Threshold of hearing for pure tone under free-field

Transcription

1 J. Acoust. Soc. Jpn. (E) 15, 3 (1994) Threshold of hearing for pure tone under free-field listening conditions Hisashi Takeshima,* Yoiti Suzuki,** Masazumi Kumagai,* Toshio Sone,** Takeshi Fujimori,*** and Hajime Miura**** *Sendai National College of Technology, 1 Kitahara, Kamiayashi, Aoba-ku, Sendai, Japan **Research Institute of Electrical Communication, Tohoku University, Katahira, Aoba-ku, Sendai, 980 Japan ***Electrotechnical Laboratory, Umezono, Tukuba, 305 Japan ****Sizuoka Institute of Science and Technology, Houzawa, Fukuroi, 437 Japan (Received 6 October 1993) The threshold of hearing for pure tone under free field listening conditions were measured to prepare basic data for a full-scale revision of the international standard ISO 226. The frequency range for the measurements was 31.5 Hz `16 khz. Number of subjects was depending on stimulus frequency. The following features are shown by comparing our data with others. (1) The thresholds of hearing do not have any great difference from other data, but some systematic deviations from ISO 226 are observed, i.e., our data are 2 `6 db higher at frequencies below 160 Hz and 1 `3 db lower at frequencies between 400 Hz and 6.3 khz. (2) A small peak between 1 khz and 2 khz is observed in our results, while the peak can not be seen clearly in other data. Supplementary experiments show that the frequency characteristic of the threshold of hearing significantly correlates with head transfer function in a frequency range between 1 khz and 8 khz, and it suggests that the small peak mentioned above is attributable to a dip in head transfer function. But this small peak was affected by head movement of a subject. Keywords: Threshold of hearing, Hearing threshold, Free-field, Pure tone, Head transfer function PACS number: Cb 1. INTRODUCTION The threshold of hearing is the minimum sound pressure level at which we can hear a sound. It is an important characteristic, useful in describing the capability of the human auditory system. For example, it is the basis for models of the auditory system, and provides reference values for audiometers and noise evaluation. It has been, therefore, standardized in ISO 226 along with the equalloudness level contours. The earliest study of the threshold of hearing under free-field listening conditions was made by Fletcher and Munson1) in They carried out measurements using earphones, then converted the data into those for free-field listening conditions. The data adopted in ISO 226 were obtained by Robinson and Dadson2,8) in Subsequently, Berger4) measured the threshold of hearing under free-field and diffuse-field conditions in the frequency range of 50 Hz to 1,000 Hz. In Berger's4) paper, he proposed revise ISO 226 for the low frequency

2 J. Acoust. Soc. Jpn. (E) 15, 3 (1994) region using estimated thresholds from his and others' published data. The standard, however, was not revised. Teranishi5) measured the threshold of hearing under free-field listening conditions in Japanese subjects in In his study, the subjects were classified according to age. At the meeting of ISO/TC43/WG1 in 1985, a new experimental finding abot equal-loudness level contours was presented by a member body.6) The data showed a significant difference from the standard of the existing ISO 226 at around 400 Hz. Referring to this, ISO/TC43/WG1 decided on a fullscale revision of ISO 226, including the threshold of hearing, as a new work item. We have been studying, therefore, the threshold of hearing as well as the equal-loudness level contours since 1986, and reported the first experimental data in ) Other studies in which the threshold of hearing and the equal-loudness level contours were measured towards the goal of revising the standard have been made by Betke et al.,8,9) Fastl et al.10) and Moller et al.11,12) In this paper, new data on the threshold of hearing for pure tones under free-field listening conditions are presented and compared with ISO 226 and other data. Furthermore, to examine the reliability of our results, we carried out two additional experiments on the relation of the head transfer function of subjects and the effect of a headrest used to fix a subject's head. 2. MEASUREMENT OF THRESHOLD OF HEARING 2.1 Sound Field and Experimental Setup The experiments were performed in an anechoic room at the Electrotechnical Laboratory in Tukuba Science City. The interior of the room measures 9.5m ~8.0m ~7.2m, and its surface is covered with about four thousand sound-absorbing wedges, each 2m in length with a low cut-off frequency of 40 Hz. The 1/3 octave band levels of ambient noise in the room") are more than 20 db below the threshold of hearing specified in ISO 226 (see Fig. 5). We carried out the experiments six times with some intervals. The loudspeaker system used to present stimuli and the distance from loudspeaker to the reference point were changed to extend the frequency range measured. The details are shown in Table 1. The details of experimental setups for EXP1 and EXP2 are described in our previous report.7) The experimental setup used in EXP3 ` EXP6 illustrated in Fig. 1. Digital quantities required for producing sound waves were calculated and recorded on a hard disk in a personal computer beforehand, and then presented through the D/A converters with 12-bit resolution. Attenuators with 0.1 db steps were used and were controlled by the personal computer. Two loudspeakers, one with a diameter of 80cm (DIATONE D-80) and the other with a diameter of 16 cm (TECHNICS 16F20) were used. The former was used for reproducing low frequency pure tones from 31.5 Hz to 160 Hz, and the latter was used for radiating middle and high frequency pure tones from 200 Hz to 20 khz. It was confirmed that the 2nd order harmonic was always less than-40 db and that the higher order harmonics were always less than-50 db relative to the fundamental. The subject sat on the frontal axis of and directly facing the 16-cm loudspeaker, and a small headrest Table 1 Experimental conditions for each experiment.

3 H. TAKESHIMA et al.: THRESHOLD OF HEARING IN FREE FIELD Fig. 1 Schematic diagram of the system used to measure the threshold of hearing. was used to fix the head position. Sound pressure levels at the ear entrance were measured with and without the headrest using a head and torso simulator, KEMER. The differences between both conditions of the headrest are shown in Fig. 2. Though there were small deviations, they were less than 1 db. The midpoint between subject's ears was defined as his/her reference point(listener's position). The loudspeakers and the reference point were always placed at a distance 3.55 m from each other. The reference point was placed at about 2 m from the wall behind and 3 m above the floor. The sound pressure levels at the reference point were calibrated under conditions with neither the subject nor the chair. To set the reference point of each subject, height and position of the chair and the headrest position were adjusted. To check if the sound field used satisfied the freesound-field listening conditions, the following measurement was carried out according to the test conditions specified by ISO/TC43/WG1.14) With neither the subject nor the chair, the sound pressure level produced by the loudspeakers at positions 0.15 m from the reference point on the left-right, the updown, and the front-back axes were measured. According to the preferred test conditions, the difference should be less than }1 db for any frequency below 4,000Hz, and less than }2 db for any frequency above 4,000Hz. On the front-back axis, Fig. 2 Effect of the headrest on the sound pressure level at the entrance of the ear canal. The ordinate shows the difference between the sound pressure level measured by KEMER with a headrest and without a headrest. the differences between the theoretical values given by the inverse distance law and the measured values were compared. The results of the measurement are shown in Fig. 3. There were a few positions that did not satisfy the conditions, but the excess was small, namely, from 0.1 db to 0.3 db. 2.2 Experimental Procedure The method of limits was used. The stimuli were

4 J. Acoust. Soc. Jpn.(E)15, 3(1994) pure tones with frequencies 1/3 octave series specified by ISO 266 between 31.5Hz and 20kHz. The level diagram of test stimuli is shown in Fig. 4. Test stimuli were presented repeatedly with a duration of 1.5 s and a pause of 0.5 s. The rise and decay time of stimuli were both 50 ms. The sound pressure level of the stimuli was changed in 1 db steps. The subject responded by a touch-switch when he/she just cannot hear the stimuli in the down staircase and when he/she just hears the stimuli in the up staircase. The initial sound pressure level of the first sequence was 20 db greater than the threshold level of ISO 226. In the next trial, the starting level was automatically set at the values shown in Fig. 4. The down staircase and the up staircase were carried out alternately for seven sequences as shown in Fig. 4. The last six sound pressure levels at the transitions(l2 to L7)in the seven trials were averaged and adopted as the threshold of hearing of a subject. Fig. 3 Distribution of sound pressure level around the reference point. The ordinate shows the difference between sound pressure levels measured at positions 0.15 m from the reference point on the left-right, up-down and front-back axis and the sound pressure levels at the reference point. 2.3 Subjects All subjects were checked to determine whether or not they satisfied the preferred conditions of ISO/ TC43/WG114) prior to the experiments; the specified monaural hearing threshold levels are less than 10 db up to 4 khz and 15 db up to 8 khz. The total number of subjects was 69, consisting of 40 males and 29 females, whose ages ranged from 19 to 25. The number of subjects, however, varied for each frequency because the frequencies measured were different from session to session.(see n in Table 2.) Fig. 4 Level diagram of test stimuli in the experiment for the threshold of hearing. The sound pressure level of stimuli decreases or increases in 1 db steps. L1, L3, L5, and L7 in the figure indicate just imperceptible sound pressure levels. L2, L4, and L6 indicate barely perceptible sound pressure levels. 2.4 Results and Discussion Table 2 shows some of the resulting statistical data. The n of the n/n column indicates the number of subjects for whom the threshold of hearing could be measured, while the N, which appears at frequencies above 10 khz, indicates the number of subjects who participated in the experiment after having passed the hearing test. Thus, two kinds of data are shown for frequencies above 10 khz. The statistics written above the dashed line of the table were calculated using N, while those written below the dashed line were calculated using n. From 10 khz to 15 khz inclusive, data of a single subject who exhibited thresholds of hearing at least 20 db higher than the present standard were omitted. Above 16 khz inclusive, the omitted subjects exhibited thresholds of hearing higher than 70 dbspl, which was the limit of the experimental apparatus. For fre-

5 H. TAKESHIMA et al.: THRESHOLD OF HEARING IN FREE FIELD Table 2 Statistical values of the threshold of hearing obtained in this study. freq:frequency[hz], mean: mean[dbspl], SD: standard deviation[db], min: minimum [dbspl], 1/4 and 3/4: quartile[dbspl], median: median[dbspl], max: maximum[dbspl], n: number of subjects measured, N: number of subjects who passed the hearing test. quencies above 18 khz, the number of subjects for whom the threshold of hearing could be measured decreased greatly. It seems, therefore, that the typical threshold of hearing for frequencies above 18 khz must be beyond the limit of the sound pressure level, 70 dbspl. In Fig. 5 shows the individual thresholds are indicated by small circles and the median values by a thick line. In the same figure, the ambient noise in the anechoic room and the standard threshold of ISO 226 are also indicated. The ambient noise level is satisfactorily lower than the thresholds of hearing at all frequencies. This figure indicates that the measurement of the thresholds of hearing was not affected by the ambient noise. Inter-subject standard deviation is shown in Fig. 6. This figure in-

6 J. Acoust. Soc. Jpn.(E)15, 3(1994) Fig. 5 Threshold of hearing for pure tones under free-field listening conditions. Small circles show measured thresholds for all subjects, the thick solid line shows their median values, and the thin solid line is the theshold given in ISO 226. The gray area indicates the ambient noise obtained by 1/3 octave band analysis in the anechoic room. Fig. 6 Standard deviation of measured thresholds of hearing. dicates that the distribution of individual thresholds of hearing is about 4 db for the frequency range between 80 Hz and 4 khz except 1.6 khz and 2 khz, and increases suddenly in the frequency ranges below 63 Hz and above 5 khz. The differences of the thresholds of hearing between ISO 226 and the present study are shown in Fig. 7. The 95 % confidence limits of our data are also shown. It can be seen that the differences of our data from ISO 226 are not so great but that there are some systematic deviations, i.e., our data Fig. 7 Comparison with the new threshold of hearing published by Betke,9) Moller et al.,11,12) Fastl et al.,10) and Teranishi.5) The ordinate shows the level relative to the threshold in ISO 226. Our data shown here are average values. are 2 `6 db higher at frequencies below 160 Hz and 1 `3 db lower at frequencies between 400 Hz and 6.3 khz, except 1.25 khz and 1.6 khz, as compared with the values specified in ISO 226. The thresholds at 1.25 khz and 1.6 khz are 3 `5 db higher than those at adjacent frequencies. The frequency characteristic between 1 khz and 2 khz in our data seems to have a small peak, which cannot be found in ISO 226. Figure 7 also shows four published data,5,9-12) which were measured after the present standard, ISO 226, was laid down. Betke9) measured the threshold of hearing at preferred frequencies of 1/3 octave series between 40 Hz and 15 khz according to ISO 266. There are some differences between Betke's thresholds and the thresholds of this study. One of the differences is that the thresholds measured by Betke9) are almost on the line of ISO 226 for frequencies below 160 Hz, while ours are 2 `6 db higher than those of ISO 226 for the same frequencies. This difference cannot be attributable to a difference in ambient noise in the anechoic rooms, because the ambient noise in our anechoic room was very low as shown in Fig. 5. Fastl et al.10) measured thresholds of hearing for frequencies between 100 Hz and 1 khz, and Moller et al.11,12) measured those between 25 Hz and 1 khz. Both set of data agree well with our results for individual frequencies. Teranishi5)measured the threshold of hearing with Japanese subjects for the frequency

7 H. TAKESHIMA et al.: THRESHOLD OF HEARING IN FREE FIELD range between 63 Hz and 10 khz. His data also agree well with the results of this study except for 6 khz. 3. EFFECT OF HEAD TRANSFER FUNCTION 3.1 Factors Affecting the Threshold of Hearing The contour of the threshold of hearing under field-listening conditions is affected by the following two factors: (1)Transfer function for a sound propagation path from a sound source to a subject's inner ear; (2)Characteristics of auditory nerve systems. Zwislocki15,16) stated that the former factor, which he called the overall acousto-mechanical transfer function of the ear, could explain, for the most part, the dependence of the threshold of hearing on frequency. The first factor is also divided into two transfer functions: One is a transfer function of a sound propagation path from a sound source to an external ear entrance, that is called a head transfer function(htf), and the other is that from the external ear to the inner ear. It is mostly impossible to measure the transfer function in the middle and inner ear of a living human, while HTF can be obtained easily by measuring sound pressure at the entrance of the ear canal. In this section, the HTF's measured to examine the reliability of our thresholds of hearing, how HTF's affect the frequency characteristics of the threshold of hearing is also examined. 3.2 Measurement of HTF The sound pressures were measured at the entrance of the ear canal and at the reference point with neither the subject or the chair by using a small-size electret-condenser microphone for hearing aids. The ratio of both sound pressures are defined as HTF. A sound reproduced from a loudspeaker to measure the sound pressures was time-stretched pulse.17) HTF's for both ears of ten subjects who participated in EXP6 were measured. 3.3 Results and Discussion Three examples of the HTF's are shown in Fig. 8 with the threshold of hearing of the corresponding subjects. In this figure, the inverse of the HTF's are shown so that they can easily be compared with the thresholds of hearing for each subject. These are superimposed on the threshold of hearing so that the average values of both data for frequencies Fig. 8 Examples of head transfer function (HTF) for comparison with thresholds of hearing. HTF's are shown inversely so that they can easily be compared with the thresholds of hearing. from 1 khz to 6.3 khz are in agreement. This figure shows that the HTF's above 1 khz are almost the same as the frequency characteristics of the

8 J. Acoust. Soc. Jpn.(E)15, 3(1994) threshold of hearing. Figure 8(a) and Fig. 8(b) show examples in which the HTF and the frequency characteristic of the threshold of hearing agree well with each other. However, a few subjects did not show a good agreement as seen from Fig. 8(c). Thus, the correlation between HTF's and threshold of hearing were examined. Figure 9 shows the results. The ordinate of the figure shows the deviations of HTF's from the average HTF's for each subject at each frequency measured. The abscissa shows the deviations of the threshold of hearing from the average thresholds of hearing for each subject. In other words, 0 db on both axes indicates the corresponding average for each subject. Figure 9(a),(b), and(c) show the correlation for frequency range from 250 Hz to 800 Hz, from 1 khz to 2.5 khz, and from 1 khz to 8 khz, respectively. The correlation coefficients are for 250 `800 Hz, 0.78 for 1 `2.5 khz, and 0.64 for 1 `8 khz. Statistical tests show that the correlation coefficients for 1 `2.5 khz and for 1 `8 khz cannot be regarded as significant beyond 0.01 point. This indicates that HTF's affect the frequency characteristics of the threshold of hearing above 1 khz. One of the features of the threshold of hearing measured in the present study is, as described in section 2.4, a small peak appearing between 1 khz and 2 khz. For HTF, a dip is often found in the same frequency range, such as shown in Fig. 8(a) and Fig. 8(c). Note that the HTF's shown in Fig. 8 are written upside down. We found clear dips in the frequency range for eight of the ten subjects. This implies that the small peak is caused due to the dip in HTF at a certain frequency. The HTF, however, could not thoroughly explain the characteristics of the threshold of hearing for frequencies above 1 khz. An impressive result was published by Sholth18,19)based on studies on otoacoustic emissions, namely, that the threshold of hearing has a several distinct maxima and minima for frequency range between only a few hundred hertz. The level of the difference between the maximum and minimum reaches 10 db or more. This result shows that the threshold of hearing in relation to frequency has a micro-structures. This micro-structure must have affected the threshold of hearing measured in this study, and the effect seems to appear as discrepancies between the thresholds of hearing and HTF. Fig. 9 Correlation diagrams of head transfer function and threshold of hearing for frequency ranges (a)200 `800 Hz,(b)1 ` 2.5 khz, and(c)1 `8 khz.

9 H. TAKESHIMA et al.: THRESHOLD OF HEARING IN FREE FIELD 4. EFFECT OF THE HEADREST 4.1 Introduction A small peak between 1 khz and 2 khz is a feature of the threshold of hearing found in this study. This small peak is not clearly seen in either ISO 226 nor in the data reported by Betke.9)The reason why the peak cannot be found in ISO 226 is simple: no experimental values were obtained for frequencies between 1 khz and 2 khz in the study of Robinson and Dason.2)Betke9)carried out his experiment at 1.25 khz and 1.6 khz, but his data do not show the peak so clearly. One possible reason for the discrepancy may be whether the subject's head was fixed by a headrest or not. We used a small headrest while Betke9)did not use any headrest. Use of a headrest might disturb the sound field around the subject's ears. From Fig. 2, we can see that the disturbance due to our headrest was very small. If no headrest is used, on the other hand, subjects would have difficulty in keeping their heads at the reference point, and thus their heads would probably move a little during an experiment. Such head movement could cause special averaging of the HTF's, resulting in an apparent threshold of hearing being different from that obtained with a headrest. In this section, therefore, the effect of the headrest as well as of an intentional head-movement on the threshold of hearing is examined. 4.2 Experiment Thresholds of hearing were measured under the following three conditions: (a)with a headrest,(b) without any headrest but with the subjects being instructed not to move their heads,(c)without a headrest and with the subjects being instructed to rotate their heads if the stimuli were hard to hear. The experimental setup and procedure are the same as described in sections 2.1 and 2.2. The thresholds were measured at frequencies of 800 Hz, 1 khz, 1.25 khz, 1.6 khz, 2 khz and 2.5 khz. The ten subjects who participated in EXP6 were also subjects in this experiment. 4.3 Results and Discussion The results are shown in Fig. 10. Each symbol shows the mean value of the results obtained for ten subjects. The means for all thresholds we measured and Betke's9)results are drawn in the same figure for a reference. It can be seen that the results Fig. 10 The average threshold of hearing under the following three conditions: (a) with a small headrest,(b)without a headrest but with the subjects being instructed not to move their heads,(c)without a headrest and with the subjects being instructed to rotate their heads if the stimuli were hard to hear. of condition(a)have a small peak at frequency 1.25 khz, while condition(b)has no peak. However, the difference between the results of condition(a) and(b)are not statistically significant. Comparison between condition(a)and condition(c)shows little difference at frequencies of 800 Hz, 1 khz, 2 khz and 2.5 khz. The differences at 1.25 khz and 1.6 khz, however, reach 5 db, and are statistically significant beyond 0.05 point at 1.25 khz, and beyond 0.01 point at 1.6 khz. This means that the thresholds of hearing for frequencies between 1 khz and 2 khz are very sensitive to movement of the head. We can assume, therefore, that the ambiguousness of the peak between 1 khz and 2 khz in the data by Betke9)might have been caused by the head movement of the subjects; the subjects could move their heads when the sound was hard to hear, so that they were able to hear the stimuli. 5. CONCLUSION The thresholds of hearing for pure tone under free-field listening conditions were measured to prepare basic data for a full-scale revision of the international standard, ISO 226. The thresholds of hearing measured in the present study do not show great differences from ISO 226, but some systematic deviations were observed, i.e., our data are 2 `6 db higher at frequencies below 160 Hz and 1 `3 db

10 J. Acoust. Soc. Jpn.(E)15, 3(1994) lower at frequencies between 400 Hz and 6.3 khz. A comparison among the data of the present study and those of other researchers shows the following tendencies:(1)the thresholds of hearing obtained by three researchers who also measured the threshold of hearing in order to revise ISO 226 are very similar to our present data except for data obtained by one of the three studies for frequencies below 160 Hz. (2) A small peak is observed at frequencies between 1 khz and 2 khz in our results, while such a peak is not seen clearly in other data. A supplementary experiment showed that the frequency characteristics of the thresholds correlate significantly with HTF in a frequency range between 1 khz and 8 khz. This result suggests that the small peak described above is attributable to a dip of HTF. Another supplementary experiment showed that head movement affects the frequency characteristics of the threshold of hearing. Possible head motion when there is no headrest may cause a characteristic different from that measured with a headrest, resulting in an ambiguousness of the small peak between 1 khz and 2 khz. ACKNOWLEDGMENTS We wish to thank Prof. Dr. Volker Mellert of the University of Oldenburg for discussing the effects of head motion, and Mr. Tamon Saeki of Mitsubishi Electric Corporation for providing the loudspeaker systems. This study was partly supported by the Sound Technology Promotion Foundation(STPF) and a Grant-in-Aid for Scientific Research(No and No )from the Ministry of Education, Science and Culture of Japan. REFERENCES 1) H. Fletcher and W. A. Munson, "Loudness, its definition, measurement and calculation," J. Acoust. Soc. Am. 5, (1933). 2) D. W. Robinson and R. S. Dadson, "A re-determination of the equal-loudness relations for pure tones," Br. J. Appl. Phys. 7, (1956). 3) D. W. Robinson and R. S. Dadson, "Threshold of hearing and equal-loudness relations for pure tones, and loudness function," J. Acoust. Soc. Am. 29, (1957). 4) E. H. Berger, "Re-examination of the low-frequency (50-1,000 Hz) normal threshold of hearing in free and diffuse sound field," J. Acoust. Soc. Am. 70, (1981). 5) R. Teranishi, "Study about measurement of loudness,-on the problems of minimum audible sound-," Res. Electrotech. Lab. No.658 (1965) (in Japanese). 6) T. Sone and H. Tachibana, "Report on the 11th plenary meetings of ISO/TC43/SC1 and ISO/TC43," J. Acoust. Soc. Jpn. (J) 41, (1985) (in Japanese). 7) S. Suzuki, Y. Suzuki, S. Kono, T. Sone, M. Kumagai, H. Miura, and H. Kado, "Equal-loudness level contours for pure tone under free field listening conditions (I)-Some data and considerations on experimental conditions-," J. Acoust. Soc. Jpn. (E) 10, (1989). 8) K. Betke and V. Mellert, "New measurements of equal-loudness level contours," Proc. Inter-noise 89, (1989). 9) K. Betke, "New hearing threshold measurements for pure tones under free-field listening conditions," J. Acoust. Soc. Am. 89, (1991). 10) H. Fastl, A. Jaroszewski, E. Schorer, and E. Zwicker, "Equal-loudness contours between 100 and 1000 Hz for 30, 50, 70 phon," Acustica 70, (1990). 11) H. Moller and J. Andressen, "Loudness of pure tones at low and infrasonic frequencies," J. Low Freq. Noise Vib. 3, (1984). 12) T. Watanabe and H. Moller, "The loudness level in the free field, and the hearing threshold in the free field and in the pressure field of pure tones at frequencies below 1 khz," Proc. Spring Meet. Acoust. Soc. Jpn., (1990). 13) T. Fujimori, "Measurement of low level sound noise by the cross-spectrum method," Tech. Rep. IEICE EA (in Japanese). 14) "3rd Draft, Preferred test conditions for the determination of the minimum audible field and the normal equal-loudness level contours," ISO/TC43/ WG1 N122 (November 1988). 15) J. J. Zwislocki, "The role of the external and middle ear in sound transmission," in The Nervous System, Vol.3: Human Communication and its Disorders, E. L. Eagles, Eds. (Raven Press, New York, 1975). 16) J. J. Zwislocki, "The acoustics and mechanics of the ear," Proc. 13th Int. Congr. Acoust., (1989). 17) Y. Suzuki, F. Asano, H. Kim, and T. Sone, "Considerations on the design of time-stretched pulses," Tech. Rep. IECE EA92-86 (1992) (in Japanese). 18) E. Schloth, "Relation between spectral compositions of spontaneous otoacoustic emissions and finestructure of threshold in quiet," Acustica 53, (1983). 19) E. Zwicker and H. Fastl, Psychoacoustics Facts and Models- (Springer-Verlag, Berlin/Heidelberg, 1990), pp

11 H. TAKESHIMA et al.: THRESHOLD OF HEARING IN FREE FIELD Hisashi Takeshima was born on 6 August 1963 in Japan. He received the B. Eng. degree in electronic engineering from Tohoku University, Sendai, Japan in He received the M. Eng. degrees in electrical and communication engineering from Tohoku University in He is now a research associate of Sendai National College of Technology, Sendal, Japan. His current research interests include equal-loudness level contours, threshold of hearing. Yoiti Suzuki was born on 11 January 1954 in Japan. He received the B. Eng. degree in electrical engineering from Tohoku University in He received the Dr. Eng. degrees in electrical and communication engineering from Tohoku University He is currently an associate professor at the Research Institute of Electrical Communication, Tohoku University. His research interest include loudness, sound localization, timbre, hearing aids and digital signal processing of a sound. He received the Awaya Kiyoshi Award and the Sato prise form the Acoustical Society of Japan in 1986 and 1992, respectively. Masazumi Kumagai was born on 19 April 1945 in Japan. He received the B. Eng. degree from University of Electro-Communications, Tokyo in 1968, and Dr. Eng. degree from Tohoku University, Sendai, Japan in He is now a professor at Dept. of Electronic Eng., Sendai National College of Technology. His research interests include equal-loudness level contours, loudness of impact sound, and evaluation of environmental noise. He is a member of ASJ, INCE/J, and JSME. Toshio Sone was born on 14 May A graduate in electrical engineering at Tohoku University, Japan in Sone did his graduate work at the same university, where he was awarded his Ph. D. in electrical and communication engineering in He is now a professor of the Research Institute of Electrical Communication, Tohoku University. Takeshi Fujimori graduated form Department of Physics, Osaka University. Since 1967, he has been with Electrotechnical Laboratory, and in currently a Senior Researcher for Acoustics Section, Metrology Fundamentals Division. Hajime Miura is a professor of Depertment of Computer Science at Shizuoka Institute of Science and Technology. He received B. S. degree from the University of Electro-Communications in 1961 and joined Electrotechnical Laboratory, and retired He recieved Dr. Eng. from Tohoku University in 1987.