基于声音的跟踪摄像的初步研究

仿照生物的趋声性,可以实现摄像头追踪拍摄一个不在其视场范围内的声源;依据图像处理技术,可以精确定位在摄像头视场内的声源。将这两种技术结合起来,建立了一套实验系统。从利用声源发出的声音信号传输到两个分开一定距离的声音探测器上的时间差初步定位声源,再利用声源的运动规律使用图像处理技术识别并精确定位声源,通过系统与PC机间的实时通信和回转部件带动声音探测器和摄像头一起做精确灵活的转动,摄像头最终对准了声源。实现了基于声音的跟踪摄像过程。     

A compound secondary source for active noise radiation control

Based on the analysis of active noise control system with multi-channel monopole secondary sources, a kind of compound secondary source is proposed in this paper. The proposed compound secondary source consists of two closely located monopole sources with their distance much smaller than a wavelength. The characteristics of the compound source are analyzed, and the performances of the active noise control system with the compound secondary sources on the monopole primary sound field and sound radiated by a plate are investigated both numerically and experimentally. It has been found that the proposed compound secondary sources control system can provide higher noise reduction for free field noise radiation control with the same number of control channels. It is shown that the better performance in noise reduction of the compound secondary sources control system is mainly due to the directivity of the secondary sources where the energy radiation pattern of the compound sources is similar to that of the primary sources.     

Interference structure of low-frequency reverberation signals in shallow water

Using numerical simulation, an analysis was conducted of the interference structure of a bottomscattered sound field generated by a wideband point source in shallow water under winter and summer conditions. The scattered signals were received from the place where the source was located and were subjected to Fourier transform with a sliding window. The paper demonstrates the possibility of estimating the waveguide invariant for backscattered signals when processing the sound intensity distributions in wide frequency and distance ranges up to the scattering area. A technique is proposed for reconstructing the twodimensional field of internal waves using variations of the interference pattern of reverberation signals. The influence of wind surface waves on the degree of interference band contrast is illustrated.     

双声源位置和强度识别的三维声强方法

目前在远场识别声源空间位置和强度缺乏行之有效的方法。针对此问题,提出采用四传声器进行三维声强测量,从而构建出声强、声源坐标和声功率的非车线性方程组,求解方程得出声源空间坐标和强度的方法。以3个三维声强探头对两个同频率单极子声源的识别为例,分别利用数值仿真和半消声室内的实验进行方法验证,并对声源的识别空间分辨率做了测试,得出角度识别最大误差为3.83°,为真实值的8.5%,距离识别最大误差0.1 m,为真实距离的10%。结果表明采用该方法空间坐标和声功率识别均具有很高的准确度,双声源的空间位置分辨力也优于远场声全息方法。      

水平阵信号压缩感知用于简正波分离

针对水平阵信号简正波分离过程中常规波束形成分辨率低以及warping模态滤波不适用于复杂声信号的问题,提出水平阵信号压缩感知用于简正波分离的方法。利用压缩感知在方位估计中的高分辨特性,通过估计水平阵接收信号在频率方位角上的二维分布,分离得到各阶简正波的方位谱,并逆Fourier变换得到时域波形。仿真孔径1 km、阵元间隔10 m水平阵接收20~200 Hz伪随机声信号和脉冲声信号,所提方法分离出的各阶简正波与理论波形的相关系数在0.97~1.0。对2011年北黄海声学实验中的海底28元水平阵接收的气枪信号,在合成至1 km孔径后使用压缩感知方法分离简正波,其与warping模态滤波分离得的前5阶简正波相关系数在0.82~0.93。仿真与实验都说明了水平阵信号压缩感知简正波分离方法的有效性。      

Mode detection in turbofan inlets from near field sensor arrays

Knowledge of the modal content of the sound field radiated from a turbofan inlet is important for source characterization and for helping to determine noise generation mechanisms in the engine. An inverse technique for determining the mode amplitudes at the duct outlet is proposed using pressure measurements made in the near field. The radiated sound pressure from a duct is modeled by directivity patterns of cut-on modes in the near field using a model based on the Kirchhoff approximation for flanged ducts with no flow. The resulting system of equations is ill posed and it is shown that the presence of modes with eigenvalues close to a cutoff frequency results in a poorly conditioned directivity matrix. An analysis of the conditioning of this directivity matrix is carried out to assess the inversion robustness and accuracy. A physical interpretation of the singular value decomposition is given and allows us to understand the issues of ill conditioning as well as the detection performance of the radiated sound field by a given sensor array.     

Point source equalised by inverse filtering for measuring ground impedance

A point source for measuring ground impedance which complies with the omnidirectionality requirements of the ANSI S1.18 standard is presented. The source consists of a small aperture connected to a conventional loudspeaker through a cone. Whilst this source radiates with deviations lesser than ±1 dB within ±45°, its frequency response exhibits strong peaks arising from resonances of air inside the cone. To equalise these resonances, inverse filtering is applied. Inverse filters pre-emphasise the electrical signal driving the sound source so that zero-phase (or minimum-phase) cosine-magnitude signals are radiated. Equalisation by inverse filtering has two main advantages in the ground impedance measurement. On the one hand, flattening the frequency response avoids large fluctuations of the excess attenuation curve, which can make difficult its inversion to ground impedance data. On the other hand, inverse filtering shorten the time response of the sound source, this in turn making easier the positioning of a time window to separate ground reflection from reflections/diffractions coming from nearby objects usually present in the acoustic scenario.     

Wave field synthesis of a sound field described by spherical harmonics expansion coefficients

Near-field compensated higher order Ambisonics (NFC-HOA) and wave field synthesis (WFS) constitute the two best-known analytic sound field synthesis methods. While WFS is typically used for the synthesis of virtual sound scenes, NFC-HOA is typically employed in order to synthesize sound fields that have been captured with appropriate microphone arrays. Such recorded sound fields are essentially represented by the coefficients of the underlying surface spherical harmonics expansion. A sound field described by such coefficients cannot be straightforwardly synthesized in WFS. This is a consequence of the fact that, unlike in NFC-HOA, it is critical in WFS to carefully select those loudspeakers that contribute to the synthesis of a given sound source in a sound field under consideration. In order to enable such a secondary source selection, it is proposed to employ the well-known concept of decomposing the sound field under consideration into a continuum of plane waves, for which the secondary source selection is straightforward. The plane wave representation is projected onto the horizontal plane and a closed form expression of the secondary source driving signals for horizontal WFS systems of arbitrary convex shape is derived.     

Intensity and space-time characteristics of the sound field in the underwater sound channel of the Black Sea

Experimental data are presented on the intensity and space-time characteristics of the sound field generated by explosions in underwater and surface sound channels of the Black Sea. The fine field structure is studied as a function of distance and positions of the source and the receiver. The discreteness of the field structure governed by the deterministic nature of the waveguide is revealed, and the destruction of this structure under the effect of the instability of the waveguide parameters is demonstrated. The effect of the rough sea surface on the sound field in the surface channel is studied, and the diffraction-caused propagation loss is estimated. The mechanism of the forerunner formation is considered. The experimentally observed sound field features are compared with the calculations. A possibility for solving the inverse problem is indicated, and the main parameters that are used in the ray method of determining the source coordinates in the underwater channel (i.e., the method earlier proposed by the author on the basis of the intrinsic structure of the sound field) are pointed out.     

Optimal source distribution for binaural synthesis over loudspeakers

When binaural sound signals are presented with loudspeakers, the system inversion involved gives rise to a number of problems such as a loss of dynamic range and a lack of robustness to small errors and room reflections. The amplification required by the system inversion results in loss of dynamic range. The control performance of such a system deteriorates severely due to small errors resulting from, e.g., misalignment of the system and individual differences in the head related transfer functions at certain frequencies. The required large sound radiation results in severe reflection which also reduces the control performance. A method of overcoming these fundamental problems is proposed in this paper. A conceptual monopole transducer is introduced whose position varies continuously as frequency varies. This gives a minimum processing requirement of the binaural signals for the control to be achieved and all the above problems either disappear or are minimized. The inverse filters have flat amplitude response and the reproduced sound is not colored even outside the relatively large "sweet area." A number of practical solutions are suggested for the realization of such optimally distributed transducers. One of them is a discretization that enables the use of conventional transducer units.     

采用分布式球形阵列的球谐波域声场分离方法

为从测得的混合声场信号中提取出需要的目标声场,提出一种基于分布式球形传声器阵列的声场分离方法。该方法依据声场的球谐波分解,利用阵列各传声器采集到的声场声压信号,获得目标声场与干扰声场的球谐波展开系数,进而估计目标声场。该方法利用声场以不同中心展开的球谐波系数之间的变换关系,直接建立传声器测量声压信号与整体坐标系下声场展开系数的方程,与传统的分布式球形阵列声场分离方法,即先求解局部坐标系下声场展开系数,再变换为整体展开系数的方法,进行比较。分别通过数值仿真和实验说明了提出方法的有效性。结果表明:该方法能够从混合声场中较准确地估计出目标声场,并且在干扰声场能量增大时,保持了较小的声场估计误差,相比于传统方法误差增加更少。     

Active resonators for sound control in narrow pipes

A method of active sound control by a local device, namely, an active resonator that provides a resonance absorption or scattering of sound in a broad frequency band, is proposed. The active resonator is modeled as a sound source with an acoustoelectric feedback, which allows the correction of its motion depending on the sound field measured in the vicinity of the resonator. The performance of the method is demonstrated by two examples.     

Limiting Capabilities of the Active Suppression of Sound Harmonic Signals

A system for the active suppression of sound with an open-loop feedback is described in relation to problems of the sound insulation of devices that radiate at discrete frequencies. Its principle of operation is based on creating a compensating field inverse to that of the primary source. The compensating field is formed at a monochromatic frequency that coincides with or is close to the frequency of the compensated signal. The efficiency and limiting capabilities of such a system are investigated.     

空气中声源激发水下声场的水平纵向相关性

声场的空间相关特性是声场的重要特征,对水下探测、水声通讯等各种设备在实际海洋环境中应用的参数选择具有重要意义,是水声工程技术研究的重要基础之一。相比于水中声源激发声场的相关特性研究,空气中声源的相关研究很少。本文推导了空气中声源激发水下声场的水平纵向相关的简正波表达式,并通过数值仿真分析,比较了声源分别位于空气中和水中时水下声场的水平纵向相关特性。对南海海域进行的一次悬挂汽笛空气声源、海底水平阵接收信号海上实验获得的数据进行分析,结果表明:空气中声源位于不同距离时,其发射的声信号激发水下声场的水平纵向相关均存在明显的起伏结构,基于本文提出的空气中声源激发水下声场的水平纵向相关系数的简正波表达式能够较好地解释该现象.      

Subjective and objective verifications of the inverse functions of binaural room impulse responses

The binaural room impulse responses (BRIRs) can be applied to 3-D sound field reconstruction, virtual reality, noise control, et al. Because the BRIRs are non-minimum phase functions, it is difficult to find the exact inverse functions of the BRIRs, especially when there are two or more sources in a reverberant space. In this work, a method was proposed to find the inverse functions of BRIRs with two sound sources in a reverberant space. The concept of time delays and the method of weighted least squares were used to find the causal, however, approximate inverse functions. The accuracy of the inverse functions was first evaluated objectively by a dummy head system. The result shows that the distortion due to crosstalk and room reverberation can be improved by 16∼18 dB. The inverse functions were also verified subjectively by 20 students. The result of subjective evaluation also shows that the inverse functions can be used successfully to reduce the crosstalk effect and the room reverberation.     

Errors in two-point sound reproduction

This paper deals with the problem of reproducing two signals at two points in space by using two acoustic sources. While much is now known about the techniques available for the design of matrices of inverse filters that enable this objective to be achieved in practice, it is still the basic physics of the sound field produced that controls the effectiveness of such systems and which ultimately dictates their design. The basic physical processes involved in producing the cross-talk cancellation that enables the reproduction of the desired signals is revisited here by using a simple two source/two field point free field model. The singular value decomposition is used to identify those frequencies where the inversion problem becomes ill-conditioned and to explain physically the origin of the ill-conditioning. As observed previously, it is found that cross-talk cancellation becomes problematic when the path length difference between the two sources and one of the field points becomes equal to one half the acoustic wavelength. The ill-conditioned frequencies are also found to be associated with a limited spatial region of cross-talk cancellation and with large source outputs manifested in the time domain by responses of long duration.     

A Sound Field Separation and Reconstruction Technique Based on Reciprocity Theorem and Fourier Transform

We show a method to separate the sound field radiated by a signal source from the sound field radiated by noise sources and to reconstruct the sound field radiated by the signal source. The proposed method is based on reciprocity theorem and the Fourier transform. Both the sound field and its gradient on a measurement surface are needed in the method. Evanescent waves are considered in the method, which ensures a high resolution reconstruction in the near field region of the signal source when evanescent waves can be measured. A simulation is given to verify the method and the influence of measurement noise on the method is discussed.     

Analysis with binaural auditory model and experiment on the timbre of Ambisonics recording and reproduction

A scheme for analyzing the timbre in spatial sound with binaural auditory model is proposed and the Ambisonics is taken as an example for analysis.Ambisonics is a spatial sound system based on physical sound field reconstruction.The errors and timbre colorations in the final reconstructed sound field depend on the spatial aliasing errors on both the recording and reproducing stages of Ambisonics.The binaural loudness level spectra in Ambisonics reconstruction is calculated by using Moore's revised loudness model and then compared with the result of real sound source,so as to evaluate the timbre coloration in Ambisonics quantitatively.The results indicate that,in the case of ideal independent signals,the high-frequency limit and radius of region without perceived timbre coloration increase with the order of Ambisonics.On the other hand,in the case of recording by microphone array,once the high-frequency limit of microphone array exceeds that of sound field reconstruction,array recording influences little on the binaural loudness level spectra and thus timbre in final reconstruction up to the highfrequency limit of reproduction.Based on the binaural auditory model analysis,a scheme for optimizing design of Ambisonics recording and reproduction is also suggested.The subjective experiment yields consistent results with those of binaural model,thus verifies the effectiveness of the model analysis.     

Ambisonics声捡拾与重放音色的双耳听觉模型分析与实验

提出用双耳听觉模型对空间声音色进行分析的普遍方法,并以Ambisonics为例进行了分析。Ambisonics是基于物理声场重构的空间声系统,其最终重构声场误差以及音色改变是由传声器捡拾和重放空间混叠误差共同引起的。采用修正的Moore双耳响度模型计算了Ambisonics重构声场的双耳响度级谱并和目标声场的情况比较,从而定量评价重构声场的音色改变。结果表明,在理想捡拾信号的情况下,无音色改变重放的上限频率和区域大小随Ambisonics的阶数而增加。而对于传声器阵列捡拾的情况,只要阵列的上限频率大于Ambisonics重放的上限频率,在重放的上限频率以下,传声器阵列空间混叠误差对最终重构声场及其感知音色的影响就可以忽略。在此基础上,提出了一种综合考虑捡拾与重放性能的Ambisonics系统优化设计方法。心理声学实验得到了和双耳听觉模型一致的结果,从而也验证了模型分析的有效性。      

Time reversal and its application to tomography with diffracting sources

An exact time-domain method is proposed to time reverse a transient scalar wave using only the field measured on an arbitrary closed surface enclosing the initial source. Under certain conditions, a time-reversed field can be approximated by retransmitting the measured signals in a reversed temporal order. Exact reconstruction for three-dimensional broadband diffraction tomography (a linearized inverse scattering problem) is proposed by time-reversing the measured field back to the time when each secondary source is excited. The algorithm is verified by a numerical simulation. Extension to the case using Green's function in a heterogeneous medium is discussed.