On the potential limitations of conventional sound metrics in quantifying perception of nonlinearly propagated noise

Abstract

The use of conventional metrics to quantify the perception of nonlinearly propagated noise has been studied. Gaussian noise waveforms have been numerically propagated both linearly and nonlinearly, and from the resulting waveforms, several metrics are calculated. These metrics are overall, A-, C-, and D-weighted sound pressure levels, perceived noise level, Stevens Mark VII perceived loudness, Zwicker loudness, and sharpness. Informal listening demonstrations indicate that perceived differences in annoyance between linearly and nonlinearly propagated waveforms are substantial. Because the metrics studied seem inadequate in representing the perceived differences, rigorous subjective testing is encouraged to properly quantify and understand these differences.

Examining the use of a time-varying loudness algorithm for quantifying characteristics of nonlinearly propagated noise (L)

“…(An example is the so-called "crackle" phenomenon, which has instead been described using the statistics of the time waveform 1 and more recently, the time waveform derivative. 2 ) The results of the Gee et al 3 study prompted further investigation to identify a metric that objectively quantifies the perceptual differences between the nonlinearly propagated signal and a signal with equivalent long-term spectrum, or, similarly, between a crackling waveform and one that has the same spectrum. After publication of the previous study, 3 it was suggested to the authors that a more suitable metric might be TVL, which incorporates both temporal and spectral features of loudness into the model.…”

“…2 ) The results of the Gee et al 3 study prompted further investigation to identify a metric that objectively quantifies the perceptual differences between the nonlinearly propagated signal and a signal with equivalent long-term spectrum, or, similarly, between a crackling waveform and one that has the same spectrum. After publication of the previous study, 3 it was suggested to the authors that a more suitable metric might be TVL, which incorporates both temporal and spectral features of loudness into the model. Thus, it was hypothesized that the model would respond more appropriately to the temporal features of the acoustic shocks.…”

“…In addition, because their average spectra were identical, none of the metrics distinguished between the nonlinearly propagated and rephased waveforms, although such differences are easily audible. 3 Such problems are not altogether uncommon in investigating the sound quality of high-amplitude jet noise, where some important qualitative characteristics are not detectable from the power spectrum of the signal alone. (An example is the so-called "crackle" phenomenon, which has instead been described using the statistics of the time waveform 1 and more recently, the time waveform derivative.…”

“…Previous work by Gee et al 3 investigated the response of stationary sound level, loudness, and annoyance metrics to a set of three waveforms that were included as multimedia content. ("Stationary" as used in this paper refers to use of the long-term average spectrum.)…”

10.1121/1.4984819.3

“…It is a startling staccato of cracks and bangs and its onomatope, 'crackle,' conveys a subjectively accurate impression." This sound quality-interesting to Ffowcs Williams because of its capacity to produce annoyance-cannot be identified by examination of the long-term spectral magnitude alone or any metric derived therefrom, as shown by Gee et al (2007b). It is instead "caused by groups of sharp compressions in association with gradual expansions" as noted by Ffowcs Williams et al (1975).…”