Skip to content
Back to search results

Acoustic Holography

Introduced by
James Maynard, Earl Williams, Yongjian Lee
Year
1985
Type
Sound field reconstruction method
Family
Acoustic imaging
Font size

What does it do?

Near-Field Acoustic Holography (NAH) is a technique for reconstructing 3D acoustic sound fields and visualizing sound radiation from sources by measuring pressure at a dense microphone array in the near field.

Pioneered by Maynard, Williams, and Lee in 1985, NAH extends holographic principles from optics to acoustics, enabling detailed acoustic source characterization, noise source identification, and acoustic field visualization that is impossible with conventional single-point or line-array methods.

The intuition behind the method

Imagine photographing a 3D scene: a camera captures a 2D projection of light from the scene. Similarly, NAH records a 2D 'snapshot' of acoustic pressure on a surface (the measurement array) and reconstructs the full 3D acoustic field around it. By moving the array or using a 3D array, you build a complete map of where sound is coming from and how it radiates into space. This enables 'seeing' sound—visualizing the acoustic patterns around machinery, musical instruments, or buildings.

History

Maynard, Williams, and Lee developed NAH in 1985, adapting holographic principles from optics to acoustics. The method was initially applied to vibration source identification and acoustic field reconstruction. Williams' foundational work in Fourier acoustics (1999) provided theoretical underpinnings. Applications expanded through the 1990s–2000s to industrial noise source identification, product testing, and aeroacoustics. Modern variants include 3D NAH, spherical harmonics-based methods (HOA), and regularized inverse methods for improved robustness.

When to use it

Use NAH when precise characterization of acoustic source radiation is needed: vibration source identification in machinery, troubleshooting radiated noise from appliances or vehicles, acoustic testing of products, understanding near-field effects around structures, and detailed acoustic design validation. NAH is particularly useful when conventional far-field measurements are impractical (small sources, noisy environments, enclosed spaces). Avoid NAH when sources are very far or low frequency (requiring very large arrays).

Limitations and considerations

  • Requires a dense, well-calibrated microphone array; cost and complexity scale with frequency (higher frequency requires tighter spacing).
  • Array must be placed very close to the source (near field); limits application to accessible sources and may be impractical for large sources.
  • Reconstruction accuracy is sensitive to measurement errors (phase misalignment, microphone mismatch); small errors propagate and amplify, especially for evanescent components.
  • Computational cost is high: 2D/3D Fourier transforms and field reconstruction require significant processing, especially for broad frequency ranges.
  • Poor performance at very low frequencies where the wavelength is large; array would need to be impractically large to achieve the near-field condition (distance < λ/2π).

Source conditions do not replace independent scientific review or validation by an analysis engine.

Strengths

  • Enables direct visualization of 3D acoustic fields; superior to far-field methods for understanding acoustic sources and radiation mechanisms.
  • Separates direct sound from surrounding noise; the near-field array isolates the source and reconstructs its field even in noisy environments.
  • High spatial resolution: can identify discrete acoustic sources (individual vibrating panels, fasteners) separated by fractions of a wavelength.
  • Provides quantitative acoustic metrics (intensity, power, directivity) derived from measured pressure and reconstructed velocity.
  • Works in complex acoustic environments (indoor, reverberant) where far-field measurements fail due to reflections and background noise.

Applications

  • Noise source identification in machinery and appliances: locate and visualize acoustic sources (bearings, fans, gears) contributing to radiated noise.
  • Automotive and aerospace noise testing: map noise radiation from vehicle surfaces and engine compartments; guide acoustic design improvements.
  • Building and structural acoustics: visualize sound radiation from building facades, HVAC systems, and mechanical vibration; assess insulation performance.
  • Hearing aid and speaker design: characterize acoustic radiation patterns of transducers and hearing aid receivers; optimize frequency response and directivity.
  • Underwater source localization: extend NAH methods to hydrophone arrays for detecting and localizing underwater acoustic sources (marine life, submarines).
  • Architectural acoustics: measure sound radiation from architectural elements (columns, facades) in concert halls and theaters; validate acoustic simulations.

How it works

How it works

1

Position a dense microphone array (planar or curved, 2D or 3D) close to the source of interest in the source's near field (distance < few wavelengths). Record acoustic pressure at all microphone locations simultaneously over a frequency range of interest.

2

Calibrate all microphones; ensure synchronization and phase alignment (< sample period); correct for microphone frequency response and environmental factors.

3

Compute the spatial Fourier transform (wavenumber spectrum) of the pressure field measured by the array. Decompose the measured field into propagating (radiating) and evanescent (near-field only) wave components.

4

Use the wavenumber spectrum and Helmholtz equation to reconstruct the acoustic field at arbitrary points in space: backward propagate through the measurement surface toward the source, forward propagate away from the measurement surface into the far field.

5

Visualize the reconstructed pressure and particle velocity fields at various planes and frequencies. Identify acoustic sources (regions of high intensity), radiation patterns, and directivity. Compute source strength, power output, and acoustic intensity.

Common pitfalls

  • Using arrays with insufficient spatial resolution; element spacing larger than λ/2 causes spatial aliasing and spurious source artifacts.
  • Neglecting the near-field condition: if the measurement plane is too far from the source, evanescent components decay and cannot be reconstructed accurately.
  • Assuming microphone mismatch is negligible; even small phase differences between adjacent microphones cause coherence loss and reconstruction errors.
  • Over-interpreting noise-induced artifacts; measurement noise (SNR < 30 dB) produces spurious sources in the reconstructed field.
  • Failing to validate reconstructions against independent measurements (e.g., accelerometers on the source); always cross-check results.

Frequently asked questions

What is the difference between NAH and far-field beamforming?

Beamforming synthesizes a directional response from an array to locate sound sources; NAH reconstructs the full 3D sound field. Beamforming works at any distance but provides limited spatial resolution. NAH requires near-field placement but delivers detailed field reconstruction with wavelength-scale resolution. NAH is superior for understanding radiation mechanisms; beamforming is better for source localization at a distance.

How close must the measurement array be to the source for NAH?

The array must be in the source's acoustic near field, typically distance < λ/2π (where λ is wavelength). Closer placement ensures evanescent waves (high spatial frequencies) are strong enough to measure accurately. At very close distances (distance << wavelength), near-field effects dominate; reconstruction becomes unstable, and far-field estimates are impossible.

What is the effect of microphone spacing on NAH performance?

Microphone spacing must not exceed λ/2 (half-wavelength) to avoid spatial aliasing. Smaller spacing (λ/4 or less) provides better resolution and is less sensitive to mismatch. At high frequencies, small spacing is needed; compact arrays with 64+ elements are common. Computational cost scales with array size, so spacing is a trade-off between resolution and cost.

How do I account for measurement noise in NAH?

Measurement noise amplifies during field reconstruction, especially for evanescent components (high spatial frequencies). Regularization techniques (Tikhonov, truncated SVD) reduce noise amplification at the cost of some spatial resolution loss. Ensemble averaging over multiple measurements improves SNR. Ensure SNR > 20–30 dB for reasonable reconstructions; lower SNR requires aggressive regularization.

Can NAH reconstruct fields in reverberant rooms?

Yes, if the near-field array is sufficiently close to the source and encompasses the source region. Reflections and room acoustics are part of the measured field, so they appear in the reconstruction. To isolate source radiation only (excluding room effects), use smaller arrays positioned very close to the source. Alternatively, use acoustic imaging in an anechoic chamber for clean source-only characterization.

Related methods

Sources

  1. Maynard, J. D., Williams, E. G., & Lee, Y. (1985). Near-field acoustic holography: I. Theory of generalized holography and the development of NAH. Journal of the Acoustical Society of America, 78(4), 1395–1413.

    Open publication reference
    Copied.
  2. Williams, E. G. (1999). Fourier Acoustics: Sound Radiation and Nearfield Acoustical Holography. Academic Press.

    ISBN: 978-0124654052
    Copied.
  3. Mueller, T. F. (2002). Aeroacoustic Measurements. Springer-Verlag.

    ISBN: 978-3540678441
    Copied.