I'll talk about ongoing efforts to model optical non-idealities in CMB experiments, with a focus on generating simulated timestreams and maps. After introducing the concept of beam convolution, I'll discuss how we can model beams using optics software. I will then show a few examples of non-idealities we have been studying in Stockholm with collaborators by inserting these beams into various simulated experiments. In particular, I'll talk about the impact of HWP non-idealities for satellite missions and about my attempts to model ground pickup for ground-based and balloon-borne telescopes.