We present an experimental study of three calibration methods for spherical reference surfaces in Fizeau interferometry. The ball average, which relies on averaging measurements of a test ball surface over sufficiently many random rotations of the ball, is theoretically an absolute technique but can be very laborious. On the other hand, a recently introduced double-pass technique, comparing the two halves of the reference surface, is able to determine the point-symmetric contribution in as few as three measurements but does not detect the point-antisymmetric portion. Finally, the point-symmetric errors of the reference surface can be captured in a single, so-called “cat’s-eye” measurement. Our study tries to answer the question of which of the techniques is preferable in practice. We find that the answer depends on the required uncertainty, and it appears that the new double-pass technique represents a practical and reasonable trade-off between expediency and accuracy.
© 2010 Optical Society of America
Equations on this page are rendered with MathJax. Learn more.