A definition of a loss factor is unique, some definitions are not

Noise control goals are often specified in terms of ratios of quadratic response quantities, among them a loss factor. Thus, the loss factor (η) is defined as the ratio of the external input power density (Πe) and the product of the stored energy density (E) and the frequency (ω), i.e., η=[Πe/(ωE)]. A unique definition of the loss factor demands that the dynamic system be enclosed so that (Πe) accounts for all the external input power density and (E) accounts for all the stored energy density generated by (Πe). Recently a number of acousticians in seeking noise control goals have defined the goals in terms of ratios that are assigned as loss factors. The definitions of these loss factors do not always conform to that uniquely defined loss factor. Those who maintain the conservative definition of a loss factor have been puzzled by reports of loss factors with values that exceed unity by several orders of magnitude. This paper attempts to examine a few examples of these less unique definitions of a loss factor and to show that a return to the unique definition is mandatory if the concept of a loss factor is not to be lost.