The use of floating-point arithmetic in digital signal processing is considered in this paper. First, an introduction to floating-point arithmetic based on the IEEE floating-point standards is given. Then, finite wordlength effects of these arithmetic systems are presented. Finally floating-point hardware is discussed. Special emphasis in this paper is on audio applications. As an example, amplitude distributions and signal-to-noise ratios of quantized music are considered, and audio standards using floatingpoint arithmetic are presented. 1784 0-7803-0593-0/92 $3.00 1992 IEEE S.H. Ardalan, "Floating-point error analysis of recursive least-squares and least-mean-squares adaptive filters," IEEE %ns.