Contribution
Acoular Workshop: Beamforming in the time domain with moving sources
* Presenting author
Abstract:
The open-source Python library Acoular provides a comprehensive framework for the analysis and simulation of microphone array data. It can be used for the investigation of moving sound sources, such as those encountered in car and train pass-by events or aircraft flyover measurements. This paper demonstrates how to use Acoular to effectively analyze such cases, highlighting the versatility of its various functionalities.Time-domain beamforming with a moving focus is employed, defined by a trajectory specifying the grid points at which sound sources are reconstructed. The approach accounts for key factors, such as convective amplification and the Doppler effect, which are essential for accurate source reconstruction. Advanced deconvolution methods are incorporated to mitigate the influence of the point spread function. This facilitates quantitative source identification through spatial integration. Band-pass filters enable the separation of sources in the frequency domain, whereas caching mechanisms optimize the process by saving partial results and preventing redundant recalculations.The overall processing workflow in Acoular is initialized as an evaluation chain, in which each processing step constitutes a link in the chain, allowing for customization to meet the requirements of specific applications. The design and implementation of such a tailored evaluation chain are demonstrated.