Back to list

Auditory Perception of Spatial Extent in the Horizontal and Vertical Plane

Authors Weger, M.
Year 2016
Thesis Type Master's thesis
Topic Psychoacoustics
Abstract In this thesis, the extent to which spatial sound can be used to represent the horizontal and vertical spatial extent of auditory objects has been investigated. To this end, the perceived spatial extent of horizontally and vertically distributed sound sources has been measured as a function of two spatial extent synthesis algorithms, three stimulus types, and three different spatial distributions of individual loudspeakers. Two spatial extent synthesis algorithms have been compared to each other. The first was a time-based algorithm which used a spatial distribution of indi- vidual grains of a granular synthesis stimulus to represent a spatially extended sound source. The second was a frequency-based algorithm in which the different spectral components of a monophonic input signal were mapped to different spatial locations. Two of the stimuli were generated by a granular synthesis algorithm, on the on hand with sound material resulting in the impression of strong rain, on the other hand with Dirac impulses as grains, leading to a so-called impulse train stimulus. An additional white Gaussian noise stimulus was used, which was not processed by the time-based algorithm, but instead a spatial distribution of statistically independent noise sources was generated. In the experiment, participants performed both absolute judgments of spatial extent and pairwise comparisons between representations of different spatial extent. Compared to the literature, smaller targets were used, which are seen as being more practicable for applications in human computer interaction. Results indicate that the variations of perceived horizontal extent judgments varied systematically with physical extent for all stimuli used in the experiment. The time-based synthesis algorithm resulted in significantly larger judgments of spatial extent irrespective of orientation, compared to the frequency-based algorithm. Perception of vertical extent was not accurate and varied less systemati- cally with actual extent, while judgments largely underestimated the actual vertical extent of sound sources. Finally, the results of the absolute judgments agreed with virtually all information contained in the time-consuming pairwise comparisons.
Supervisors Marentakis, G., Eckel, G.