Abstract

High-density optical disk data storage requires high numerical aperture objectives to achieve a minimum spot size at focus (N.A. & 0.4). At these numerical apertures the standard approximations of the classical diffraction theory fail, and it becomes necessary to introduce corrections for the neglected factors. In addition, the classical theory is a scalar theory which must be corrected for the effects of polarization. We show, starting from the fundamental postulates of diffraction theory, that it is possible to obtain exact formulas for the distribution of light at and near the focus of high numerical aperture objectives. This is different from previous work^1–3 in that our computations require only a sequence of 2-D fast Fourier transforms. We show that when the incident beam is linearly polarized, one can compute the field distribution in a plane parallel to the object plane with either three (Fraunhofer diffraction), four (Fresnel diffraction), or five (high N.A. tens action) fast Fourier transforms. The technique has been used in computing the distribution of light at various cross sections in the focusing/tracking arm of a magne-tooptical disk system. We discuss the relationship between groove structure (used for tracking) and astigmatic optics (used for focusing) and their combined effect on the error signals. Finally, we show that the same technique can be applied to analyzing the frequency response of the magne-tooptic readout channel. (12 min)

© 1986 Optical Society of America