JPH05265477

Patent Translate
Powered by EPO and Google
Notice
This translation is machine-generated. It cannot be guaranteed that it is intelligible, accurate,
complete, reliable or fit for specific purposes. Critical decisions, such as commercially relevant or
financial decisions, should not be based on machine-translation output.
DESCRIPTION JPH05265477
[0001]
BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a
sound field correction apparatus which artificially creates a good acoustic space by adding
reflected sound and reverberation sound to an original signal sound.
[0002]
2. Description of the Related Art When the reverberation characteristic in an acoustic space such
as a concert hall or the like is indicated by an impulse response, for example, the characteristic is
as shown in FIG. Although this impulse response changes with the environment of acoustic space,
it can be divided into an early reflection part and a back reverberation part from the level and
density. The early reflection sound is considered to be a factor that determines a spatial image
such as a spreading feeling in a room such as a concert hall, and becomes high in level. The rear
reverberation sound following the initial reflection sound increases in density in proportion to
the square of time, and the high frequency component tends to decrease because the reflection
on the wall surface in the room is repeated.
[0003]
On the other hand, there is a sound field correction device for artificially creating an acoustic
space such as a concert hall by adding a reflected sound or a reverberation sound to an original
08-05-2019
1
signal sound in a home or a vehicle interior. In the conventional sound field correction apparatus,
as shown in FIG. 2, a delay unit group 1 in which a plurality of delay units are connected in series
is provided, and an input digital audio signal is supplied to the first delay unit. The audio signal
delayed by the delay group 1 is supplied to a coefficient multiplier group 2. The coefficient
multiplier group 2 multiplies the delayed audio signal by the coefficient in each coefficient
multiplier and supplies the multiplication result signal to the adder 3. The signal resulting from
the addition in the adder 3 is further supplied to the adder 5 via the coefficient multiplier 4 as a
reflected sound signal. The input digital audio signal which is the original signal is supplied to the
adder 5 through the coefficient multiplier 6, and the reflected sound signal is added to the
original signal. This is the output signal.
[0004]
The impulse response characteristic when such a conventional sound field correction apparatus
is used is as shown in FIG. When this characteristic is compared with the impulse response in the
actual acoustic space shown in FIG. 1, both of the initial reflection sound portion have
substantially the same characteristics. However, in the rear reverberation part, the density is low
when the sound field correction device is used. This is caused by the limited number of stages of
delay units and coefficient multipliers. Since the density is low, the residual sound of the auditory
sense does not decrease smoothly, and a so-called spread feeling as felt in the concert hall can
not be obtained, and there is a problem that the sense of incongruity is felt.
[0005]
On the other hand, it is conceivable to use, for example, a cyclic filter represented by a comb
filter in order to increase the density of the rear reverberation, but in this case, the signal is
degraded because a very large number of calculations are repeated. As a result, the sound quality
is adversely affected. In addition, when a cyclic filter is used, there is a disadvantage that a limit
cycle in which the output signal does not become 0 even when the input signal becomes 0
occurs.
[0006]
SUMMARY OF THE INVENTION An object of the present invention is to provide a sound field
correction apparatus which can obtain a sufficiently wide sense of expansion.
08-05-2019
2
[0007]
The sound field correction device according to the present invention comprises a plurality of
cascade-connected delay units, a plurality of coefficient multipliers connected to each of the
outputs of the plurality of delay units, and a plurality of coefficient multipliers. At least two of the
reflected sound generation circuits comprising adders for adding and outputting output signals,
connection means for connecting the reflected sound generation circuits in cascade connection,
and addition means for adding the output signals of the respective reflected sound generation
circuits It is characterized by
[0008]
In the sound field correction apparatus according to the present invention, the plurality of
reflected sound generation circuits are cascaded to generate a plurality of initial reflected sound
signals for the original signal, and the plurality of initial reflected sound signals are further
generated. A large number of reflected sound signals are generated and it is made to obtain them
as reverberant sound signals.
[0009]
Embodiments of the present invention will be described in detail below with reference to the
drawings.
FIG. 4 shows a sound field correction device according to the invention, in which a delay group
11 is provided to which an input digital audio signal is supplied via an input terminal IN.
The delay group 11 is formed by connecting a plurality of delay elements 111 to 11a (a is an
integer of 2 or more) in series.
A coefficient multiplier group 21 consisting of coefficient multipliers 211 to 21a corresponding
to the number of delay elements in the delay group 11 is connected to each delay output of the
delay group 11. An adder 31 is connected to the outputs of the coefficient multipliers 211 to
21a. The adder 31 adds the output signals of the coefficient multipliers 211 to 21a. A circuit
consisting of the delay unit group 11, the coefficient multiplier group 21 and the adder 31 is
called a reflected sound generation circuit A1. A reflected sound generation circuit A2 is
connected to the output of the adder 31, which is the output of the reflected sound generation
circuit A1, via the filter B1. The filter B1 is a band limiting filter such as a band pass filter or a
08-05-2019
3
low pass filter, and is composed of, for example, an IIR (Infinite Impulse Response) type digital
filter. It is preferable that the IIR filter has a large number of bits. A filter and a reflected sound
generation circuit are repeatedly connected to the output of the reflected sound generation
circuit A2 to the filter Bm-1 (m is an integer of 2 or more) and the reflected sound generation
circuit Am. Similarly to the reflected sound generation circuit A1, the reflected sound generation
circuits A2 to Am also include a delay group, a coefficient multiplier group, and an adder. The
number of stages of delay units and coefficient multipliers of each of the reflected sound
generation circuits A1 to Am is not a fixed number, and the delay time of each delay unit and the
coefficient of the coefficient multipliers are not constant values. Each delay time is a time
sufficiently longer than the sampling interval of the input digital audio signal.
[0010]
A coefficient multiplier 400 to 40m is connected to the output of the input terminal IN and each
of the reflected sound generation circuits A1 to Am, and the output signals of the coefficient
multipliers 400 to 40m are supplied to the adder 41. The output of the adder 41 is connected to
the output terminal OUT. In such a configuration, the reflected sound signal generated in the
reflected sound generation circuit A1 for the input digital audio signal is further supplied to the
reflected sound generation circuit A2, and a reflected sound signal is generated for the reflected
sound signal. Therefore, in the reflected sound generation circuits A2 to Am, the generation of
the reflected sound signal with respect to the reflected sound signal is repeated. When repetition
is performed in this manner, it includes elements as reverberation signals that the density of the
reflected sound components in the signal is high.
[0011]
The filters B1 to Bm-1 attenuate the high frequency components in the output signal of the
reflected sound generation circuits A1 to Am-1 in consideration of the decrease of the high
frequency components every time the sound is reflected in the actual sound field. Provided for
Assuming that all the filters B1 to Bm-1 are low pass filters, the relationship at the cut-off
frequencies f1 to fm-1 is f1> f2>...> Fm-1.
[0012]
FIG. 5 shows a configuration having three reflected sound generation circuits A1 to A3 in order
08-05-2019
4
to explain the operation of the sound field correction apparatus according to the present
invention shown in FIG. The filters B1 and B2 are omitted. Here, the reflected sound generation
circuit A1 includes three delay units 111 to 113, coefficient multipliers 211 to 213, and one
adder 31, and the reflected sound generation circuit A2 includes two delay units 121 and 122
and a coefficient multiplier. The reflected sound generation circuit A3 comprises two delay units
131 and 132, coefficient multipliers 231 and 232, and one adder 33. The delay times of the
delay units 111 to 113 are τ11 to τ13, the delay times of the delay units 121 and 122 are
τ21 and τ22, the delay times of the delay units 131 and 132 are τ31 and τ32, and the
coefficient multipliers 211 to 213 The coefficients of each of the coefficients are g11 to g13, the
coefficients of the coefficient multipliers 221 and 222 are g21 and g22, the coefficients of the
coefficient multipliers 231 and 232 are g31 and g32, and the coefficients of the coefficient
multipliers 400 to 403 are G0 to G3. Do. These delay times and coefficients are set, for example,
to the following values. τ11 = 3 msec, τ12 = 4 msec, τ13 = 4.5 msec, τ21 = 12 msec, τ22 =
2 msec, τ31 = 5 msec, τ32 = 0.5 msec, g11 = 0.8, g12 = 0.7, g13 = 0.5, g21 = 0.8, g22 = 0.5,
g31 = 0.8, g32 = 0.5, G0 = 1.0, G1 = 1.0, G2 = 0.5, G3 = 0.3 FIG. As shown, if the impulse P0 is
supplied to the input terminal IN, the impulse P0 is immediately output from the output terminal
OUT through the coefficient multiplier 400 and the adder 41. The impulse P0 is delayed by time
.tau.11 in the delay unit 111 and then multiplied by the coefficient g11 in the coefficient
multiplier 211. FIG. P0 .times.g11 passes through the adder 31 and is further multiplied by the
coefficient G1 in the coefficient multiplier 401 and then output through the adder 41. As a result,
an impulse response X (= P0.times.g11.times.G1) is obtained after a time .tau.11 has elapsed
from the input time of the impulse P0. The impulse P 0 delayed by the delay unit 111 is delayed
by time τ 12 in the delay unit 112 and then multiplied by the coefficient g 12 in the coefficient
multiplier 212. P0 × g12 is further multiplied by the coefficient G1 in the coefficient multiplier
401 via the adder 31, and then output through the adder 41.
As a result, an impulse response Y (= P0.times.g12.times.G1) is obtained after a time .tau.11 +
.tau.12 has elapsed since the input of the impulse P0. Similarly, an impulse response Z (=
P0.times.g13.times.G1) is obtained after a time .tau.11 + .tau.12 + .tau.13 has elapsed from the
input of the impulse P0. This impulse response X, Y, Z becomes an initial reflection sound.
[0013]
The initial reflected sound component of the impulse P0 output from the adder 31 is delayed by
time .tau.21 in the delay unit 121, and then multiplied by the coefficient g21 in the coefficient
multiplier 221, and then through the adder 32 the coefficient multiplier 402 It is multiplied by
G2. The component delayed by the delay unit 121 is delayed by time .tau.22 in the delay unit
122, then multiplied by the coefficient g22 in the coefficient multiplier 222, and further
08-05-2019
5
multiplied by the coefficient G2 in the coefficient multiplier 402 through the adder 32. Ru.
Therefore, after time .tau.21 and .tau.21 + .tau.22 have elapsed since the generation of the
impulse response X, impulse responses X1 (= P0.times.g11.times.g21.times.G2) and X2 (=
P0.times.g11.times.g22.times.G2) are generated. After time .tau.21 and .tau.21 + .tau.22 have
elapsed from the generation of the impulse response Y, impulse responses Y1 (=
P0.times.g12.times.g21.times.G2) and Y2 (= P0.times.g12.times.g22.times.G2) are generated.
After time .tau.21 and .tau.21 + .tau.22 have elapsed from the generation of the impulse response
Z, impulse responses Z1 (= P0.times.g13.times.g21.times.G2) and Z2 (=
P0.times.g13.times.g22.times.G2) are generated.
[0014]
The component output from the adder 32 is delayed by time .tau.31 in the delay unit 131,
multiplied by the coefficient g31 in the coefficient multiplier 231, and further multiplied by the
coefficient G3 in the coefficient multiplier 403 through the adder 33. The component delayed by
the delay unit 131 is delayed by time .tau. 32 in the delay unit 132, multiplied by the coefficient
g32 in the coefficient multiplier 232, and further multiplied by the coefficient G3 in the
coefficient multiplier 403 through the adder 33. Ru.
[0015]
Therefore, after time .tau.31 and .tau.31 + .tau.32 have elapsed from the generation of the
impulse response X1, impulse responses X11 (= P0.times.g11.times.g21.times.g31.times.G3) and
X12 (= P0.times.g11.times.g21.times.g32.times.G3) are generated. After time .tau.31 and .tau.31
+ .tau.32 have elapsed from the generation of the impulse response X2, impulse responses X21
(= P0.times.g11.times.g22.times.g31.times.G3) and X22 (=
P0.times.g11.times.g22.times.g32.times.G3) are generated. Similarly, impulse responses Y11,
Y12, Y21, Y22, Z11, Z12, Z21, Z22 are generated. As a result, as shown in FIG. 6, 21 impulse
responses are generated for the impulse P0. Among them, the impulse responses X1, Y1, Z1,
X11, X12, X21, X22, Y22, Y12, Y21, Y22, Z11, Z12, Z21, Z22 become rear reverberation, and
the density becomes higher than that of the early reflection part. .
[0016]
The reflected sound generation circuit may be at least two stages including an initial reflected
08-05-2019
6
sound generation stage and a rear reverberation generation stage. As an actual design example in
the case of two stages, the number of delay units and coefficient multipliers in the reflected
sound generation circuit of the initial reflected sound generation stage is, for example, 23 stages.
The number of stages of the coefficient multiplier can be, for example, 21. The delay time of the
delay unit of the initial reflection sound generation stage is set to several msec to 10 msec, and
the total delay time is set to about 150 msec. The delay time of the delay unit of the rear
reverberation generation stage is also set to several milliseconds to 10 milliseconds, and the total
delay time is set to about 500 milliseconds. Also, the filter between the initial reflection sound
generation stage and the rear reverberation generation stage may be a band pass filter of 100 Hz
to 8 KHz.
[0017]
Further, in the above τ, g and G, the impulse response characteristic envelope decreases
exponentially as a whole, and the impulse occurrence probability is 2 times more so that the level
and time interval of each impulse response become random to some extent. It may be set to
increase in proportion to the power. It is also necessary to set the delay time τ so that the
responses do not overlap as much as possible.
[0018]
Further, coefficient multipliers 400 to 40m included in the addition means together with the
adder 41 are for setting the levels of the original sound, the initial reflection sound and the rear
reverberation sound. That is, since the levels can be set relatively by the coefficient multipliers
400 to 40m without changing the coefficients of the coefficient multipliers of the coefficient
multiplier groups 21 to 2m, when changing the characteristics according to the environment It is
convenient for
[0019]
The configuration in the above-described embodiment may be achieved using a DSP (Digital
Signal Processor). Furthermore, in the above embodiment, the filters B1 to Bm-1 are provided on
the input side of the reflected sound generation circuits A2 to Am, but they may be provided on
the output side of the reflected sound generation circuits A2 to Am. .
08-05-2019
7
[0020]
According to the sound field correction apparatus of the present invention, at least two reflection
sound generation circuits are connected in cascade, and the output signals of the reflection
sound generation circuits are added. Thereby, a plurality of initial reflection sound signals for the
original signal are first generated, more reflection sound signals are generated for the plurality of
initial reflection sound signals, and the many reflection sound signals are obtained as the rear
reverberation portion. The thing is done. As a result, the density of the rear reverberation portion
is increased, so that the natural hearing feeling in which the residual sound of the sound
smoothly decreases is obtained, and the feeling of expansion can be obtained more sufficiently
than in the past. Also, there is almost no signal degradation that occurs when using a recursive
filter, and there is no occurrence of limit cycles.
08-05-2019
8