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

Abstract

Subjective ratings of acoustical quality in concert halls were determined from interviews of conductors, musicians, music critics, and informed listeners during which they were asked to judge the quality of the acoustics and to rank the order of those halls each knew. Subjective ratings on 34 halls were divided into three principal groups: excellent, very good, and less good. No bad halls were studied. The use of the interaural cross-correlation coefficient (IACC) and lateral efficiency (LF) as correlates with the subjective ratings were analyzed in depth. In order to make IACC sensitive to quality ratings, a multi-octave-band average was developed, based on Blauert et al. [Acustica 59, 292 (1986)] and on a subjectively derived set of equal apparent source width (ASW) contours that showed the 0.5-, 1.0-, and 2.0-kHz octave bands to be of equal and principal importance to ASW. This IACC3 was divided into two components, IACCE3, integrated over the first 0.08 s after arrival of the direct sound, and IACCL3, integrated over 0.08–1 s. Subjective judgments were also performed to determine the effects of increased sound levels of symphonic music at frequencies above or below 355 Hz on ASW. It was found that changes in the low-frequency levels (GL) made greater changes in ASW than changes in the high-frequency levels (GH). Because the difference between the low- and high-frequency levels in real halls does not exceed ±5 dB and because both are inversely related to [EDT/Volume], it suffices to measure GL. Thus IACCL3 and GL appear valuable for determining ASW and the degree of sound-field diffusion in the frequency range from 100 to 3000 Hz for symphonic music in concert halls. The measured quantity [1−IACCE3] alone was found to separate exactly the 17 concert halls for which data were available into three category groups. The constituents of these three groups were the same as those determined from the interviews. The measured lateral fraction (LF) for the 24 halls for which data were available was also compared to the subjective groupings. The average measured values of LF for these halls covered a small range and there were so many overlaps among halls when separation of the halls into three rating groups was attempted that LF was judged not to be suitable for rating the acoustical quality of occupied concert halls.