To get a useful spectrograph, it would be good to know the requirements in advance. But finding out the requirements appeared to be difficult. For the wanted spray result, we could model the scattering with mie-scattering models for spherical particles. But when particles stick together, they form a new geometry, for which there are no known scattering models. One plan, right now, is to use models of scattering for ellipsoidal particles, to give an approximation of several particles stuck together. The animation below, shows the scattering intensities in different directions (light is incident from the left) and for 5 different wavelengths (the five pictures in the animation). The red, green and blue shows the different polarization components. One good thing is that the intensity for different wavelength shifts, but another thing that complicates matters, is that the intensity changes rapidly with angle. We thus also need to get an accurate value of the scattering angle of a certain spectrum.