Prediction of subjective response from objective measurements applied to walking sound

This paper discusses prediction of the subjective response to walking sound - also called drum sound - based on differences in objective measurements. 'Walking sound' refers to the sound heard when someone is walking in the same room as the listener. Walking sound has attracted interest in recent years, particularly due to an increased use of thin floating floor constructions, such as veneer or laminate flooring, which can produce loud and sharp walking sound. A paired comparison test was performed in laboratory where listeners were asked which of the walking sounds was most disturbing. The response was analysed using a modified Bradley and Terry model allowing ties. Various measures, such as loudness according to ISO 532B, were tested against the subjective response using linear regression. The difference in 10-percentile loudness, N-10, between two stimuli was shown to predict the subjective response better than, for example, A-weighted sound pressure level. A difference of about 8% in N-10 resulted in 50% of the subjects noticing a difference. The methodology used is applicable in situations when objective measures that have subjective counterparts are sought. Although the method is based on relative observations, an absolute ranking can be obtained by using a reference or a well-defined recording situation. (Less)