Sorting through the lore of phase mask options: performance measures and practical commercial designs

Abstract

Phase masks are needed in Fourier-transform holographic data storage systems (HDS) to reduce the range of light intensities found in the Fourier plane. The range of light intensities must match the dynamic range of the holographic storage medium and of the full HDS system. Descriptions, mathematical models, and tests of a variety of phase mask types have been reported in the literature: pixelated phase masks, non-pixelated phase masks, and axicons. Lacking, however, has been a systematic way of comparing the relative merits of phase mask types in order to make sound choices. To address this problem, performance criteria are proposed for both the Fourier plane and for the output image plane (e.g. the margin by which 1’s can be distinguished from 0’s). The criteria are useful both for comparisons and for design optimization. A new numerical model has been developed enabling quantitative comparisons to be made between the predicted performance of the various phase mask types. The model reported here enables more extensive investigations than could be carried out with previously reported models, including investigation of systems in which multiple bits of data are encoded by each pixel using light intensity modulation. The viability of using non-pixelated phase masks integrated with spatial light modulators is also examined. The use of non-pixelated (continuous random) phase masks instead of the more common pixelated phase masks would eliminate the need for costly precision lateral alignment, and integration eliminates the need for precise positioning in an image plane. These advantages would enable smaller, cheaper, high performance HDS optical systems.