The ultimate criterion for acoustical design is given by the human listener. Therefore, the investigation of perceptual attributes and physiological effects evoked by sounds with respect to specific contexts is fundamental. We link them to signal properties by means of subjective assessments and auditory modeling. This knowledge in the area of applied psychoacoustics enables a successful application in sound engineering and sound design.
We develop new audio reproduction methods that provide an authentic impression of recorded acoustic scenes and environments. We use approaches that are perception driven and which take into account the acoustic properties of the reproduction room. To advance these methods, we study the fundamentals of the perception of sounds in rooms and develop models of room acoustical perception.
Human perception provides us with a means to reliably extract information about our environment even in highly complex situations. Different processing stages are involved in establishing this information extraction. Gaining knowledge about these stages is the focus of this research area. For example, in the cocktail party effect, we investigate the perceptual mechanisms that are the basis for our ability to segregate a sound source in a complex environment. In addition, we are interested to know how other modalities interact with auditory perception.
Sound reaching our ears conveys a wealth of information about our surroundings. For many applications extracting this information from an acoustical scene is highly useful. We develop algorithms and applications of computational auditory scene analysis (CASA), for localization of sources in the auditory scene, classification of sounds, and sound enhancement.