Comparison of techniques for extracting signals from a strong background

The useful data from experiments are sometimes superimposed on a large-amplitude, slowly changing continuum that masks the information of interest. Techniques for extracting the information are analyzed and compared. We show that the processing comprises a tradeoff between removal of the continuum and detectability, distortion, and the signal-to-noise ratio of structures. It is found that wide structures are the hardest of all to extract without distortion and with a good signal-to-noise ratio, and that for a given set of experimental parameters, there is a limit to the class of structures that may be extracted from the data.