The pdf mentions using a cfl to calibrate the position -> wavelength mapping, but wouldn't the peaks depend on the particular phosphors used for that lamp?