Update wavelength windows to use for stellar typing

We have been using multipel wavelength windows for stellar typing. These windows are chosen to avoid telluric absorptions and the blue edge of the FLUXSTD spectra, where the whitening is difficult.  We have been using the same windows for both LR and MR, but it clearly does not work for MR. See an example plot here. The shaded regions in the top panel are the masked regions and we use the white regions for typing. For instance, Halpha is on the edge of the b-arm and should probably be excluded from the typing.  We should use different windows for different resolutions. I would also like to take this opportunity to update the windows for LR a bit.