An introduction to digital audio

Discuss some of the principles of digital audio and then outlines example areas that are currently topical within the industry. However, it is assumed that the principles of uniform sampling, uniform amplitude quantization and dither are understood and that when used together, realise a digital channel with analogue-like characteristics where the signal is distorted only by bandlimitation and additive noise. Central to a digital audio system are the processes of filtering and of noise shaping, where for example, oversampling and noise shaping can influence a system by enabling an interchange of complexity between the analogue and digital domains. This leads to alternative structures for digital-to-analogue and analogue-to-digital conversion where system related errors can be traded. Also, if these processes are combined with auditory data to form models of human hearing, then there is an opportunity to achieve optimum performance for a given data rate, which at one extreme can lead to a performance beyond the requirement for human hearing, or at the other extreme, reasonable audio quality with a low rate of data communication. This paper reviews some methods of data conversion that depends upon digital signal processing (DSP) and also considers aspects of loudspeaker systems to illustrate how DSP can radically affect an audio system