Traditional Representation of Audio Signal

Audio signals representation is a foundational pillar of audio signal processing, critically shaping the analysis, manipulation, and interpretation of data across a variety of applications. This chapter provides a systematic overview of foundational paradigms for audio signal representation, emphasizing the discretization processes that convert continuous-time domains into discrete-time implementations, as well as multidomain representations that encompass temporal, spectral, and joint time-frequency contexts. It explores the mathematical definitions and fundamental concepts underlying continuous-time signals and their conversion into discrete-time signals via sampling. The historical evolution of audio signal representation is presented to contextualize paradigm shifts from analog-era signal conditioning to digital-domain feature engineering, thereby establishing the theoretical foundations for subsequent exploration of representation learning using deep learning techniques.