Title: Psychoacoustic metrics for assessing the quality of automotive HMIs’ impulsive sounds
Abstract: Designing an automotive HumanMachine Interface (HMI) falls into a resource intensive and iterative prototyping process, where design inputs also emerge from subjectively evaluating the products. These inputs are defined from the outcome of the automaker's brand sense appraisal process: achieving a desired and holistic perception of product experience. This work aims at characterizing the acoustic feedback of mechanical push–buttons' prototypes, parametrizing their sound by an array of acoustic and psychoacoustic parameters. A review of literature published in automotive sound quality was performed, specially addressing impulsive sounds. From this review, a list of metrics for both temporal and spectral content stands as a contribution to methodologies addressing impulsive sounds. This research made use of a Design of Experiment (DOE), producing 18 minor variations of an automotive push–button, designed for affording different haptic feedback. Yet, the DOE impact on acoustic emissions remained unclear. An experimental procedure allowed the prototypes' sound characterization with acoustic and psychoacoustic metrics. Parametric data was analysed to assess feedback consistence within each design and sound differences among design variations. The work includes an interpretation of results with perceptual thresholds, including Just–Noticeable Differences (JNDs) for the hearing sense. This work also found discrepancies and reports different interpretations of results when comparing SPL with loudness models, and standardized loudness models with a psychoacoustic model specific for impulsive sounds. It was concluded that both sound duration and the nature of fast-decaying impulsive sounds explains the different results among loudness models. Despite numerical differences, the introduction of JNDs in the interpretation of results showed that the design variations of the prototypes almost do not introduce noticeable differences in the psychoacoustic feedback. Exceptions were found in some designs for a single metric, sound duration.