Active Loudspeakers Research Articles

Amplitude-decorrelation avoiding spectral artifacts of sound field synthesis in the anechoic chamber or free sound field

Sounds to appear in directions between the loudspeakers will typically activate neighboring loudspeakers with the same signal when spatially rendered. The resulting interference causes frequency- and position-dependent cancellation in the superimposed sound field. While keeping the number of active loudspeakers uniformly small is often enough to have the few acoustic reflections of a studio environment conceal the unwanted cancellations, they will remain a problem when synthesizing sound fields in the anechoic chamber or free field. Our contribution investigates short filters for amplitude decorrelation to reduce this problem. The filters supply neighboring loudspeakers via frequency responses that are mutually exclusive and hereby avoid cancellation in the superimposed soundfield. This only works at the expense of requiring a minimum spread to ensure a spectrally balanced synthesis, i.e., by avoiding to activate single filtered loudspeaker, alone. We present a listening experiment investigating in how far the proposed filter strategy is able to improve the spectral stability within a reasonable range of motion for a seated listener, in the anechoic chamber. Other thinkable approaches are investigated for comparison.

Time-domain sound field reproduction using the group Lasso.

A time-domain sound field reproduction (SFR) method based on the group Lasso (GL) is presented. The proposed method optimizes the positions and the number of active loudspeakers in the time domain by iteratively solving a mixed-norm constrained optimization problem. Simulation results demonstrate that the proposed method achieves accurate broadband SFR using a small number of active loudspeakers and significantly outperforms the least-squares method when the desired sound field is under-sampled. Moreover, the GL algorithm can also be used to optimize the loudspeaker placement for sound field control systems.

Sound Field Reproduction via the Alternating Direction Method of Multipliers Based Lasso Plus Regularized Least-Square

This paper proposes a 3-D sound field reproduction (SFR) approach through the combination of alternating direction method of multipliers (ADMM)-based least-absolute shrinkage and selection operator (Lasso) and regularized least square (LS). The proposed SFR method is split into two parts through the pressure matching optimization of loudspeaker positions and the computation of driving signals. At the first part, a plurality of candidate positions of loudspeakers in planar array is given and, then, the active speaker selection method is proposed based on ADMM complex Lasso algorithm for selecting the optimal loudspeaker positions. Afterwards, regularized LS is adopted to calculate the selected loudspeaker weights and control the total power. The numerical simulation experiments demonstrate that the proposed SFR scheme outperforms the existing sparse loudspeakers’ placement and weight optimization algorithms especially in under-sampled sound fields. Meanwhile, the evaluations also confirmed that the proposed method could significantly reduce the computational complexity of the active loudspeaker selection compared to the state-of-the-art Lasso-based SFR. Effectively, the proposed method uses a relatively small number of loudspeakers for a satisfying reproduction quality.

Sparse $\ell _{1}$-Optimal Multiloudspeaker Panning and Its Relation to Vector Base Amplitude Panning

Panning techniques, such as vector base amplitude panning (VBAP), are a widely used practical approach for spatial sound reproduction using multiple loudspeakers. Although limited to a relatively small listening area, they are very efficient and offer good localization accuracy, timbral quality, as well as a graceful degradation of quality outside the sweet spot. The aim of this paper is to investigate optimal sound reproduction techniques that adopt some of the advantageous properties of VBAP, such as the sparsity and the locality of the active loudspeakers for the reproduction of a single audio object. To this end, we state the task of multiloudspeaker panning as an ${\ell _{1}}$ optimization problem. We demonstrate and prove that the resulting solutions are exactly sparse. Moreover, we show the effect of adding a nonnegativity constraint on the loudspeaker gains in order to preserve the locality of the panning solution. Adding this constraint, ${\ell _{1}}$ -optimal panning can be formulated as a linear program. Using this representation, we prove that unique ${\ell _{1}}$ -optimal panning solutions incorporating a nonnegativity constraint are identical to VBAP using a Delaunay triangulation for the loudspeaker setup. Using results from linear programming and duality theory, we describe properties and special cases, such as solution ambiguity, of the VBAP solution.

A Lasso-LS Optimization with a Frequency Variable Dictionary in a Multizone Sound System

This paper presents an approach for multizone wideband sound field generation using an efficient harmonic nested (EHN) dictionary for sparse loudspeakers’ placement and weight. Effectively, the nested arrays provide a priori knowledge of prospective loudspeaker locations based on the frequency bands of interest. The nested arrays are then further optimized in the Lasso stage to form an efficient loudspeakers’ location dictionary. The final loudspeaker locations and weightings are estimated by a two-stage Lasso-LS pressure matching optimization. In the first stage Lasso algorithm, the center band frequencies of octave bands from 1 kHz to 8 kHz were used to select active loudspeakers. A second stage then optimizes reproduction using all selected loudspeakers on the basis of a regularized LS algorithm. The results demonstrate that the proposed approach provides a solution for the multizone sound system with the mean squared error (MSE) under $-30 \text{dB}$ across the targeted frequency range (500 Hz-16 kHz) using a linear array e.g. 13 loudspeakers. While, the single-stage LS approach generates the MSE peaks of $-10 \text{dB}$ and $-9 \text{dB}$ at 9 kHz within the active and silent zones respectively using an identical number of loudspeakers and array length.

Compensating acoustic feedback in feed-forward active noise control systems using spline adaptive filters

Implementation of a feed-forward active noise control (ANC) system in a short duct may cause acoustic feedback between the active loudspeaker and the reference microphone. The conventional filtered-x least mean square (FxLMS) algorithm based ANC systems are not designed to handle this situation. Similarly, an FxLMS algorithm based ANC system fails to effectively mitigate noise when non-linearities are present in the system. In an endeavor to overcome these two limitations of traditional ANC systems, in this paper, we propose an adaptive infinite impulse response (IIR) spline filter for non-linear ANC. The adaptive nature of the spline activation function in the proposed scheme enables the filter to effectively switch from a linear to non-linear nature and vice versa, depending upon the scenario in which the ANC system is implemented.

Effect of aided hearing in the nonimplanted ear on bimodal hearing.

This study was designed to evaluate the binaural effects from bimodal hearing according to the aided hearing threshold in the nonimplanted ear. Subjects included 17 individuals who continued to use a hearing aid (HA) in the nonimplanted ear for more than 6 months postoperatively. Speech perception and sound localization were tested with unilateral cochlear implantation (CI) and bimodal hearing with and without background noise. Materials were presented at an average of 70 dB sound pressure level from a front loudspeaker in a quiet condition and then with background noise at a signal-to-noise ratio of +10 dB HL. Speech perception scores were based on percent-correct performance of repeating a spoken word under each condition. Sound localization scores were obtained by averaging the sum of angle differences between the active loudspeaker and the loudspeaker indicated by the subject. Speech perception scores (mean ± SD) of unilateral CI and bimodal hearing were 63.3% ± 17.7% and 73.1% ± 18.5% under the quiet condition (p = 0.029) and 65.5% ± 21.9% and 70.9% ± 23.6% under the noisy condition (p = 0.01), respectively. Angle differences (mean ± SD) of unilateral CI and bimodal hearing were 72.8 ± 27.4 degrees and 84.1 ± 29.9 degrees under the quiet condition (p = 0.052) and 79.3 ± 26.9 degrees and 77.3 ± 22.0 degrees under the noisy condition (p = 0.906), respectively. Patients were divided into two groups according to their aided hearing thresholds: Group 1 (aided hearing threshold ≤50 dB HL; n = 8) and Group 2 (aided hearing threshold >50 dB HL; n = 9). The speech perception scores of bimodal hearing in each group were 85.3% ± 13.3% and 60.8% ± 17.5% (p = 0.023) under the quiet condition and 82.7% ± 9.0% and 59.4% ± 26.8% under the noisy condition (p = 0.052), respectively. For sound localization, the angle differences of bimodal hearing in each group were 54 ± 28.6 degrees and 83.9 ± 11.9 degrees under the quiet condition (p = 0.042) and 63.0 ± 23.5 degrees and 89 ± 22.6 degrees under the noisy condition (p = 0.049), respectively. Based on the relationship between the aided hearing level and bimodal hearing performance, this current study suggests that bimodal benefits for sound localization and speech perception in noise are significant but only when sound detection is adequate for the hearing aids. Therefore, bimodal hearing could be applied to selective patients with favorable aided hearing levels.

A Framework for the Calculation of Dynamic Crosstalk Cancellation Filters

Dynamic crosstalk cancellation (CTC) systems commonly find use in immersive virtual reality (VR) applications. Such dynamic setups require extremely high filter update rates, so filter calculation is usually performed in the frequency-domain for higher efficiency. This paper proposes a general framework for the calculation of dynamic CTC filters to be used in immersive VR applications. Within this framework, we introduce a causality constraint to the frequency-domain calculation to avoid undesirable wrap-around effects and echo artifacts. Furthermore, when regularization is applied to the CTC filter calculation, in order to limit the output levels at the loudspeakers, noncausal artifacts appear at the CTC filters and the resulting ear signals. We propose a global minimum-phase regularization to convert these anti-causal ringing artifacts into causal artifacts. Finally, an aspect that is especially critical for dynamic CTC systems is the filter switch between active loudspeakers distributed in a surround audio-visual display system with 360° of freedom of operator orientation. Within this framework we apply a weighted filter calculation to control the filter switch, which allows the loudspeakers' contribution to be windowed in space, resulting in a smooth filter transition.

Source Width of Frontal Phantom Sources: Perception, Measurement, and Modeling

Abstract Phantom sources are known to be perceived similar to real sound sources but with some differences. One of the differences is an increase of the perceived source width. This article discusses the perception, measurement, and modeling of source width for frontal phantom sources with different symmetrical arrangements of up to three active loudspeakers. The perceived source width is evaluated on the basis of a listening test. The test results are compared to technical measures that are applied in room acoustics: the inter-aural cross correlation coefficient (IACC) and the lateral energy fraction (LF). Adaptation of the latter measure makes it possible to predict the results by considering simultaneous sound incidence. Finally, a simple model is presented for the prediction of the perceived source width that does not require acoustic measurements as it is solely based on the loudspeaker directions and gains.

The Influence of the Measurement Chain on the Impulse Response of a Reverberation Room and Its Application Listening Tests

There are many objective parameters of the room impulse response which we can calculate that describeits acoustic field. The room impulse response can also be used to provide listening tests using the convolutiontechnique. The aim of this work is to consider the influence of the impulse response measurement chain onobjective numerical parameters and on the subjective perception of listeners. This article shows the impulseresponsemeasurementresultsobtainedinthereverberationroomusingnotonlycommonlyknownmeasurementchainelementsbutotherelements,too. Standardcomponentssuchasanomni-directionalsourceandmeasurementmicrophone, as well as a consumer active loudspeaker and various audio recording stereo microphone pairs wereused. Theresultsarecomparedtakingintoconsiderationcalculatednumericalparametersandtheapplicationofcalculatedimpulseresponsestosubjectivelisteningtests. Todeterminetheapplicabilityofimpulseresponsestopsychoacoustic listening tests, a preliminary listening test using headphones was carried out on a selected groupoflisteners. Theaudiotestmaterialconsistedofsamplesofchoralmusicandspeechrecordedinananechoicroomandconvolvedwithpreviouslyanalyzedimpulseresponses. Thisarticleshowshowbigthedifferencesofsubjectivelistenerfeelingsarefordifferentacoustictransducers.PACS:43.55.Hy,43.55.Mc,43.66.Lj

Method And Apparatus For Wireless Digital Audio/Video Playback For Computers

A computer-based digital wireless audio and video system for consumer audio, home theater and recording studio applications is disclosed. The system distributes synchronized 2-channel stereo and multi-channel home theater video and audio digital signals via wireless broadcast to a closed network including a personal computer, a display, and one or multiple digital active loudspeakers or headphones. Each loudspeaker has a separate digital amplifier dedicated to each transducer within it. The system provides a data link capable of sending an all-digital, full-bandwidth signal from the original digital source material to each separate transducer in the system without using sound degrading lossy data compression. This system is designed to read, broadcast, and reproduce with accurate audio and video channel synchronization in the effort to maintain the integrity of the entire audio and video signal.

Localization by Postlingually Deafened Adults Fitted With a Single Cochlear Implant

: The main purpose of the study was to assess the ability of adults with unilateral cochlear implants to localize noise and speech signals in the horizontal plane. : Six unilaterally implanted adults, all postlingually deafened and all fitted with MED-EL COMBI 40+ devices, were tested with a modified source identification task. Subjects were tested individually in an anechoic chamber, which contained an array of 43 numbered loudspeakers extending from -90 degrees to +90 degrees azimuth. On each trial, a 200 millisecond signal (either a noise burst or a speech sample) was presented from one of nine active loudspeakers, and the subject had to identify which source (from the 43 loudspeakers in the array) produced the signal. : The relationship between source azimuth and response azimuth was characterized in terms of the adjusted constant error (C). C for three subjects was near chance (50.5 degrees ), whereas C for the remaining three subjects was significantly better than chance (35 degrees -44 degrees ). By comparison, C for a group of normal-hearing listeners was 5.6 degrees . For two of the three subjects who performed better than chance, monaural cues were determined to be the basis for their localization performance. : Some unilaterally implanted subjects can localize sounds at a better than chance level, apparently because they can learn to make use of subtle monaural cues based on frequency-dependent head-shadow effects. However, their performance is significantly poorer than that reported in previous studies of bilaterally implanted subjects, who are able to take advantage of binaural cues.

Horizontal-Plane Localization of Noise and Speech Signals by Postlingually Deafened Adults Fitted With Bilateral Cochlear Implants*

The main purpose of the study was to assess the ability of adults with bilateral cochlear implants to localize noise and speech signals in the horizontal plane. A second objective was to measure the change in localization performance in these adults between approximately 5 and 15 mo after activation. A third objective was to evaluate the relative roles of interaural level difference (ILD) and interaural temporal difference (ITD) cues in localization by these subjects. Twenty-two adults, all postlingually deafened and all bilaterally fitted with MED-EL COMBI 40+ cochlear implants, were tested in a modified source identification task. Subjects were tested individually in an anechoic chamber, which contained an array of 43 numbered loudspeakers extending from -90 degrees to +90 degrees azimuth. On each trial, a 200-msec signal (either a noise burst or a speech sample) was presented from one of 17 active loudspeakers (span: +/-80 degrees ), and the subject had to identify which source from the 43 loudspeakers in the array produced the signal. Subjects were tested in three conditions: left device only active, right device only active, and both devices active. Twelve of the 22 subjects were retested approximately 10 mo after their first test. In Experiment 2, the spectral content and rise-decay time of the noise stimulus were manipulated. The relationship between source azimuth and response azimuth was characterized in terms of the adjusted constant error (ĉ). (1) With both devices active, ĉ for the noise stimulus varied from 8.1 degrees to 43.4 degrees (mean: 24.1 degrees ). By comparison, ĉ for a group of listeners with normal hearing ranged from 3.5 degrees to 7.8 degrees (mean: 5.6 degrees ). When subjects listened in unilateral mode (with one device turned off), ĉ was at or near chance (50.5 degrees ) in all cases. However, when considering unilateral performance on each subject's better side, average ĉ for the speech stimulus was 47.9 degrees , which was significantly (but only slightly) better than chance. (2) When listening bilaterally, error score was significantly lower for the speech stimulus (mean ĉ = 21.5 degrees ) than for the noise stimulus (mean ĉ = 24.1 degrees ). (3) As a group, the 12 subjects who were retested 10 mo after their first visit showed no significant improvement in localization performance during the intervening time. However, two subjects who performed very poorly during their first visit showed dramatic improvement (error scores were halved) over the intervening time. In Experiment 2, removing the high-frequency content of noise signals resulted in significantly poorer performance, but removing the low-frequency content or increasing the rise-decay time did not have an effect. In agreement with previously reported data, subjects with bilateral cochlear implants localized sounds in the horizontal plane remarkably well when using both of their devices, but they generally could not localize sounds when either device was deactivated. They could localize the speech signal with slightly, but significantly better accuracy than the noise, possibly due to spectral differences in the signals, to the availability of envelope ITD cues with the speech but not the noise signal, or to more central factors related to the social salience of speech signals. For most subjects the remarkable ability to localize sounds has stabilized by 5 mo after activation. However, for some subjects who perform poorly initially, there can be substantial improvement past 5 mo. Results from Experiment 2 suggest that ILD cues underlie localization ability for noise signals, and that ITD cues do not contribute.

Interior noise control through tuned loudspeaker absorption

In this work we describe the development and application of an interior noise control approach based on active loudspeaker tuning for sound absorption. Loudspeakers are tuned to minimize global sound pressure at specific acoustic modes through diaphragm velocity feedback. An experimental model of the acoustic space used for controller design is obtained through system identification. H2 optimal controllers are designed to minimize a performance microphone output by feeding back the tuned loudspeaker diaphragm velocity. By sensing speaker velocity rather than far-field sound pressure, the loudspeaker is actively tuned to minimize reflected sound at specific acoustic modes and therefore reduce the global pressure field. Experimental results, obtained from an acoustic enclosure modeled after a rocket fairing, are used to demonstrate the effectiveness of the tuned loudspeaker at global attenuation. Details of the design procedure and experimental applications are discussed.

Navier-Stokes Simulation of Harmonic Point Disturbances in an Airfoil Boundary Layer

Laminar-turbulenttransitionmechanismsinducedbyaharmonicpointsourcedisturbanceinae at-plateboundary layerwithadversepressuregradientareinvestigated by fourth-orderaccuratespatialdirectnumericalsimulation based on the complete three-dimensional Navier ‐Stokes equations for incompressible e ow. The disturbance is introduced into thetwo-dimensionalbasee owbytime-periodicsimultaneousblowing andsuction within acircular spot at the wall to quietly mimic the momentum input by an active loudspeaker below a hole in the surface in respective experiments. Thus a wave train consisting of pure Tollmien ‐Schlichting waves of a singlefrequency and a large number of obliqueness angles is stimulated, and its downstream evolution in both physical and spectral space is investigated. A breakdown scenario dominated entirely by oblique modes is observed that shows a spanwise peak/valley amplitude splitting with the valley plane at the centerline of the wave train. Dominant vorticity structures develop off centerline, and a clear-cut -shaped structure is formed in e nal stages related to the wall shear being a footprint of a e rst pair of K -vortices in the e ow.

Sound localization after transection of the commissure of Probst in the albino rat.

1. The ability of rats to localize sounds in space was determined before and after cutting of the commissure of Probst. The commissure of Probst was transected at its midline decussation with a microknife inserted into the brain according to stereotaxic coordinates. Six animals were tested after extensive lesions that destroyed all of the commissure of Probst fibers. An additional animal was tested after a smaller lesion that destroyed most of the commissure of Probst but left some fibers intact. Three control animals were tested before and after surgical intervention that did not involve the commissure of Probst. 2. The animals were tested in a semicircular apparatus with loudspeakers located on the right or left of midline. They were trained to make a response toward the left or right in the direction of the active loudspeaker. Correct reponses were rewarded by delivery of a small quantity of water from spouts located at +30 and -30 degrees azimuth. Tests of sound localization were conducted with a single broadband noise burst, 45 ms in duration, presented at the beginning of each trial. The position of the active loudspeakers was varied from trial to trial and performance at different speaker angles was calculated to determine psychometric curves. Minimum audible angles were estimated by interpolation from a performance level of 75% correct. 3. After postoperative testing was completed, the effectiveness of the lesions was confirmed by cell counts to determine the extent of retrograde degeneration in the dorsal nucleus of the lateral lemniscus (DNLL). These data showed that most of the contralaterally projecting neurons in DNLL underwent retrograde degeneration and the number of neurons was reduced by 60-65%. Ninety to 95% of the contralaterally projecting neurons in the DNLL disappeared within 7 wk after transection of the commissure of Probst. 4. The condition of damaged commissural fibers was further confirmed by tract tracing methods. A unilateral Fluoro-Gold injection was made into the inferior colliculus and the auditory brain stem was examined for retrograde label. No labeled neurons were seen in the contralateral DNLL in cases with complete transection of the commissure of Probst. In addition, the distribution of Fluoro-Gold labeling in other brain stem auditory structures was similar to that seen in normal animals. This result confirmed that the transection of the commissure of Probst was successful and that projections to other auditory structures remained intact. 5. Transection of the commissure of Probst produced marked deficits in midline sound localization. Although sound localization was still possible, there was a degradation in the ability of rats to localize sounds in the horizontal plane. The mean minimum audible angle was elevated 22.1 degrees after destruction of the commissure of Probst compared with a shift of only 2.0 degrees in control animals.

Electro-acoustic transducers

An electro-acoustic transducer is formed of a loudspeaker enclosure having a form of revolution about an axis each end of which is closed by a loudspeaker. The two loudspeakers are mounted back to back in the axis of the enclosure, one of the loudspeakers being fed from an electric power source and being in this sense "active", whereas the other loudspeaker is "passive" in that it receives no electric supply, its movements resulting from the rear wave of the active loudspeaker. The enclosure is maintained vertically along its axis of revolution, and, on the other hand, the side wall of the enclosure is formed of a flexible material, unlikely to resonate with the front and rear waves of the loudspeakers.

Auditory Localization: Effects of Reflecting Surfaces

Based on anatomical and evolutionary conceptions of the human ear, an experiment was conducted in which forty-eight human subjects were asked to localize sounds (a human voice) emitted by one of twenty-seven stationary loudspeakers in an anechoic chamber. The position of the active loudspeaker varied with respect to azimuth, distance, and elevation in three steps each. The position of a single sound-reflecting surface (about 6 m2) was varied: on the floor, on the ceiling, to the left, and to the right. The accuracy of identifying the active loudspeaker for each position of the sound-reflecting surface was compared intraindividually with the absence of reflection. The results show an overall increase in correct localizations with a sound-reflecting surface on the floor. Especially the elevation of the sound source can be detected with greater precision. Additionally, the percentage of correct localizations decreased systematically with the presence of a sound-reflecting ceiling, while the presence of sound-reflecting walls did not systematically affect the localization performance. Judgments in the horizontal plane and those of distance were not systematically influenced by the presence of a sound-reflecting surface.