Author´s reply

The arguments in the note by Janssen fail to make the distinction between a class of time-frequency distributions with particular parameterizations of them. The approach used is to satisfy two conditions by making independent "good choices." We point out that all constraints must be satisfied jointly, not separately, and demonstrate how an infinite number of good choices can readily be made. In addition, we question the view that signal analysts are "willing to sacrifice quite a few of the mathematical properties" and we argue that these "mathematical properties" are the essence of the subject.