Acoustic Transfer Function Research Articles

Influence of Voice Focus on Auditory Feedback Control of Speech Using Long-Term Average Spectrum, Phon Spectrum, and Accelerometry.

This study investigates the effect of voice-focus adjustments on oronasal balance and auditory feedback control of speech via analyzing spectral distribution, perceived loudness, and nasal vibrations during sustained phonation and passage reading. Twenty-five speech-language pathologists sustained /a/ and read passages with forward, backward, and natural voice focuses in quiet and noisy conditions. The low-frequency power (LFP) below 3 Hz of vocal fundamental frequency was analyzed to access audio-vocal feedback control. Long-term average spectra of speech were converted to phon spectra based on equal-loudness contours ISO 226:2003 to estimate perceived loudness of self-voice across different conditions. Nasal vibrations were also recorded using a digital accelerometer to measure oronasal coupling. Forward-focused voice enhanced the nasal acceleration and reduced LFP, suggesting that increasing the degree of oronasal coupling can improve audio-vocal feedback control compared to natural and backward-focused voices. Voice-focus adjustments were most related to average power and phon values in the 0.2- to 0.3-kHz band. In noise, perceived loudness in the 0.5- to 2.3-kHz band effectively predicted LFP, outperforming the average spectral power of the same band. Voice-focus adjustments significantly impact both the acoustic transfer function of the vocal tract and the spontaneous fine-tuning of audio-vocal integration. This influence becomes particularly pronounced when sound intensity or perceived loudness is changed within the frequency range of 0.2-2.3 kHz, depending on the voice focus selected by the speakers. https://doi.org/10.23641/asha.27183483.

A development on the interior noise prediction technology using active noise control algorithm

When developing NVH of a vehicle, identifying a noise transmission vulnerable path is a very important task. The traditional TPA(Transfer Path Analysis) is based on a physical theoretical equation and calculates the contribution of each path by multiplying the ATF(Acoustic Transfer Function) of each path by the driving load. In this case, it is often difficult to calculate the driving load, and M/H consumption is also high. OTPA(Operational Transfer Path Analysis) performs transfer path analysis only with driving signals for each route, but has a disadvantage of insufficient physical causality. In this study, we proposed an efficient indoor noise TPA method that modifies the ANC algorithm by combining the reference signals while driving with the secondary path of Filtered-X LMS. The proposed method uses the secondary transfer function of Fx-LMS as the ATF for each path. It is not necessary to obtain physical load by predicting target noise with the Ref signal for each route. Therefore, it is possible to predict indoor noise with low M/H. In this study, it can be confirmed that the proposed method shows similar results compared to the existing methods and efficiently identifies main noise path for interior noise.

Realization of global audio telepresence via a learning-based model-matching approach with an acoustic array system

A Global Audio Telepresence (GOAT) system requires a microphone array to capture the spatial audio signals in the far end and a loudspeaker array to reconstruct the sound field in the near end. This seamlessly immerses near-end users in remote audio scenes with full ambience. In this paper, we use a learning-based GOAT system (L-GOAT) based on the model-matching principle, where a deep neural network (DNN) acts as non-linear filters for the GOAT system. The network training attempts to minimize the matching error between the signals reproduced by the DNN and the desired signals filtered by the far-end acoustic transfer functions (ATFs). Extensive simulations were carried out for multi-source scenarios in two different rooms with different reverberation times. To implement the L-GOAT system, a five-microphone linear array was adopted in the far-end room, while a six-loudspeaker array was utilized in the near-end room. The objective evaluation matrices, including the Perceptual Evaluation of Speech Quality (PESQ), Short-Time Objective Intelligibility (STOI), and the matching errors, were conducted to validate the efficacy of the GOAT systems. The proposed learning-based approach has demonstrated superior performance compared to a conventional digital signal processing (DSP)-based method.

A time-domain approach to predict sound radiation from track vibrations

With the intended shift of transport from road to railway, noise exposure due to railway noise will increase. Auralisation of the noise is a key tool for assessing noise perception beyond equivalent sound pressure levels. Researching the connection between, for example, track design parameters and perceived sound quality holds potential for perception-centered track design. A computationally efficient model is required to render the sound field generated by the track vibration. In this paper, using simulated time-domain surface velocities of a rail and sleepers as input, such a model is presented. The sound pressure signal in a position on the side of the track is efficiently predicted by convolving these velocities with precalculated acoustic transfer functions. For a typical ballasted track, the necessary spatial discretisation of the evaluated velocities is determined in a convergence study. Further, the required length of the finite section of the track to be included in the prediction is specified. Finally, pass-by signals generated by track vibration are demonstrated.

Physics-constrained adaptive kernel interpolation for region-to-region acoustic transfer function: a Bayesian approach

A kernel interpolation method for the acoustic transfer function (ATF) between regions constrained by the physics of sound while being adaptive to the data is proposed. Most ATF interpolation methods aim to model the ATF for fixed source by using techniques that fit the estimation to the measurements while not taking the physics of the problem into consideration. We aim to interpolate the ATF for a region-to-region estimation, meaning we account for variation of both source and receiver positions. By using a very general formulation for the reproducing kernel function, we have created a kernel function that considers both directed and residual fields as two separate kernel functions. The directed field kernel considers a sparse selection of reflective field components with large amplitudes and is formulated as a combination of directional kernels. The residual field is composed of the remaining densely distributed components with lower amplitudes. Its kernel weight is represented by a universal approximator, a neural network, in order to learn patterns from the data freely. These kernel parameters are learned using Bayesian inference both under the assumption of Gaussian priors and by using a Markov chain Monte Carlo simulation method to perform inference in a more directed manner. We compare all established kernel formulations with each other in numerical simulations, showing that the proposed kernel model is capable of properly representing the complexities of the ATF.

Source characterization of a ducted source using acoustic free velocity

A common method to characterize linear time invariant ducted acoustic sources is the electroacoustic analogy. The sources are typically defined by their internal source strength and source impedance. Various methods have emerged to determine the source characteristics, broadly categorized into direct and indirect methods. Direct methods involve using a secondary source with significantly higher amplitude to measure source impedance, but those methods are unable to determine source strength. Indirect methods vary acoustic loads to obtain both source strength and source impedance, but accuracy depends on appropriate load combinations. In this research, we propose an inverse method to determine the acoustic source strength from the measured acoustic transfer function and the operational acoustic pressure. This approach yields the “acoustic free velocity,” corresponding to a volume velocity source that characterizes the original sound source. Subsequently, the source impedance is obtained using the acoustic free velocity in conjunction with the load impedance and acoustic pressure in the acoustic circuit. The method is demonstrated using a compressor driver source attached to an impedance tube. Validation of the measured source strength and source impedance is conducted using a two-load method. Prediction of muffler insertion loss shows good correlation with experimental measurements.

Uncovering avalanche sources via acceleration measurements

Avalanche sources describe rapid and local events that govern deformation processes in various materials. The fundamental differences between an avalanche source and its associated measured acoustic emission (AE) signal are encoded in the acoustic transfer function, which undesirably modifies the properties of the source. Consequently, information about the physical characteristics of avalanche sources is scarce and its exposure poses a great challenge. We introduce a novel experimental method based on acceleration measurements, which eliminates the effect of the transfer function and distills the avalanche source. Applying this method to deformation twinning in magnesium shows that the amplitudes and characteristic times of avalanche sources are unrelated by a clear physical law. Conversely, the amplitudes and durations of AE signals are related by a power law, which is attributed to the transfer function. Using our method, we identify and compute a new feature of avalanche sources, which is directly linked to the growth rate of the twinned volume. This feature displays a power-law distribution, implying an unpredicted behavior at dynamic criticality. Simultaneously, the characteristic times of avalanche sources possess an intrinsic upper bound, indicating a predicted limit that relates to the underlying physical process of twinning.

Online adaptation of fourier series-based acoustic transfer function model and its application to sound source localization and separation

In this paper, we propose an online adaptation method for Fourier series-based acoustic transfer function (FS-ATF) models for robot audition systems using microphone array signal processing. The ATF represents the characteristics of signal propagation from a sound source to a microphone, which is essential for sound source localization (SSL) and sound source separation (SSS). For practical applications in dynamically changing real-world environments, ATF models must satisfy two criteria: (1) they must be adaptable to changes in the acoustic environment; and (2) they must be lightweight to be suitable for resource-constrained systems, such as robots with limited memory and computational capacity. The proposed method addresses these challenges using Fourier series expansions for interpolation, which reduces the memory footprint of the ATF model and facilitates online adaptation to acoustic environmental changes. The experimental results demonstrate that the proposed online adaptation method both improves the SSL and SSS performance while reducing the size of the ATF model, which represents a significant improvement over existing online ATF adaptation methods.

Source characterization of a ducted source using acoustic free velocity

A common method to characterize linear time invariant ducted acoustic sources is the electro-acoustic analogy. The sources are typically defined by their internal source strength and source impedance. Various methods have emerged to determine the source characteristics, broadly categorized into direct and indirect methods. Direct methods involve using a secondary source with significantly higher amplitude to measure source impedance but cannot determine source strength. Indirect methods vary acoustic loads to obtain both source strength and source impedance, accuracy depends on appropriate load combinations. In this research, we propose an inverse method to determine the acoustic source strength from the measured acoustic transfer function and the operational acoustic pressure. This approach yields what we term as the "acoustic free velocity," corresponding to a plane monopole source that characterizes the original sound source. Subsequently, the source impedance is obtained using the acoustic free velocity in conjunction with the load impedance and acoustic pressure in the acoustic circuit. The method is demonstrated using a compressor driver source attached to a 1 3/8" impedance tube. Validation of the measured source strength and source impedance is conducted using a two-load method. Additionally, prediction of muffler insertion loss shows a good correlation with experimental measurements, affirming the efficacy of the proposed methodology.

Analytical modeling of thermoacoustic instability influences in gas turbine combustors: A detailed parameter sensitivity analysis

In this study, the effects of various system parameters on combustion instability were numerically identified through a network-based thermoacoustic model using the concept of a transfer function. The acoustic transfer function was derived from the wave equation and various conservation laws in a two-duct system consisting of a nozzle and combustion chamber that is a simplified form of a gas turbine combustor. From the current numerical approach, the major parameters affecting the system acoustics, such as the duct length, area ratio, sound speed ratio, and Mach number, as well as the inlet and outlet acoustic boundary conditions, were defined explicitly, and a method for intuitively examining their influence was presented. The gain (nf) of the flame transfer function was noted to have a complex effect in combination with factors such as temperature ratio and area ratio, and the time delay (τf) was one main dominant factor that determined the main characteristics of instability, resulting in mode transition as well as changes in frequency and growth rate of the instability.

Multizone sound field reproduction using pressure matching with sparse equivalent source

Multizone sound field reproduction aims to create different acoustic environments in multiple spatial zones, allowing listeners to enjoy their individual sounds without being disturbed by sound from other zones. The conventional pressure matching method is used in practice to reproduce the desired sound field by minimizing the sound field error with a limited number of control points, whose reproduction performance decreases above the cutoff frequency. In this paper, a multizone sound field reproduction method based on a sparse equivalent source method is proposed to reproduce the sound field in the whole target region. The acoustic transfer function of each loudspeaker and the desired sound field are represented by the sparse equivalent sources. The whole target region is controlled by interpolating fine virtual control points within the region of the uncontrolled points, and the acoustic transfer functions between the virtual control points and the loudspeaker array are interpolated by the sparse equivalent sources. An optimization reproduction model with the objective of reproducing a desired sound field in the bright zone and constraining the acoustic energy in the dark zone is formulated to find the sparse loudspeaker weights. The simulation results and experimental results show that the proposed method achieves better performance than the conventional pressure matching method in free field and reverberant environments.

Search domain dimension expansion of robust matched-field source radiated power estimation

Compared with the traditional source radiated power estimation method, the matched-field source radiated power estimation method can obtain more accurate estimation results. However, its performance is greatly influenced by environmental mismatch. This paper defines the impact factor of environmental mismatch to quantify the impact of environmental mismatch on the acoustic transfer function and the matched-field source radiated power estimation method. It determines that the key to improve the robustness of environmental mismatch is to accurately estimate the acoustic transfer function. In the traditional matched-field source radiated power estimation method, the search domain of the acoustic transfer function is a distance-depth two-dimensional plane and does not contain the true acoustic transfer function when the environment is mismatched. Therefore, the estimation results of the acoustic transfer function are bound to deviate from those of the true acoustic transfer function. The paper expands the search domain dimension to a three-dimensional search domain that contains distance, depth and uncertain parameter sets. The true acoustic transfer function is included in the search domain. Thus the optimal alternative acoustic transfer function is equal to the true acoustic transfer function. The objective function for the three-dimensional search domain is constructed. The smaller the absolute value of the objective function, the less the acoustic transfer function estimation deviates from the true acoustic transfer function. The source radiated power is estimated more accurately when the alternative acoustic transfer function estimation corresponding to the minimum absolute value of the objective function. The simulation results verify the theoretical analysis results.

Efficient calculation of the three-dimensional sound pressure field around a slab track

The wavenumber domain Boundary Element (2.5D BE) method is well suited to calculate the acoustic sound field around structures with a constant cross-section along one dimension, such as noise barriers or railway track. By expressing the sound field along this dimension in wavenumber domain, the numerical model is reduced from a 3D model to 2D model at each wavenumber. A consequence of the required discrete Fourier domain representation is that the sound field is represented by periodically repeating sections, of which only one section is physically meaningful. The resolution and the number of required wavenumbers increases with the desired length and spatial discretisation of this section. Describing the sound field adequately to auralise the sound without disturbing artefacts requires a large number of wavenumbers (and thus 2D BE computations), which is not feasible for large geometries. Here, a method is introduced that allows the calculation of the 3D sound field by solving a single 2D BE problem for a dense frequency spectrum and interpolating at higher wavenumbers. The calculation efficiency is further increased by precalculating the acoustic transfer functions between each BE surface element and receiver positions. Combining these two methods allows for efficient calculation of the 3D sound field around acoustically rigid structures such as slab tracks. The numerical approach is validated by comparison with a standard 2.5D BEM calculation and an analytical solution. Precalculated transfer functions to calculate the sound radiation from railway track, which are made available online, are illustrated. An example application is presented.

Determination of railway noise contribution based on noise source identification and acoustic transfer function

Environmental noise standards of noise for Japan's bullet trains, known as Shinkansen, are among the most stringent in the world. To reduce train noise and meet these standards, the contribution ratio of each noise source to the total noise at the observation point provisioned by the Japanese government, 25 m from the nearest track must be determined. Although previous studies have determined the contribution ratio using multiple omnidirectional microphones and a 1-dimensional microphone array positioned at the observation point, we propose a new method to evaluate the contribution ratio by combining the noise source identification with an acoustic experiment. The microphone array clearly measures the noise source distribution around a vehicle using signal processing of the deconvolution algorithm, indicating that bogies and pantographs are the dominant aerodynamic noise sources. The transfer function of the radiated sound between the sources and observation point was determined in the acoustic model experiment with a viaduct and noise barrier. The contribution ratio obtained should be useful in assessing the noise reduction from shape modification.

Acoustic contrast control in two regions in car cabins

Personal audio systems that generate different sound zones in different regions have received much interest recently, especially in car cabins. The feasibility of creating a listening zone at the driver's seat and a quiet zone at the passenger's seat with the headrest speakers, door loudspeakers, and the combination of them is investigated in this manuscript. A finite element model of a car interior is created in COMSOL Multiphysics to obtain the acoustic transfer functions and the acoustic contrast control performances of 4 different speaker configurations are compared with each other. It is found that the 4 headrest speakers achieve better acoustic contrast control performance than the door speakers at low frequencies, while the sound pressure level distribution tends to be more uniform in the bright zone when using the door speakers.

Robustness optimization of active headrest with virtual sensing to head rotation

The movement and rotation of the human head will cause changes of acoustic transfer functions, leading to the performance deterioration of active headrests with virtual microphones. This paper uses a transfer response model optimization method to improve the performance robustness of active headrests with remote microphones to the head rotation. The variation of transfer functions with head rotation and the improvement of the system performance by the optimization method are analyzed for three different acoustic environments, including an anechoic chamber, a normal room and a lightly damped room. The experimental results verify that the optimization method can improve the performance robustness of the active headrest to head rotation under different acoustic conditions. When the head rotates from -90 deg. to 90 deg., the minimum noise reduction at 125Hz, 250Hz and 500Hz is improved by 7.2~11.0 dB. The noise reduction of the optimized system thus remains above 10 dB in the anechoic chamber and in the ordinary room, while in the lightly damped room the noise reduction is above 6 dB due to strong reflections.

Validation of rolling-stock interior noise simulation in viaduct and tunnel environments

Acoustic comfort for modern rolling stock is a standard design requirement. The use of standard measurement conditions (ballast and free-field conditions) is the best way to identify quiet vehicles but operators (specially metro operators and some suburban) prefer to set noise targets in their own infrastructure. Metro infrastructures are mainly today done with ballastless tracks either in elevated track (viaduct) or tunnels. Therefore it is critical for rolling stock suppliers to develop methodologies to simulate interior noise performances on those scenarios to be able to propose design solutions and to control the risks. In this paper, the methodology used by Alstom to simulate interior noise in tunnels and viaduct is described. Methodology uses databases of acoustic transfer functions between sound power level of noise sources and acoustic wall pressures around vehicles. These transfer functions are modified in function of the environment (viaduct or tunnel) taking into account the type of acoustic field. Then, analytical or numerical models of the train structure plus interiors are calculated. Simulation cases and validation of the models are discussed to evaluate the pertinence of using virtual validation instead of interior noise type tests.

Near-field modeling of the head-related transfer function by using the dummy head sound source

Various approaches based on direct measurement have been used to obtain the precise headrelated transfer function (HRTF). However, several problems persist, such as scattering of sound source and distance effect, which are related to the measurement configuration. In this work, the modified dummy head having two sound sources located at the ear positions is applied as the sound source reciprocally to overcome the trouble. By using the near-field acoustic holography, the sound sources located in the ear holes are reconstructed from the measured data in the near-field hologram. One can predict the acoustic transfer function between the ear holes and the field positions at any point using the recovered source information described in spherical harmonics. The obtained transfer function is valid for predicting the binaural response from any sound source position. This configuration is possible to measure the near-field response with the relatively small scattering effect because of the low Helmholtz number of microphones. The far-field reconstruction of HRTF is also possible with the near-field measurement data using acoustical holography. The proposed method is demonstrated with the modified B&K HATS system, in which the microphones in the ear holes are replaced by pipe sources connected to the loudspeaker.

Enhancing acoustic contrast while tracking acoustic transfer functions: A simultaneous control approach

In the context of sound zone control (SZC) systems, the ability to effectively track variations in acoustic transfer functions is essential due to the inevitable variation of acoustic environments. This paper addresses the non-uniqueness problem encountered in time domain SZC, arising from the coupling observed in multichannel transfer function modeling. To address this issue, the utilization of time-varying control filters is proposed, which offers a viable solution. In order to ensure a stable acoustic contrast performance regardless of environmental changes, as well as to mitigate extreme distortion resulting from maximal acoustic contrast (AC) in the time domain, the paper focuses on the objective of maintaining a desired level of AC while tracking the acoustic transfer functions. Simulations are carried out to evaluate the proposed method, demonstrating its capability to effectively track environmental changes and continually update the control filters, consequently providing a stable acoustic contrast performance.

Improving the acoustic contrast control performance in car cabins

The car cabin is an enclosure where personal audio is very important because different occupants may have individual listening requirements. Potential approaches to improve the acoustic contrast control performance in enclosures are discussed in this paper, including the configuration of loudspeakers and the sound absorption of sidewalls. The effect of the number of loudspeakers on the acoustic contrast control performance is first investigated, and the locations of a fixed number of loudspeakers are then optimized to maximize the acoustic contrast achieved with the system. The sound absorption of sidewalls also has an effect on the acoustic contrast control performance and it is discussed in detail. The simulations are based on the acoustic transfer functions obtained with a finite element model of a car cabin. The findings in this paper may serve as a guide to future designs of personal audio systems in car cabins.