Best to check the documentation and the references therein. POP uses both Fraunhofer (otherwise known as FFT) diffraction and Huygens (also FFT-based, but not referred to as such) to compute beam diffraction in any space irrespective of whether it is imaging or not. I can't speak for what CodeV does, but my guess is that their BSP is similar.

The Knowledge Base has lots of good articles on POP and its applicability. Remember to use POP only for those cases where rays don't describe the optical physics well. When you use it for cases that rays handle well, you just spend a much longer time calculating the results that rays gave you almost instantaneously.

THT,

- Mark

Code V sequential raytracing implements some FFT / diffraction as in Zemax (FFT MTF, Huygens). I don't think Code V has anything really similar to POP which is closer to what you would encounter is e.g. VirtualLab. On the other hand, Code V BSP uses Gausslets, it is similar to techniques found in ASAP and FRED, but not available in Zemax.

Approximately, BSP is best at:


- speed & memory


- large incidence angles (but locally low divergence)


- large smooth objects

and POP is best at:


- complicated apertures and obscurations


- profiles with higher irregular gradients (lens arrays, diffusers)


 

A few years back, POP could do multimode beams but BSP could not (it's more a software limit that a fundamental limit, they could add it).

Thank you for this information. Very interesting. Let us know if you’d like further details. I think the limitations of POP are listed in the help files under The Analyze Tab (sequential ui mode) > Laser and Fibers Group > About Physical Optics Propagation > Algorithm Assumptions.