E Elements of Signal Theory
Pages 8

AnN-dimensional continuous signal can be sampled to produce anN-dimensional discrete signal. The sampling procedure involves the selection of the appropriate points in the signal domain and the recording of sample values at these locations. When the samples acquired are digitized, they are quantized to match the resolution of the discrete storage type used (e.g., 8, 16, 24 bits per sample). Quantization is the conversion of a continuous range of values to discrete levels (Figure E.1(b)). The continuous signal is sampled using a sampling pattern. The most common pattern of signal discretization is uniform sampling, where each sample is taken at regular intervals in the continuous domain (Figure E.1(a)).