Search

The acoustic emission sensors were tailored by using Piezoelectric Lead Zirconate Titanate (PZT) ceramic as sensing element and different materials as backing layers. The influences of backing layers on electroacoustic properties of the sensors were investigated based on the pencil lead break test and simulated noise experiments. The results show that compared with PZT ceramic, the resonant frequency of the acoustic emission sensors decreases due to damping effects from backing layer and the metal shell. The backing layer of air has the largest acoustic reflection coefficient, and epoxy resin/tungsten powder has the largest acoustic attenuation performance, which is benefit to improve the bandwidth of the acoustic emission sensor. The acoustic emission sensors made of different backing layers own good reception ability to acoustic emission signals, and the characteristic parameters analysis results prove their good identifying ability to effective acoustic emission signals and noise signals.

Generally, the better sound absorption characteristics can be achieved in the low frequency band by embedding resonance units into the conventional anechoic coating. However, in this way, the frequency range of effective sound absorption is relatively narrow, which limits the application of anechoic coating in engineering. In this paper, in order to optimize this sound absorption characteristics, a new type of underwater sound absorption metamaterials with multi-resonator are proposed based on the coupled resonance mechanism. The underwater sound absorption characteristics of the proposed metamaterials with steel plate backing are investigated by finite element method (FEM). The effect of coupled resonance between the resonators on the sound absorption characteristics of metamaterials is investigated. In addition, the effect of the geometric structure of steel plate backing on the sound absorption characteristics is studied as well. Numerical results show that the proposed metamaterials have many advantages over the conventional anechoic coating in underwater sound absorption due to the coupled resonance effect. Moreover, the steel plate backing has a significant effect on the sound absorption characteristics of the proposed metamaterials, especially, in the low frequency range.

The spurious sound in the measurement of the parametric loudspeaker is resultant from the nonlinearity at the receiving system caused by the ultrasonic waves with high amplitudes. It may provide a misleading result, especially in the nearfield, where the parametric array effect of the parametric loudspeaker cannot be easily observed. In this study, to eliminate or alleviate the effect of the spurious sound, an in-depth investigation into the spurious sound in the parametric loudspeaker is carried out experimentally by adopting an index of a ratio between the spurious sound and the demodulated difference-frequency sound. A relation between the above ratio and four main parameters, i.e., the amplitude of the primary waves at the source, the radius of the ultrasonic emitter array, the frequency value and the observation distance, is obtained. It exhibits a good match with the well-known theory described by the Gaussian-beam expansion technique. The threshold to ignore the spurious sound in the measurement of the parametric loudspeaker is also established.

A highly improved set-membership normalized least mean square (SM-NLMS) algorithm is carried out to further strengthen the estimation behavior of previously presented SM-NLMS algorithm for estimating sparse multi-path and acoustic echo channels. The newly developed sparse SM-NLMS algorithm utilizes an approximating l 0 function (AL0) to modify the SM-NLMS’s cost function to devise a zero attractor. Then, a zero attractor is realized to provide a quick zero attraction for estimating the sparse signals. The proposed soft parameter functioned SM-NLMS (SPFSM-NLMS) algorithm is derived by means of the unconstrained stochastic gradient minimization method to form the finalized SPFSM-NLMS (FSPFSM-NLMS) algorithm. The FSPFSM-NLMS algorithm is derived mathematically, and whose estimation behavior is considered and analyzed on a multi-path sparse channel as well as an acoustic echo channel. Moreover, our presented FSPFSM-NLMS algorithm is also studied under various sparsity levels. The simulated results illustrate that the developed FSPFSM-NLMS algorithm provides a superior sparse channel estimation behavior compared to the zero-attracting SM-NLMS (ZASM-NLMS), reweighting ZASM-NLMS (RZASM-NLMS) and NLMS in terms of the steady-state performance and convergence speed.

The study of the sound propagation in duct systems containing a porous material is important because they are used in many industrial applications. The acoustic behavior of this kind of ducts which are supposed two-port systems can be modeled by many matrices like the reflection, transmission, mobility, transfer and scattering matrices. In this work, Theoretical and experimental methods are used to evaluate the acoustic performance of duct systems containing one or two porous materials. To reach this goal, the Transmission coefficient was calculated by determining transfer and scattering matrices. The comparison between theoretical and experimental results shows a good agreement between results with some differences due to the no respect of the hypothesis of the rigidity of the duct. A parametric study allowed estimating the more influent parameters on the acoustic behavior of duct systems containing porous materials.

In this study, a modified two-thickness method to measure acoustic properties of porous materials using surface acoustic impedance is proposed and compared to those obtained by the four-microphone impedance tube and traditional two-thickness methods. In the two-thickness method, the complex propagation constant and characteristic impedance are estimated from the measured surface acoustic impedances, and the sample thicknesses should have a ratio of two. Through the proposed method, the complex characteristic impedance and wavenumber of a porous material are calculated under arbitrary sample thicknesses condition. The surface acoustic impedances for samples of different thicknesses are measured via the two-microphone impedance tube method to estimate their complex acoustic properties. Sensitivity analysis is performed to analyze the measurement uncertainty on the resulting values. The sensitivity depended on the sample thickness and mount edge constraint in the tube. Using the method proposed in this study, acoustic properties are measured accurately with small differences in thickness and surface acoustic impedance.

© 2018 Elsevier Ltd Speech enhancement algorithms is to improve speech quality, naturalness and intelligibility by eliminating the background noise and improving signal to noise ratio. There are several objective measures predicting the quality of noisy speech enhanced by noise suppression algorithms, and different objective measures capture different characteristics of the degraded signal. In this paper, the multiple linear regression analysis is used to obtain a composite measure which has high correlation with subjective tests, and the performance of several speech enhancement algorithms under car noise conditions is compared. The uncertainty of the results of the proposed measures on different speech enhancement algorithms is analyzed, and the reliability of the results is discussed.

A crucial stage in unsupervised speech enhancement algorithm is the estimation of noise related parameters which usually needs prior models for noise. However, estimation of such parameters is a challenging task at low signal-to-noise ratios or in nonstationary noisy environments. In this paper, without knowing the prior models, an unsupervised and iterative speech enhancement algorithm is proposed which assumes speech spectrogram and its temporal gradient as sparse components. The quasi-harmonic description of the speech signals justifies this assumption. The speech enhancement is performed by decomposing the spectrogram of noisy speech into sparse matrix, enforcing the sparsity and temporal gradient regularizations. The Kullback–Leibler divergence is incorporated to minimize the distance between the observation and reconstructed components with nonnegativity constraints. Alternating direction method of multipliers is used to optimize the algorithm. The proposed algorithm is different from many speech enhancement approaches as it reduces background noise in an uncomplicated manner without need of a noise estimation algorithm to find noise-only excerpt. In addition, the proposed algorithm obtains an improved performance in adverse environments without knowing the exact distribution of noise. The experimental results demonstrate that the proposed algorithm outperforms the competing algorithms in terms of the speech quality and intelligibility. Moreover, the composite objective measure reinforced better performance in terms of residual noise and speech distortion in strong noise.

The common methods to calibrate microphone array are those that calibrate each microphone. Different from these methods, this article proposes a simulated point sound source calibration method without calibrating each microphone. Suppose there is a microphone array in the sound field of the point source, according to the positional relationship between the array and the source, the dynamic sound signal of every microphone can be calculated. A microphone array calibration device (MACD) is designed to input the dynamic signals calculated into the actual microphone array by the coupling cavity standard sound sources. By comparing the position of the simulated point source with that located by array, position error correction curves are fitted to calibrate the positioning accuracy of microphone array. The experimental results reveal validity of the proposed method.

Plane-wave-enriched finite element method (FEM), an efficient wave-based prediction method, uses shape functions incorporating a set of plane waves propagating in various directions to enhance the sound field approximation capability. The method can be an efficient wave-based acoustic solver for room acoustic simulations because it can simulate wave phenomena in rooms accurately with markedly fewer finite elements in spatial discretization than those used for standard FEM. However, several aspects remain to be addressed before its use for practical room acoustic simulations. Accurate sound absorber modelings able to address their frequency and incident angle dependence of absorption characteristics present extremely important issues. This paper presents a proposal of a method of implementing an extended-reaction model of microperforated panel (MPP) and permeable membrane (PM) sound absorbers into a room acoustic solver with plane-wave-enriched FEM. First, we demonstrate the validity of the proposed method in comparison with theoretical values in which impedance tube problems including three sound absorbers composed of MPP and PM are used. Then, the effectiveness over standard FEM is demonstrated via 2D real-scale office problems with the three MPP-PM sound absorbers.

The inaccurate estimates of the speech and noise linear prediction coefficients (LPCs) introduce bias in augmented Kalman filter (AKF) gain, which impacts the quality and intelligibility of enhanced speech. Although current tuning methods offset the bias in AKF gain, particularly in colored noise conditions, they do not adequately address nonstationary noise conditions. This paper introduces a new tuning algorithm of the AKF gain for speech enhancement in real-life noise conditions. Due to this purpose, a speech presence probability (SPP) method first estimates the noise power spectral density (PSD) from each noisy speech frame to compute the noise LPC parameters. A whitening filter is constructed with the noise LPCs to pre-whiten each noisy speech frame prior to computing the speech LPC parameters. The AKF is then constructed with the estimated speech and noise LPC parameters. To achieve better noise reduction, the robustness metric is employed to dynamically offset the bias in AKF gain during speech absence of the noisy speech to that of the sensitivity metric during speech presence. The speech activity is obtained through adopting the speech and noise production model parameters. It is shown that the reduced-biased AKF gain achieved by the proposed tuning algorithm addresses speech enhancement in real-life noise conditions. Objective and subjective scores on the NOIZEUS corpus demonstrate that the proposed method produces enhanced speech with higher quality and intelligibility than the competing methods in real-life noise conditions for a wide range of signal-to-noise ratio (SNR) levels.

Coprime acoustic sensor arrays have been recently developed to estimate the direction-of-arrival (DOA) of multiple sound sources and may be needed in many acoustic applications because they can provide greater degrees of freedom and better estimation performance. However, most existing DOA estimation algorithms are derived under the assumption that the number of sources is known and have poor robustness due to unknown noise. This paper proposes a robust DOA estimation algorithm without estimating the number of sources using a coprime acoustic sensor array. The solution is based on the multiple signal classification (MUSIC)-like DOA estimation algorithm framework, in which a new spatial covariance model via spatial smoothing of the coprime array output signal is designed. The proposed spatial smoothing generalized MUSIC-like (SS-G-MUSIC-like) algorithm utilizes the diagonal loading technique to reconstruct the spatial smoothed covariance matrix. Results related to one-sound source and two-sound sources DOA estimation experiments show that the proposed algorithm can provide more focused source tracks over the entire data segment and better clutter suppression.

At low frequencies the assessment of the sound reduction index of building elements in the laboratory according to the standard ISO 10140-2:2010 is burdened by a large variation in the measurement results. This is due to the fact that at low frequencies the acoustic field is not sufficiently diffuse. This paper discusses a measurement procedure in which a diffuse field is created in the source room by means of an array of loudspeakers positioned closely to the building insulation element under test. This procedure exploits the acoustic near field of the loudspeaker array. The problems related to the breakdown of the diffuse field assumption of the receiving room are eliminated by measuring the structural response of the building insulation element under test by means of laser Doppler vibrometry and the application of the Rayleigh integral to compute the radiated sound power. The sound reduction index is determined from the ratio of the incident sound power, created by the loudspeaker array, and the radiated sound power. The measurement approach is validated by means of a measurement of the sound reduction index of a single layer glazing. Comparisons are made with an analytical model and with a standardized ISO 10140-2:2010 measurement. Although the method offers clear, strong points in terms of removing room acoustic effects from the measurements in the lower frequency range, a point of concern is the measurement effort. ispartof: APPLIED ACOUSTICS vol:140 pages:225-235 status: published

Microphone positions have to be calibrated in distributed microphone array applications. A constrained total least squares calibration method for distributed microphone arrays is proposed in this paper. All the source event positions are first estimated by the weighted multidimensional scaling algorithm. Then the suitable source events are picked up by the TDOA selection strategy at each node. Finally, the node microphones are calibrated by the total least squares, and further refined based on constrained total least squares when the estimated results suffer from large errors. The proposed method can obtain higher calibration accuracy and works well in noise and reverberation conditions. Simulation and real-world experiment results reveal the validity of the proposed method.

The analysis of acoustic data typically discusses the problem of segmenting the acoustic events into non-overlapping acoustically compact categories. In Acoustic Event Detection (AED), an acoustic event is categorized into speech and non-speech events. Detection of non-speech sounds such as scream, gun shots, explosions, and glass break events is very helpful in acoustic surveillance, multimedia information retrieval, and acoustic forensic applications. In this paper, we propose global statistical features-based representation for multi-variate varying length acoustic data. A discriminative model-based classifier is then used to classify different acoustic events. The proposed representation is of very less dimension. The proposed approach is evaluated on surveillance-oriented AED datasets such as CICESE (recorded from a smart room scenario), Environmental Sound Classification (ESC), and IEEE AASP/DCASE2013 (Office environment) datasets. The proposed approach gives a better performance when compared with the conventional Hidden Markov Model (HMM) and Gaussian Mixture Model (GMM) approaches.

This paper aims to examine sound fields in extra-large spaces, which are defined in this paper as spaces used by people, with a volume approximately larger than 125 , 000 m 3 and absorption coefficient less than 0.7. In such spaces inhomogeneous reverberant energy caused by uneven early reflections with increasing volume has a significant effect on sound fields. Measurements were conducted in four spaces to examine the attenuation of the total and reverberant energy with increasing source-receiver distance, which was then validated by the simulations with image-source method. Results show that the reasons for the total energy’s exponential decrease are not only the direct sound, but also the reverberant energy. The prediction difference of total sound pressure level (SPL) between classical formula and the image-source method increases with the volume and decreases with the surface absorption, based on which a critical line separating extra-large spaces from large ones is proposed. Moreover, a newly modified model based on the importance of first reflection from floor is proposed, showing more advantages of sound level prediction in extra-large spaces.

In this paper, we propose a modified second-order nonlinear infinite impulse response (IIR) filter for equalizing the frequency response and compensating nonlinear distortions of an electrodynamic loudspeaker. A problem of electrodynamic loudspeakers is the generation of nonlinear distortions, which degrade the sound quality. One of the approaches to reducing nonlinear distortions is to use a second-order nonlinear IIR filter. This filter is based on an equivalent circuit model of an electrodynamic loudspeaker and its coefficients are determined by physical parameters. However, it is difficult to compensate nonlinear distortions when a loudspeaker has a high Q factor at the lowest resonance frequency, at which the displacement of the diaphragm and nonlinear distortions become large. The Q factor determines the linear frequency response of the electrodynamic loudspeaker. Although it is necessary to compensate the Q factor of the loudspeaker, a nonlinear IIR filter cannot compensate the Q factor because it does not have a linear filtering feature. In this paper, we propose a modified second-order nonlinear IIR filter that can not only compensate the nonlinear distortions caused by the nonlinearities of the force factor and stiffness but also equalize the frequency response by employing the linear characteristics of the loudspeaker with the desired Q factor. Experimental results show that the proposed filter can compensate the linear and nonlinear distortions more effectively than a conventional filter.

The feedback cancellation performance of behind the ear hearing aids can be improved by employing two microphones in the feedback cancellation process. A frequency domain implementation of feedback cancellation in a two microphone behind the ear hearing aid has been proposed in this paper. A frequency domain step size control scheme has been further introduced to improve the convergence behaviour. The implementation of a the frequency domain step size control method requires the microphone input signal as well as the loudspeaker output signal to be uncorrelated. In order to reduce the correlation effects and enable such an implementation, we have incorporated frequency shifting in the forward path. The proposed approach has been shown to provide improved convergence behaviour as well as offer enhanced speech quality. Further, the proposed scheme is expected to offer a reduced computational load for longer feedback paths.

A dual-microphone noise reduction method is proposed in this paper. It is based on the coherence function and the bionic wavelet transform using Kalman filter. This filter is applied to the Bionic wavelet coefficients obtained through the application of the BWT to the input signal. The adopted method can treat noisy signals using two closely spaced microphones in different noise scenarios and outperforms other dual-channel speech enhancement methods referring to the perceptual evaluation of speech quality, the spectrogram analysis and the time domain waveforms.

In this paper, we propose a novel method for estimating clean speech from a single channel transient noise corrupted speech. In the proposed method we assume that speech spectrogram is both sparse and has temporal continuity property, and transient noise spectrogram is both sparse and has spectral continuity property. Based on these assumptions, we define a novel regularization model with sparsity and continuity imposing regularization terms for transient noise reduction. Then we solve the proposed model via alternating direction method of multipliers (ADMM) and derive an efficient iterative algorithm. Based on the assumption that transient noise spectrogram is low rank, we construct a binary mask that specifies locations of the transients and apply it in the proposed algorithm to achieve better separation results. Our method straightforwardly estimates speech and is free of noise power spectral density (PSD) estimation and does not need any pre-trained models of speech or noise. Experiments with various types of transient noises demonstrate effectiveness of the proposed method.

The average sound absorption coefficient of a composite absorber from 0 to 1.6 kHz was maximized using a teaching–learning-based optimization algorithm, where the geometric parameters were used as optimization variables. The effective-medium model and transfer-matrix method were used to calculate the sound absorption coefficient, and optimized sound absorption coefficient was maximized at 0.938. The specific acoustic impedance of the material is almost perfectly matched with air thanks to the optimized size of the lateral plates, micro-perforated plates, and cavities. Experimental results confirm the theoretical calculation, showing the investigated passive acoustic design method of designing sound absorber.

The vibro-acoustic environment generated during a rocket launch induces dynamic loads which may damage critical components of spacecraft and scientific instruments. The research is focused on the vibro-acoustic response of the Wide Field Imager (WFI) instrument subjected to an intense diffuse sound field. Noise-induced structural vibrations in low- to mid-frequency can be computed by means of coupled FEM-BEM numerical models. Nevertheless, this approach is unsuited to calculations of cavity-structure interactions of light-weight and thin structures. In this paper the application and assessment of the elasto-acoustic FEM-FEM coupling approach for the cavity - optical blocking thin foil filter interactions is investigated. The diffuse sound field excitation is reconstructed by matching values of the sound pressure level to these obtained from experimental measurements. The viscosity effects on the wave propagation in thin layers are included by using the visco-thermal fluid model. The accuracy and efficiency of this approach are investigated and results obtained are compared to measured data from a reverberant test chamber. The results show that a very good agreement is obtained in terms of the acoustic pressure in the vicinity of the optical blocking filter. The behaviour of mechanical vibrations is correctly reproduced, although some considerable discrepancies in the levels of the vibrations are observed. The case of strong coupling between the cavity acoustic field and the filter membrane is finally discussed.

The sound generation mechanisms of vocal folds’ self-induced oscillations have been investigated using vocal fold replicas. In this study, a compressible flow simulation coupled with a one-dimensional beam equation was conducted to study a reed-type artificial vocal fold. The vibrating reed was expressed by the volume penalization method in the aeroacoustic fields. The sound of the artificial vocal fold was measured in a flow experiment and compared with the simulation. The results showed that the predicted acoustic characteristics agreed well with those measured in the experiment. The flow fields during reed oscillation indicated that the rapid change in the flow rate downstream from the reed formed a strong source in the high-frequency harmonics.

A joint dereverberation and denoising framework is proposed for enhancement of distant speech in a conference room. The sparsity of speech multipath propagation is exploited by a microphone array for suppressing late reverberation. The dereverberation speech is transformed to solve a minimization problem where the sparse prediction coefficients are estimated by a least-square iteration algorithm. The residual noise components are removed by iteratively reconstructing the time–frequency spectrum of speech signals. Detection of speech presence is carried out over each cell of the spectrum based on the Bayes rule. The data analysis and the real-time processing results have shown that the proposed joint framework outperforms the existing approaches in terms of objective metrics of speech quality.

The article focuses on an optimized method for Helmholtz resonators (HR) design to be used in room acoustics. They are resonators in the form of acoustic wall facings and ceilings formed by perforated boards with circular apertures. In order to verify available calculation methods for sound absorption coefficient, an experiment was performed in a reverberation room. Several variants of the internal absorbent and internal arrangement of the HR cavity for the experiment were measured. The measured data were compared with two available methods of sound absorption coefficient calculation for normal incidence using a 1D transfer matrix and electro-acoustic analogy. The calculation optimization was performed with the use of the measured data with the aim to reach better conformity of calculated courses of the sound absorption coefficient and the reverberation room measurements.

Due to the existence of the Doppler effect, a frequency shift and band expansion occur in the measured signal that belongs to the moving sound source. To estimate the radiated sound power of the moving sound sources, this study proposes a method that is based on linear array beamforming. By applying this method, the influence of the Doppler effect on the received signal for each sensor is eliminated by interpolating the de-dopplerisation. The estimated result is obtained according to the conversion relation between the linear array beamforming output and the radiated sound power of the source. The sound power estimation error for the different motion states can be studied by considering the sound source frequency, amplitude, test distance, and other influencing factors by performing a simulation. The sound power estimation experiment of the double-moving sound sources verifies the effectiveness of the proposed method. Overall, the proposed methodology provides a simple and effective way to measure the radiated power of the moving sound sources.

Targeted at solving the problem in extracting the line spectrum features of ship radiated noise in complex marine environments, this paper proposes a new deep convolutional neural network model for underwater acoustic target radiated noise timbre perception, which is inspired by the timbre perception mechanism of the auditory system. The model first realizes the adaptive decomposition of the line spectrum of the underwater acoustic target radiated noise through one-dimensional deep convolution with multiple receptive fields, and then models the energy distribution structure of the line spectrum through the normalization of the decomposed signals and the point-by-point convolution, which expresses the frequency domain characteristics of ship radiated noise timbre. Next, a two-dimensional deep convolutional network is used to extract the deep representation of the difference in frequency domain characteristics of radiated noise timbres of different targets. And finally, the softmax layer is used to predict the target category. Validation experiments were carried out using measured data of five types of underwater acoustic targets. The results show that the proposed model is able to perceive the frequency domain characteristics of the radiated noise timbres of underwater acoustic targets, selectively enhance the signal energy of the radiated noise line spectrum, suppress the energy of interference noise, and extract stable line spectrum features of underwater acoustic targets. The recognition accuracy of the proposed model reaches up to 78.2%, which is 2.1%∼13.5% higher than the 6 control deep learning models.

Based on wave theory, the propagation equations of sound waves in single layer and multilayer media are established. In the traditional multi-layered structure and the microfiber multi-layered structure, the sound transmission characteristics of air in the middle and the back are analyzed by numerical simulation and acoustic parameter experiment. The effects of bending, viscous characteristic length and thermal characteristic length on the acoustic performance of multilayer structures are compared. At the same time, the sound transmission characteristics of two multilayer structures are compared and analyzed. The results show that the sound insulation performance of the multi-layer structure increases when the air sandwich appears in the middle of the composite structure. The sound insulation performance of the single composite structure, the sound insulation performance of the double wall structure and the superfine fiber structure are better than the traditional PU (Polyurethane, PU) foam and EVA (Ethylene vinyl acetate copolymer, EVA) composite structure.

In this study, to estimate the acoustic field properties inside a cavity with a flexible plate, a fast analytical framework based on the modal interaction approach in frequency domain is introduced. In order to elicit such an algorithm, an in-depth investigation on self-cancellation and mode coupling phenomena is presented to show how to determine effective vibrational and acoustic modes in the acoustic-structure interaction problem. The outline of this method is a compromising between accuracy and computational cost. Unlike methods require high computational cost such as FEM and BEM, the present method is suitable to obtain a rapid acoustic field estimation which is applicable for optimization problems as well as the design of active sound control systems. The results are validated with the experimental measurements available in literature and FEM solutions.

Recently, the use of speech processing in a time–frequency domain that handles the phase spectrum in addition to the amplitude one has been increasing because many studies have revealed the importance of phase. Inspired by this motivation, this paper presents a new speech signal phase processing. The contributions of this paper include the following two points: a detailed analysis of a speech signal by considering a novel phase feature, a derivative of instantaneous frequency (DIF), and demonstrating a new phase-based voice activity detection (VAD) algorithm as one of the DIF’s applications. These contributions develop on our previous work that briefly introduced the DIF into VAD. In the analysis part, we investigate the DIF from the theoretical aspect and research the statistical distribution of the DIF of speech signals under various conditions. We also propose a new phase-based VAD algorithm via the statistical likelihood ratio, and use the DIF as an auxiliary feature to improve a conventional amplitude-based VAD method. The experimental results confirm the efficacy of the phase feature in the VAD application and the possibility of combining the phase and amplitude for better performance.

In a recent work by the authors, an optimal sound absorber with a compact scale consisting of multiple inhomogeneous Helmholtz resonators with extended necks (HRENs) was designed and demonstrated for effective sound absorption in a prescribed frequency range. The development of HREN-based absorber provides a promising potential for sound absorption in a compact dimension. The present study aims to line the compact optimal absorber in a circular duct for noise attenuation. A rigid duct with the same configuration is fabricated for comparison. The acoustic performances of the lined duct and the rigid duct are experimentally evaluated by using two linear microphone arrays in an anechoic chamber. Two types of noise sources are considered: a point sound source and a rotating propeller. The experimental results reveal that a significant noise reduction is achieved by the lined duct for both the point source and the propeller noise source in the target frequency range when comparing to the rigid duct.

We report a detection method of an acoustic source by the convolutional neural network (CNN) that utilizes analytic predictions of sound radiation. The analytic predictions with various source conditions are implemented to effectively collect a large-annotated training dataset, allowing straightforward utilizations of the CNN in the acoustic domain. The data conversion from the synthetic audio signals into the pseudo-images is presented to secure compatibility with actual audio signals in terms of the direction of angle (DOA) estimation. Our source localization network fully trained with the synthetic pseudo-images were verified with various source conditions in a semi-reverberant room. The verifications demonstrate remarkable robustness and noise resistance to estimate the DOA regardless of source conditions. Moreover, considering our network has implemented a small number of short-time audio signals (i.e., three audio signals for 0.1 s), the proposed algorithm can be a breakthrough in a real-time tracking of the acoustic source by hybridizing analytical and data-driven approaches.

Porous materials are widely used for noise control treatments. Sound-absorb cotton is commonly used in the transportation industry such as automobiles and airplanes because of its excellent sound absorption performance. In this paper, two typical acoustical models were used to evaluate the sound absorption coefficient of sound-absorb cotton. By comparing the measurement results with those of acoustical models, suitable model for sound-absorb cotton can be found. Physical parameters are measured and calculated by experimentally and inverse methods. Porosity and density are directly measured, while airflow resistivity is measured by changing boundary conditions between sample and container. Tortuosity and characteristic length are identified by using material acoustic parameters identification software, when samples considered as limp and rigid skeleton-type respectively. When all parameters of the samples are obtained, then the sound absorption coefficient of the sample can be simulated by material acoustic simulation software. This paper, also investigated the effect on sound absorption coefficient by airflow resistivity due to different boundary conditions. The conclusion of the theoretical model applicable to sound-absorb cotton can be drawn from the results obtained, also proved the best method of airflow resistivity measure.

The average sound-absorption coefficient of an ultra-broadband parallel sound absorber, between 0 and 2000 Hz, is optimized in certain conditions. This is accomplished using a teaching–learning-based optimization algorithm, where the geometric parameters are used as optimization variables. The optimized ultra-broadband parallel sound absorber has an average sound-absorption coefficient of 0.9255. The surface impedance of its structure is almost perfectly matched with the surrounding air – thanks to the optimized lengths of both the lateral and micro-perforated plates. The effective-medium model, the transfer-matrix method, and the finite-element method are applied and show consistent results. The acoustic-impedance tube measurements confirm that the optimized ultra-broadband parallel sound absorber shows high sound-absorption from 200 to 1715 Hz. The investigated passive acoustic design opens new possibilities for new absorber types in the future.

Acoustic absorbing materials are used in various applications to reduce noise. In air ducts and pipes, they are used to attenuate noise from devices such as fans. When an absorbing material is installed across the entire cross-section of a duct, sound wave attenuation is large, but so is pressure loss in the flow. For greater effectiveness with less acoustic absorbing material in a one-dimensional tube system, we propose a method in which layers of acoustic absorbing material are separated by air layers and we derive the transmission loss in this method. A parametric study was performed to confirm the effectiveness of the air layers and investigate characteristics of the system. The results show that the method’s effectiveness is due to the antiresonance properties of the air layers. An experiment with an acoustic tube was performed to assess the validity of the proposed method and the derived transmission loss. The experimental and numerical results for transmission loss agree well, indicating the proposed method is effective when acoustic absorbing material layers are used.

A speech enhancement (SE) technique improves the quality and intelligibility of the noise degraded speech. The conventional MMSE-based SE technique minimizes the Bayesian cost function by modifying the magnitude-spectra, whereas keeping the phase-spectra intact. In this current study, both magnitude and phase spectra are modified in order to enhance the noisy speech using a phase compensated perceptually weighted β-order Bayesian estimator. In the proposed SE technique, we first modified the phase of noisy speech spectra alone. Secondly, we used perceptually motivated β-order Bayesian estimator to modify the magnitude-spectra. The estimator holds the advantages of the perceptually-weighted and the β-order spectral amplitude estimator to get a better gain function. The compensated phase spectra and the estimated magnitude spectra are combined to reconstruct the noise attenuated enhanced speech signals. The proposed speech enhancement technique is evaluated for different noise levels (0 dB to +10 dB) in terms of the objective quality and intelligibility measures using the NOIZEUS and AURORA databases. It is observed that the proposed enhancement technique considerably improves the quality and maintains intelligibility in non-stationary and stationary noisy environments. The experimental results show that the average short-time objective intelligibility (STOI), normalized covariance metric (NCM), perceptual evaluation of speech quality (PESQ), segmental SNR (SegSNR), and Likelihood-Ratio (LLR) are improved by 12.2%, 9.25%, 19.69%, 31.25%, and 39.06% over the noisy unprocessed speech signals. The proposed technique also outscored the competing techniques.

The Deep Neural Networks (DNNs) have recently shown a high performance applied to speech classification tasks. In this paper, we argue that the improved accuracy generated by the Deep Convolutional Neural Network (DCNN) classifier is the result of their ability to extract discriminative representations. They are efficient to the different sources of variability in speech signals. We propose, in this study, a new algorithm, called ST-DCNN in order to classify normal and pathological voices. We demonstrate the improvement of recognizing voices theory with advances in speech features in order to improve the identification pathological voices. The proposed approach operates in two steps: First, we extract scatter wavelet features. Then, we introduce the DCNN for voices classification. The performance of the proposed system is evaluated based on silent and noisy environments using various Signal-to-Noise Ratio (SNR) levels. The results underscore that our proposed system shows better performance using scattering wavelet and DCNN in a silent environment with 99.62% of recognition rate.

To clarify the mechanism of the enhancement of heat transfer by sound wave. We have experimentally studied the convective heat transfer characteristics of heated copper spheres under sound waves with different frequencies and sound pressure levels. The experimental results showed that the existence of an acoustic field made strengthen the cooling process of the copper sphere to some extent, and the convective heat transfer coefficient increased nonlinearly with increasing sound pressure level. The contribution of the second-order nonlinear acoustic streaming generated by the interaction between sound waves and the heated copper sphere in the process of enhancing heat transfer was theoretically analyzed. Sound frequency determines the relative strength of acoustic streaming and oscillatory flow. When the sound frequency increases, the influence of the acoustic-induced flows on the heat transfer process of the copper sphere can be divided into four stages in turn: (1) acoustic streaming control stage, (2) acoustic streaming and oscillatory flow coordinated control stage, (3) oscillatory flow control stage, (4) stable stage. There is an optimal sound frequency to enhance heat transfer. Furthermore, the flow field characteristics of the acoustic streaming outside the sphere were numerically calculated. Calculations showed that low-frequency and high-intensity sound waves can form strong acoustic streaming motion outside the sphere. This proves that acoustic streaming in the low-frequency region is the predominant mechanism affecting heat and mass transfer.

Deep learning and Neural Networks strategies have become very popular in the last year as tools for image and data processing. As for acoustics, neural network-based approaches have been typically used to recognize audio patterns or features or to spatially localize a single emitting source like a speaker. More recently, some authors used deep learning to localize multiple-sources exploiting the grid-based approach typical of sound source localization methods or to filter/improve acoustic maps obtained by more traditional techniques like conventional beamforming. This paper wants to propose the use of artificial neural networks (ANNs) for localizing and quantifying multiple sound sources in a grid-less way. The approach uses the microphones Cross-Spectral-Matrix (CSM) as input to the network and provides as output both the location and strength of sources contributing to the acoustic field. The grid-less strategy targets improving spatial resolution and computational efficiency. The proposed solution is discussed on simulated data for assessing its accuracy and sensitivity. Preliminary investigations on real data are also reported.

In this paper, a microphone is arranged to move along a line at a low speed while collecting sound pressure signals, then the mathematical relation is researched to locate the source and the localization processes are as follows. The collected continuous-time signals are segmented and each segment is transferred to the frequency domain by integral transformation. As a result, according to the Riemann Lebesgue theorem, the interference from the negative frequency term is eliminated, a correlation relationship between the real sound source and virtual sound source (a hypothetical sound source) is set up. By using the formed correlation, the position of the sound source is located by founding the maximum of the inner product module through optimization. Analyzing the details of the segmentation, it is found that the correlation will be enhanced when segment time length is integer multiples of signal's half period, and it benefits the localization accuracy. This method is also extended to more complicated multi-point source identification. Simulation examples and experiments show that the proposed method can locate the sound source. This method uses only one microphone, there is no demanding phase matching and calibration work like the of a microphone array, the cost of use is low.

It has been known that the application of diffuse field theory to the sound field with a non-single-exponential sound decay did not comply well with listeners’ experience. Recent researches have concentrated on quantifying and controlling the non-single-exponential decay processes in rooms. However, the effect of non-single-exponential sound decay on listeners’ reverberance has not been clearly known in spite of its importance. The present work tried to explore the relationship between the non-single-exponential decay of sound in rooms and the listener’s reverberance, and further to develop a simple way of predicting subjective reverberance caused by non-single-exponential decay processes. A series of listening experiments were carried out with various computer generated non-single-exponential sound decay patterns using a two alternative forced choice (2AFC) paired comparison method. A simple way of predicting listeners’ reverberance for non-single-exponential sound decays was proposed by combining the results of subjective tests and the analysis of Schroeder decay curves using decay rates at multiple fixed intervals and their effects on listeners’ reverberance. The correlation analyses suggested that the proposed method could predict listeners’ reverberance caused by non-single-exponential sound decays better than conventional parameters within the range of non-single-exponential decay patterns examined in the present work.

Cancelling the effect of acoustic feedback is a challenging task in the design of a behind the ear digital hearing aid. In traditional behind the ear digital hearing aids, feedback cancellation is usually achieved using an adaptive finite impulse response filter, the weights of which are updated using a suitable learning rule. However, the impulse response of the acoustic feedback path in a hearing aid is sparse in nature and traditional feedback cancellation systems are not designed to utilize this sparseness. An adaptive feedback canceller, which is trained using a set of sparse adaptive algorithms is designed in this paper to take advantage of the sparseness. Further, an attempt has been made to enhance the convergence of the feedback cancellation mechanism by introducing an adaptive de-correlation filter as well as using the concept of probe noise injection. The proposed feedback cancellation schemes are shown to provide improved and accurate feedback cancellation over traditional feedback cancellation mechanisms.

This paper focuses on improving the noise attenuation performance of a ducted Helmholtz resonator (HR) system and fully utilizing an available space. The transmission loss achieved by a periodic ducted HR system is depended on the structure and the number of HRs. However, the number of HRs is restricted by the available space in longitudinal direction of the duct. Moreover, such system will occupy a large space and may have some spare space in the transverse direction of the duct. By adding HRs on the available space in the transverse direction, a modified ducted HR system is therefore proposed. The wave propagation in the periodic ducted HR system and the modified HR system are investigated theoretically and numerically. The transfer matrix method is developed to conduct the investigation. The predicted theoretical results fit well with the Finite Element Method (FEM) simulation results. The results indicate that both the noise attenuation band and peak amplitude are increased by adding HRs on arbitrary side of the cross-section of the duct. The proposed modified ducted HR system can improve the noise attenuation performance and fully utilize the available space, and it is practical to be used in an actual ventilation ductwork system.

This paper presents a new method that estimates the fundamental frequency in the case of a real noisy environment when many persons speak at the same time and considers the case of two speakers. It essentially gives an accurate estimation of the pitch characterizing the second speaker. The first pitch is determining by detecting the Autocorrelation of the Multi-scale Product (AMP) of the mixture signal. Then a multiple-comb filters is applied to eliminate the dominant signal. After subtracting the resulting signal from the mixture, we obtain the residual signal. Next, we reapply the AMP to the obtained signal to estimate the second pitch. We get a matrix of the second pitch candidates. We classify its elements into three groups. After, we calculate the mean of each column of the appropriate selected group. Finally, the intrusion pitches of each frame are obtained. Experiments are performed using Cooke database. The results show the robustness and effectiveness of the proposed approach.

In this paper, a novel single microphone channel-based speech enhancement technique is presented. While most of the conventional nonnegative matrix factorization-based approaches focus on generating a basis matrix of speech and noise for enhancement, the proposed algorithm performs an additional process to reconstruct speech from noisy speech when these two elements are highly overlapped in selected spectral bands. This process involves a log-spectral amplitude based estimator, which provides the spectrotemporal speech presence probability to obtain a more accurate reconstruction. Moreover, the proposed algorithm applies an unsupervised learning method to the input noise, so it is adaptable to any type of environmental noise without a pre-trained dictionary. The experimental results demonstrate that the proposed algorithm obtains improved speech enhancement performance compared with conventional single channel-based approaches.

This paper experimentally investigates a new technique for measuring the modal amplitude distribution, sound power transmission and radiation, and far field directivity of the broadband noise from hard walled ducts. The innovative aspect of this method is that it only requires the measurements of the two-point complex coherence function between the acoustic pressures at two closely spaced points on the duct wall. This method is therefore very useful when direct measurements of sound power and directivity are not possible. This paper describes detailed measurements of the sound power spectrum and coherence function from a hard walled circular duct excited at one end by a diffuse sound field. The other open end is terminated within an anechoic chamber with which to measure the radiated sound field at 11 microphones distributed over a polar arc. Measurements of the complex coherence were made at the duct and used to infer the sound power spectrum and far field directivity. This paper demonstrates generally good agreement between direct measurements of sound power and directivity and those inferred from the coherence function. The method is restricted to broadband noise in large ducts in the frequency range where many modes are able to propagate and the modal amplitudes are mutually uncorrelated.

The pulse-echo nonlinear acoustic wave fields generated by focused transducers are calculated numerically and verified experimentally. Simulation results show that in the reflection direction, phase differences between newly generated second harmonic and reflected waves are about π/2, regardless of the rigid and stress-free boundaries; thus, the total second harmonic wave will neither be strongly enhanced nor be cancelled. Experiments are conducted with a focused transducer working in pulse-echo mode and the measured amplitudes of nonlinear waves reflected after water-air and water-aluminum/steel interfaces agree with the simulated results. This work helps advance applications using pulse-echo nonlinear acoustics.

The reciprocity principle provides a possible means of measuring the low-frequency volume velocity of a sound source in arbitrary surroundings; however, the auxiliary transducer in the reciprocity experiment is required to be absolutely reciprocal, which means that the radiation property and the receiving property of the transducer are identical. However, it is not always easy to obtain such a low-frequency reciprocal transducer, because some issues like nonlinearity would lead the transducer to be non-reciprocal. In this study, a modified method is proposed to conduct the reciprocity measurement using an ordinary loudspeaker instead of a strictly reciprocal one as the auxiliary transducer. Firstly, the relationship is established between the measurement error and the inconsistency of auxiliary loudspeaker’s radiation property and receiving property. Furthermore, a modified reciprocity method is proposed to compensate the inconsistency. Finally, experiments are carried out with the modified reciprocity method in both an anechoic room and a reverberation room with two different auxiliary loudspeakers. The experimental results of the modified reciprocity method agree with the results obtained with the free field method very well.

A parametric loudspeaker utilizes nonlinearity of a medium and is known as a super-directive loudspeaker. In this paper, the sound localization in the vertical direction using the upper and lower parametric loudspeakers is confirmed by listening tests and physical measurements. The differences in levels between the upper and lower parametric loudspeakers are varied as a parameter. The direction of sound localization in the vertical plane can be controlled not only when the acoustical axis is set to the right ear but also when it is set to at 5 deg to the right of the right ear. The effect of the level difference between the upper and lower loudspeakers is weaker than the differences observed when using ordinary loudspeakers. We obtained interesting characteristics of the left-right sound localization in the horizontal plane with the upper and lower parametric loudspeakers in the vertical plane. It is found that by setting the parametric loudspeaker at the right ear (that is, the horizontal angle of a listener to it is only 3 deg to the right), the direction of sound localization in the horizontal plane moved approximately 10 deg to the right. Moreover, by setting the parametric loudspeaker 5 deg to the right, the direction of sound localization moves approximately 20 deg to the right. The ILD (Interaural Level Difference) using a dummy head is calculated from the measured left and right sound signals. It is determined that ILDs of the parametric loudspeaker are larger than those of the ordinary loudspeaker. A simple geometrical acoustic model is introduced and analyzed. The analysis helps to explain the measured characteristics.

Visualization of sound field using optical techniques is a powerful tool for understanding acoustical behaviors. It uses light waves to examine the acoustical quantities without disturbing the sound information of the field under investigation. Schlieren imaging is an optical method that uses a camera to visualize the density of transparent media. As it uses a single shot to capture the information without scanning, it can observe both reproducible and non-reproducible sound field. Conventionally, the Schlieren system is applied to high-pressure ultrasound and shock waves. However, since the density variation of air caused by the audible sound field is very small, this method was not applicable for visualizing these fields. In this paper, a spatio-temporal filter bank is proposed to overcome this problem. As the sound is a very specific signal, the spatio-temporal spectrum (in two-dimensional space and time) of the audible sound is concentrated in a specific region. The spatio-temporal filter bank is designed for extracting the sound field information in the specific region and removing noise. The results indicate that the visibility of the sound fields is enhanced by using the proposed method.