- Introduction on spherical beamforming
- What is spherical beamforming?
- Why are spatial resolution and dynamic range important for spherical beamforming?
- Introducing the spherical beamforming 3D acoustic camera
- More information on spherical beamforming
Spherical beamforming: an introduction to interior sound source localization
Interior sound source localization is a method that lets you precisely localize the source of a sound in a vehicle’s interior, such as a car or truck cockpit, the wagon of a train or the passenger cabin of an airplane.
In cavities, the sound field is called a reflective field. It is a complex sound field. Taking the example of a car’s interior, the sound is there partially reflected at some locations, such as the windows, and absorbed again at other places, such as the lining on the roof.
Traditional methods for interior sound source localization are based on acoustic troubleshooting techniques. They are very time-consuming, and often do not give conclusive results. A new technique based on spherical beamforming provides convincing results in a shorter timeframe.
What is spherical beamforming?
The adequate technique for interior sound source localization is called spherical beamforming. Spherical beamforming uses a far-field beamforming technique, suited for free-field conditions, in a cavity where the sound field is reflective. Spherical beamforming does not make use of a flat, 2-dimensional array, but employs a spherical array. Therefore, the spherical beamforming technique helps identify the exact position of a sound source in the surrounding space.
Knowing that spherical beamforming requires a spherical array to scan surrounding sounds, one still has to make a further design choice by selecting an open or a closed sphere. Naturally, one might tend to select an open sphere, since 2D arrays that are used for traditional beamforming are also open arrays. However, verifiable experiments point out that, compared to closed spheres, open spheres suffer from worse spatial resolution in the lower frequencies and from lower dynamic range in the mid and high frequency range.
Why are spatial resolution and dynamic range important?
These are two important criteria to assess the validity of sound source localization methods:
- Spatial resolution is the ability to separate 2 sound sources. It is usually expressed in centimeters. It is the closest distance between two sources, where these still appear to be different sound sources and do not merge into a single source. The lower the spatial resolution, the better the source localization. In beamforming techniques, spatial resolution is proportional to sound wavelength. In the lower frequencies where sound has high wavelengths, one will obtain a higher (worse) spatial resolution.
- Dynamic range expresses the difference in sound level in dB between real sound sources and their surrounding mathematical artifacts that are inherent to beamforming techniques. The higher the dynamic range, the better the source localization. In beamforming techniques, the dynamic range is also proportional to the frequency: the lower the frequency, the higher the dynamic range.
Spatial resolution // Dynamic range
Introducing the 3D acoustic camera
The 3D acoustic camera is a closed sphere equipped with 36 microphones and a laser that automatically scans the interior’s geometry. The 3D acoustic camera enhances spherical beamforming by offering sound propagation on the actual geometry and improved spatial resolution and dynamic range with novel localization method
3D Acoustic camera on the left and scanned geometry on the right
After measuring sound with the 36-microphone spherical array, spherical beamforming allows sound back propagation to any geometry, including the real interior geometry.Verifiable experiments show that propagating sound to a location that is further or closer to the spherical array than the actual sound source leads to errors and lower dynamic range. Using an actual geometry of the interior rather than a virtual sphere has therefore clear advantages. The 3D acoustic camera is able to scan the vehicle’s interior in a fair amount of time. It then automatically creates a 3D model on which it performs accurate sound propagation, yielding easily interpretable results. It is also possible to project sound sources on existing geometry models, such as NASTRAN models that you will have loaded into the software.
Low frequency spherical beamforming analysis on the left (mirror), and mid-high frequency analysis on the right (B-pillar).
Novel localization method: identification
Next to the spherical beamforming method, the 3D acoustic camera introduces a novel technique called Identification. Identification is a so-called Inverse Method that provides a quantitative estimation of the source strength on the geometry. This technique significantly improves both spatial resolution and dynamic range, especially in the low frequency range where (spherical) beamforming has an intrinsic bad spatial resolution.
Low frequency analysis (500Hz) with spherical beamforming (left) and identification (right).
More information on spherical beamforming
Download the technical papers:
- “Acoustic source localisation inside cabins”, M. Robin and B. Béguet, Automotive and Railway Comfort congress, Le Mans, France, November 19 and 20, 2008.
- “Capturing a noise source in an interior enclosure”, F. Deblauwe and M. Robin, NAG/DAGA, Rotterdam, The Netherlands, March 2009.
- “Noise mapping and sound quantification in the space using spherical array”, L. Lamotte, M. Robin, and F. Deblauwe, Euronoise, Edinburgh, UK, October 2009.