Self-report measurement of segmentation, mimesis and perceived emotions in acousmatic electroacoustic music

Abstract

Acousmatic Electroacoustic Music is a form of art in which any sound is regarded to have the potential to be used as a basic unit to build a musical discourse. This study addresses the measurement of three important aspects of perception of Acousmatic Electroacoustic Music: Segmentation, which is the separation of the audio stream into units that could have meaning by themselves or that could be part of larger meaningful structures; Mimesis, which in this context is defined as the perceptual ability of identifying a possible and meaningful source that produced a sound and the action that produced it; Perception of Emotions, which shapes the meaning of perceived events. This document reports the development and adaptation of tools and techniques to perform self-report measurements of these three aspects and the use of them in perceptual experiments with human participants. The results of these experiments suggest that participants' strategies to segment sounds can correspond both to their acoustic features and to their meaning inferred by source and action. Participants' identification of sound source and action can greatly vary and narrative responses suggest that sounds trigger meaning rather than having meaning by themselves. Post-hoc and continuous measurements of perceived emotions were remarkably similar, suggesting that perceived emotions in discrete sounds are established constructs that might not change considerably when acting in context. Finally, it is shown that continuous measurement of Segmentation and Perceived Emotions is reliable only for sounds with a duration over 4 seconds, which leaves an open window for further investigation on the measurement of perception of shorter sonic events.