This is the second of two papers introducing the main topics in digita
l signal processing. The subject of this paper is the estimation of th
e spectra of signals and both classical estimation methods and modern
model-based methods are discussed. These techniques are compared and r
esults are presented which demonstrate the superior resolution which i
s obtained with modern autoregressive spectral analysis methods. Examp
les are given of the application of these techniques in the areas of s
peech processing and biomedical signal analysis.