Very often, the extracted positive and negative beams show a certain amount of ripple. Seeing conditions usually result in slight changes in the amount of radiation getting through the slit. Given the sampling procedure used with CGS4, this simply means that integrations at different detector positions will display a slightly different number of counts. The result of interleaving the integrations to produce the observations is a ripple seen across the spectra. For a sampling of two the period of the ripple is two pixels. The amplitude of the period two ripple is determined by averaging the data from a window where the spectrum should be flat. Having the period and the amplitude of the ripple allows it to be determined. Dividing each beam by the respective ripple spectrum eliminates the ripple (see Figure 3.3). Having de-rippled the two beams they can be combined by subtracting the negative beam from the positive beam.
Figure 3.3: Top panel - Optimally extracted
spectrum of the positive beam from the spectral image seen in Figure 3.2. Bottom panel - Same spectrum as above
after de-rippling.