Hello Ryoji,

The difference is in how the input beams are setup.

In the ray-based model, you have uniform pupil apodization, so the pupil is uniform and circular. When you take the FFT of a disk, you get the Bessel function shape we recognise as a 'diffraction pattern'.

In the POP model you are modeling the input beam as a Gaussian, and the Fourier transform of a Gaussian is still a Gaussian. A diffraction-limited Gaussian is still smooth and has no secondary structure.

So both are 'correct' but they are set up differently. If you want the Gaussina illumination in your ray-based moel, change the apodisation type from Uniform to Gaussian. If you want the uniform pupil in the POP analysis, change the Beam Definition to 'Top Hat' from its current Top Hat setting.

In this system, the FFT results are fine so the POP results will be the same but will take longer to calculate.

- Mark

Hi, I post my question here since it is similar to the one asked in this thread. I followed the recommendations given here, as well as in other forum threads or tutorials from Zemax. My optical model consists of a standard Ritchey-Chretien telescope such as f#=6 and fl~500 [mm]. However the PSF obtained using the "Huygens" method (as it is a non sequential system) and POP are quite different. In both cases, the illumination of the entrance pupil is uniform (="Top Flat Apodization"). In addition, I don't see any artifacts in both results that would point to a "bad choice of algorithm" or settings (e.g. under-sampling). Below, please find attached, the spot diagram, the diffraction pattern calculated with "Huygens", and the one calculated by POP, for the outermost position in the FOV (but this is also true for intermediate positions in the FOV). Do you have any comments to explain this discrepancy? Thank you for your help.

SPOT DIAGRAM

“Huygens” DIFFRACTION PATTERN:

“POP” DIFFRACTION PATTERN: