The implementation in #2516 suggests that perhaps a mid-level efficiency tool could be added to simtools/sim_telarray as well. The description of the features of the suggested tool are best summed up in the e-mail excerpt from @kbernloehr below.
I think there is a mid-level efficiency tool missing, something between
the super quick and dirty testeff and the full simulation, like
deterministic ray-tracing (without random scattering and random
mirror mis-alignment), then throwing light along 10k optical paths or so
into it (on some sort of square or hex grid pattern perhaps), evaluate the
mirror (segment) hits together with reflection angles, and the incidence
angle on the camera. Keeping the list of angles, evaluation for different
wavelengths does not need re-doing the ray-tracing, just evaluating
the reflectivity in a different spot of the 2D table for reflection leading
to a camera hit. Multiply efficiency factors along each path, add up
all the paths. No pixelation and no gaps and no random per-pixel QE/PDE
on the camera end of it. No p.e. and signal traces. But with the actual
mirror geometry and the actual efficiency tables. By re-using parts from
sim_telarray, it should be easy to use the same set of configuration
lines (or config files to be pre-processed) like in the 'real' simulation.
(And ignore most of those parameters.) What do you think?
If such a tool could be added entirely in simtools with not too much "human" effort, then perhaps it is worth considering.
The implementation in #2516 suggests that perhaps a mid-level efficiency tool could be added to simtools/sim_telarray as well. The description of the features of the suggested tool are best summed up in the e-mail excerpt from @kbernloehr below.
If such a tool could be added entirely in simtools with not too much "human" effort, then perhaps it is worth considering.