EFFECTS OF FREQUENCY CHARACTERISTICS OF REVERBERATION

TIME ON LISTENER ENVELOPMENT

PACS No 43.55.Hy

Morimoto, Masayuki; Jinya, Munehiro; Nakagawa, Koichi; Sakagami, Kimihiro

Environmental Acoustics Laboratory, Faculty of Engineering, Kobe University
Rokko, Nada
657-8501 Kobe
Japan
Tel: +81-78-881-6035
Fax: +81-78-881-2508
E-mail: [email protected]

ABSTRACT

It is well-known that reverberation contributes to listener envelopment (LEV). In this paper, three
listening tests were performed to make clear effects of the reverberation time (RT) and its
frequency characteristics on LEV. The result of the first test with RT as a parameter shows that it
affects LEV significantly. The results of the second and third tests with frequency characteristics of
RT as a parameter of show that RT at low frequencies affects LEV as well as RT at high
frequencies.

INTRODUCTION

Listener envelopment (LEV) is one of spatial impressions which is the most important in concert
halls. Many pieces of research on LEV have reported that reverberation (late reflections)
contributes significantly to the perception of LEV[1-7]. The parameters relating to reverberation
which contribute to LEV can be divided into the spatial structure and the temporal structure of
reverberation.

With the spatial structure, the degree of interaural cross-correlation[1], lateral energy fraction[3,4]
front/back energy ratio of reverberation[2,5] and SBTs[6] are proposed as the physical measure
for LEV. Meanwhile, with the temporal structure, the effects of C80[3-5] and reverberation time
(RT)[4,6] on LEV are demonstrated.

However, the effect of frequency characteristics of RT on LEV is not clarified. Considering that the
frequency components of a direct sound and reflections affect apparent source width (ASW) which
is the other spatial impression[9-11], it is estimated that the frequency of characteristics of RT
affects LEV, too. In this paper, in the first experiment the effect of RT on LEV is investigated and in
the second and the third experiments the effects of frequency characteristics of RT on LEV is
clarified.

METHOD
In the experiments, 7s section of the first movement of Mozart's "Divertiment" recorded in an
anechoic chamber was used as a music motif. The motif was reproduced with a limited frequency
range from 100Hz to 10kHz.

Figure 1 shows the impulse response of a stimulus. The sound field used as the stimulus
consisted of a direct sound, two early discrete reflections and five reverberant signals. Figure 2
shows the arrangement of loudspeakers. Six loudspeakers, each of which is installed in a
cylindrical enclosure (diameter: 108mm, length: 350mm), were arranged at azimuth angles of 0° ±
45° from the median plane, that is, they were arranged symmetrically to the aural axis, in an
anechoic chamber. The distance between the center of subject's head and the loudspeakers was
1.5m.

Paired comparison tests were performed in the experiments. Each subject was tested individually
and 10 times for each pair, while seated, with head fixed. The task of the subject was to judge
which LEV is greater. The psychological scales of LEV were obtained using Thurstone Case V
model[12] and Gulliksen method[13].

The following must be considered in interpreting the psychological scales obtained using
Thurstone Case V model: The difference of 0.68 on the psychological scale means that the
probability of discrimination of difference between two stimuli is 75%. Therefore, it is generally
considered that the difference of 0.68 on the psychological scale corresponds to the just
noticeable difference (jnd).

EXPERIMENT 1: EFFECTS OF REVERBERATION TIME ON LEV

Experimental Conditions
Table 1 shows the stimuli used in this experiment. RT changes as well as C80, keeping ÄL of all
stimuli constant at -26.3dB. The frequency characteristics of RT were flat. C80 was controlled by
changing the density of reverberant signals. The binaural sound pressure levels[14] were constant
at 80.0 ± 0.1dBA slow, peak, measured at two ears of KEMAR dummy head without an artificial ear
simulator. The degree of interaural cross-correlation (DICC)[15] of the reverberant signals were
constant at 0.43 ± 0.03 measured by KEMAR dummy head without an artificial ear simulator.
Seven students with normal hearing sensitivity acted as subjects for the experiment.

0•‹
D

-45•‹ +45•‹
ER2+Rev3 ER1+Rev2

-6dB ƒ ¢LdB
Direct

ER1

ER2
REV1

-6dB
111 REV4
115 REV5
195 REV2
REV3

-135•‹ +135 •‹
Rev4
140

140

170
180

103

t [ms] Rev5
180•‹
Rev1

Figure 1. Schematic diagram of impulse Figure 2. Arrangement of loudspeakers in

response of the stimulus used in the the experiments.
experiments.
 Table 1. Kinds of stimulus in the experiment 1.

Parameter
Stimulus
RT [s] C80 [dB] ÄL [dB]

1 1.0 6.0 -26.3

2 1.0 2.8 -26.3
3 1.0 0.9 -26.3
4 2.0 5.5 -26.3
5 2.0 2.7 -26.3
6 2.0 0.7 -26.3

2
RT
=2
.0s
1
LEV

0 RT
=1
.0s
-1

-2
0 1 2 3 4 5 6 7
C80 [dB]

Figure 3. Psychological scale of LEV vs. C80 as a parameter of RT in the experiment 1.

Experimental Results and Discussion

Figure 3 shows the psychological scale of LEV in the experiment 1, that is, LEV with a parameter
of RT and as a function of C80. For each C80 value, LEV for RT=2.0s is greater than that for
RT=1.0s. The difference between them exceeds 0.68 for each C80 value. This means that RT
significantly affects LEV which the listener perceives. Namely, LEV increases as RT becomes
longer. Meanwhile, for each RT value, LEV increases as C80 decreases. The difference between
the maximum and the minimum LEV exceeds 0.68 for each RT. This means that C80 significantly
affects LEV, too.

EXPERIMENT 2: EFFECTS OF REVERBERATION TIME AT LOW FREQUENCIES ON LEV

Experimental Conditions
In the experiment 2, the effects of RT at low frequencies on LEV were investigated for RT=1.0s
and 2.0s with flat frequency characteristics. Table 2 shows the frequency characteristics of RT of
stimuli used in the experiment. The stimulus of No.3 for each RT value has flat frequency
characteristics of RT. C80 of all stimuli were constant at 0dB. ÄL for the stimulus of No3 was
constant at -25.6dB and -27.6dB for RT=1.0s and 2.0s, respectively. The binaural sound pressure
levels of all stimuli were constant at 79.9 ± 0.1dBA and 80.0dBA for RT=1.0s and 2.0s, respectively.
DICC of the reverberant signals of all stimuli were constant at 0.32 ± 0.03. A paired comparison
test was carried out separately for each RT. Six students with normal hearing sensitivity acted as
subjects for the experiment.
 Table 2. Frequency characteristics of RT of stimuli in the experiment 2.

Center frequency of 1/3 oct. band [Hz]

Stimulus
125 250 500 1k 2k 4k 8k
Experiment 2(a) (RT=1.0s)
1 0.50 0.63 0.75 0.85 0.93 1.00 1.00
2 0.66 0.77 0.77 0.91 0.94 1.00 1.01
3 1.13 1.00 1.04 1.02 0.96 1.00 1.01
4 1.82 1.28 1.19 1.06 1.01 1.06 0.99
5 2.11 1.58 1.30 1.10 1.01 1.02 1.00
6 2.34 1.82 1.41 1.10 1.03 1.02 0.99
Experiment 2(b) (RT=2.0s)
1 0.49 0.92 1.21 1.61 1.81 1.91 1.93
2 1.16 1.18 1.46 1.87 1.90 1.91 1.93
3 2.20 1.93 1.91 2.03 1.96 1.92 1.94
4 2.92 2.67 2.18 2.12 1.99 1.93 1.94
5 3.33 3.22 2.45 2.17 1.98 1.91 1.94
6 3.82 3.33 2.63 2.20 2.03 1.94 1.94
(unit : sec.)

1
LEV

-1

(a) RT=1.0s (b) RT=2.0s

-2
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Stimulus Stimulus

Figure 4. Psychological scal of LEV vs frequency characteristics of RT for

(a) RT=1.0s and (b) 2.0s in the experiment 2.

Experimental Results and Discussion

Figure 4 shows the psychological scales of LEV in the experiment 2, that is, LEV vs. RT at low
frequencies for each RT with flat frequency characteristics. For each RT value, LEV increases as
RT at low frequencies becomes longer and vice versa, comparing with LEV for the stimulus (No.3)
with flat frequency characteristics of RT. Furthermore, the difference between the maximum and
minimum LEV exceeds 0.68 for each RT. This means that RT at low frequencies affects
significantly LEV.

EXPERIMENT 3: EFFECTS OF REVERBERATION TIME OF LOW AND HIGH FREQUENCIES

ON LEV

Experimental Conditions
Table 3 shows the frequency characteristics of RT of stimuli used in the experiment. The stimulus
of No. 3 has a flat frequency characteristics of RT. C80 of all stimuli were constant at 0dB. ÄL for
 Table 3. Frequency characteristics of RT of stimuli in the experiment 3.

Center frequency of 1/3 oct. band [Hz]

Stimulus
125 250 500 1k 2k 4k 8k
1 1.05 1.54 1.68 1.89 1.97 2.01 2.00
2 1.44 1.57 1.73 1.91 1.97 2.01 2.01
3 2.00 1.97 2.02 1.99 1.98 1.99 1.98
4 2.03 2.05 2.03 1.99 1.82 1.63 1.34
5 2.04 2.04 2.01 1.97 1.84 1.57 1.23
(unit : sec.)

1.0

0.5
LEV

0.0

-0.5
0 1 2 3 4 5 6
Stimulus

Figure 5. Psychological scal of LEV vs frequency characteristics of RT in the experiment 3.

the stimulus of No.3 was constant at -26.3dB. Th e binaural sound pressure levels were constant at
79.9 ± 0.1dBA. DICC of the reverberant signals of all stimuli were constant at 0.43 ± 0.03. A paired
comparison test was carried out separately for each RT. Five students with normal hearing
sensitivity acted as subjects for the experiment.

Experimental Results and Discussion

Figure 5 shows the psychological scales of LEV in the experiment 3, that is, LEV vs. RT at low and
high frequencies. LEV decreases as RT at low frequencies becomes shorter as well as high
frequencies, comparing with LEV for the stimulus (No.3) with flat frequency characteristics of RT.
The decrease exceeds 0.68. This means that RT at not only low frequencies but also high
frequencies affect significantly LEV. This phenomenon is different from the perception of ASW
which is affected by low frequency components but high frequency components.

CONCLUSIONS

The results of three subjective experiments on LEV clarify:

(1) The frequency characteristics of reverberation time affect LEV significantly as well as C80.
LEV increases as RT becomes longer and C80 decreases.
(2) Reverberation times at not only low frequencies but also high frequencies affect LEV
significantly. LEV decreases as they decrease.
REFERENCES

[1] Morimoto M, Maekawa Z. Auditory spaciousness and envelopment. Proc 13th ICA (Belgrade)
1989; 2: 215-218.
[2] Morimoto M, Iida K. A new physical measure for psychological evaluation of a sound field;
front/back energy ratio as a measure of envelopment. J. Acoust. Soc. Am 1993; 93:2282.
[3] Bradley JS, Soulodre GA. The influence of late arriving energy on spatial impression. J Acoust
Soc Am 1995; 97: 2263-2271.
[4] Bradley JS, Soulodre GA. Objective measures of listener envelopment. J Acoust Soc Am 1995;
97: 2590-2597.
[5] Morimoto M, Iida K, Sakagami K. The role of reflections from behind the listener in spatial
impression. Applied Acoustics 2001; 62: 109-124.
[6] Hanyu T, Kimura S. A new objective measure for evaluation of listener envelopment focusing
on the spatial balance of reflections. Applied Acoustics 2001; 62: 155-184.
[7] Bradley JS, Reich RD, Norcross SG. On the combined effects of early- and late-arriving sound
on spatial impression in concert halls. J Acoust Soc Am 2000; 108: 651-661.
[8] Barron M, Marshall AH. Spatial impression due to early lateral reflections in concert halls: The
deviation of a physical measure. Journal of Sound and Vibration 1981; 77: 211-232.
[9] Blauert J, Lindemann W. Auditory spaciousness: Some further psychoacoustic analysis. J
Acoust Soc Am 1986; 80: 533-542.
[10] Blauert J, Mobius U, Lindemann W. Supplementary psychoacoustical results on auditory
spaciousness. Acustica 1986; 59: 292-293.
[11] Morimoto M, Setoyama H, Iida K. Consistent physical measures of auditory source width for
various frequency components of reflections. Proc. of 3rd Joint meeting of ASJ and ASA
1996: 83-86.
[12] Thurstone LL. A law of comparative judgment. Psychological Review 1927; 34: 273-286.
[13] Gulliksen H. A least squares solution for paired comparison with incomplete data.
Psychometrika 1956; 21: 125-134.
[14] Robinson DW, Whittle LS. The loudness of directional sound field. Acustica 1960; 10: 74-80.
[15] Morimoto M, Iida K. A practical evaluation method of auditory source width in concert halls. J
Acoust Soc Jpn (E) 1995; 16: 59-69.