Timbre Recognition and Source Identification

The ability to recognize many sounds in everyday soundscapes is a useful and impressive feature of auditory perception in which timbre likely plays a key role. This chapter discusses what is known of timbre in the context of sound source recognition. It first surveys the methodologies that have been used to characterize a listener’s ability to recognize sounds, and then examines the types of acoustic cues that could underlie the behavioral findings. In some studies, listeners were directly asked to recognize familiar sounds or versions of them that were truncated, filtered, or distorted by other resynthesis methods that preserved some cues but not others. In other studies, listeners were exposed to novel sounds and the build-up of cues over time or the learning of new cues was tracked. The evidence currently available raises an interesting debate that can be articulated around two qualitatively different hypotheses: are sounds recognized through distinctive features, unique to each sound category but of which there would need to be many to cover all recognized categories; or rather, are sounds recognized through a relatively small number of perceptual dimensions, in which different sounds have their own recognizable position?