This is an account of 8 years of experience with the same computer-aided methods (IRIS program; http://128.119.70.193/Lab/IRIS.html). The analysis of dynamical mechanical data consists of two main steps, conversion from the frequency to the time domain and check for consistency. Both are equally important. The first step is pretty much consolidated and some important progress can be reported with the second step which, however, has not received the attention deserved. The quality of the input data determines the limits of the calculated results. The analysis starts out with the postulate that there exists a continuous spectrum H(lambda) which then can be expressed in N discrete modes to express G＇, G＇＇ data. The ＇sampling frequency＇, N per decade, of the discrete representation cannot exceed a value of about 1.5-2 because of the noise in the data. To conclude the analysis, the discrete modes are converted into a continuous spectrum. The method of data analysis is outlined and characteristics of the solution are explored.