S-Distortion

The Figaro routines described in this section originated in the need to straighten the distorted spectra produced by image tube detectors. However, even with CCD detectors-which do not suffer the geometrical distortion of the images tubes-some instruments, particularly échelle spectrographs, still produce curved spectra, and the techniques described here can be used to correct these as well. Indeed, nowadays the main application of these routines is to échelle data.

Instruments such as the 2D-Frutti and IPCS that use image intensifiers suffer from various distortions, in particular S-distortion. The effect-and the reason for the name-can be seen clearly by displaying any 2D-Frutti image of a point object, such as a star, on the display. Instead of being a perfectly horizontal line, the spectrum snakes across the image in the shape of a horizontal letter S.

Note that this distortion is actually a two-dimensional distortion-a picture of cartwheel taken through an image tube will show all the spokes bent into S shapes in a radially symmetric manner. However, the difference in pixel scale in the two dimensions for spectral detectors means that-to a first approximation-the distortion can be treated as though it were simply a vertical displacement in the data whose magnitude varies along the spectrum (and to a lesser extent, with position along the slit). In any case, the two dimensional distortion can be corrected by two orthogonal one dimensional corrections: the S-distortion correction described here is one, and the other can be performed as a side effect of a two-dimensional re-binning to a linear wavelength scale. It may be that such full two-dimensional `scrunching' is regarded as overkill-see the section on wavelengths for more details.

The process described here is a one-dimensional correction, in which data is re-binned in the direction perpendicular to the dispersion, in such a way as to straighten the spectra.