Continuous Measurement of Spatio-temporal Impulse Responses

• Applicant: Professor Dr.-Ing. Sascha Spors
• Subject Area: Acoustics ; Electronic Semiconductors, Components and Circuits, Integrated Systems, Sensor Technology, Theoretical Electrical Engineering
• Term: from 2016 to 2020
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 318346346

• Final Report Year: 2020

Final Report Abstract

Impulse responses play an important role in virtual acoustics for the characterization of acoustic environments. They are commonly used for the auralization of existing environments, as for instance renowned concert halls. Impulse responses are captured from a sound source to various receiver positions in order to capture the spatio-temporal structure of a sound field. The measurement of spatially dense datasets of acoustic impulse responses is a tedious task. Traditionally these measurements are performed in a spatially sequential manner by measuring one position after each other. The duration of such measurements can be considerable. Dynamic system identification techniques provide an interesting alternative to sequential measurements. These techniques aim at identifying the time-variant impulse response from a source to a receiver in a partially or full dynamic scene. For instance, from a moving source to a static microphone or from a static source to a moving microphone. The latter lays the grounds to significantly lower the measurement duration when capturing impulse responses in a spatially dense grid. This project developed novel techniques for the identification of time-variant acoustic systems. We initially concentrated on the analysis of known single-input/single-output (SISO) techniques. It was shown that the dynamic system identification problem can be interpreted in terms of sampling and interpolation. The algorithms differ mainly in their inherent or explicit interpolation scheme. These indepth insights layed the grounds for the development of various improved algorithms. For instance, a multi-band technique which explicitly takes the typical length of room impulse responses for low/high frequencies into account to speed up the measurement. Furthermore, various extensions towards multiple-input/single-output (MISO) and multiple-input/multiple-output (MIMO) dynamic systems have been developed. For instance, an algorithm for the measurement of spatial impulse responses on the surface of a sphere by a constantly moving microphone and a decomposition into surface spherical harmonics. As applications, data-based binaural and sound field synthesis where considered. For the former an novel in-situ measurement technique for individual BRIRs was investigated. The research was accompanied by a detailed instrumental evaluation of the derived methods.

Publications

• Comparison of continuous measurement techniques for spatial room impulse responses. In European Signal Processing Conference, September 2016
  N. Hahn and S. Spors
  
• Continuous measurement of spatial room impulse responses using a non-uniformly moving microphone. In IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, pages 205–208, New Paltz, USA, October 2017
  N. Hahn and S. Spors
  
• Dynamic measurement of binaural room impulse responses using an optical tracking system. In International Conference on Spatial Audio, September 2017
  N. Hahn, W. Hahne, and S. Spors
  
• Continuous measurement of spatial room impulse responses on a sphere at discrete elevations. In 144th Convention of the Audio Engineering Society, May 2018
  N. Hahn and S. Spors
  
• Improved time-efficiency in continuous measurement of spatial room impulse responses by dual-band excitation. In German Annual Conference on Acoustics (DAGA), March 2018
  N. Hahn and S. Spors
  
• Simultaneous measurement of spatial room impulse responses from multiple sound sources using a continuously moving microphone. In European Signal Processing Conference, September 2018
  N. Hahn and S. Spors