“…9 and earlier studies, while we were aware of a spatial-frequency-dependent loss of phase range, we did not model this as phase blurring, nor did we consider the related nonlinear effects on the diffraction pattern. 5,8,9 Instead, we attempted to minimize phase blurring by making the pixel spacing of the discretely sampled CGH's large with respect to the maximum phase slope anticipated. At the time, the major source of errors between the experimentally measured and the simulated results was assumed to be various point nonlinearitiese.g., the inaccurate setting of the mapping between input intensity and phase, variations in response across the SLM, and quantized phase levels.…”