Markus Müller-Trapet

Effect of boundary diffusers in a reverberation chamber: Standardized diffuse field quantifiersa)

The sound field inside a reverberation chamber must have a high degree of diffusivity to allow for the accurate measurement of various acoustic quantities. Typically, hanging or rotating diffuser panels are installed in the chamber in an effort to achieve this diffusivity. However, both of these diffuser types have certain limitations, and adequate sound field diffusivity is often difficult to realize. A 1:5 scale reverberation chamber has been used to systematically analyze the relative effectiveness of hanging diffusers versus an alternative diffuser type referred to as a boundary diffuser. To characterize sound field diffusivity, three quantifiers from the ASTM E90, ASTM C423, and ISO 354 standards have been used: maximum absorption coefficient, standard deviation of decay rate, and total confidence interval. Analysis of the quantifier data reveals that boundary diffusers and hanging diffusers produce roughly equivalent diffusion in the sound field. The data also show that the standards have certain inconsistencies that can obfuscate the characterization of sound field diffusivity, which may explain reproducibility and repeatability issues previously documented in the literature.

On the in situ impedance measurement with pu-probes—Simulation of the measurement setup

High-quality numerical simulations in room acoustics require a detailed knowledge of the acoustic reflection characteristics of the materials in the room, in order to realistically model the interferences between multiple sound reflections at the room boundaries. While different standardized measurement methods exist for the determination of the absorption coefficient and reflection factor these methods can generally not be applied in situ. Thus time-consuming laboratory measurements and the supply of material samples are required. Driven by the obvious demand for a reliable in situ measurement technique, a pu-probe based method has emerged during the last years, which derives the reflection factor based on the simultaneous measurement of sound pressure and velocity. However, previous investigations of the setup and publications by other authors have shown that the measurement results are affected by various uncertainty factors. The present study aims at the identification, separation, and quantitative assessment of the uncertainty factors related to reflection and diffraction effects at the loudspeaker, sensor, and the absorber geometry. Therefore, a purely simulative approach will be used that replicates the actual measurement situation in every detail, including the geometries of sensor, loudspeaker, and absorber. The simulation setup is validated by measurements and is used to systematically separate the different uncertainty factors.

Uncertainty analysis of standardized measurements of random-incidence absorption and scattering coefficients

This work presents an analysis of the effect of some uncertainties encountered when measuring absorption or scattering coefficients in the reverberation chamber according to International Organization for Standardization/American Society for Testing and Materials standards. This especially relates to the uncertainty due to spatial fluctuations of the sound field. By analyzing the mathematical definition of the respective coefficient, a relationship between the properties of the chamber and the test specimen and the uncertainty in the measured quantity is determined and analyzed. The validation of the established equations is presented through comparisons with measurement data. This study analytically explains the main sources of error and provides a method to obtain the product of the necessary minimum number of measurement positions and the band center frequency to achieve a given maximum uncertainty in the desired quantity. It is shown that this number depends on the ratio of room volume to sample surface area and the reverberation time of the empty chamber.

Signal processing for hemispherical measurement data

To realistically model the sound propagation in rooms, a detailed knowledge of the reflection properties of the surrounding surfaces is required. In this context, the reflection properties include both the sound absorption as well as scattering. In order to be able to measure the angle-dependent reflection properties of surfaces in-situ, a hemispherical microphone array was recently designed and built. For a reduction of the required hardware an efficient, rotationally symmetric sampling was chosen, so that 24 microphones on two concentric semicircles are employed to measure a total of over 2300 positions on a hemisphere in 20 minutes. This contribution will give an overview over the required signal processing steps to process the measurement data from such a microphone array. Special emphasis will be placed on the determination of the microphone positions and the special case of data available on a hemispherical surface. Preliminary results will be presented.

Prediction and measurement of the random-incidence scattering coefficient of periodic reflective rectangular diffuser profiles

Periodic reflective rectangular diffuser profiles have been previously studied by Embrechts through an analytical method calculating their random incidence scattering coefficients. This paper presents the results of further investigations obtained by real-scale and 1:5th scale random incidence scattering coefficient measurements on these profiles. All measurements were performed according to the corresponding ISO 354 and ISO 17497-1 standards. The measurement results are compared with the calculation results derived by the analytical method. Although some periodic profiles could not be treated by the analytical model, fairly good agreements are obtained between the different approaches. The real scale and 1:5th scale measurement results are also presented and compared.

A revised scale model reverberation chamber for measurements of scattering coefficients.

For measurements of random‐incidence scattering coefficients according to ISO 17497‐1, scale models have proven to yield results with reasonable time and effort. Especially, handling the samples becomes much easier so that several measurements, which may be difficult or even unfeasible in a full‐scale reverberation chamber, turn out to be possible. Despite these advantages, using a scale model environment poses other difficulties that are related to the extended frequency range. In order to achieve a better repeatability and stability of the measurements, a revised scale model room is proposed. As a further development from the ordinary scale model rooms, the turntable can now be lowered so that the rotating base plate is mounted flush with the floor. The previously used hanging panels have been replaced with wall‐panel diffusers. Furthermore, the chamber includes temperature and humidity sensors to enable correction for air absorption. Measurements in the new model room are performed to evaluate the quality of the results compared with reference data from calculation and the repeatability of the measurements. Results will be compared for different room setups.

Acoustic source localization with microphone arrays for remote noise monitoring in an Intensive Care Unit

An approach is described to apply spatial filtering with microphone arrays to localize acoustic sources in an Intensive Care Unit (ICU). This is done to obtain more detailed information about disturbing noise sources in the ICU with the ultimate goal of facilitating the reduction of the overall background noise level, which could potentially improve the patients’ experience and reduce the time needed for recovery. This paper gives a practical description of the system, including the audio hardware setup as well as the design choices for the microphone arrays. Additionally, the necessary signal processing steps required to produce meaningful data are explained, focusing on a novel clustering approach that enables an automatic evaluation of the spatial filtering results. This approach allows the data to be presented to the nursing staff in a way that enables them to act on the results produced by the system.

Importance of correlation between reverberation times for calculating the uncertainty of measurements according to ISO 354 and ISO 17497-1

The calculation of measurement uncertainties follows the law of error propagation as described in the Guide to the Expression of Uncertainty in Measurements (GUM). The result can be expressed as a contribution of the variances of the individual input quantities and an additional term related to the correlation between the input quantities. In practical applications, the correlations are usually neglected. This has, e.g., led to the expression included in Annex A of ISO 17497-1 to calculate the precision of the measurement of random-incidence scattering coefficients. To determine whether it is actually justified to neglect the input correlations, this contribution investigates the correlations between the reverberation times used to determine the random-incidence absorption coefficient (ISO 354) and scattering coefficient (ISO 17497-1) in a reverberation chamber. The data used here are taken from measurements in a real-scale and a small-scale reverberation chamber. It is found that for ISO 354 correlations...

Machine listening in combination with microphone arrays for noise source localization and identification

In a recent project, a large microphone array system has been created to localize and quantify noise sources in an Intensive Care Unit (ICU). In the current state, the output of the system is the location and level of the most dominant noise sources, which is also presented in real-time to the nursing staff. However, both staff as well as patients have expressed the need for information about the types of noise sources. This additional source identification can also help to find means of reducing the overall noise level in the ICU. To accomplish the source identification, the approach of machine listening with a deep neural network is chosen. A feed-forward pattern recognition network is considered in this work. However, it is not clear which types of features are best suited for the given application. This contribution thus examines the problem from a practical point of view, comparing different features including those related to sound perception, such as specific loudness, Mel-frequency cepstral coefficients, as well as the output of a gamma-tone filter bank. Additionally, the concept of time-delay networks is tested to see whether a better classification of the signals can be achieved by including their time history.In a recent project, a large microphone array system has been created to localize and quantify noise sources in an Intensive Care Unit (ICU). In the current state, the output of the system is the location and level of the most dominant noise sources, which is also presented in real-time to the nursing staff. However, both staff as well as patients have expressed the need for information about the types of noise sources. This additional source identification can also help to find means of reducing the overall noise level in the ICU. To accomplish the source identification, the approach of machine listening with a deep neural network is chosen. A feed-forward pattern recognition network is considered in this work. However, it is not clear which types of features are best suited for the given application. This contribution thus examines the problem from a practical point of view, comparing different features including those related to sound perception, such as specific loudness, Mel-frequency cepstral coeffi...

Array processing methods for the determination of reflection properties of architectural surfaces

It has become popular in architectural acoustics to use microphone arrays, very often in spherical arrangements, to capture and analyze the sound field in rooms. This contribution will present a hemispherical receiver array, designed to obtain information about sound reflection from architectural surfaces, preferably in-situ. Similarly to previous studies, the analysis in the spherical harmonics (SH) domain is favored, with the additional challenge of data available on a hemisphere instead of a complete sphere. This problem is solved by obtaining orthonormal base functions on the hemisphere. As application cases, spherical beamforming for the determination of reflection factors will be presented as well as scattering near-field holography in order to determine the diffusion coefficient of small samples. Results from numerical and experimental case studies will be discussed.