How are complex signals defined (and generated)? The following sections show how a number of kinds of complex signal are defined and generated by means of basic arithmetic operations, and how transformations of signals into other representations are defined and generated.
A DSP can often be best explained as a `black box' which performs operations and transformations on two input signals, yielding one output signal. A function which which has both arguments and values in the time domain will be expressed as an operation. A function which has arguments in the time domain and values which are not in the time domain (e.g. correlations, spectral analysis) will be expressed as a transformation.
The operations may be quite simple, such as addition, or they may be extremely complicated, such as the PSOLA operation for frequency modulation.
The transformations may also be relatively simple, such as autocorrelation, or relatively complicated, such as Fourier analysis
.