Normalization and Fitting

A frequently used procedure, alternative to the correction for the chromatic response, is to normalise the continuum to unity by dividing the observed spectrum by a smooth approximation of its continuum. This approximation can be obtained either interactively with the graphic cursor, or from a table (command NORMALIZE/SPECTRUM) or by dividing the raw data by itself after filtering or smoothing. Median filtering and running average algorithms are well suited for this purpose (command FILTER). A spline fit can be made to the points defined interactively as well as to the filtered data (command CONVERT/TABLE). The MIDAS Fitting Package permits to go further and perform a more advanced modelling of the continuum.