The role of spectrum in subjective interpretation of speech privacy estimates: an analysis of prominent metrics

Abstract

Categorization, acceptable-level and reverberation-time schemes are prevalent throughout standards and building codes; however, these strategies refer to ‘objective’ metrics that quantify the performance of particular facets (e.g., full- or partial-height partition assemblies, furniture, surface finishings) of the acoustical environment and, therefore, do not offer insight into occupants’ perception of the real—and dynamic—acoustic conditions within a space. Consequently, newer acoustical frameworks are turning to psychoacoustic metrics—tools that can be used to define and assess occupants’ needs and expectations. Because it is a strong indicator of acoustical satisfaction, and its methods are well-defined and standardized, speech privacy is being promoted as the foundational pillar for designing and commissioning spaces that are better able to support health and wellbeing. Exploring the differences and defining the limitations of current speech privacy metrics presents opportunities to further our understanding of these tools. In this investigation, different metrics are used to assess speech privacy at locations throughout an open-plan office. ‘Spectrum’ is found to be the reason for differences in their subjective (i.e., in relation to the occupant) interpretation of results. Given architecture’s impact on the magnitude and spatial heterogeneity of the spectral variability of sound within the built environment, as long as the methodological differences between metrics persist, so too will the risk of disagreement.