Sound Source Identification Research Articles

The estimation of distance and power of multiple sound sources by combining three-dimensional sound intensity and beamforming

The problem of identifying multiple sound sources occurs in various fields. This study proposes a method for multi-source localization and sound power estimation based on three-dimensional sound intensity (3DSI) measurement combined with high-precision beamforming. First, according to the propagation law of the point source in the free field and 3DSI measurement method, a mathematical model of the sound intensity vector, sound power, and source location in 3D space for multiple sources is derived. The 3DSI components can be measured by 3DSI probes. To reduce the computational cost required to solve the mathematical model, the angles of the sources are estimated via compressive sensing beamforming orthogonal matching pursuit for the method. Moreover, the sound intensity components and angles are used as inputs in the multiple sound source identification model, and the dimensions of the model are significantly reduced. The model was solved to obtain sound source distance and sound power. Two tetrahedral intensity probes and a 16-channel microphone array were used for experiments in a semi-anechoic chamber. The results demonstrate that the method can accurately identify the locations of multiple sound sources and determine the sound power. Considering three sound sources, the maximum angular error is 1.41°, while the maximum distance error is 0.08 m. This method can use a small number of probes to obtain the sound sources spatial position and sound source power, and it improve the resolution and calculation efficiency.

Long-term average spectrum measures of consecutive snore sounds from different sources determined by drug-induced sleep endoscopy.

The goal of this study was to investigate the value of the long-term average spectrum in the acoustic analysis of snore sounds arising from different sources in the upper airway. Long-term average spectrum was used to analyze sequences of 10 consecutive snore sounds that had been divided into 2 groups, soft-palate type and lateral-wall type, according to the vibration site generating the snore sounds and the patterns of soft tissue collapse in the upper airway as identified by drug-induced sleep endoscopy. We calculated the first spectral peak, mean spectral energy, high-frequency energy, 0-1 kHz spectral energy, 1-5 kHz spectral energy, and 0-1 kHz/1-5 kHz difference from each group and compared the differences between them. All parameters except mean spectral energy showed significant differences between the 2 groups. The first spectral peak of less than 265.53 Hz, and the 0-1k/1-5 kHz difference of less than -11.6 dB strongly suggests soft-palate-type snore sounds. Long-term average spectrum has potential application for snore sound source identification. We recommend using first spectral peak and a 0-1 kHz/1-5 Hz difference to identify soft-palate-type snore sounds. Peng H, Xu H, Xu Z, Jia R, Yu H. Long-term average spectrum measures of consecutive snore sounds from different sources determined by drug-induced sleep endoscopy. J Clin Sleep Med. 2023;19(1):145-150.

A review for the noise source identification methods based microphone array

Sound source identification is an important prerequisite for noise control. In recent years, new methods of sound source identification have been developed to improve the robustness of source identification. In this paper, a comprehensive review of noise source identification methods is developed to summarize its developing status and engineering applications. At first, a microphone array which is to acquire the sound pressure field is described. And then, the sound field visualization methods (beamforming and acoustic holography) and their advantages and disadvantages are reviewed. Finally, to improve the identification accuracy of the moving sound source, the hybrid acoustic holography methods with more generality and robustness are reviewed. Through the comparison and summary of the sound source identification methods, we expect this work may provide a potential guidance for the subsequent research studies in the field of sound source identification.

A deconvolution sound source identification method based on twice weak selection orthogonal matching pursuit

The deconvolution sound source identification algorithm based on orthogonal matching pursuit has high identification accuracy and spatial resolution. Still, it has a significant defect in that the sparse degree of sound source needs to be known in advance, so it often has significant limitations in practical engineering applications. In this paper, a deconvolution sound source identification algorithm with twice weak selection orthogonal matching pursuit (TWSOMP-DAMAS) is proposed. It can delete the wrong atoms according to twice weak selection criteria in the iterative process, gradually narrow the range of sound sources, and finally find the location of real sound sources. The simulation results show that the TWSOMP-DAMAS algorithm can effectively reduce the main lobe width and has a higher spatial resolution than the deconvolution algorithm with sparse constraints (SC-DAMAS). And the deconvolution sound source identification algorithm with orthogonal matching pursuit (OMP-DAMAS) has the same identification effect; The TWSOMP-DAMAS algorithm is proved to have good adaptability to noise environment, and the recognition results show that the algorithm has high recognition stability.

Performance Enhancement of Functional Delay and Sum Beamforming for Spherical Microphone Arrays

Functional delay and sum (FDAS) beamforming for spherical microphone arrays can achieve 360° panoramic acoustic source identification, thus having broad application prospects for identifying interior noise sources. However, its acoustic imaging suffers from severe sidelobe contamination under a low signal-to-noise ratio (SNR), which deteriorates the sound source identification performance. In order to overcome this issue, the cross-spectral matrix (CSM) of the measured sound pressure signal is reconstructed with diagonal reconstruction (DRec), robust principal component analysis (RPCA), and probabilistic factor analysis (PFA). Correspondingly, three enhanced FDAS methods, namely EFDAS-DRec, EFDAS-RPCA, and EFDAS-PFA, are established. Simulations show that the three methods can significantly enhance the sound source identification performance of FDAS under low SNRs. Compared with FDAS at SNR = 0 dB and the number of snapshots = 1000, the average maximum sidelobe levels of EFDAS-DRec, EFDAS-RPCA, and EFDAS-PFA are reduced by 6.4 dB, 21.6 dB, and 53.1 dB, respectively, and the mainlobes of sound sources are shrunk by 43.5%, 69.0%, and 80.0%, respectively. Moreover, when the number of snapshots is sufficient, the three EFDAS methods can improve both the quantification accuracy and the weak source localization capability. Among the three EFDAS methods, EFDAS-DRec has the highest quantification accuracy, and EFDAS-PFA has the best localization ability for weak sources. The effectiveness of the established methods and the correctness of the simulation conclusions are verified by the acoustic source identification experiment in an ordinary room, and the findings provide a more advanced test and analysis tool for noise source identification in low-SNR cabin environments.

Acoustic far-field prediction based on near-field measurements by using several different holography algorithms.

Near-field acoustical holography (NAH) is a useful tool for sound field reconstruction and sound source identification. In NAH, a basis model is first selected to represent the physical sound field, and then a near-field measurement is made with a microphone array. Next, the parameters in the selected model can be estimated based on the measurements by using an inverse approach, resulting in the sound field near the source being reconstructed so that the sound source location can be identified. But, in addition to being able to reconstruct the near-field of a source, the far-field can also be predicted with the identified sound source model. A significant amount of work has been performed to study the near-field reconstruction capability of different NAH algorithms, but there has been a limited number of publications in which the far-field prediction accuracy, based on the near-field measurement constructed model, has been considered. In the present experimental work, two multi-transducer loudspeakers were placed side-by-side to create a multi-component sound source, and two sets of measurements were conducted: an intensity probe scanned the sound field generated by the loudspeakers in both the near-field (0.05 m) and far-field (0.48 m) such that the true near- and far-field intensity spatial distributions and total sound power could be identified. Then, based on the near-field pressure measurements, four acoustical holography algorithms, statistically optimized near-field acoustical holography, wideband acoustical holography, l1-norm minimization, and a hybrid compressive sampling method were used to predict the near- and far-field sound intensity distributions. The near- and far-field prediction results were compared with the direct measurement, and the sound field reconstruction accuracy was studied. It was found that all of the algorithms were able to reconstruct the near-field well when the near-field measurements were used to construct the model. It was found that with the abovementioned models, far-field reconstructions could correctly predict the spatial sound field distribution, but in all of the cases, the total sound power was underestimated.

Analysis of error induced by the Doppler effect in the ROSI beamforming for rotating sound source identification

Analysis of error induced by the Doppler effect in the ROSI beamforming for rotating sound source identification

Sound Source Identification Based on Orthogonal Matching Pursuit Algorithm Guided by Source Strengthen Prior

Sound Source Identification Based on Orthogonal Matching Pursuit Algorithm Guided by Source Strengthen Prior

Compatibility of quantitative and qualitative data-collection protocols for urban soundscape evaluation

This study investigates the compatibility of soundscape evaluation results based on the data-collection protocols proposed in ISO 12913-2, Method A (questionnaire), Method B (questionnaire and open-answer), and Method C (narrative interview), as well as guidelines for evaluating urban soundscapes. We assessed the soundscapes of 10 multi-functional sites in an urban environment through the responses of 50 participants to questions on sound source identification, perceived affective quality, and overall quality. Using virtual reality technology, we reproduced a laboratory environment similar to the actual assessment sites. The responses to sound source identification were similar for each protocol. Regarding perceived affective quality, the “pleasantness-eventfulness model” derived from Method A was also found in the text-mining results of Methods B and C; additional emotional responses were discovered. Regarding overall quality, the preference for each assessment site was similar for each protocol. Method C revealed the influence of non-acoustic factors on soundscape perception. The quantitative data protocol was appropriate for a large group and for deriving a generalized model, whereas the qualitative data protocol was effective for small groups or an in-depth analysis of certain sites. The findings of this study are useful for assessing urban soundscapes and informing urban planning.

Urban soundscape categorization based on individual recognition, perception, and assessment of sound environments

This study proposes soundscape recognition models by clustering people based on differences in sound source perceptions. We investigated the effect of sound source identification differences on urban soundscape perception by categorizing people’s environmental sound recognition in outdoor environments. Virtual reality technology employing audio-visual stimuli collected in various urban environments replicated actual environments. Fifty participants’ subjective responses regarding sound source identification, perceived affective quality (8 typical (ISO scale) and 116 extensive attributes (Swedish rating scale)), and overall quality were surveyed. Their categorizations by sound source identification were divided into three clusters: Cluster 1–Attentive to traffic noise and other noises, Cluster 2–Less attentive to the sound environment, and Cluster 3–Attentive to natural and human sounds. Even in identical spaces, participants identified different sound sources, as each cluster focused on different sounds. The soundscape perceptual components were derived differently for each cluster; Cluster 2 extracted additional perception dimensions, i.e., tranquil and relaxed soundscapes. The results showed that each sound source that received an attentive reaction had a positive effect on soundscape perception, showing that appropriate human activities can be encouraged to improve relaxation via soundscape enhancements. The overall quality assessment by cluster revealed similar results, but the resulting indicators’ effects varied. The study’s different soundscape recognition models for each cluster, based on the relationship between soundscape indicators and descriptors, present a new perspective for interpreting urban soundscape perception and can also be used effectively in urban planning design.

Sound arrival direction and acoustic scene analysis for the monitoring of airport noise

The authors have developed a sound direction detection method based on the cross-correlation method and applied it to automatic monitoring of aircraft noise and identification of sound sources. As aircraft performance improves, noise decreases, and people are interested in and dissatisfied with low-level noise aircraft, especially in urban areas where environmental noise and aircraft noise combine to complicate the acoustic environment. Therefore, it is necessary to monitor and to measure not only aircraft noise but also environmental noise. Since our surveillance is aircraft noise, it is important to analyze noise exposure from acoustic information rather than trucks or images. In this report, we will look back on the development process of this sound direction detection technology, show examples of helicopters and application examples of acoustic scene analysis to high-altitude aircraft, and consider the latest situation realized as acoustic environment monitoring. We believe that this analysis will make it easier to understand the noise exposure situation at the noise monitoring station. It also describes the future outlook for this method.

A basic study on estimating location of sound source by using distributed acoustic measurement network

The sounds from childcare facilities are often a cause of noise problems with neighbors, however since the sound power levels of children's play and other sounds in child-care facilities have not become clear, evaluation methods have not been established, making countermeasures difficult. In order to evaluate the noise, it is necessary to model the location of the sound source and the sound power level. We have been developing a sound source identification system that uses multiple Raspberry Pi-based recording devices to estimate the location of a sound source and sound power levels. By using GPS for time synchronization, the system can be distributed and placed without connecting cables, which is expected to expand the measurement area significantly. As a method of estimation, the arrival time difference is calculated by cross-correlation from the signals input to each recording device, and the sound source location is estimated from the calculated arrival time difference and the location information of the device. The effectiveness of this system was verified in an anechoic room and outdoor fields.

Machine Learning-Based Tools for Wind Turbine Acoustic Monitoring

The identification and separation of sound sources has always been a difficult problem for acoustic technicians to tackle. This is due to the considerable complexity of a sound that is made up of many contributions at different frequencies. Each sound has a specific frequency spectrum, but when many sounds overlap it becomes difficult to discriminate between the different contributions. In this case, it can be extremely useful to have a tool that is capable of identifying the operating conditions of an acoustic source. In this study, measurements were made of the noise emitted by a wind turbine in the vicinity of a sensitive receptor. To identify the operating conditions of the wind turbine, average spectral levels in one-third octave bands were used. A model based on a support vector machine (SVM) was developed for the detection of the operating conditions of the wind turbine; then a model based on an artificial neural network was used to compare the performance of both models. The high precision returned by the simulation models supports the adoption of these tools as a support for the acoustic characterization of noise in environments close to wind turbines.

Filter-and-sum based high-resolution CLEAN-SC with spherical microphone arrays

High-resolution CLEAN-SC (HR-CLEAN-SC) with spherical microphone arrays is a promising technology for panoramic sound source identification. To further enhance the ability to resolve multiple closely spaced sources at low frequencies, filter-and-sum based HR-CLEAN-SC is proposed in this paper. Simulations and experiments demonstrate that the proposed approach is superior to the existing spherical harmonics beamforming based HR-CLEAN-SC. Its advantages benefit from the use of sound pressure weight vector with strong sidelobe suppression ability. Besides, the proposed approach is further combined with diagonal denosing algorithm to achieve strong anti-interference ability. Finally, two different source marker constraints are compared. Dynamic constraint outperforms fixed constraint with respect to sound source identification performance, but it requires the preparation and storage for the frequency spectrum of maximum sidelobe level as well as more elapsed time.

Potential restorative effects of urban soundscapes: Personality traits, temperament, and perceptions of VR urban environments

This study investigated the potential restorative (PR) effects of urban soundscapes. To establish a new set of PR criteria for urban environments, this study recreated 10 urban sites in virtual reality, instead of using the typical method of questionnaires. A list of restoration-related semantic terms was derived from narrative interviews of 50 subjects who freely expressed their perceptions of the space. PR criteria were determined based on the derived list and two groups were found: the PR group and potential non-restorative group. The objective characteristics of urban soundscapes—acoustic and visual characteristics and characteristics of moving objects—were analyzed. Participants provided their subjective responses regarding sound source identification, perceived affective quality, and overall quality. The Temperament and Character Inventory-Revised-Short Version and Satisfaction with Life Scale were utilized to analyze participants’ individual characteristics. Based on it, a model was developed to predict whether PR criteria could be satisfied, which was 82.6% accurate (85.9% specificity, 71.8% sensitivity). It was determined that the overall level of preference regarding an urban soundscape is a significant factor in satisfying PR criteria. In particular, the design and planning of urban spaces should aim to reduce a space’s loudness and increase the revisitation rate to promote the restoration of the urban residents’ mental health. Furthermore, for individuals, a trait based on acquired experience has a greater impact on the PR effects of urban soundscapes than innate temperament. This study’s findings can serve as supporting data to design sustainable and health-promoting urban soundscapes.

Source identity shapes spatial preference in primary auditory cortex during active navigation

Information about the position of sensory objects and identifying their concurrent behavioral relevance is vital to navigate the environment. In the auditory system, spatial information is computed in the brain based on the position of the sound source relative to the observer and thus assumed to be egocentric throughout the auditory pathway. This assumption is largely based on studies conducted in either anesthetized or head-fixed and passively listening animals, thus lacking self-motion and selective listening. Yet these factors are fundamental components of natural sensing1 that may crucially impact the nature of spatial coding and sensory object representation.2 How individual objects are neuronally represented during unrestricted self-motion and active sensing remains mostly unexplored. Here, we trained gerbils on a behavioral foraging paradigm that required localization and identification of sound sources during free navigation. Chronic tetrode recordings in primary auditory cortex during task performance revealed previously unreported sensory object representations. Strikingly, the egocentric angle preference of the majority of spatially sensitive neurons changed significantly depending on the task-specific identity (outcome association) of the sound source. Spatial tuning also exhibited large temporal complexity. Moreover, we encountered egocentrically untuned neurons whose response magnitude differed between source identities. Using a neural network decoder, we show that, together, these neuronal response ensembles provide spatiotemporally co-existent information about both the egocentric location and the identity of individual sensory objects during self-motion, revealing a novel cortical computation principle for naturalistic sensing.

Unsteady correlation between shear layer vorticity and acoustic refraction in low speed open-jet wind tunnel

There is a close relationship between jet shear layer and acoustic refraction, but research on the unsteady characteristics between them is relatively inadequate. Firstly, the relationship between the sound ray direction and the vorticity, which is the classical ray tracing method, was simply derived, and the applicable conditions were focused on to extend it to the unsteady conditions. Then, the concept of effective area of sound refraction was proposed for beamforming sound source identification array. Sound propagation experiment was carried out in a 3/4 open-jet acoustic wind tunnel by using microphone array and controllable sound source with low flow resistance and self-noise, and the unsteady acoustic drift of initial source with different frequencies was measured. After that, the flow structure of same wind tunnel, especially the jet shear layer, was simulated and verified, which proved that the shear layer formed by the straight section of one nozzle side is quasi two-dimensional. In the simulation, 30 monitoring areas were divided along the shear layer, and the variation of vorticity integral value in each area was analyzed. Finally, similar to the Amiet method, the volume of the effective region was ignored, and several conditions that could be satisfied in this paper were introduced to establish a linear relationship between regional vorticity and acoustic drift. Through the comparison between the acoustic drift of direct measurement and vorticity calculation, the validity of the linear relationship at the time average and instantaneous levels is proved. The research of this paper show that the relationship between vorticity and acoustic ray is still valid at unsteady level as long as the conditions of low-speed flow and high frequency sound are satisfied. At the same time, it provided a new idea for non-contact flow measurement in shear layer of jet structure.

An investigation into high-speed train interior noise with operational transfer path analysis method

Passengers' demands for riding comfort have been getting higher and higher as the high-speed railway develops. Scientific methods to analyze the interior noise of the high-speed train are needed and the operational transfer path analysis (OTPA) method provides a theoretical basis and guidance for the noise control of the train and overcomes the shortcomings of the traditional method, which has high test efficiency and can be carried out during the working state of the targeted machine. The OTPA model is established from the aspects of "path reference point-target point" and "sound source reference point-target point". As for the mechanism of the noise transmission path, an assumption is made that the direct sound propagation is ignored, and the symmetric sound source and the symmetric path are merged. Using the operational test data and the OTPA method, combined with the results of spherical array sound source identification, the path contribution and sound source contribution of the interior noise are analyzed, respectively, from aspects of the total value and spectrum. The results show that the OTPA conforms to the calculation results of the spherical array sound source identification. At low speed, the contribution of the floor path and the contribution of the bogie sources are dominant. When the speed is greater than 300 km/h, the contribution of the roof path is dominant. Moreover, for the carriage with a pantograph, the lifted pantograph is an obvious source. The noise from the exterior sources of the train transfer into the interior mainly through the form of structural excitation, and the contribution of air excitation is non-significant. Certain analyses of train parts provide guides for the interior noise control.

Application of machine learning in ocean data

In recent years, machine learning has become a hot research method in various fields and has been applied to every aspect of our life, providing an intelligent solution to problems that could not be solved or difficult to be solved before. Machine learning is driven by data. It learns from a part of the input data and builds a model. The model is used to predict and analyze another part of the data to get the results people want. With the continuous advancement of ocean observation technology, the amount of ocean data and data dimensions are rising sharply. The use of traditional data analysis methods to analyze massive amounts of data has revealed many shortcomings. The development of machine learning has solved these shortcomings. Nowadays, the use of machine learning technology to analyze and apply ocean data becomes the focus of scientific research. This method has important practical and long-term significance for protecting the ocean environment, predicting ocean elements, exploring the unknown, and responding to extreme weather. This paper focuses on the analysis of the state of the art and specific practices of machine learning in ocean data, review the application examples of machine learning in various fields such as ocean sound source identification and positioning, ocean element prediction, ocean biodiversity monitoring, and deep-sea resource monitoring. We also point out some constraints that still exist in the research and put forward the future development direction and application prospects.

Potential restorative effect of the urban soundscape experience

This study investigated the potential restorative effect of various urban soundscape experiences. The evaluation stimulus was implemented using virtual reality technology in a laboratory environment. Restoration-related semantic expression words were selected from the interview responses of 50 subjects, and the presence or absence of the expression of the words was newly proposed as a restorative potential criteria. The objective characteristics of urban soundscapes were analyzed based on auditory, visual, and object aspects, and subjects responded to sound source identification, perceived affective quality, and overall quality. The temperament and character inventory-revised short version and satisfaction with life scale were used to investigate individual characteristics. Based on this, a prediction model having an accuracy of 83% was proposed for the restorative potential criteria. As a result, it was found that the overall perception of soundscape was a crucial factor in increasing the restorative potential effect of the urban soundscape experience. In addition, it was confirmed that the restorative potential effect by individual characteristics varied significantly even in the same space. The findings of this study are expected to serve as a basis for sustainable and healthy urban soundscape design.