Improve description of spectral_bin_width.

[hugobuddel]

I'm not really happy about this but I guess it's acceptable as a band aid. Ideally, some more explanation should be included on what spectral_bin_width is and how it ought to be matched to the resolution of the source spectrum and the dispersion of the MICADO spectra. Personally, I make simulations doable by restricting to a detector window, mostly coincident with the central detector, but the notebooks should probably show the entire array. My laptop cannot do a full array simulation, and I haven't run any of the notebooks to compare the results.

• 4_scopesim_SPEC_pipe_galaxy.ipynb: I don't know what spectra are used here, but it's quite possible that their intrinsic line widths do not even require the full MICADO resolution, hence lowering spectral_bin_width should be fine.
• 6_basic_spectroscopy.ipynb: This uses a sky spectrum from skycalc at a spectral resolution of 5000, which doesn't quite reach what we expect for MICADO. I assume that this is fine as well.
• MICADO_line_lists.ipynb: That's the most tricky one as the lines should be unresolved, i.e. have very small intrinsic line widths. However, for the interpolation in Scopesim to work, the lines have to be presmoothed. For the purpose of this instructional notebook this ought to be explained, and it might be necessary to increase the presmoothing along with spectral_bin_width. I'd been meaning to look at the line_list function, anyway, so I'll keep the notebook in mind.

I don't object to merging for testing whether they run, but the notebooks should be refined a bit more after that.

Originally posted by @oczoske in #120 (comment)