straight fringes - null Two Mirror Offner Compensator - spacing issues did the Hubble mirror in).

have found, for my mirrors (22" f/4) the Bath interferometer images has many, many densely spaced fringes which were not useful, and as a result developed what i call Slit Image Test SIT a computer (no human judgment) fast (less than 30 minutes) repeatably accurate Foucault results which then input to FigureXp or Sixtests (which i prefer) gives the surface profile.?? i simply then work on the high spots.

the SIT is a derivative of the LWT (elegantly simple).?? but with the LWT could not get repeatable results which i attribute to room temperature changes while taken the measurements.?? as a result developed the SIT which takes an image of the returning zone rays and the process that image with free ImageJ which automatically determines the LATERAL position of each of the zone rays to a 3rd of a pixel or better - ImageJ computes the center of gravity of the recorded ray photons - ray image is bucketS of recorded photonS - the LWT in turn is a derivative of the Caustic, but being mask tests measuring a diameter strip and can't detect or deal with astigmatism.

when the image and light source are in the same plane the SIT equation is just a factor of 2 greater (just as the fixed Foucault is approximately 2 times greater than the moving Foucault) than the simple LWT equations.?? but when the image plane and source are not in the same plane (as with a digital camera) then need the SIT equations (non linear algebraic equations?? using Excels Solver for solution).???? the SIT equations are exact, not an approximation.

this paper i think might have the answer for us ATM'ers the Shark Hartmann
but instead of using a Shark Hartmann wave front sensor (same as used in adaptive optics) would use a Hartmann mask.?? but then instead of trying to numerically solve a 2d partial differential equations (which is the BIG draw back to the Hartmann) instead as this paper describes fit the the terms of the Zernike polynomial - doing a least squares fit instead of trying to numerically solve a 2d partial differential equations.

On 11/12/2019 11:58 PM, Rien wrote:
Thank you Bill, that was exactly the kind of information I was looking for!

Join to automatically receive all group messages.