Interaural Cross-correlation Research Articles

Perceived spatial impression from various musical motifs

Most of the objective measures which are currently used to quantify spatial impression in auditoria, such as lateral energy fraction (LF), interaural cross-correlation coefficient (IACC) and late lateral relative sound level (GLL), are calculated from an impulse response. For subjective testing, these impulse responses are often later convolved with music recorded in an anechoic environment to simulate music performance in the hall. One spatial impression measure, though, interaural level fluctuations (IALFs) which was first proposed by Griesinger [Aud. Eng. Soc., Preprint No. 3292 (1992)] and further developed by L. M. Wang and A. C. Gade [J. Acoust. Soc. Am. 107, 2891(A) (2000)], is calculated from the musical source signal instead of from the impulse response. Preliminary investigations showed that the IALF, calculated from simulated sound fields generated by ODEON, changed relatively the same for different pieces of music across various halls. If IALF is a valid measure for spatial impression, then this indicates that the musical source is not critical in evaluating the difference in spatial impression between halls. In this paper, results from subjective listening tests are presented to determine if the same trend is perceived by listeners across different musical motifs, which would aid in validating IALF.

MEASUREMENT OF SOUND TRANSMISSION BY HALL DOORS

Acoustical properties of sound transmission through to all doors were investigated by measuring and calculating temporal and spatial factors in sound fields based on the model of the auditory–brain system (see reference [4]). To describe acoustical properties of the effect of doors, measurements were carried out obtaining five orthogonal factors. These are (1) sound pressure level (SPL), (2) the magnitude of the interaural cross-correlation (IACC), (3) subsequent reverberation time (Tsub), (4) interaural time delay (τIACC), and (5) the width of the IACC function (WIACC). In order to describe subjective responses in the receiving room, not only the difference of SPL but also other factors, for example IACC, and Tsub, can be measured at the same time.

A COMPUTATIONAL SOFTWARE FOR NOISE MEASUREMENT AND TOWARD ITS IDENTIFICATION

The computational system outlined here is based on a model of the human auditory– brain system including the autocorrelation and interaural crosscorrelation mechanisms, and the specialization of the cerebral hemispheres [1]. It consists of a binaural receiver, a laptop computer, and software designed to measure the physical factors of noise fields and to identify environmental noise. The temporal factors τe,τ1 , and φ1as well as Φ (0), are extracted from the autocorrelation function (ACF) of the noise source, and the spatial factors LL, IACC, τIACC, andWIACC are extracted from the interaural crosscorrelation function (IACF). These factors may be utilized for subjective evaluations of a source of noise and a noise field.

ACOUSTICAL PROPERTIES OF AIRCRAFT NOISE MEASURED BY TEMPORAL AND SPATIAL FACTORS

Acoustical properties of aircraft noise were investigated by means of temporal and spatial factors in sound fields based on the model of auditory–brain system (see reference [10]). The model consists of the autocorrelation and crosscorrelation mechanisms for sound signals arriving at two ears and the specialization of human cerebral hemisphere. There are four temporal factors extracted from the autocorrelation function (ACF): (1) sound energy Φ (0); (2) effective duration of ACF, τe; (3) delay time of the first peak, τ1; and (4) its amplitude φ1. From the interaural crosscorrelation function (IACF), three spatial factors are extracted as (1) magnitude of the interaural crosscorrelation IACC (2) interaural delay time at IACC,τIACC , and (3) width of the maximum peak of the IACF, WIACC. It is found that the acoustical properties are well represented by the factors extracted from the ACF and the IACF.

MEASUREMENT OF REGIONAL ENVIRONMENTAL NOISE BY USE OF A PC-BASED SYSTEM. AN APPLICATION TO THE NOISE NEAR AIRPORT “G. MARCONI” IN BOLOGNA

Measurements of aircraft noise were made at the airport “G. Marconi” in Bologna by using a measurement system for regional environmental noise. The system is based on the model of the human auditory–brain system, which is based on the interplay of autocorrelators and an interaural cross-correlator acting on the pressure signals arriving at the ear entrances, and takes into account the specialization of left and right human cerebral hemispheres (see reference [8]). Measurements were taken through dual microphones at ear entrances of a dummy head. The aircraft noise was characterized with the following physical factors calculated from the autocorrelation function (ACF) and interaural cross-correlation function (IACF) for binaural signals. From the ACF analysis, (1) energy represented at the origin of delay,Φ (0), (2) effective duration of the envelope of the normalized ACF, τe, (3) the delay time of the first peak, τ1, and (4) its amplitude, φ1were extracted. From the IACF analysis, (5) IACC, (6) interaural delay time at which the IACC is defined, τIACC, and (7) width of the IACF at the τIACC, WIACCwere extracted. The factorΦ (0) can be represented as the geometrical mean of the energies at both ears. A noise source may be identified by these factors as timbre.

Relation of acoustical parameters with and without audiences in concert halls and a simple method for simulating the occupied state.

Five acoustical parameters-reverberation time RT, early decay time EDT, clarity C80, strength G, and interaural cross-correlation coefficient IACC-were measured using identical procedures with and without audiences in six concert and opera halls. Reverberation times without audiences were measured in 15 additional halls using the same measuring techniques as for the six halls above, but for full occupancy the data were taken from musical stop chords at symphonic concerts. This paper shows that in all halls (1) the occupied RT can be predicted from the unoccupied RT using a linear regression equation, y = a - b exp(x), within acceptable limits, at low- and mid-frequencies. It is also shown for the six halls that (2) occupied C80's are predicted accurately from unoccupied values by the newly proposed equation; (3) G's with and without audiences are highly correlated by a first degree linear regression equation; and (4) IACCs have nearly the same value in both occupied and unoccupied halls. As a separate subject, the successful use of a cloth covering for seats in a concert or opera hall to simulate the occupied condition has been developed.

SUBJECTIVE EVALUATION OF FLOOR IMPACT NOISE BASED ON THE MODEL OF ACF/IACF

Floor impact noises generated by bang and tapping machines were measured through dual microphones on the ear of a human head. Matching and magnitude estimation techniques for noise evaluation were used to investigate the perceptual differences of the impact noises from apartment floors. Measurements of noise were also conducted by a diagnostic system based on the model of the human auditory–brain system [1]. Physical factors in the model were calculated by the use of the auto-correlation function (ACF) and interaural cross-correlation function (IACF) of binaural signals. From the results, it was found that perceived loudness of floor impact noise could be represented by the ACF/IACF factors. It became apparent that, at the beginning of each impact of intermittent noise (“bang noise”), the spatial impression of the sound field in an apartment bedroom corresponds to a high value of the IACC. Results also show that bang machine noise is perceived to be louder and noisier than that of a tapping machine.

EFFECTS OF TELEPHONE RING ON TWO MENTAL TASKS RELATIVE TO AN OFFICE

In many cases, there are a lot of noise sources in an office and particularly, telephone ringing often irritates the office workers. Effects of aircraft noise on the mental work of pupils were reported by Ando et al.[1]. In spite of its serious effect, it has not yet been found how the physical parameters of the wave form influence the perception of noise. The purpose of this study is to investigate the effects of telephone ringing on two mental tasks. This investigation is based on the human auditory–brain model consisting of the auto-correlation function (ACF) of sound source, the interaural cross-correlation function (IACF) for sound signals arriving at the two ears, and the specialization of the cerebral hemispheres. Under the stimulus of a telephone ringing, an adding task and a drawing task were performed. Results show that telephone ringing influences differently the two tasks: the V-type relaxation was observed only during the drawing task. It is revealed that the interference effect between the drawing task and the noise may occur in the right hemisphere.

A THEORY OF PRIMARY SENSATIONS AND SPATIAL SENSATIONS MEASURING ENVIRONMENTAL NOISE

A theory of primary sensations and spatial sensations to environmental noise is proposed here. The model of the auditory–brain system includes the autocorrelation function (ACF), the interaural cross-correlation function (IACF) mechanisms and the specialization of the human cerebral hemispheres [1]. In addition to primitive sensations to the noise source-loudness, pitch and timbre-perception of duration is introduced here as a fourth. They are described by the temporal factors extracted from the ACF. It is worth noting that the information in the power density spectrum of a signal is identical to the ACF of the signal. In order to describe spatial sensations to the noise field, the spatial factors extracted from the IACF are taken into account. Timbre is defined as an overall sensation including the primitive sensations and spatial sensations.

PRELIMINARY STUDY ON RECOMMENDED TIME DURATION OF SOURCE SIGNALS TO BE ANALYZED, IN RELATION TO ITS EFFECTIVE DURATION OF THE AUTO-CORRELATION FUNCTION

A workable model of human auditory–brain system including the auto-correlation function (ACF) and the interaural cross-correlation function (IACF) has been reported [1]. Subjective evaluations can be described by physical factors extracted from both the ACF of sound source and the IACF. The purpose of this study is to find a recommended integration time (2 T)rby using several noise and sound sources existing in an office. Results show the most suitable integration interval from the viewpoints of the loudness and the pitch can be determined in relation to the effective duration (τe) of the sound source.

Manipulating the "straightness" and "curvature" of patterns of interaural cross correlation affects listeners' sensitivity to changes in interaural delay.

The purpose of this study was to test the hypothesis that stimuli characterized by "straight" trajectories of their patterns of cross correlation foster greater sensitivity to changes in interaural temporal disparities (ITDs) than do stimuli characterized by more "curved" trajectories of their patterns of cross correlation. To do so, sensitivity to changes in ITD was measured, as a function of duration, using a set of "reference" stimuli that yielded differing relative amounts of straightness within their patterns of cross correlation while keeping the dominant trajectory at or near midline. The relative amounts of straightness were manipulated by employing specific combinations of bandwidth, ITD, and interaural phase disparity (IPD) of Gaussian noises centered at 500 Hz. The results were consistent with expectations in that the patterning of the threshold ITDs revealed increasingly poorer sensitivity as greater and greater curvature was imposed on the dominant, "midline," trajectory. The variations in threshold ITD across the stimulus conditions can be accounted for quite well quantitatively by assuming either that the listeners based their judgments on changes in the position of the most central peak of the cross-correlation function or that they based their judgments on changes in the centroid of a second-level cross-correlation function. In a second experiment, binaural detection was measured using a subset of the reference stimuli as maskers. As expected, sensitivity was poorest with the maskers characterized by the greatest curvature, which were also those having the lowest interaural correlation.

Resolution of front-back confusion in virtual acoustic imaging systems.

A geometric model of the scattering of sound by the human head is used to generate a model of localization cues based on interaural time delay (ITD). The ITD is calculated in terms of the interaural cross-correlation function (IACC) for sources placed at a series of azimuthal angles in the horizontal plane. This model is used to simulate the pressures generated at the ears of a listener due to real sources and due to a two-channel and a four-channel virtual source imaging system. Results are presented in each case for the variation of ITD with head rotation. The simulations predict that the rate of change of the ITD with head rotation produced by a real source and replicated by the four-channel virtual source imaging system, cannot be replicated by the two-channel system. These changes to the ITD provide cues which allow resolution of front-back confusion. The results of subjective experiments are also presented for the three cases modeled. These results strongly support the findings from the modeling work indicating that, for the systems described here, front-back confusion is resolved through changes to the ITD arising from head motion.

Image shift caused by strong lateral reflections, and its relation to inter-aural cross correlation

Conditions of sound fields were studied to avoid the occurrence of image shift, an acoustical phenomenon caused by extremely strong lateral reflections. Three types of image shift are defined as variations of the normal condition, in which there is only one auditory image perceived in the source direction. From a series of listening tests in electrically reproduced sound fields, each of the three types of image shift was found to have its own specific frequency range in which the image shift occurred. They were also found to relate to each of three measures which are all derived from the inter-aural cross correlation function (IACF) of binaural impulse responses of a sound field. The results of measurement made in three existing concert halls suggested acceptable ranges for the measures, in which the image in the source direction is not lost in the 500 Hz and higher frequency bands. However, the 250 Hz and lower frequency bands were found not so important in evaluating sound fields of concert halls in relation to the image shift. Finally, 0.85 is found to be the upper limit for the preferable range of [1 - IACC(E3)].

A model to simulate human sound localization and detection in presence of a distracter

To simulate human auditory localization performance of a target in the presence of a distracter, existing models focus the analysis of the targets on frequency bands or time gaps where the distracter is not present. However, human localization also works quite well if the frequency ranges of target and distracter are identical and the distracter is played during the complete duration of the target—as long as the onsets of target and distracter are separated in time. This is taken into account in the proposed model. For each frequency band, the interaural time differences of the target are calculated from the difference of the interaural cross correlation (ICC) of the total sound (target+distracter) and the ICC of the distracter alone (ΦT=3DΦTD−ΦD, assuming that target and distracter are uncorrelated). The latter is calculated from the part of the distracter that precedes the target. A similar procedure is used to measure the interaural level differences of the target. Auditory events are simulated assuming an ideal observer and internal noise. The model is able to mimic a number of psychoacoustic effects, including spatial shifts of the auditory events for low SNR situations. [Work supported by DFG.]

The effects of cylindrical diffusers on sound diffusion in a rectangular concert hall

Acoustical measurements in a 1:20 scale model were conducted to investigate the influence of a 1/4-cylinder diffuser on sound diffusion in a rectangular concert hall. Diffusion was evaluated regarding temporal and directional distribution of acoustical energy at individual seats as well as energy distribution over all seats. With volume of 19,000 m3 and seating area of 1,200 m2, the size and shape of the hall were determined based on three classical rectangular concert halls. Overall, diffusers around the stage were more effective for the source at the center, front of the stage than the source at the side, rear of the stage. The hall average values of late interaural cross correlation (IACCL3) were reduced by using transversal diffusers on ceiling or longitudinal diffusers on the side walls. Although the influences on the hall average values of echo criteria (EC) and temporal diffusion (TD) were relatively small, diffusers on the rear stage wall were effective in reducing EC values for the seats near the side walls. Strength factor (G) at the seats under the balcony was significantly increased by placing diffusers on the ceiling. On-going research is conducted to evaluate the effects of diffusers in the audience section. [Work supported by the National Science Council of Taiwan.]

Subjective relevance of objective measures for spatial impression

Several objective measures have been proposed to describe the feeling of spatial impression in concert halls, including Lateral Energy Fraction (LF) and Interaural Cross-Correlation Coefficient (IACC). However, previous studies have shown that LF and IACC values did not highly correlate with each other at individual seat positions in real halls [J. S. Bradley, J. Acoust. Soc. Am. 96, 3525–3535 (1994)]. To investigate the listener envelopment aspect of spatial impression further, subjective paired-comparison tests have been run using signals which have various values for LF, early IACC (from 5–80 ms), late IACC (from 80 ms–1 s), and late lateral relative sound level (GLL). Another proposed measure, called Interaural Level Fluctuations (IALF), has also been included, which is based on the rate of change over time of the level difference between the ears. The binaural test signals were generated by the room acoustics simulation program ODEON, using four different musical motifs, and were presented to test subjects via headphones. Results from the experiments are presented. [Work supported by ASA Hunt Postdoctoral Research Fellowship.]

A DIAGNOSTIC SYSTEM MEASURING ORTHOGONAL FACTORS OF SOUND FIELDS IN A SCALE MODEL OF AUDITORIUM

Based on the model of auditory–brain system which consists of the autocorrelation mechanism, the interaural cross-correlation mechanism between both the auditory pathways, and the specialization of human cerebral hemispheres (Y. Ando 1998 Architectural Acoustics, Blending Sound Sources, Sound Fields, and Listeners New York: AIP Press/Springer-Verlag), a new diagnostic system was developed. After obtaining the binaural impulse response, four orthogonal factors including the SPL, the initial time-delay gap between the direct sound and the first reflection, the subsequent reverberation time and the IACC can be analyzed for the calculation of the scale values of both global and individual subjective preferences. In addition, two more factors extracted from the interaural cross-correlation functionτIACC and WIACC, can be figured out. Also, the sound energy,Φ (0), the effective duration, τe, and fine structures of autocorrelation function of sound signals including the magnitude of first maximum, φ1, and its delay time,τ1 , can be analyzed. As an example of the measurement, effects of reflectors' array above the stage in a 1/10 scale model of auditorium at each seat are discussed here.

FORMULAE DESCRIBING SUBJECTIVE ATTRIBUTES FOR SOUND FIELDS BASED ON A MODEL OF THE AUDITORY-BRAIN SYSTEM

This article reviews the background of a workable model of the auditory–brain system, and formulae of calculating fundamental subjective attributes derived from the model. The model consists of the autocorrelation mechanisms, the interaural cross-correlation mechanism between the two auditory pathways, and the specialization of the human cerebral hemispheres for temporal and spatial factors of the sound field. Typical fundamental attributes, the apparent source width, the missing fundamental, and the speech intelligibility of sound fields, for example, in opera houses, are described in terms of the orthogonal spatial factors extracted from the interaural cross-correlation function, and the orthogonal temporal factors extracted from the autocorrelation function, respectively. Also, other important subjective attributes of the sound fields, the subjective diffuseness, and subjective preferences of both listeners and performers for single reflection are demonstrated here.

AN EVALUATION OF THE EFFECTS OF SCATTERED REFLECTIONS IN A SOUND FIELD

In this paper we outline and apply a procedure to evaluate sound fields in a concert hall which involve scattered reflections. We adopted an experimental method and used a 1/10 scale model of the concert hall. Arrays of circular columns were placed in front of its walls to act as the scattering obstacles. The acoustic properties of the hall were measured both with and without the arrays of circular columns. Here, the quality of the scattered sound field was evaluated in terms of four orthogonal physical factors: sound pressure level (SPL), initial time-delay gap between the direct sound and the first reflection (Δt1), subsequent reverberation time (Tsub), and magnitude of the interaural cross-correlation function (IACC). The IACC at central seats near the stage and seating area near the columns were improved by the arrays of circular columns.

Objective and subjective evaluations of twenty-three opera houses in Europe, Japan, and the Americas

The room acoustical parameters, reverberation time RT, early decay time EDT, clarity factor C80, bass ratio BR, strength G, interaural cross-correlation coefficient IACC, and initial-time-delay gap ITDG [definitions in Hidaka et al., J. Acoust. Soc. Am. 107, 340-354 (2000) and Beranek, Concert and Opera Halls: How They Sound (Acoustical Society of America, New York, 1996)], were measured in 23 major opera houses under unoccupied conditions in 11 countries: Argentina, Austria, Czech, France, England, Germany, Hungary, Italy, Japan, The Netherlands, and the USA. Questionnaires containing rating scales on the acoustical quality of 24 opera houses were mailed to 67 conductors, 22 of whom responded. The objective measurements were analyzed for reliability and orthogonality, and were related to the subjective responses. Presented are (a) the rankings of 21 opera houses each rated by at least 6 conductors for acoustical quality as heard by them both in the audience areas and in the pit; (b) relations between objective room acoustical parameters and subjective ratings; (c) findings of the most important of the parameters for determining acoustical quality: RT (or EDT), G(M), ITDG, [1 - IACC(E3)], texture (appearance of reflectrograms in the first 80-100 ms after arrival of the direct sound), a lower limiting value for BR, and major concern for diffusion and avoidance of destructive characteristics (noise, vibration, echoes, focusing, etc.); (d) the differences between average audience levels with and without enclosed stage sets; and (e) the differences between average levels in audience areas for sounds from the stage and from the pit.