What do your footsteps sound like? An investigation on interactive footstep sounds adjustment

This paper presents an experiment where participants were asked to adjust, while walking, the spectral content and the amplitude of synthetic footstep sounds in order to match the sounds of their own footsteps. The sounds were interactively generated by means of a shoe-based system capable of tracking footfalls and delivering real-time auditory feedback via headphones. Results allowed identification of the mean value and the range of variation of spectral centroid and peak level of footstep sounds simulating various combinations of shoe type and ground material. Results showed that the effect of ground material on centroid and peak level depended on the type of shoe. Similarly, the effect of shoe type on the two variables depended on the type of ground material. In particular, participants produced greater amplitudes for hard sole shoes than for soft sole shoes in presence of solid surfaces, while similar amplitudes for both types of shoes were found for aggregate, hybrids, and liquids. No significant correlations were found between each of the two acoustic features and participants’ body size. This result might be explained by the fact that while adjusting the sounds participants did not primarily focus on the acoustic rendering of their body. In addition, no significant differences were found between the values of the two acoustic features selected by the experimenters and those adjusted by participants. This result can therefore be considered as a measure of the goodness of the design choices to synthesize the involved footstep sounds for a generic walker. More importantly, this study showed that the relationships between the ground-shoes combinations are not changed when participants are actively walking. This represents the first active listening confirmation of this result, which had previously only been shown in passive listening studies. The results of this research can be used to design ecologically-valid auditory rendering of foot-floor interactions in virtual environments.This work was supported partly by a grant from the Danish Council for Independent Research awarded to Luca Turchet (Grant No. 12-131985), and partly by a grant from the ESRC awarded to Ana Tajadura-Jiménez (Grant No. ES/K001477/1)