The authors are now making cooperative work to investigate room acoustic characteristics of various types of auditoriums. Up to this time, we have made surveys of concert halls and opera theaters in several European countries and Japan. In this article, the outlines of the acoustic measuring methods applied in these surveys are introduced and some examples of the measured results are illustrated.

View full abstract

It is a common belief among room acousticians that reverberation later than 80ms after the direct sound does not contribute to auditory spaciousness. However there is no definitive proof for this opinion. In this paper, two psychoacoustical experiments were performed. The first one compared the degree of spaciousness as caused by reverberation with the degree of spaciousness as caused by early reflections. The second one investigated whether spaciousness as caused by reverberation is suppressed by early reflections. The results indicated that the contribution of reverberation to spaciousness is the same as that of early reflections.

View full abstract

This thesis explains the calculation theory of a revised finite sound ray integration method and the algorithm based on the integration equation, which is obtained by the solution of Kirchhoff, from the wave equation. It also explains the calculation of small rectangular space pulse response by using the revised finite sound ray integration method, and correspondence of the values with model test values. The theoretical equation which is the basis of this calculation method is the transformed equation of Kirchhoff's integration equation, that is assuming a point source of sound without dimensions. In order to calculate this equation, the integral domain on the boundary surface, which changes in accordance with the lapse of time, and the sound wave propagation distance, etc. have to be obtained. We, therefore, simulated the spherical progressive wave by utilizing a multiple number of sound rays, which were radiated from the sound source at equal solid angle, and then obtained the various necessary information for the calculation. Up to this time, the said calculation method is programmed on the condition that the pursuing sound rays are only mirror reflecting sound rays, and a small computer can be used for the calculation. When we measured the pulse response by using a 1/4 scale rectangular room model, which was composed of complete reflecting surfaces and a complete sound absorbing surface, and the results were compared with the calculation values by the revised finite sound ray integration method, we were able to obtain very reasonable results.

View full abstract

When we estimate a sound field in a room, it is important to grasp the spatial informa-tion, especially of the early reflection periods. In this paper we'll discuss a way to grasp the spatial information of sound fields from impulse responses measured at closely located four points, the origin and three points of the same distance (3∼5cm) from the origin on the rectangular coordinate axes. From these four impulse responses the coordinates and powers of virtual image sources are calculated by correlation technique or intensity technique. Concert halls, opera theaters and many other sound fields are measured by this technique. The distributions of virtual image sources and directivity patterns of some concert halls are shown.

View full abstract

As a method to extend the efficiency of scale model experiment of room acoustics, the hybrid simulation technique has been developed, in which impulse responses are measured in a 1/10 scale model auditorium and they are convolved with arbitrary anechoic source signals like music and speech by the digital convolution technique. Through these nrocesses, the “hall tones” can be heard in advance of construction of the real auditorium. A spark discharge sound source with high repeatability and broad band spectra has been developed for the impulse response measurement, and a simple model dummy head was used for binaural recording. As the result of various contrivances for computer programming, considerably fast convolution calculation has been realized.

View full abstract