Reconstruction of the signal produced by a directional sound source from remote multi-microphone recordings

A mathematical method for reconstructing the signal produced by a directional sound source from knowledge of the same signal in the far field, i.e., microphone recordings, is developed. The key idea is to compute inverse filters that compensate for the directional filtering of the signal by the sound source directivity, using a least-square error optimization strategy. Previous work pointed out how the method strongly depends on arrival times of signal in the microphone recordings. Two strategies are used in this paper for calculating the time shifts that are afterward taken as inputs, together with source directivity, for the reconstruction. The method has been tested in a laboratory environment, where ground truth was available, with a Polaroid transducer as source. The reconstructions are similar with both strategies. The performance of the method also depends on source orientation.     

Rendering moving sound with the doppler effect in sound space

In this report, a new rendering method of a moving sound with the Doppler effect is proposed. In the conventional rendering method of moving sound, Head Related Impulse Responses (HRIRs) are simply changed according to a sound position. However, the Doppler effect cannot be added to a sound in this method. The pitch of a sound object must be controlled using some other rendering method when a sound object moves at high speed. In our method, each HRIR is divided into two components, such as an initial delay and a main wave form. Two initial delays of both right and left ears are recalculated, respectively, based on relative speeds and a propagation path. These new initial delays are used in rendering. Therefore, the Doppler effect is added to a sound automatically only when a sound position is set in this algorithm. Details related to this algorithm are discussed in this report.     

The wavelet response as a multiscale NDT method

We analyze interfaces by using reflected waves in the framework of the wavelet transform. First, we introduce the wavelet transform as an efficient method to detect and characterize a discontinuity in the acoustical impedance profile of a material. Synthetic examples are shown for both an isolated reflector and multiscale clusters of nearby defects. In the second part of the paper we present the wavelet response method as a natural extension of the wavelet transform when the velocity profile to be analyzed can only be remotely probed by propagating wavelets through the medium (instead of being directly convolved as in the wavelet transform). The wavelet response is constituted by the reflections of the incident wavelets on the discontinuities and we show that both transforms are equivalent when multiple scattering is neglected. We end this paper by experimentally applying the wavelet response in an acoustic tank to characterize planar reflectors with finite thicknesses.     

深海脉冲信号簇到达结构特征及其在水下声源定位中的应用

为了实现对深海水下声源的定位,对典型深海环境中的脉冲信号到达结构特征进行了理论分析,给出了声源和接收位置位于近海面时到达信号的簇信号形式近似表达式。当声源和接收位置处于近海面深度时,接收到的信号呈簇状结构形式。在提取到达信号中各簇信号到达时间、幅值等特征参数的基础上,提出了一种通过对到达信号中簇信号特征参数匹配搜索进行水下脉冲信号定位的方法,仿真分析了不同距离和不同深度处的簇信号特征,并利用一次南海海域爆炸声源的声传播实验数据进行了实验验证。结果表明:各簇信号对应的到达时间、幅值等特征参数可用于对声源位置的匹配定位;海上实验中利用单水听器得到的水下声源的距离估计结果与实测距离结果较为一致,对实验中2.0 km至90.4 km内声源的距离估计误差不大于8%。     

利用海底反射信号进行地声参数反演的方法

针对现有反演方法的缺点,提出了一种基于海底反射信号的地声参数高分辨反演方法.它利用短距离声源在不同深度上发射宽带线性调频信号,采用垂直阵进行接收,首先通过匹配滤波方法提取多径到达信息,然后利用海底反射损失曲线,反演海底表层的声速和密度,最后利用浅底层反射信号估计沉积层参数.由于海水中直达波受到内波的强烈影响,选择海底表面反射作为参考,用以可靠地计算浅底层反射的相对到达时间和幅度,从而估计出沉积层的厚度、速度和衰减系数.通过海上实验,验证了利用浅底层反射信号反演参数的有效性. 关键词： 海底参数 反演 浅底层反射信号     

利用冲击声辨识声源材料的特征提取

声源辨识属于环境声识别的范畴,是模式识别的一个重要研究方向。冲击声携带了大量的声源物理信息,因此利用冲击声提取特征进行声源材料辨识是提高声目标识别分类性能的重要途径。对球-板撞击物理模型合成的冲击声连续统,提出使用基函数学习法提取目标特征,同时利用短时傅里叶变换和小波包变换进行特征提取,以此为基础完成被击平板的材料识别。研究结果表明,利用基函数学习法获得的特征,对于冲击声分类的效果明显优于短时傅里叶变换和小波包变换方法,说明利用该方法进行冲击声声源材料辨识的可行性和优势。     

Audible Threshold of Early Reflections with Different Orientations and Delays

Early reflection is an important component in an enclosed sound field. Due to the precedence effect, the early reflection may not be the dominant factor in sound source localization; however, it still has obvious influences on spatial position, loudness, timbre, and etc. Till now, there have sparse studies on evaluation of the audible threshold of early reflections with lacking of general and quantitative results. This work investigated the audible threshold of early reflection with a simplified sound field model under various experimental conditions including the combination of eight incidence angles and five time delays. Three-down-one-up adaptive strategy with three-interval three-alternative forced-choice (3I-3AFC) paradigm was used due to its efficiency and robustness. Results indicate that (1) the audible threshold of early reflections decreases monotonically with increasing time delay relative to the direct sound. Furthermore, a linear equation between early reflection threshold and time delay is established with correlation coefficient higher than 0.9; (2) When the direct sound and the reflection locate in the same half-plane, the audible threshold of early reflections decreases with increasing angle deviation between the direct sound and the reflection. Moreover, a front-back symmetry of early reflection threshold is observed for stimuli below 5 kHz; (3) Considerable variations in early reflection threshold are found among individuals, especially at large angle deviation and time delay of early reflections relative to the direct sound.     

Localization of distinct reflections in rooms using spherical microphone array eigenbeam processing

This paper presents an experimental and comparative study of several spherical microphone array eigenbeam (EB) processing techniques for localization of early reflections in room acoustic environments, which is a relevant research topic in both audio signal processing and room acoustics. This paper focuses on steered beamformer-based and subspace-based localization techniques implemented in the spherical EB domain, including the plane-wave decomposition, eigenbeam delay and sum, eigenbeam minimum variance distortionless response, eigenbeam multiple signal classification (EB-MUSIC), and eigenbeam estimation of signal parameters via rotational invariance techniques (EB-ESPRIT) methods. The directions of arrival of the original sound source and the associated reflection signals in acoustic environments are estimated from acoustic maps of the rooms, which are obtained using a spherical microphone array. The EB-domain-based frequency smoothing and white noise gain control techniques are derived and employed to improve the performance and robustness of reflection localization. The applicability of the presented methods in practice is confirmed by experiments carried out in real rooms.     

Feature extraction for sound source material recognition with impact sounds

Sound source recognition is a part of environmental sound recognition,which is one of the most important research areas in pattern recognition.Impact sounds carry much physical information associated with the sound sources,which makes impact sound based sound source recognition an important approach to improve recognition performance.In this study,the impact sound continuum synthesized with a ball-plate collision model is used for material recognition of the impacted plates.The basis function learning method and time-frequency representation methods,including the short time Fourier transform and the wavelet packet transform,are applied into classification and the recognition results are compared.The result shows that the features obtained by using the basis function learning perform better for material classification of the impacted plates than that by using the short time Fourier transform and the wavelet packet transform.This demonstrates the high efficiency and superiority of this method in material recognition of sound sources.     

The influence of early reflections on the identification and lateralization of vowels

Sound coming directly from a source is often accompanied by reflections arriving from different directions. However, the "precedence effect" occurs when listeners judge such a source's direction: information in the direct, first-arriving sound tends to govern the direction heard for the overall sound. This paper asks whether the spectral envelope of the direct sound has a similar, dominant influence on the spectral envelope perceived for the whole sound. A continuum between two vowels was produced and then a "two-part" filter distorted each step. The beginning of this filter's unit-sample response simulated a direct sound with no distortion of the spectral envelope. The second part simulated a reflection pattern that distorted the spectral envelope. The reflections' frequency response was designed to give the spectral envelope of one of the continuum's end-points to the other end-point. Listeners' identifications showed that the reflections in two-part filters had a substantial influence because sounds tended to be identified as the positive vowel of the reflection pattern. This effect was not reduced when the interaural delays of the reflections and the direct sound were substantially different. Also, when the reflections were caused to precede the direct sound, the effects were much the same. By contrast, in measurements of lateralization the precedence effect was obtained. Here, the lateral position of the whole sound was largely governed by the interaural delay of the direct sound, and was hardly affected by the interaural delay of the reflections.     

Incorporating source directionality into outdoor sound propagation calculations

Many outdoor sound sources, such as aircraft or ground vehicles, exhibit directional radiation patterns. However, long-range sound propagation algorithms are usually formulated for omnidirectional point sources. This paper describes two methods for incorporating directional sources into long-range sound propagation algorithms. The first is the equivalent source method (ESM), which determines a compact distribution of omnidirectional point sources reproducing a given directivity pattern in the far field. This method can be used with any propagation algorithm because it explicitly reconstructs a source function as a set of point sources with certain amplitudes and positions. The second is a directional starter method (DSM), which is developed specifically for the parabolic equation (PE) algorithms. This method derives narrow- or wide-angle directional starter fields, corresponding to a given source directivity pattern, without reconstructing the equivalent source distribution. Although the ESM can also be used for the PE, the DSM is simpler and can be more convenient, especially if the sound propagation is calculated only for one or a few azimuthal directions. While these two methods are found to produce generally distinct starter fields, they nonetheless yield identical directivity patterns.     

High-resolution plane-wave decomposition in an auditorium using a dual-radius scanning spherical microphone array

The spatial and temporal distribution of early reflections in an auditorium is considered important for sound perception. Previous studies presented measurement and analysis methods based on spherical microphone arrays and plane-wave decomposition that could provide information on the direction and time of arrival of early reflections. This paper presents recent results of room acoustics analysis based on a spherical microphone array, which employs high spherical harmonics order for improved spatial resolution, and a dual-radius spherical measurement array to avoid ill-conditioning at the null frequencies of the spherical Bessel function. Spatial-temporal analysis is performed to produce directional impulse responses, while analysis based on the windowed Fourier transform is employed to detect direction of arrival of individual reflections at selected frequencies. Experimental results of sound-field analysis in a real auditorium are also presented.     

Estimation of ultrasound attenuation and dispersion using short time Fourier transform

Determination of the acoustic attenuation and dispersion has important applications in ultrasound tissue characterization and non-destructive material testing. Current signal processing methods Fourier transform of ultrasound signals to get the spectra of amplitude and phase to estimate respectively the attenuation and dispersion of a given medium. These methods are frequency domain method and obsessed with ambiguity issue in the phase unwrapping calculation. Conventional ultrasound velocity measuring method detects the time of arrival of a pulse (or echo) signal, which is a time domain method to compute group velocity (not phase velocity). This paper presents a novel approach based on the short time Fourier transform (STFT)--a time-frequency analysis, to estimate the ultrasonic dispersion and attenuation. Only the amplitude information of the pulse-signal spectra is used. Based on the time-frequency presentation, the attenuation coefficient of the signal is obtained by computing the amplitude decay of pulse spectrum in time domain, while phase velocities are obtained based on the "time-of-flight" (TOF) of the mono frequency component of the pulse signals. As a result, we eliminate the ambiguity issue in phase angle calculation. Furthermore, the proposed method makes the phase velocity pedagogically intuitive for novice users. The paper presents experiments to evaluate demonstrate the performance of the proposed method.     

Manatee position estimation by passive acoustic localization

Passive sound source localization with sensor arrays is based on the estimation of the time difference of arrival (TDOA), and precise TDOA is required to achieve accurate position estimation. For a majority of practical localization systems (based on TDOA estimation with four sensors in two dimensions), only three time delays are computed to determine the location of interest. This paper presents an approach to determine the position of a manatee by using four hydrophones and all the combinations of the TDOAs available. With four hydrophones, six TDOAs are computed and then combined three by three to get 20 possible points for each position to estimate. Experimental results using the Hilbert envelope peak technique to estimate the TDOAs and the least square method to estimate the position are presented. For the tests conducted it is shown that for a manatee call having a high signal-to-noise ratio, the individual position estimated for each of the 20 combinations of TDOAs lies on a straight line, providing a good estimation of the direction of arrival approximately 85% of the time. However, a good estimation of the position is obtained for a manatee near the hydrophone array approximately 55% of the time.     

基于矢量水听器的深海直达波区域声传播特性及其应用

对于深海近水面声源产生的声场, 处于较大深度处的接收器在一定水平距离范围内能接收到直达波. 2014年在某深海海域进行的水声考察实验中, 应用深度为140 m的拖曳声源发射实验信号, 布放在水下3146 m深处的矢量水听器成功地接收到了直达波信号. 本文应用射线理论, 分析了深海直达波区域声场的传播特性, 得出了水平振速与垂直振速的传播损失与声线到达接收点处的掠射角以及收发水平距离之间的关系. 在以上分析的基础上, 提出了一种利用水平振速与垂直振速的能量差估计声源距离的方法, 并结合2014年实验数据对实验中两条航线上8 km范围内的目标声源进行了测距, 测距结果与目标的GPS数据符合得较好.     

Source localization in complex listening situations: selection of binaural cues based on interaural coherence

In everyday complex listening situations, sound emanating from several different sources arrives at the ears of a listener both directly from the sources and as reflections from arbitrary directions. For localization of the active sources, the auditory system needs to determine the direction of each source, while ignoring the reflections and superposition effects of concurrently arriving sound. A modeling mechanism with these desired properties is proposed. Interaural time difference (ITD) and interaural level difference (ILD) cues are only considered at time instants when only the direct sound of a single source has non-negligible energy in the critical band and, thus, when the evoked ITD and ILD represent the direction of that source. It is shown how to identify such time instants as a function of the interaural coherence (IC). The source directions suggested by the selected ITD and ILD cues are shown to imply the results of a number of published psychophysical studies related to source localization in the presence of distracters, as well as in precedence effect conditions.     

深海声影区稀疏时延估计与声源测距

研究了深海声影区中经一次海底反射的多途声线到达垂直双水听器的时延差与声源位置的关系,提出了一种稀疏时延估计与声源测距方法。首先利用近海面布放的短间距垂直双水听器接收一定频带的声信号,然后计算接收信号的广义互相关函数,并利用频谱搬移和稀疏解卷积技术提取时延差,最后通过时延差匹配,估计水下声源的距离。仿真实验表明,在4300 m深海中,所提方法能够正确提取多途到达时延差,估计声影区内的声源距离。海试结果表明,当垂直接收孔径分别为21 m和30 m时,声源测距误差分别小于13.6%和8.1%。上述结果表明,所提出的时延估计方法可适应带宽较窄的接收信号,多途到达时延估计参数可用于实现声影区中的水下声源测距。     

Optimal wavelet filter for suppression of coherent noise with an application to seismic data

Wavelet analysis is combined with the Karhunen-Loève (KL) transform into an innovative hybrid method for locally filtering coherent noise. In applying our method, the original time series is first decomposed with wavelet transform, the scales more contaminated with noise are reduced by an attenuation factor A_f, and the signal is reconstructed using the inverse wavelet transform. Then the KL transform is applied to the reconstructed signal and the behavior of the first energy modes is analyzed as a function of A_f. The point corresponding to a minimum in the first mode is identified with the maximum extraction of the coherent noise. Our methodology is applied with success to seismic data with the aim of locally extracting the relevant coherent noise, namely the ground roll noise. The procedure can be easily extended to other situations where an undesirable signal is associated with a specific set of energy modes.     

A new definition of boundary point between early reflections and late reverberation in room impulse responses

The early reflections in the room impulse response are usually defined as those observed within the initial 80 ms after the arrival of the direct sound, after which time the sound field is called reverberant. This number was chosen from measurements of other functions in a limited number of halls. In order to give an objective foundation to this time separation and to establish a physical indicator for it, a new method is proposed that defines a "transition time t(L)," which is the time at which the energy correlation between the direct plus initial sound and the subsequent decaying sound first achieves a specified low value. For various halls this number is shown and its relevance as a new parameter is discussed.     

运动声源快速定位的声达时差法

针对声达时差法只能用于非运动声源定位的问题,本文提出一种运动声源快速定位方法。该方法以声达时差为基本定位原理,基于声源计算位置对多普勒效应进行解耦并进行声信号多普勒效应修正,根据三角定位方法构建声传播空间矩阵,以声源位置偏差度为目标基于单纯形优化搜索算法进行声源位置快速逼近,实现了对匀速直线运动的单声源的定位追踪,提高定位实时性。该方法将声达时差法拓展到运动声源的定位,同时解决了消除多普勒效应带来的计算过程复杂、运算量大的问题,仅用4个传声器就可实现运动声源的快速定位,突破了传统运动声源识别中对大传声器阵列的依赖。仿真实验和实车运动声源识别实验结果证明了该方法的有效性,本研究为短时发声运动声源的识别提供了一种简便、高效的方法。