[Paper Review] Indoor Sonic Boom Reproduction Using ANC

Nicolas Épain, Emmanuel Friot+1 more|arXiv (Cornell University)|Jun 14, 2005

Aerodynamics and Acoustics in Jet Flows5 references4 citations

TL;DR

This paper proposes using Boundary Surface Control (BSC) with Active Noise Control (ANC) to reproduce low-frequency sonic booms in a controlled indoor environment, enabling accurate, head-motion-insensitive psychoacoustic testing. Simulations show that reproducing only the incident wave fails at low frequencies; instead, the full scattered field—including head diffraction—must be captured for accurate reproduction, with optimal microphone placement within 50 cm of the listener to minimize error.

ABSTRACT

The European programs for development of supersonic air-flights involve new studies on the human perception of sonic boom. Because this noise includes high-level components at very low-frequency, the usual psycho-acoustic tests with headphones are not relevant; instead, the original sound-field can be reproduced with many loudspeakers in a small room, but the loudspeakers must be controlled for an accurate reproduction, both in time and space, in an area large enough to enclose a listener's head. In this paper, Active Noise Control is applied to sonic boom reproduction through Boundary Surface Control (as named by S.Ise) of the acoustic pressure around a listener. A small room was built at LMA with sixteen powerful low-frequency acoustic sources in the walls. Frequency and time-domain numerical simulations of sonic boom reproduction in this room are given, including a sensitivity study of the coupling between a listener's head and the incident sonic boom wave which combine into the effective sound-field to be reproduced.

Motivation & Objective

• To develop a method for accurate indoor reproduction of low-frequency sonic booms, which are unsuitable for headphone-based psychoacoustic testing due to spectral and spatial characteristics.
• To overcome the limitations of headphone reproduction, which fails to replicate whole-body and head-related transfer function (HRTF) effects critical for low-frequency perception.
• To enable listener movement within a 3D sound field by using boundary surface control instead of fixed-point headphone reproduction.
• To evaluate the impact of a listener’s head on sound field reproduction accuracy and determine optimal microphone placement.
• To assess whether the sound field recorded around one listener can be used to reproduce the same field around another listener, minimizing inter-person variability.

Proposed method

• Formulates sonic boom reproduction as an Active Noise Control (ANC) problem, where the desired sound field is treated as the primary disturbance and the secondary sources (loudspeakers) are used to cancel it in the error region.
• Applies Boundary Surface Control (BSC), a technique that controls sound pressure on the boundary surface of a volume to reproduce a desired sound field inside it, based on Kirchhoff-Helmholtz integral theory.
• Uses frequency- and time-domain numerical simulations to model sonic boom reproduction in a 16-loudspeaker room at LMA, with microphones placed on the walls to monitor error signals.
• Incorporates adaptive filtering and error signal feedback to improve robustness against environmental changes and listener presence.
• Simulates sound fields around rigid spheres and detailed finite element head models (with ears and nose) to assess the influence of head diffraction on reproduction accuracy.
• Compares reproduction performance using: (1) the incident wave only, (2) the total scattered field around a listener, and (3) the field without any diffractive object, to isolate the effect of head scattering.

Experimental results

Research questions

• RQ1 Can Boundary Surface Control (BSC) with ANC accurately reproduce a low-frequency sonic boom in a small room, enabling realistic psychoacoustic testing?
• RQ2 How does the presence of a listener’s head affect the accuracy of sound field reproduction, particularly in terms of scattered wave contributions?
• RQ3 Is it possible to use the sound field recorded around one listener (e.g., listener A) as a reference to reproduce the same field around another listener (listener B) without significant error?
• RQ4 What is the optimal distance between control microphones and the listener’s head to minimize reproduction error while maintaining a large control area?
• RQ5 Does reproducing only the incident plane wave (without accounting for head diffraction) lead to significant errors in the reproduced sound field at low frequencies?

Key findings

• At low frequencies (below 300 Hz), the head’s dimensions are acoustically small relative to the wavelength, so its diffraction effects are minimal, allowing for simplified modeling.
• The error between the incident wave and the total pressure field (including head scattering) decreases rapidly within 0–50 cm from the head, but decreases slowly beyond 50 cm.
• Reproducing only the incident wave leads to large errors around the listener, even at low frequencies, proving that the scattered field must be included in the reproduction.
• Using the sound field recorded around a listener A to reproduce the field around listener B is feasible at low frequencies, as the influence of head diffraction is minimal.
• Microphone placement beyond 50 cm from the listener yields diminishing returns in error reduction, while increasing the control surface leads to reduced reproduction accuracy with a fixed number of microphones.
• Numerical simulations confirm that eigenfrequencies of the Dirichlet problem do not hinder BPC performance for sonic boom reproduction, supporting the feasibility of real-world implementation.