Toshiyuki Okano

Interaural cross‐correlation, lateral fraction, and low‐ and high‐frequency sound levels as measures of acoustical quality in concert halls

Interaural cross‐correlation coefficient (IACC) and lateral efficiency (LF) as correlates with subjective ratings [see companion paper, L. Beranek, ‘‘Determination of categories of acoustical quality in concert halls’’ (this session)] were analyzed. To make IACC sensitive to quality ratings, a multi‐octave‐band average was developed, based on Blauert et al. [Acustica 59, 292] and on a subjectively derived set of equal ASW contours that showed the 0.5, 1.0, and 2.0 octave bands to be of equal importance. This IACC3 was divided into IACCE3 integrated over first 0.08 s after arrival of the direct sound, and IACCL,3 same, 0.08 to 1 s. Subjective judgments were also made to determine the effects of increased sound levels of symphonic music at frequencies above and below 355 Hz on apparent source width ASW. IACCE3 and GL both appear important for determining ASW. The measured quantity [1−IACCE3] alone was found to separate the 17 concert halls of this study for which IACC data were available into the same three category groups as those determined from the interviews. Lateral fraction (LF) was found to extend over a very small range and there were so many overlaps among halls when separation into three rating groups was attempted that LF was judged not to be suitable for rating acoustical quality.

Relations among interaural cross-correlation coefficient (IACCE), lateral fraction (LFE), and apparent source width (ASW) in concert halls

Relations are determined between one of the important subjective attributes of concert hall acoustics, the apparent source width, ASW, and three acoustical measures, interaural cross-correlation coefficient IACCE, LFE, and strength factor G. Although these measures previously have been found to correlate with ASW, their relations with it have not been examined sufficiently, especially in respect to their frequency characteristics. Herein, ASWs are directly determined for electronically reproduced musical sound fields with extensive ranges of values for IACCE and LFE. Investigated as parameters are angles of incidence, the time delay difference between a pair of symmetric early lateral reflections, and the number of early lateral reflections. These studies indicate the relative efficacy of IACCE and LFE for determining ASW under conditions that are realistically encountered in concert halls. The results were compared with measured IACCEs, LFEs, and also the strength factor Gs in existing concert halls. It is concluded that the arithmetic average of [1-IACCE]s at 500, 1 k and 2 k Hz combined with the strength factor Glow of the sound field at frequencies below 250 Hz are physical measures highly correlated with the subjective rank ordering of concert halls and that they cover the effects on ASW of the entire octave-band frequency range from 125 to 4 k Hz.

The Running Autocorrelation Function of Different Music Signals Relating to Preferred Temporal Parameters of Sound Fields

This paper defines the minimum value of the effective duration of the running autocorrelation function of a music signal used in subjective preference testing. A short duration signal, which may relate to ‘‘psychological present,’’ is more suitable for calculating the most preferred time delay of the single reflection than is the long‐time autocorrelation function. This fact is found particularly when passages of Japanese music are being played, which include a wide‐range‐noise component at the very beginning of the signal and the so‐called ‘‘MA.’’ MA is a stationary sound passage without any rapid movement. This paper also shows that, for subsequent reverberation time, a flat frequency response may be in the range of preferred conditions. Such preferred conditions may be calculated by the minimum effective duration of the running autocorrelation function of source signals after passing through the A‐weighting network.

Acoustical design of the Tokyo Opera City (TOC) concert hall, Japan

The TOC Concert Hall, called ‘‘Takemitsu Memorial’’ in honor of the late composer Toru Takemitsu, was opened September 10, 1997. With a seating capacity of 1636 and a volume of 15 300 m3, the hall is sized to cover the musical range from recitals to orchestral concerts. The plan is rectangular in shape but, by request of the architect Takahiko Yanagisawa, the ceiling is a distorted pyramid, with its peak nearer the stage than the rear of the wall. This unique shape had to be analyzed using a CAD model and a 1:10 scale model so that all interior surfaces would be adjusted in shape and absorption to yield optimum values for RT, EDT, IACCE3, surface diffusion, initial time delay gap, and loudness [Beranek, Concert and Opera Halls (ASA, New York, 1996)]. To provide a better ensemble condition for the musicians on stage and to provide early reflections to several other regions, a square canopy, almost 10 m on a side, is suspended above the stage. The pyramidal ceiling has diffusing elements added to simulate c...

Judgments of noticeable differences in sound fields of concert halls caused by intensity variations in early reflections

In concert halls early reflections combine with the direct sound and with reverberation to determine the subjective rating of a rooms acoustics. Of interest is how variations in the amplitudes of these early reflections are related to the subjectively just-noticeable differences (jnd) in several important acoustical parameters for their wide range encountered in existing halls. Investigated were four subjective parameters, apparent source width (ASW), loudness, intimacy and clarity, which are related to the physical measurements, [1-IACC(E3)], G, ITDG, and C80, defined mathematically in Beranek [Concert and Opera Halls: How They Sound (Acoustical Society of America, New York, 1996)]. Forty-eight types of sound fields were chosen in which to make variations in the amplitudes of early reflections and were reproduced electro-acoustically by multiple loudspeakers in an anechoic chamber. The results indicate that ASW and loudness are more sensitive to changes in the levels of early reflections, and were the primary parameters investigated. Although the number of subjects available with enough experience in listening classical music is limited and the measured jnd is an initial estimation, the jnd of [1-IACC(E3)] is measured as 0.065+/-0.015 in variations of sound field structures and the jnd of G was measured as 0.25+/-0.15 dB, which is consistent with the results of previous studies.

Relations between the apparent source width (ASW) of the sound field in a concert hall and its sound pressure level at low frequencies (GL), and its inter‐aural cross correlation coefficient (IACC)

The influence of GL (amplifier gain the low‐frequency range below 355 Hz) and IACC on ASW was determined by psychoacoustic experiments with simulated concert‐hall sound fields using anechoic symphonic music presented to subjects by multiple loudspeakers. ‘‘Equal ASW curves’’ were determined for 1/1 octave band filtered source signals with mid‐frequencies from 125 to 4000 Hz. The ASW’s for the upper four bands are found to be equal for the same IACC and SPL band values, indicating equal importance of those bands in determining overall ASW’s. Combinations of GL’s and IACCE3’s (average of IACC’s in the 500, 1 and 2 kHz bands) for wide‐band musical source signals were determined that produced the same ASW’s. The early sound was comprised of 2 to 11 early ‘‘reflections’’ and judgments were made with and without later reverberation. It was found that both larger values of GL and smaller values of IACCE3 result in larger values of the subjectively determined ASW’s. It is shown that GL and IACCE3 jointly are phys...

Interaural cross correlation (IACC) as a measure of spaciousness and envelopment in concert halls

Spaciousness, subjective broadening of the musical source, and envelopment, subjective immersion of the listener in the reverberant sound field, are measured by IACCEarly and IACCLate, respectively, in concert halls using binaural recordings from small microphones placed at the entrance of a person’s ear canal. IACC’s replace LF measured with a figure‐8 microphone, which fails to evaluate reflections from suspended panels and lateral reflections at small angles. Validity of real heads in comparison with artificial heads is confirmed. Measurements reveal difference between early and late as 80 ms, consistent with other orthogonal room measurements. An hearing importance curve is constructed to justify averaging the measured IACC’s in four 500‐ to 4000‐Hz octave bands, yielding, IACCE4 and IACCL4. The IACC’s must be derived from exact measurement of impulse response; the authors use a stretched pulse signal from a omnidirectional source. Preferred source positions and seats for measurements in halls are sta...

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)].

Necessary accuracy for early reflections in sound field auralization

Necessary accuracy for early reflection levels was investigated in order to derive criteria to verify the accuracy of a sound field simulator. Early reflections are the most important in controlling the subjective impression of a room’s sound field. The accuracy of the sound field simulator depends primarily on the accuracy of early reflections. The just‐noticeable level difference (jnld) of early reflections defines the maximum of allowable level difference between real and simulated reflections. Also, the masked threshold level (MTL) of early reflections defines the level of the smallest reflections which a simulator should generate. However, jnld and MTL of early reflections depend on the sound field’s configuration. It is desirable that criteria are independent of the configuration of the sound field. The jnld and MTL were measured for lateral and ceiling reflections in various types of rooms with simulated sound fields. Measured jnlds and MTLs were converted into the difference in objective parameter...

Multi‐dimensional analysis of subjective acoustical ratings and acoustical measures in existing concert halls

Correlations between subjective acoustical ratings and hall‐averaged values of acoustical measures are studied among existing worldwide major concert halls. It was shown that the classified acoustical ratings by Beranek [Concert and Opera Halls, How They Sound (ASA, 1996)] are discriminated correctly by combining binaural quality index (BQI) with some other acoustical measures. BQI is determined by the arithmetic average of inter‐aural cross correlation coefficient in three octave bands of 500, 1000, and 2000 Hz, subtracted from unity, calculated from the early 80‐ms part of binaural impulse response. Considering that the upper limit value of BQI not to cause disturbing image shift is approximately 0.85 at individual seat [Okano, J. Acoust. Soc. Am. 2219–2230 (2000)], the values of 0.6 or higher in hall averaged value of BQI, 0.85 or smaller in individual seat value of BQI, and approximately 5 dB or higher in strength factor at middle frequencies are proposed as design objectives to attain a high acoustic...