Detailseite
Projekt
Models of room acoustical perception
Antragsteller
Professor Dr. Steven van de Par
Fachliche Zuordnung
Akustik
Förderung
Förderung von 2011 bis 2018
Projektkennung
Deutsche Forschungsgemeinschaft (DFG) - Projektnummer 174776315
This project will develop models that can predict the perception of room acoustical perceptual attributes and just noticeable differences in these attributes. By developing such models, a better understanding will be gained about how specific physical room acoustical properties contribute to perception. This knowledge can be used for supporting the development of more effective room acoustical simulation methods.Special focus will be on spatial and spectral aspects of room acoustical perception and simulation. Spatial properties of room acoustics are known to be important for the judged quality of room acoustics, spectral properties can be affected by simulation methods which can impair the quality of the room acoustical simulation.For spatial perception our perceptual model will be further developed to account for room acoustical perception of real rooms, simulated rooms, and real rooms with artificially manipulated room acoustics. This broad range of evaluation conditions will provide a critical test for the model and based on the manipulated room acoustical signals we expect to get a better understanding of the technical signal parameters that contribute to the perception of spatial room acoustics. It will also be investigated whether the perceived plausibility of a room acoustical simulation can be predicted from statistical models of binaural cues. According to this hypothesis, a room is perceived as spatially plausible when the observed sets of binaural cues are likely to occur in realistic acoustic settings.For the perception of spectral properties of room acoustical simulation we will first make a characterization of the types and magnitudes of spectral distortions that occur due to various room acoustical simulation methods. We will then do extensive listening tests regarding the perceptibility of such distortions and develop a model that predicts whether a specific reproduction method will be perceived to be authentic (i.e. very similar to the original room that is simulated). In addition, it will be investigated what approaches could be used to model the perceived plausibility of a room acoustical simulation when spectral distortions are present. A room simulation is judged to be plausible when it is perceived to be a real room without comparing it to a reference. To conclude, the effect of spectral distortions on spatial perception will be investigated. This is motivated by the fact that different spectral distortions at both ears can lead to distortions of the spatial percept also. The role of head movements on room acoustical perception will be investigated. When moving our head, we typically perceive a constant acoustic scene without changes in the spectral coloration of the sound sources in the scene. The perceptual mechanisms that support this perceptual constancy will be investigated.
DFG-Verfahren
Forschungsgruppen
