A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Quality assessment of spherical microphone array auralizations
The thesis documents a scientific study on quality assessment and quality prediction in Virtual Acoustic Environments (VAEs) based on spherical microphone array data, using binaural synthesis for reproduction. In the experiments, predictive modeling is applied to estimate the influence of the array on the reproduction quality by relating the data derived in perceptual experiments to the output of an auditory model. The experiments adress various aspects of the array considered relevant in auralization applications: the influence of system errors as well as the influence of the array configuration employed. The system errors comprise spatial aliasing, measurement noise, and microphone positioning errors while the array configuration is represented by the sound field order in terms of spherical harmonics, defining the spatial resolution of the array. Based on array simulations, the experimental data comprise free-field sound fields and two shoe-box shaped rooms, one with weak and another with strong reverberation. Ten audio signals served as test material, e.g., orchestral/pop music, male/female singing voice or single instruments such as castanets. In the perceptual experiments, quantitative methods are used to evaluate the impact of system errors while a descriptive analysis assesses the array configuration using two quality factors for attribution: Apparent Source Width (ASW) and Listener Envelopment (LEV). Both are quality measures commonly used in concert hall acoustics to describe the spaciousness of a room. The results from the perceptual experiments are subsequently related to the technical data derived from the auditory model in order to build, train, and evaluate a variety of predictive models. Based on classification and regression approaches, these models are applied and investigated for automated quality assessment in order to identify and categorize system errors as well as to estimate their perceptual strength. Moreover, the models allow to predict the array’s influence on ASW and LEV perception and ...
Quality assessment of spherical microphone array auralizations
The thesis documents a scientific study on quality assessment and quality prediction in Virtual Acoustic Environments (VAEs) based on spherical microphone array data, using binaural synthesis for reproduction. In the experiments, predictive modeling is applied to estimate the influence of the array on the reproduction quality by relating the data derived in perceptual experiments to the output of an auditory model. The experiments adress various aspects of the array considered relevant in auralization applications: the influence of system errors as well as the influence of the array configuration employed. The system errors comprise spatial aliasing, measurement noise, and microphone positioning errors while the array configuration is represented by the sound field order in terms of spherical harmonics, defining the spatial resolution of the array. Based on array simulations, the experimental data comprise free-field sound fields and two shoe-box shaped rooms, one with weak and another with strong reverberation. Ten audio signals served as test material, e.g., orchestral/pop music, male/female singing voice or single instruments such as castanets. In the perceptual experiments, quantitative methods are used to evaluate the impact of system errors while a descriptive analysis assesses the array configuration using two quality factors for attribution: Apparent Source Width (ASW) and Listener Envelopment (LEV). Both are quality measures commonly used in concert hall acoustics to describe the spaciousness of a room. The results from the perceptual experiments are subsequently related to the technical data derived from the auditory model in order to build, train, and evaluate a variety of predictive models. Based on classification and regression approaches, these models are applied and investigated for automated quality assessment in order to identify and categorize system errors as well as to estimate their perceptual strength. Moreover, the models allow to predict the array’s influence on ASW and LEV perception and ...
Quality assessment of spherical microphone array auralizations
Nowak, Johannes (author) / Brandenburg, Karlheinz / Rafaely, Boaz / Melchior, Frank
2019-07-18
Theses
Electronic Resource
English
How dry does the recording for auralizations need to be?
| British Library Conference Proceedings | 2008
Predicting an individual’s preference in concert halls from measurement-based auralizations
| TIBKAT | 2019
Study of Spherical Near-Field Acoustic Holography with Rigid Spherical Microphone Array
| British Library Online Contents | 2013
| SAGE Publications | 2022