Perforated facings are now extensively used in practical applicaitons as acoustic material. In this paper are described measurements of the acoustic impedance of fairly thick porous materials with perforated facings, in particular the effect of the air-space between the two components upon the acoustic reactance; and the date have been compiled into a simple design chart. The acoustic reactance of perforated panels of various dimensions (Table 1) placed before a rigid wall is shown in Fig. 2 with the width of the air-space as parameter. The solid curves in this figure were calculated from Eq. (1) and are found to be in fairly good agreement with experiment. From these results we can conclude that the acoustic reactance of these structures may be considered to be consituted of two components: the mass reactance of the holes and the compliance of the air cavity. Secondly, the acousitc impedance of preforated panels supplemented by backing of porous material was measured for several combinations of panel perforation, porous material and air-space (Table 2). The results of measurement are as follows:(1) The acoustic resistance of this strucure is shown in Fig. 5. Here R__0 represents the resistance when the perforated panel is placed immediately in front of the porous material without air-space. It seems to be reasonable to consider as an approximation that, independently of the frequency of the sound and the variety of the porous material, the ratio R__L/R__0 decreases as the distance between the panel and the porous material increases, though the value of R__0 itself depends on the variety of the porous material and the arrengement of the preforated facings. (2) As examples of caluculation of acoustic reactance, Fig. 6(a) shows the measured results for a perforated panel(hole diameter d=0. 45cm, thickness l=0. 5cm, open area ratio p=0. 078) backed by a padding of hairfelt(1. 2cm thick) and separeated by a 1cm air-space. In this figure X__<m0> and X__<c0> are the mass and stiffness reactance of the holes and the air cavity respectively as caluculated from Eq. (1) and X' is the reactance of the hair-felt itself. The composite reactance X caluculated by means of the equivalent circuit of Fig. 6(b) and by using the values of X__<m0>, X__<c0>, X'(for this example) also is shown, by the solid curve, and this is in good agreement with measurement. Measured and caluculated results for other combinations of panel and porous material are shown in Figs. 7 to 10, while the absorption coefficents calculeted from the acoustic impedance are represented in Figs. 11 to 14. Combined absorption characteristics can thus be estimated when the acoustic impedance of porous materials are known.
Measurements were made of the frequencies and modes of resonant vibrations of fibrous acoustic tiles which were set up into a scale model so named by the Japan Broadcasting Corporation the "B-type standard studio wall structure". Each piece of the tilework of the model showed a different resonant characteristic due to the non-uniformity of the texture of the tile piece and to slight differences originating from accidental constructional irregularities. The frequency of occurrence of resonance was therefore measured for band-widths of 25 c/s below 300 c/s, and for band-widths of 50 c/s above 300 c/s on a large number of pieces of tile in the model. As a result of these measurements, the occurring frequency of resonance was found to be dominant around and below 200 c/s. The frequency characteristics of the reverberant sound absorption coefficient of this type of wall structure were also measured. From these results, it can be qualitatively interpreted that the increase of sound absorption below about 200c/s generally found in such types of wall structure is due to the most dominant occurrence in this frequency band of the resonant panel vibrations of the component pieces of tiling.
Reverberation time is generally used in the estimation of the acoustical properties of a room. However, the decay curve representing a room shows complex irregularities according to the nature of the source of sound and the writing speed of the recording system. To find the effects of the source and the writing speed of the recorder on the reverberation curve, the author obtained the reverberation curves of several halls of differernt sizes and has set forth a tentative standard method for the determination of the reverberation time. When we use white noise as the source of sound, the amplitude fluctuations of the source appear on the slope of the decay curves; several such curves obtained at a given position in the room will not appear identical, due to the fact that white noise has a random distribution of amplitudes and the source is cut off at arbitrary instants. (Examples are shown in Fig. 1). In curves where a warbling tone is adopted as source a similar fluctuation is observed that depends on the frequency of the tone being emitted when the source is cut off. The decay curves obtained with a pistol shot source and those with white noise of short duration such as 30 m. s. , 50 m. s. , and 100 m. s. , were compared with those resulting from white noise lasting 1 second and from a warbling tone. To get relatively high levels at low frequency, the pistol shot was fired in a small felt covered box, the energy being radiated from a small hole on the box. The fluctuation of the decay curves are influenced also by writing speed of the recorder: the experimental results with a white noise pulse of 100 milliseconds are shown in Fig. 5. Letting the decay constant of the sound be β and that of the recording system α, we obtain theoretically the decay curve (x/x_0)=<α/α-β^e>^-<βt>-<β/α-β^e>^αt where x__0: maximum ampliteude of the signal, x: amplitude at time t. Calculated examples are shown for the cases of writing speeds of 70, 140 and 300 db/sec, and for reverberation times, of 0. 7, 1 and 1. 5 secs. From this it will be seen that the reverberation time must be determined from the tangent of the courve, particulary when the writing speed is not adequate in relation to teh decay rate of the sound. The shape of the decay curve is affected by the directional properties of the source also, a steep decay in the initial part is often observed at high frequencies. It has nothing to do with the properties of the room and this part must be excluded in determining the reverberation time. For the study of the acoustical properties of a room, the detailed structure of the decay curve must be investigated, and the author has proposed a method of annotating irregularities in the charts representing the frequency response of the reverberation time.
The pulse method has become increasingly popular for experimental researches on room acoustics, since speech and music may usually be regarded as pulsive tones emitted in succession. We have made a convenient apparatus for this purpose. The A. C. pulse which is used as the sound source is composed on an audio frequency carrier of arbitrary duration, which carrier, if required, can be modulated in frequency. The oscilloscope may be made to display alternatively any two of either the transient wave form, the envelope of the reflected pulse tones, the timing, or the wave form of the source in two elements. A general description of the apparatus and some examples of measurements are given.
The steady-state transmission characteristics of a fan-shaped room, taken up as an example of a splayed room, were analysed mathematically. In this analysis, the wave type factor "eN", developed by Morse and others, was used. The factors eN together with the distribution of the normal modes of the model were compared with those of a rectangular room. The results were as follows: The fan-shaped room with ρν⋍1. 2〜1. 4 has good characteristics in the frequenccy region of less than about 200 c/s where ρν is the ratio of the outer to the inner radius of the room. And the fan-shaped room with ρν⋍1. 2〜1. 4 corresponds to a room with l/a=0. 09〜0. 17, where l is given by l=b/2tan φ__0, φ__ν and b being angle of splay and the distance between the parallel walls of the room respectively. These results show good agreement with facts known from experience and with the results of our experiment to be described in our next paper.
The steady-state transmission characteristics in model rooms of various shapes were measured. The shape of the model was varied from the rectangular to the trapezoidal and the quadrangle with non-parallel walls, while maintaining the interior volume of the room constant. The experimental results obtained on the decrease of the factor F__L in the transmission characteristics of a room of trapezoidal shape agree well with the results of theoretical analysis described in the preceding paper.
The "echo-stripe diagram" was proposed in order to investigate the "echo-time diagram" with respect to the requencies of sound in rooms. If we plot as abscissa the lapse of time between the instant of observation of the direct sound and that of one of the peaks or dips in an individual reverberation curve, and ordinate the frequency of sound, then connect vertically the points thus plotted, an "echo-stripe diagram" such as shown in Figs. 1 and 2 is obtained. A study was made of the diagram thus obtained from the pulsed glide displays observed at the Daiichi Seimei Hall in Tokyo. It was clarified that the continuous vertical stripes conspicious in the diagram was the combined effect of the focussing action on sound of the concave hard wall of "Horizont" on the stage and the multiple reflection of sound between the "Horizont" and the plane plywood panel placed behind the audience. A method was also devised of obtaining the "echo-stripe diagram" directly on photographic film with the same apparatus as used for obtaining the pulsed glide display. The display obtained by this method was named the "longitudinal pulsed glide display" to distinguish it from the pulsed glide display presented by Mr. Somerville. The two devices should both have their useful purposes in the investigation of the acoustics of rooms.
The Denki Hall is accepted today as one of the best in this country from the point of acoustics judged from subjective evaluations. This paper reports on data from objective measurements made on its acoustical properties together with the results of a survey of audience impression. The interior volume of this hall is about 5700 m^3 and the seating capacity 1164 persons. In preparing for our investigation, we aimed at grasping the general characteristics of the hall with a minimum of time spent in actual measurements. With this point in mind, the measuring set up was arranged as shown in Fig. 3. The sources of sound used for reverberation measurements were the octave band noise (1 sec and 50 millisec), previously recorded on magetic tape, and pistol shots. The reverberant sound in both cases were also recorded on tape. The recorded sounds were then reproduced, analyzed and recorded on a Bruel high speed level recorder. From these data we studied the reverberation characteristics, the steady state pressure distribution and other factors. The results of measurements may be summarized as follows. (1) The frequency chracterisitics of reverberation time (abbrebiated R. T. ) for the empty hall are shown in Fig. 5 for six different stage conditions. (It is the usual practice now in Japan to use such halls for a variety of pruposes such as for music, lectures, operas, plays, movies and so forth. The stage arrangement is changed according to these purposes and this influences the acoustics of the hall considerably. ) In this figure, i)〜iii) are for cases where a sound reflecting board was set on the stage and iv)〜vi) for cases without the board. The sound reflecting board lengthens the R. T. in the medium frequency region and shortens it in the lower frequencies (Fig. 6). Deviation of R. T. from mean values at each point are represented in Fig. 7. (2) The R. T. with the audience half full is shown in Fig. 8. It is 1. 3 sec at 500 c/s which is 0. 2 sec shorter than the Knudsen-Harris optimum value for a hall of this size. The R. T. of other halls in Japan are in general considerably shorter than that of this hall, and so it is probable that this is the important reason accounting for the superiority of this hall. (3) The steady state octave band noise levels were also measured at several points in this hall and the results are shown in Fig. 9. The comments from the audience are collected in Fig. 10. Except for the low-frequency reverberation, the acoustics of this hall were thought to be good by the greater part of the audience. In the lower frequencies, 30% of the audience comments mention insufficient reberberation. This could be predicted from the results of our objective measurements.
This report is divided into three parts. The method of articulation test is described in chapter 1. The word lists compiled by a random enumeration from the " Table of One Hnndred Japanese Syllables " were tape-recorded by a standard speaker, and the conditions of the sound source were then maintained constant throughout the tests. Chapter 2 gives an explanation of the analysing method for the test results. The articulation νs. speech level characteristics in a given rooom is analysed by using the articulation νs. speech characteristics in a dead room, the characteristic relation between articulation and the noise and speech levels, and the articulation νs. reverberation time characteristics. Using the method described in chapters 1 and 2, practical studies on the test results of each Hall are presented in chapters 3 to 5. In these chapters, average percentage-articulation and distribution of articulation in the Hall were compared with the standard criteria, and effective reverberation time and noise level of each Hall then estimated by reference to the articulation νs. speech level characteristics of the Hall.
Following are the summarized results of measurements on reverberation time and on sound pressure distribution in two auditoriums in Tokyo, the Hibiya City Hall having a seating capacity of 2600, and the First Life Insurance Auditorium seating 650. Both are used as public halls and are amongst the top-ranking auditoriums in Tokyo. (1) The First Life Insurance Auditorium: The sound pressure was measured at eighteen seats employing a sound source situated on the stage and producing a field intensity of 93 phones at the front edge of the stage. The result is shown in Fig. 2. A sound reflecting borad provided in the background of the stage seems to have little effect in increasing the sound pressure reaching the furthest parts of the auditorium. Noise entering from near-by streets in 44-49 phones when the double windows are closed, and 58-65 phones when the windows are open. By the use of questionnaires we have also surveyed the opinion of the audience regarding the acoustics in this auditorium. The results are presented in §5. We have discussed in the present article the date thus obtained and have formulated a plan for the remodelling of the auditorium. For measuring reverberation time we have used white noise, pulse sound and pistol sound emitted from the cneter of the stage. Average reverberation time is 1. 28 seconds on the first floor and 1. 32 seconds on the second floor, and there is little variation in the freauency characteristics. (6) Hibiya City Hall: this auditorium has an interior volume of 11, 872 cubic meters, the floor plane being as shown in Fig. 6. The sound pressure distributions is represented in Fig. 7, 8 and 9. The sound pressure of higher frequencies decreases much faster then that of frequencies as the ditance increases from the stage. The frequency characteristics of the reverberation curves in the auditorinm is almost flat thoughout the range 0. 5-2 kilocycles. The average reverberation time is 1. 4 seconds, when the auditrium is empty and 0. 9 second when full.