日本音響学会誌
Online ISSN : 2432-2040
Print ISSN : 0369-4232
27 巻, 1 号
選択された号の論文の10件中1~10を表示しています
  • 小橋 豊
    原稿種別: 本文
    1971 年 27 巻 1 号 p. 1-2
    発行日: 1971/01/10
    公開日: 2017/06/02
    ジャーナル フリー
  • 山本 武夫
    原稿種別: 本文
    1971 年 27 巻 1 号 p. 3-10
    発行日: 1971/01/10
    公開日: 2017/06/02
    ジャーナル フリー
    In modern halls and theaters, so called proscenium loudspeaker systems are used for sound reinforcement. Most of the proscenium loudspeaker systems consist of many medium size component loudspeakers distributed over the front part of the ceiling to obtain a desired directivity. This paper shows a design method for anti-acoustic feedback loudspeaker system. The sound pressure level frequency response of the arrayed loudsperker cluster in the direction of the stage have peaks in the medium frequency range as shown in Fig. 2. It follows that an acoustic feedback can easily happen at some peak frequency. This is caused by the fact that sound radiating from the component loudspeakers in the direction of the stage are in nearly the same phase in this frequency. Therefore, to eliminate the acoustic feedback, it is necessary to suppress the peaks in the frequency response. As a result of the study of several methods to suppress the peaks, it becomes evident that an unequally spaced array of the component loudspeakers is the most useful. In this study, sound pressure level frequency responses of unequally spaced linear array point sound sources, with geometrical or arithmetical progression spaces, were investigated. Fig. 4 and Fig. 6 are the calculated sound pressure level frequency responses of the unequally spaced linear array point sound sources. with geometrical and arithmetical progression space, (n=6, minimum space=20cm) at a point far distant from the center of the array and in a direction of 60-degrees off axis from the main lobe. If the larger common ratio or common difference is used, the lower peaks in the medium frequency range can be obtained. If a more larger common ratio or common difference were used, however, small peaks would appear in the high frequency range, and the total size of the loudspeaker system would be large. Therefore, there are most suitable common ratios and common differences for the numbers of component loudspeakers. Table I shows the suitable common ratios and common differences. Fig. 9 shows an example of a proscenium loudspeaker cluster with equally spaced linear array. Fig. 10 shows the measured loop gain frequency response just below acoustic feedback of the sound reinforcement system with the proscenium loudspeaker cluster shown in Fig. 9. The peaks at 2, 500 Hz and 4, 000 Hz show that acoustic feedbacks can easily happen at these frequencies. Fig. 13 shows an example of the proscenium loudspeaker cluster with arithmetical progression spaced curved array. The calculated sound pressure level frequency response in the direction of the stage in this case does not have large peaks as shown in Fig. 14. Consequently, the loop gain frequency response is almost flat, as shown in Fig. 15. If the sound pressure level frequency reponses in the direction of the stage shown in Fig. 12 and Fig. 14 are compared, it is evident that the acoustic feedback stability can be improved 5 to 6 dB by the use of the unequally spaced array.
  • 三輪 俊輔, 米川 善晴
    原稿種別: 本文
    1971 年 27 巻 1 号 p. 11-20
    発行日: 1971/01/10
    公開日: 2017/06/02
    ジャーナル フリー
    Problems of measurement and evaluation of hand vibration as well as of the whole body vibration have recently been raised in the field of public unisance and industrial health. Many researchers have published various kinds of papers concerning the problems, however it is somewhat hard to derive clear evaluation methods from their results (Figs, 1, 2 and 3), because of difficulty of systematization. Psychological experiments were carried out by using new vibration tables of electro-dynamic type in vertical and horizontal directions, to establish a new systematic method of evaluation. Threshold level defined as the minimum perceptible vibration acceleration level and equal sensation level defined as the vibration level of the other frequency of the same vibration sensation as that at 20 Hz were observed by the method of paired comparisons on the whole body in sitting, standing and lying postures and on the hand (Fig. 6, 8 and 9). It is shown in Fig. 6 that there is no difference in the threshold curves (T curves) between sitting and standing postures, and that the T curve of the horizontal vibration is higher by 10 dB than that of the vertical vibration in the domain above 5 Hz. Characteristic feature of the T curve of the lying posture in Fig. 8 is its dip near 80 Hz for the vertical vibration due to skull resonance. In Fig. 9, the T curves of hand for both vibrations are in sufficient agreement. No clear differences in the equal sensation curves (ES curves) between sitting and standing postures were observed in Fig. 6, and those for vertical and horizontal vibrations have the same slope in the domain above 7 Hz. The ES curve of the vertical vibration in lying posture (Fig. 8) is lower by 11 dB than that of the horizontal vibration near 80 Hz. ES curves of the hand vibration are almost similar to those of the horizontal whole body vibration. Thus, by the value on these ES curves we define the vibration greatness level which has unit of VGL corresponding to the loudness level (phon). Next, the sensation difference between vertical and horizontal vibrations was observed. It is seen in the domain above 10 Hz that the sensation of the vertical vibration is severer by about 10 dB in sitting and by about 13 dB in standing posture than that of the horizontal vibration of the same amplitude. This difference gradually decreases below 10 Hz till it vanishes near 2 Hz, while there is no sensation difference between the two vibrations on the hand. Finally, correlation between the VGL values and unpleasant or intolerable level was defined as Eq. 1. This expression was determined by carrying out short time exposure experiments with 10 min. and referring to long exposure results in the draft of ISO/TC 108/WG 7.
  • 三輪 俊輔, 米川 善晴
    原稿種別: 本文
    1971 年 27 巻 1 号 p. 21-32
    発行日: 1971/01/10
    公開日: 2017/06/02
    ジャーナル フリー
    Compound sinusoidal vibrations with several frequency components and random vibrations of a continuons wide frequency band are often encountered in the environmental and industrial investigations. Although a few evaluation methods for these vibrations have been proposed, they are not available to all the cases, and the new method widely applicable has been required. A ratio scale called vibration greatness in the unit of VG was experimentally determined using the corrected ratio method proposed by W. R. Garner. This corresponds to loudness (sone) in psycho-acoustics. The object was to sum up the VG values of their frequency components. The relations between vibration greatness level (VGL) and vibration greatness (VG) on the whole body and on the hand, for the vertical and the horizontal vibrations, were obtained as in Fig. 2 and Table 1. For the evaluation of the compound sinusoidal vibrations, the vibration acceleration levels (VAL) of their frequency components were converted into VG values (VG_i) by Table 1 through their VGL values using Figs. 6 and 9 in Part 1. Next, these VG_i values were summed up in a similar method to one proposed by S. S. Stevens devised for assessment of noise. That is, VG_T=VG_M+0. 3((Σ__i VG_i)-VG_M where VG_M is the maximum value among VG_i values. These calculated values were compared with the observed ones which were sensationally obtained by the comparison of the compound sinusoidal vibration with the vibration of 20 Hz as the standard frequency. These two values are in good agreement with each other below the level of compound vibration of 55 VGL. For the assessent of random vibrations, threshold and equal sensation curves were determined for the random vibration with 1 and 1/3 octave-band. Their results on the whole body and the hand are shown in Fig. 6 and 7. VG values of the frequency components of the random vibrations of continuous wide frequency bands were obtained using these figures and Table 1, and were summed up with weighting factor 0. 3 for the data analyzed by the 1-octave band and 0. 13 for 1/3 octave band, in the same method as in compound sinusoidal vibrations. The calculated values were in good agreement with the observed values in which the random vibration was directly compared with the vibration of 20 Hz. The sensation difference between vertical and horizontal vibrations was observed for the 1-octave band ranbom vibration. Sensation of the vertical is severer by 10 dB than that of the horizontal vibration at the same vibration level above 16 Hz.
  • 三輪 俊輔, 米川 善晴
    原稿種別: 本文
    1971 年 27 巻 1 号 p. 33-39
    発行日: 1971/01/10
    公開日: 2017/06/02
    ジャーナル フリー
    Pulsed vibrations seen in forging and pile driving works have almost occupied the vibration problems concerning human sensation. Results derived by Reiher &amp Meister as shown in Fig. 1 are helpful for assessment of the puled vibrations. However, they have treated damped vibrations alone, although there are several other kinds of pulsed vibrations such as pulsed sinusoidal and built-up vibrations (Fig. 2). Sensation of the pulsed sinusoidal (PS) vibration of a certain constant level was matched to continuous (CS) vibration of variable level having the same frequency as the fundamental frequency of the PS vibration. Duration of PS vibration was changed between 6 and 0. 007 sec. The results of the matching experiments on the whole body and the hand for vertical or horizontal vibration are shown in Fig. 3, a, b and c. The ordinate indicates VAL_p-VAL_j (dB) and the abscissa the duration of the PS vibrations (t)(sec). VAL_p means vibration acceleration level (VAL) of the CS wave which has the same maximum amplitude and the same fundamental frequency as the PS vibration, and VAL_j means the VAL value of the CS vibration judged as the equal sensation with the PS vibration. These reults are approximated by three lines formulated as, VAL_p-VAL_j=7 log_&lt10&gt T_0/t, where T_0 is the critical time limit which was nearly 2 sec for 2-60 Hz, 0. 8 sec for 60-200 Hz and 0. 5 sec above 200 Hz. Then, the sensation of the damped or the built-up vibration was equalized to that of the PS vibration having the same fundamental frequency and the maximum p-p amplitude by changing the duration of the PS vibration every half period of its fundamental frequency. It was found that the damped vibration, regardless of decay processes, is equalized to one period of the PS vibration and that the built-up vibration is equalized to the PS vibration of the duration which corresponds to its duration just when the amplitude of the built-up vibration is smaller by 1. 5 dB from its maximum value. The matched duration of the PS vibration is defined as equivalent duration(Table 1). Thus, the vibration greatness sensation for the damped or the built-up vibration can be estimated from the equivalent PS vibration. On the other hand, the estimated values were compared with the observed values which were determined by the sensational comparison between the damped or the built-up vibration and the CS vibration having the same fundamental frequency. Both values are in good agreement as seen in Table 2. Evaluation methods for these three kinds of pulsed vibrations were eventually established.
  • 佐藤 幸平
    原稿種別: 本文
    1971 年 27 巻 1 号 p. 40-41
    発行日: 1971/01/10
    公開日: 2017/06/02
    ジャーナル フリー
  • 青木 一郎
    原稿種別: 本文
    1971 年 27 巻 1 号 p. 42-50
    発行日: 1971/01/10
    公開日: 2017/06/02
    ジャーナル フリー
  • 早川 尚夫, 御子柴 宣夫
    原稿種別: 本文
    1971 年 27 巻 1 号 p. 51-53
    発行日: 1971/01/10
    公開日: 2017/06/02
    ジャーナル フリー
  • 藤崎 博也, 城戸 健一, 中田 和男, 玄地 宏, 早坂 寿雄, 藤村 靖
    原稿種別: 本文
    1971 年 27 巻 1 号 p. 54-64
    発行日: 1971/01/10
    公開日: 2017/06/02
    ジャーナル フリー
  • 原稿種別: 本文
    1971 年 27 巻 1 号 p. 68-
    発行日: 1971/01/10
    公開日: 2017/06/02
    ジャーナル フリー
feedback
Top