Reassessing the theoretical accuracy of the crossed-cylinder aberroscope technique

We replicated results of the theoretical accuracy of the Howland crossed-cy linder aberroscope for measuring the monochromatic aberration of the human eye [J. Opt. Sec. Am. A 15, 2477 1998]. Two Simple modifications to the pre vious mathematical analysis techniques, namely, the use of a new method for estimating the size of the unaberrated retinal grid element length and the correction of the aberroscope grid spacing size for magnification differen ces introduced during predistortion, almost completely eliminate the errors introduced into 4th-order wave-front-aberration coefficients for a centere d predistorted aberroscope grid over a large range of back vertex distances . The use of the crossed cylinders in the aberroscope introduces only small amounts of wave-front aberration to the measurements of the schematic eye. Wave-front aberrations above the 4th order in a Taylor polynomial expressi on can introduce errors that vary as a function of pupil area. Results prod uced by orthogonal polynomial analysis confirm the least-squares analysis r esults. We believe that the Howland crossed-cylinder aberroscope can be use d in its objective form to make accurate measurements of at least the 4th-o rder components of the wave-front aberration of the human eye. We note that cre must be exercised both in the use of the equipment and in the analysis of the results to ensure that accuracy is maintained. (C) 1999 Optical Soc iety of America [S0740-3232(99)01210-7].