Skip to main content
. Author manuscript; available in PMC: 2008 Nov 7.
Published in final edited form as: J Opt Soc Am A Opt Image Sci Vis. 2007 Feb;24(2):391–414. doi: 10.1364/josaa.24.000391

Fig. 5.

Fig. 5

Comparison of traditional LS estimation of wavefront coefficients from centroid data versus direct ML estimation from photodetector outputs. Parameters used in the simulation include: λ=680 nm; pupil diameter=24 μm×128=3072 μm; lenslet size=192 μm; CCD pixel size=24 μm; and focal length=9.9 mm. The wavefront was sampled at 1726 points across the pupil diameter, and 322 rows and columns of zeros were used to pad the wavefront function to a 2048×2048 array before computing the FFT. The markers represent the mean, and the error bars represent the standard deviation of the residual wavefront rms of the 50 estimations for each light level.