In this paper, the results of model room experiments, carried out to determine the effect of the room shape on the steady-state and transient characteristics of the normal modes of vibration, are mentioned. In order to carry out accurate measurements of absorption coefficient by Sabine formula, the assumption of the diffuse distribution of sound energy in the reverberation chamber must be fulfilled. The absorption coefficient problems have been discussed by many authors, and thus, fairly good explanations on the sound field of the rectangular chamber have been made available up till present. Also, many attempts have been made to obtain the diffuse condition in the chamber (such as, non-parallel walls, cylindrical pillars, rotating vanes, etc. ) But, most of them were merely product of empirical drocedures, and it is deemed necessary to investigate more extensively the influence of irregular room shape on the sound field of the reverberation chamber. From this point of views, it was decided to start this work from the basic research of the sound field in the model room. In the first place, the influence of the room shape on the normal modes of vibration was investigated by using the two dimensional models such as shown in Fig. 1. The following results were obtained. (1) At low frequencies, normal modes of vibration existed even the room shape were made adequately irregular. (2) Sound pressure level at the corner of the room at each normal mode depends on the relative position of loudspeaker and microphone, except in the case of rectangular room. (3) The standing wave pattern of each normal mode was investigated using the dust figures (The details of this method will be reported before long). Some of the results are shown in Figs. 5 and 6. When the parallel plane or symmetry exists in the room shape, the nodal lines still appear regularly. But, as the room shape becomes irregular, the nodal lines of successive normal modes distribute at random in space. Thus, in such an irregular room, the number of normal modes involved in the decay of sound will be irrelevant to the position in the room. (4) For each normal mode, spatial variation of sound pressure does not depend on room shape. Then, we constructed a three dimensional model of irregular pentagon (Fig. 8) and compared with a rectangular room. The result were: (1) As for the pressure distribution of each normal mode, the results were the same as in the case of two dimensions. (2) In this irregular room, the variation of decay rate for each normal mode was much smaller than that of a rectangular room (Fig. 12). It seems to be reasonable to consider that the well-defined axial mode decrease in such an irregular room. (3) When the warble tone is used as the sound source, its decay rate becomes the average of those of normal modes involves in its frequency range. At low frequencies, considerable bent of decay curve occurs for rectangular room. From these results, it is concluded that a certain irregular room is superior than the rectangular room as a reverberation chamber.
In this paper, the details of the new reverberation chamber with non-parallel walls are mentioned. To determine the shape and volume of this reverberation chamber, this results of model room experiments (previously reported) were used. Details of new reverberation chamber are as follows: Chamber No. 1 Chamber No. 2 Volume 513 m^3 120 m^3 Surface area 382 m^2 148 m^2 Shape Irregular pentagon Rectangle Wall thickness 40 cm 26 cm The inner surface of the wall and floor was finished by polished concrete. The warble tone was used as the sound source. The reverberation time-frequency characteristics of empty chamber are shown in Fig. 6. At 500 c/s, the reverberation time of chamber No. 1 is 22 sec, and that of chamber No. 2 is 14 sec.
In this paper, several problems concerning the absorption coefficient measurements are mentioned. The new reverberation chamber No. 1 (Volume, 513 m^3; reverberation time at 500 c/s, 22 sec) was used for this research. The test sample used are fibrous acoustic tiles and mineral wool. These materials were selected from the commercial acoustic materials according to their frequency characteristics of absorption coefficient. The following results were obtained: (1) Decay curve shows nearly perfect logarithmic decay even when the test materials are present. (2) Absorption coefficient of test materials was unaffected by the position of the microphone for the sample area ranging from 10- m^2 to 30 m^2. (3) For the sample area 10 m^2, absorption coefficient was not dependent on the position of the test materials. There results seem to confirm the presumption of model room experiments (previously reported). In closing the accuracy of absorption coefficient measurements is discussed. It is stated that, in this chamber, the variation of repeatedly measured decay curves is very small and the reverberation time is scarcely affected by the observers. Thus, it is possible to measure the absorption coefficient with relatively small test materials.
The heating phenomenon of bone and other materials by ultrasonic irradiation has been investigated. Metals, whose internal losses are negligible, do not exhibit the temperature rise in usual liquids in the ultrasonic field, although a slight temperature rise originating in the surface layer is measured in viscous liquids such as castor oil. On the other hand, in the case of the bone and the Indian rubber, whose sizes are over a certain dimension, the internal absorption of ultrasonic waves plays the major role of the heating ; and the surface heating or interface heating is considered only as a secondary effect. Moreover, the differences of the temperature rises in various liquids indicates that the cooling effect in the ultrasonic field depends on the viscosity. The coefficient λ describing the rate of cooling will lead to the following empirical formula λ∝1/(A+Blogη). The existence of bone cortex and muscles is not so essential as to the bone heating. In the frequency range from 500 kc to 2 Mc, the tendencies of the temperature distributions are very similar in the bones whose thicknesses are about 1 cm, and the temperature rise is about 2 times higher in the front layer facing the sound beam than in the rear layer. The above measurements leads to the estimation that the ultrasonic waves are mainly absorbed near the superficial layer, and the temperature distributions are influenced by the heat conductivities.
In order to know the abnormalities in the speech sounds of the deaf children, the experiments were performed to analyse with the Sonagraph the vowels of the deaf children and also as the control those of the normal hearing children. The following results were obtained: I. Concerning the pitches of vowels, (1) The individual differences of the pitches of vowels could be found either in the normal group or in the deaf group, but no significant difference could be ascertained between both groups by F-test. (2) Though there were negative correlations between the pitches and ages in both groups, the significant difference could not be found between the normal and deaf groups. (3) Among five vowels (a, e, i, o, u) the significant differences of the pitches was found in the normal group, but not in the deaf group. (4) the height of the pitches of vowels in the deaf group were in close correlation with the degree of hearing loss. As the hearing loss increased, the pitches of vowels decreased. II. Concerning the spectral of vowels, (1) The frequency regions of the formants in the normal group were higher than those in the group of normal adults. The differences in the spectra of vowels was found between the male and female group, but these differences were not so great as those found between the normal and deaf group. (2) The spectra structure of vowels was observed to be much influenced by the degree of hearing loss in the deaf group. The frequency regions of the formants of vowels were shifting to lower frequency region, and moreover the characteristic differences in the spectra among five vowels were decreasing. (3) In the deaf group with severe hearing loss, the individual differences in the spectra of vowels were remarkable, but the irregularities in the spectra were not always proportional to the degree of hearing loss. (4) As regards to the deaf children, whose hearing had been impaired after three years of age, it was noticed that the spectra of their vowels resembled to those of the normal hearing children.