film is also modeled in this simulation, both the pre-distortion and

film distortion must be included. To represent the overall effect,

the pre-distortion must be the same as the one to be used in the

production of the CGH. This pre-distortion model needs only to

provide the accuracy desired in the CGH. However, the model of the

film distortion must be as accurate as possible to provide an accurate

simulation of the overall transfer function. Thus, while a third or

fifth order fit may suffice for the pre-distortion, the model for the

film distortion may require a higher order or a spline fit.

 The type of modulation utilized in the hologram is also modeled.

 Recall that the pre-emphasis or phase-only filtering of the

 information used to create the hologram is independent of the

 modulation technique incorporated in the physical hologram. That is,

 the hologram may be produced as an amplitude hologram with constant

 phase but variable transmission, or produced as a phase hologram with

 constant amplitude but variable phase. The proper choice of hologram

 is modeled in the simulation by mapping the hologram pattern to

 transmission or phase. This modified amplitude or phase pattern is

 used as the reference and is multiplied by the Fourier transform of

 the test image. The following steps of the correlation and analysis

 are identical with the ideal correlation simulation with the exception

 that the correlation peak is no longer centered on-axis.

 Figure 7.8 shows the continuous-tone hologram of a square,

produced by the simulation. The carrier frequency was maximized in

this case as half the sampling frequency to produce the greatest

separation of the various output terms. The Fourier transform shown

in Figure 7.2, a since function, is spatially modulated by a sinusoidal