Microphone Signals Research Articles

Time-Aligned Spatial Upsampling of Spherical Microphone Array Recordings

With the Spherical Array Interpolation by Time Alignment (SARITA) method, the authors introduced an approach for spatial upsampling of spherical microphone array (SMA) signals (T. Lübeck, J. M. Arend, and C. Pörschmann, IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 31, pp. 1163–1174 [2023]). The basic idea of this method is to perform the interpolation after time-aligning adjacent microphone signals. The upsampled SMA signals can be represented as spherical harmonic coefficients of much higher spatial order than is possible with the sparsely measured signals. Instead of impulse responses, the method is now applied to SMA recordings. Binaural decoding of upsampled SMA recordings is compared technically and perceptually to a baseline Ambisonics decoding. The results show that the SARITA method can also be applied to low-order array recordings and significantly improves their binaural reproduction.

High-altitude balloon-launched uncrewed aircraft system measurements of atmospheric turbulence and qualitative comparison with infrasound microphone response

Abstract. This study investigates the use of a balloon-launched uncrewed aircraft system (UAS) for the measurement of turbulence in the troposphere and lower stratosphere. The UAS was a glider which could conduct an automated descent following a designated flight trajectory and was equipped with in situ sensors for measuring thermodynamic and kinematic atmospheric properties. In addition, this aircraft was equipped with an infrasonic microphone to assess its suitability for the remote detection of clear-air turbulence. The capabilities of the UAS and sensing systems were tested during three flights conducted in New Mexico, USA, in 2021. It was found that the profiles of temperature, humidity, and horizontal winds measured during descent were in broad agreement with those made by radiosonde data published by the US National Weather Service, separated by up to 380 km spatially and by 3 to 5 h temporally. Winds measured during controlled flight descent were consistent with the winds measured by global-positioning-system-derived velocity during balloon ascent. During controlled descent with this particular payload, a nominal vertical resolution on the order of 1 m was achieved for temperature, relative humidity, and pressure with a nominal vertical resolution of the wind velocity vector on the order of 0.1 m; the aircraft had a glide slope angle from 1 to 4° during this time. Analysis approaches were developed that provided turbulent kinetic energy and dissipation rate, but it was found that the corresponding Richardson number was sensitive to the methodology used to determine the vertical gradients from a single flight. The low-frequency content of the infrasonic microphone signal was observed to qualitatively align with long-wavelength wind velocity fluctuations detected at high altitude. Moreover, the microphone measured more broadband frequency content when the aircraft approached turbulence produced by the boundary layer.

An Experimental Performance Assessment of Temporal Convolutional Networks for Microphone Virtualization in a Car Cabin.

In this paper, the experimental results on microphone virtualization in realistic automotive scenarios are presented. A Temporal Convolutional Network (TCN) was designed in order to estimate the acoustic signal at the driver's ear positions based on the knowledge of monitoring microphone signals at different positions-a technique known as virtual microphone. An experimental setup was implemented on a popular B-segment car to acquire the acoustic field within the cabin while running on smooth asphalt at variable speeds. In order to test the potentiality of the TCN, microphone signals were recorded in two different scenarios, either with or without the front passenger. Our experimental results show that, when training is performed in both scenarios, the adopted TCN is able to robustly adapt to different conditions and guarantee a good average performance. Furthermore, an investigation on the parameters of the Neural Network (NN) that guarantee the sufficient accuracy of the estimation of the virtual microphone signals while maintaining a low computational complexity is presented.

Localization and Sound Power Estimation of Sound Sources in Industrial Plants using the Sound Field Scanning Method

Reducing sound emissions from industrial plants requires an initial step of localizing key contributing sources. Acoustic cameras are pivotal for quickly identifying these sources, yet challenges persist in low-frequency source localization and presenting results for effective impact assessment. This contribution introduces the Sound Field Scanning technology, utilizing a rotating linear microphone array covering a measurement surface with diameter of 2.5 meters for low-frequency source localization from 125Hz. The underlying sound imaging method compensates for Doppler distortions in the moving microphone signals and evaluates coherence spectra with a non-moving reference microphone to compute an acoustic image. For analysis, a method quantifying sound power contributions from specific regions in the optical image is presented. Comparing sound power metrics of the full optical field of view with metrics of regions of interest enables users to quickly rank source contributions and assess the impact of reducing these contributions on total sound power. In complex scenarios, preliminary investigations can be streamlined with a few measurements from diverse angles, expediting the problem-solving process with confidence. The method is exemplified through the analysis of sound emissions from a production plant.

Postfilter for Dual Channel Speech Enhancement Using Coherence and Statistical Model-Based Noise Estimation.

A multichannel speech enhancement system usually consists of spatial filters such as adaptive beamformers followed by postfilters, which suppress remaining noise. Accurate estimation of the power spectral density (PSD) of the residual noise is crucial for successful noise reduction in the postfilters. In this paper, we propose a postfilter utilizing proposed a posteriori speech presence probability (SPP) and noise PSD estimators, which are based on both the coherence and the statistical models. We model the coherence-based a posteriori SPP as a simple function of the magnitude of coherence between two microphone signals and combine it with a single-channel SPP based on statistical models. The coherence-based estimator for the PSD of the noise remaining in the beamformer output in the presence of speech is derived using the pseudo-coherence considering the effect of the beamformers, which is used to construct the coherence-based noise PSD estimator. Then, the final noise PSD estimator is obtained by combining the coherence-based and statistical model-based noise PSD estimators with the proposed SPP. The spectral gain function is also modified, incorporating the proposed SPP. Experimental results demonstrate that the proposed method led to more accurate noise PSD estimation and perceptual evaluation of speech quality scores in various diffuse noise environments, and did not degrade the speech quality under the presence of directional interference, although the proposed method utilizes the coherence information.

Vehicle Control by using Human Voice for Physically Disabled and Aged People

The most crucial form of human communication is speech signals. Almost all encounters use speech signals in conversations. With a microphone, sounds and different speech signals can be transformed into electrical form. Physical disability can result from a variety of causes, such as age-related issues, health issues, and accident-related injuries. For these challenged individuals who suffer from impairments in their hands and legs, vehicles are used as a means of transportation. It might be challenging for anyone with certain conditions, such as paralysis, to manually push a wheelchair or use a remote assembly. For these individuals, the project is made to work with voice-based commands, enabling the paralyzed or impaired person to offer directions simply by speaking into the provided microphone. Additionally, the device has directional buttons for remote vehicle control. An Atmega328-based circuit interfaced with a voice recognition module builds up the system. The voice recognition module receives user commands and translates them into digital data, which the micro-controller analyzes to obtain directional orders. The transmitter circuit and the receiver circuit make up the whole system. The voice recognition module is part of the transmitter circuit, and the motor and driver assembly are part of the one that receives circuit. For communication, an NRF trans-receiver module is used. The command issued to the vehicle is shown on a 16*2 LCD. Keywords: Bluetooth module, Arduino, wireless night vision camera, robotic arm.

IMAR in Frequency Domain for Separating Speech Signals by STFT

Blind Source Separation is a statistical method for isolating signals coming from unidentified sources that have been picked up by a variety of sensors. In the fields of biological signal analysis and speech analysis technology, it is an active researcher. The quantity of noise that is present in the signal throughout the speech signal separation process is a disadvantage. This noise affects the signal that has been separated, and it is often loud melodic noise. For the purpose of separating speech signals from BSS, the suggested method makes use of a network that is configured with the parameters of Instantaneous Mixing in addition to an Auto Regressive model and a maximum-likelihood function. In order to separate the individual voice signal from the input mixed signals, the Back Propagation Network is put into operation. The Short Time Fourier Transform is then utilized for the purpose of dividing the spoken stream into extremely precise time periods. Next, Maximum-likelihood method is implemented to determine the optimum values of the IMAR model's W and G parameters. This step is done after the ideal values have been estimated. The suggested model was tested using a combination of voice signals and microphone signals, and the results demonstrated that it provided superior execution in comparison to other algorithms that are currently in use.

On the Regularization of Recursive Least-Squares Adaptive Algorithms Using Line Search Methods

Stereophonic acoustic echo cancellation (SAEC) requires the identification of four unknown impulse responses corresponding to four loudspeaker-to-microphone pairs. Recent developments in the field of adaptive filters for SAEC setups have allowed for the handling of a single complex-valued adaptive impulse response, instead of the four classical real-valued adaptive filters. With the simplified framework provided by the widely linear (WL) model, more advanced versions of recursive least-squares (RLS) were employed in order to take advantage of their superior tracking speeds when working with highly correlated input signals (such as speech). Despite the performances and numerical stability provided by using exponentially weighted versions of the RLS family in combination with line search methods (LSMs), the SAEC configurations have limited capabilities in mitigating the negative effects caused by high-level disturbances affecting the two microphone signals. Such is the case of double-talk scenarios, which considerably reduce the update accuracy of the adaptive system. This paper analyzes a regularization technique for the named WL-RLS-LSM adaptive filters by adjusting the correlation matrix associated with the input signals and creating a reaction in the update process. The proposed method is designed to considerably slow (or even freeze) the adaptation process while the disturbance is manifested. Simulation results are discussed in order to validate the theoretical content.

Doppler frequency analysis for sound-field sampling with moving microphones

Moving microphones allow for the fast acquisition of sound-field data that encode acoustic impulse responses in time-invariant environments. Corresponding decoding algorithms use the knowledge of instantaneous microphone positions for relating the dynamic samples to a positional context and solving the involved spatio-temporal channel estimation problem subject to the particular parameterization model. Usually, the resulting parameter estimates are supposed to remain widely unaffected by the Doppler effect despite the continuously moving sensor. However, this assumption raises issues from the physical point of view. So far, mathematical investigations into the actual meaning of the Doppler effect for such dynamic sampling procedures have been barely provided. Therefore, in this paper, we propose a new generic concept for the dynamic sampling model, introducing a channel representation that is explicitly based on the instantaneous Doppler shift according to the microphone trajectory. Within this model, it can be clearly seen that exact trajectory tracking implies exact Doppler-shift rendering and, thus, enables unbiased parameter recovery. Further, we investigate the impact of non-perfect trajectory data and the resulting Doppler-shift mismatches. Also, we derive a general analysis scheme that decomposes the microphone signal along with the encoded parameters into particular subbands of Doppler-shifted frequency components. Finally, for periodic excitation, we exactly characterize the Doppler-shift influences in the sampled signal by convolution operations in the frequency domain with trajectory-dependent filters.

IMAR in Frequency Domain for Separating Speech Signals by STFT

Blind Source Separation is a statistical method for isolating signals coming from unidentified sources that have been picked up by a variety of sensors. In the fields of biological signal analysis and speech analysis technology, it is an active researcher. The quantity of noise that is present in the signal throughout the speech signal separation process is a disadvantage. This noise affects the signal that has been separated, and it is often loud melodic noise. For the purpose of separating speech signals from BSS, the suggested method makes use of a network that is configured with the parameters of Instantaneous Mixing in addition to an Auto Regressive model and a maximum-likelihood function. In order to separate the individual voice signal from the input mixed signals, the Back Propagation Network is put into operation. The Short Time Fourier Transform is then utilized for the purpose of dividing the spoken stream into extremely precise time periods. Next, Maximum-likelihood method is implemented to determine the optimum values of the IMAR model's W and G parameters. This step is done after the ideal values have been estimated. The suggested model was tested using a combination of voice signals and microphone signals, and the results demonstrated that it provided superior execution in comparison to other algorithms that are currently in use.

Detection of air- and water-filled cavities beneath concrete plate using electromagnetic and acoustic waves

Underground excavation and tunneling may result in cavities filled with air or water. This study aims to detect cavities beneath underground concrete walls or tunnel linings and assess their risks by estimating the cavity thickness and width using electromagnetic and acoustic waves. In this study, air- and water-filled cavities are simulated, and nondestructive testing methods such as ground penetrating radar (GPR), time domain reflectometry (TDR), and microphones are used to investigate the thickness and width of a cavity beneath a reinforced concrete plate. The results show that GPR signals converted from temporal to spatial scales, combined with TDR measurements, can be used to accurately examine the cavity thickness. However, the examination of the cavity width based on GPR signals remains challenging because of the strong parabolic patterns caused by steel bars. Therefore, this study employs flexibility, which is the ratio of the particle velocity of the structure to the load signal according to the frequency. The flexibility of the concrete plate calculated using microphone signals increases in the cavity section for both air- and water-filled cavities because the flexural vibration mode prevails at the cavity–concrete plate interface. The characteristics of the microphone signals analyzed in the context of flexibility enabled the prediction of the cavity width beneath the concrete plate in both air- and water-filled cavities. GPR and TDR signatures combined with microphone signals can be used for the detection and risk assessment of cavities beneath underground structures when subjected to variations in the groundwater level.

Seismometers as infrasound sensors

An experiment was conducted in West-Central Mississippi in which five explosive charges were detonated. The TNT equivalent sizes of the charges ranged from 0.57 to 10.91 kg (1.25 to 24 lb). Among the arrays of sensors deployed, seismometers were deployed near microphones at distances of 0.5, 2.1, and 8.4 km from the source. The blast wave at these distances had decayed in amplitude to an acoustic wave. The coherence between the seismometer and microphone signals showed that the seismometer provided reasonable representation of the acoustic wave over limited frequency bands; however, these bands changed between sensor locations. In addition, two 3-component seismometers were deployed near each other 8.4 km from the source. These seismometers were of different types, one having a resonance frequency of 1 Hz and the other at 4.5 Hz. The signals from the horizontal components of these seismometers were analyzed to determine their effectiveness as vector sensors. The results showed that the back-azimuth determined from the seismometers agreed reasonably well with ground truth for the first arrival of the acoustic wavefront; however, the results degraded as the trailing part of the wavefront passed. Permission to publish was granted by the Director, Geotechnical and Structures Laboratory.

Phase-shifted beamforming for denoising acoustic reflective signals

When the beamforming algorithm is used to locate the fault of the internal structure, the effective fault acoustic signal will be reflected by the structure shell during its transmission process. This results in the microphone signals being polluted, and then the positioning accuracy being low. In this paper, a phase-shifted beamforming algorithm is proposed to segment the microphone signal, extract the time delay of the reflected noise by using the cross correlation property, and then subtract the reflected noise from the raw data to obtain the microphone signal with weakened reflected signal. The newly generated signals are utilized to calculate the cross-spectral matrix of beamforming, which is then used to locate vibration faults. Simulation results show that the phase-shifted beamforming algorithm can more accurately locate the vibration faults of reflective structures.

Characterization of Source-Filter Interactions in Vocal Vibrato Using a Neck-Surface Vibration Sensor: A Pilot Study.

Vocal vibrato is a singing technique that involves periodic modulation of fundamental frequency (fo) and intensity. The physiological sources of modulation within the speech mechanism and the interactions between the laryngeal source and vocal tract filter in vibrato are not fully understood. Therefore, the purpose of this study was to determine if differences in the rate and extent of fo and intensity modulation could be captured using simultaneously recorded signals from a neck-surface vibration sensor and a microphone, which represent features of the source before and after supraglottal vocal tract filtering. Nine classically-trained singers produced sustained vowels with vibrato while simultaneous signals were recorded using a vibration sensor and a microphone. Acoustical analyses were performed to measure the rate and extent of fo and intensity modulation for each trial. Paired-samples sign tests were used to analyze differences between the rate and extent of fo and intensity modulation in the vibration sensor and microphone signals. The rate and extent of fo modulation and the extent of intensity modulation were equivalent in the vibration sensor and microphone signals, but the rate of intensity modulation was significantly higher in the microphone signal than in the vibration sensor signal. Larger differences in the rate of intensity modulation were seen with vowels that typically have smaller differences between the first and second formant frequencies. This study demonstrated that the rate of intensity modulation at the source prior to supraglottal vocal tract filtering, as measured in neck-surface vibration sensor signals, was lower than the rate of intensity modulation after supraglottal vocal tract filtering, as measured in microphone signals. The difference in rate varied based on the vowel. These findings provide further support of the resonance-harmonics interaction in vocal vibrato. Further investigation is warranted to determine if differences in the physiological source(s) of vibrato account for inconsistent relationships between the extent of intensity modulation in neck-surface vibration sensor and microphone signals.

On the Origin of Cardiovascular Sounds Recorded from the Ear.

A study was conducted on 10 subjects, who were instructed to stand up, lie down and exercise with a body-coupled microphone mounted in each ear, using personalised ear-pieces. The subjects were additionally instructed to follow a guided breathing session. Recorded body-coupled microphone signals were evaluated against standard measures, such as electrocardiography, photoplethysmography, and respiration flow and effort. All signal modalities were synchronised to a common trigger signal. For the analysis, the body-coupled microphone and photoplethysmography signals were epoched in accordance with the electrocardiography R-peak. Cluster-based permutation test showed that the body-coupled microphone signals were time-locked to the electrocardiography R-peak. No difference was found between ears. Comparing the timing of the electrocardiography and photoplethysmography events suggests that signals recorded using body-coupled microphones in the ears reflect blood pressure waves in the arteries in the proximity of the bodycoupled microphones.

A nested generalized sidelobe canceller for source counting, localization, and signal separation in reverberant fields

This paper proposes a nested generalized sidelobe canceller (NGSC) for typical array signal processing tasks, including source counting, localization, and signal separation. Multiple blocking matrices are arranged in a nested structure to successively eliminate dominant sources until the remaining signal is predominantly incoherent noise. The microphone signals are dereverberated using a multichannel weighted prediction error algorithm. In source counting, the number of sound sources is determined by tracking the average power of the blocked signal. In source localization, the direction-of-arrival (DOA) is estimated via cosine similarity in conjunction with the golden section search. In signal separation, the estimated DOA enables speech separation in a linearly constrained minimum variance beamformer with postfiltering (LCMV-PF). Monte Carlo simulations are performed to compare the proposed NGSC approach with four baselines, minimum description length, second order statistic of the eigenvalue, multistage Wiener filtering, and multiple signal classification. The results show that NGSC achieves at least 28.80% higher source counting accuracy with a 1.04° lower root mean square degree error than the baselines. The signal-to-distortion ratio achieved by LCMV-PF is 1.52 dB higher than that achieved by the linearly constrained minimum power beamformer and the multichannel Wiener filter.

Counting and Localization of Multiple Unknown Harmonic Sound Sources in Entire 3D Space Using Just Five Microphones

Counting and localization of multiple sound sources in an unknown environment has many applications, namely fault diagnosis of complex machineries, hearing aids' development, and localization of enemy vehicles in defense. However, multiple sound source localization is extremely challenging because it requires numerous microphones, which increases the cost, computational time, and installation difficulties. Here, we present a new method to count the number of stationary harmonic sound sources and localize them in a free-field environment using a novel oblique square pyramid shaped array of only 5 microphones. We find FFT peaks of microphone signals and compare the ratio of the magnitude of the most dominant peak with remaining peaks' one-by-one. If these ratios are below 25 dB difference, they are considered as sound source and counted. A new time delay and intensity ratio computation is applied to the segregated source signals to calculate exact coordinates of each source. Simulation studies of four concurrent sources show that our method predicts all sources' coordinates at every octant location with mean error of 3.4%, and less than 5% error beyond 20 dB Signal-to-Noise ratio. Mean computational time is 0.92 second. Therefore, our method is low-cost, quicker, more accurate and more reliable than existing techniques.

An analysis of a feedback active noise control system using the remote microphone technique

This paper examines strategies in virtual sensing feedback active noise control systems and compares the noise reduction performance of three strategies: using physical microphones for direct feedback control, using virtual microphones for direct feedback control, and using physical microphones for indirect virtual feedback control. For a primary sound field with incident waves in a single direction, simulation results indicate that employing a physical microphone for indirect virtual feedback control consistently yields superior noise reduction performance than that using a physical microphone for direct feedback control. However, if the causal requirement between physical and virtual microphones is not satisfied, the increase in the equivalent secondary path delay deteriorates noise reduction performance while utilizing physical microphones for indirect virtual feedback control. For a primary sound field containing waves from multiple directions, the simulation results show that the noise reduction performance is affected by both the coherence and causality between the physical and virtual microphone signals.

Intermittency of Flame Structure and Thermo-Acoustic Behavior in a Staged Multipoint Injector Using Liquid Fuel

Abstract This study investigates the impact of the staging factor, the ratio between the fuel injected through the pilot stage and the multipoint injection, on the flame dynamic. The BIMER combustor is an atmospheric pressure rig equipped with two corotating swirling air injections (a fixed amount of around 87% of the air goes inside the multipoint stage) and two fuel injection paths for staged combustion. Liquid dodecane is injected with air preheated at 437 K with a global equivalence ratio of 0.6 and a thermal power of around 72 kW. The change of the staging factor from 100% (pilot-only injection) toward 0% (multipoint-only injection) generates changes in the flame-shape which bifurcates from an anchored V-flame into a lifted flame. This flame shape bifurcation appears at a staging of factor around 25%. Around this staging factor, one can witness multistable flames where the flame structure transits randomly between five different states. Processing microphone signals recorded in the chamber provides an understanding of the flame dynamics. The attached flame presents limited pressure fluctuations level at 270 Hz, while the lifted flame features high-pressure fluctuations at 323 Hz. The intermittency between the five states (including the two stable states) is investigated.

Beamformed Envelope Spectrum of acoustic signals for bearing diagnostics under varying speed conditions

Rolling element bearings are critical components in rotating machinery. Early damage detection methods are usually based on vibration signals from accelerometers. Microphones provide an interesting alternative, as they acquire system level information and could allow reducing the number of sensors. However, the bearing signature is mixed in these signals with noise coming from other nearby sources. Beamforming methods can enhance the targeted bearing signature while suppressing the contributions of other noise sources in the environment. The bearing impulses related to the localized surface defects are function of speed, and thus under varying speed conditions become nonstationary. Angular resampling based methods are therefore needed to compensate for the speed fluctuations in order to detect bearing damages. The goal of this paper is to propose a method for bearing diagnostics under varying speed conditions using an array of microphones and an encoder. The method is based on angular resampling of the microphone signals in function of the shaft speed and enhancing them using delay-sum beamforming based on the known positions of the microphones relative to the tested bearing. The method is validated on an experimental dataset acquired at a KU Leuven bearing test rig, under varying rotating speed conditions.