Non-parallel rays for source broadening

[lemmatum]

I've been making a simulation for a Johann spectrometer based on the example in the repository, and I noticed that the rays from the GeometricSource to the crystal are parallel. Like so:

When I reduce this to a simple example of a GeometricSource and a detector it confirms that the rays do not diverge, and the size of the source is simply imprinted directly onto the detector

Is there a way to simulate diverging (non-parallel) rays in order to simulate line broadening due to the finite size of the source?