Note: Descriptions are shown in the official language in which they were submitted.
CA 02241000 1998-06-19
DESCRIPTION
AUDIO SIGNAL WAVEFORM EMPHASIS PROCESSING DEVICE AND
METHOD
TECHNICAL FIELD
The present invention relates to an audio signal
waveform emphasis processing device and method wherein the
sound quality of audio devices of various types is improved
by emphasis processing of the audio signal waveform, and in
particular relates to an audio signal waveform emphasis
processing device and method wherein sound quality is
enormously improved in the low frequency range and in the
high frequency range by emphasizing the low frequency range
and the high frequency range without destroying the
characteristics of the audio signal waveform.
BACKGROUND ART
Typically, techniques for improving the sound quality of
audio signals output from audio devices concentrate
exclusively on the frequency-gain characteristic; scarcely
any take into consideration the frequency-phase
characteristic of the audio signal. This is because the
phase characteristic of an audio signal is not considered to
be an important element in human auditory perception.
In regard to evaluation on the sound quality of audio
devices, conventionally, so long as basic performance such as
frequency-gain characteristic, waveform distortion, and S-N
1
CA 02241000 1998-06-19
ratio of the audio signal satisfied certain standards, there
was no further evaluation of performance. Therefore,
evaluation depending on the subjective perceptions of
individuals was often relied on.
For this reason, conventionally, improvements in the
sound quality of audio devices were only made in terms of
improvements in the frequency-gain characteristic of the
amplifiers and filters; scarcely any improvements in sound
quality were made taking into consideration the frequency-
phase characteristic.
Thus, with conventional techniques, if improvement in
the frequency-gain characteristic of the amplifiers and
filters constituting the audio device was sought, there was a
concomitant change in the frequency-phase characteristic,
giving rise to the problem that a high level of benefit in
terms of improved sound quality satisfying the user was not
obtained in particular in the high frequency range and low
frequency range close to the middle frequency range .
DISCLOSURE OF THE INVENTION
Accordingly, an object of the present invention is to
provide an audio signal waveform emphasis processing device
and method whereby a high level of sound quality improvement
is made possible by enabling the desired frequency-gain
characteristic to be obtained without destroying the phase
relationship of the frequency components constituting the
audio signal, in particular the phase relationship between
the frequency components of the middle frequency range and
2
CA 02241000 1998-06-19
the frequency components of the low frequency range and high
frequency range close to the middle frequency range.
In order to achieve this object, an audio signal
waveform emphasis processing device according to the present
invention comprises integration means comprising one or a
plurality of even-order integrators for performing
integration of even-number order on an input audio signal;
differentiation means comprising one or a plurality of even-
order differentiators for performing differentiation of even-
number order on the input audio signal; and addition means
for adding output of the integration means and output of the
differentiation means to the input audio signal in the same
phase or in reverse-phase with the input audio signal.
The integration means may comprise a plurality of double
integrators for performing successive double integration on
the input audio signal; the differentiation means comprises a
plurality of double differentiators for performing successive
double differentiation on the input audio signal; and the
addition means comprises a plurality of first coefficient
means for respectively multiplying outputs of a plurality of
the double integrators by a first coefficient; a first adder
for adding outputs of a plurality of the first coefficient
means; a plurality of second coefficient means for
respectively multiplying outputs of a plurality of the double
differentiators by a second coefficient; a second adder for
adding outputs of a plurality of the second coefficient
means; third coefficient means for multiplying output of the
first adder by a third coefficient; fourth coefficient means
3
CA 02241000 1998-06-19
for multiplying output of the second adder by a fourth
coefficient; and a third adder for adding output of the third
coefficient means and output of the fourth coefficient means
to the input audio signal.
In the audio signal waveform emphasis processing device
according to the present invention, a first group of
coefficients including the first coefficient with which
multiplication is effected by the first coefficient means and
a second group of coefficients including the second
coefficient with which multiplication is effected by the
second coefficient means may be determined in correspondence
with a desired frequency-gain characteristic.
The third coefficient with which multiplication is
effected by the third coefficient means may be determined in
accordance with the degree of emphasis in a low frequency
range of the input audio signal, and the fourth coefficient
with which multiplication is effected by the fourth
coefficient means is determined in accordance with the degree
of emphasis in the high frequency range of the input audio
signal.
The integration means may comprise a single double
integrator; the differentiation means comprises a single
double differentiator, and the addition means comprises an
adder for inverting output of the single double integrator
and output of the single double differentiator, respectively
multiplying these outputs by arbitrary coefficients and
adding the inverted outputs to the input audio signal.
4
CA 02241000 1998-06-19
Also, an audio signal waveform emphasis processing
device may comprise a plurality of cascade-connected double
integrating circuits for performing successive double
integration on an input audio signal; a plurality of cascade-
s connected double differentiating circuits that perform
successive double differentiation on the input audio signal;
a plurality of first coefficient generators for multiplying a
first coefficient respectively with outputs of even-number
double integrating circuits of the plurality of double
integrating circuits; a first addition circuit for inverting
and adding outputs of the plurality of the first coefficient
generators; a plurality of second coefficient generators for
respectively multiplying a second coefficient with outputs of
odd-number double integration circuits of the plurality of
the double integration circuits; a second addition circuit
for inverting and adding outputs of the plurality of the
second coefficient generators and output of the first
addition circuit; a third coefficient generator for
multiplying a third coefficient with the output of the first
addition circuit; a fourth coefficient generator for
multiplying a fourth coefficient with output of the second
addition circuit; and a third addition circuit for adding
output of the third coefficient generator and output of the
fourth coefficient generator to the input audio signal.
Also, an audio signal waveform emphasis processing
device may comprise a waveform emphasis circuit consisting
solely of passive elements, for emphasizing waveform in a low
frequency range and a high frequency range of an input audio
5
CA 02241000 1998-06-19
signal; an addition circuit for adding output of the waveform
emphasis circuit and the input audio signal; and variation-
means for controlling ratio of the output of the waveform
emphasis circuit to be added by the addition circuit and the
input audio signal.
Further, an audio signal waveform emphasis processing
method may comprise a first step of performing even-order
integration on an input audio signal; a second step of
performing even-order differentiation on the input audio
signal; and a third step of adding the even-order integrated
input audio signal produced in the first step and the even-
order differentiated input audio signal produced in the
second step to the input audio signal.
The first step may comprise a step of performing double
integration on the input audio signal successively a
plurality of times; the second step comprises a step of
performing double differentiation on the input audio signal
successively a plurality of times; and the third step
comprises: a fourth step of multiplying a first coefficient
respectively with values obtained by double integration each
of the plurality of times in the first step; a fifth step of
adding the double-integrated values that are multiplied by
the first coefficient in the fourth step; a sixth step of
multiplying a second coefficient respectively with values
obtained by double differentiation each of the plurality of
times in the second step; a seventh step of adding the
double-differentiated values that are multiplied by the
second coefficient in the sixth step; and an eighth step of
6
CA 02241000 1998-06-19
multiplying a third coefficient and a fourth coefficient
respectively with the sum that is obtained by the addition in
the fifth step and the sum that is obtained by the addition
in the seventh step, and adding results to the input audio
signal.
Also, a first coefficient group containing the first
coefficient and second coefficient group containing the
second coefficient may be determined in accordance with a
desired frequency-gain characteristic.
Also, the third coefficient and fourth coefficient may
be determined in accordance with the degree of emphasis in a
low frequency range and a high frequency range of the input
audio signal.
BRIEF DESCRIPTION OF THE DRAWINGS
Fig. 1 is a block diagram illustrating a typical
configuration of an audio signal waveform emphasis processing
device constituted by applying an audio signal waveform
emphasis processing device and method according to the
present invention;
Fig. 2 is a view showing an example of the frequency-
gain characteristic obtained with the audio signal waveform
emphasis processing device illustrated in Fig. 1;
Fig. 3 is a view showing the frequency-phase
characteristic of the audio signal waveform emphasis
processing device illustrated in Fig. 1 when the frequency-
gain characteristic shown in Fig. 2 is realized;
7
CA 02241000 1998-06-19
Fig. 4 is a view showing the frequency-gain
characteristic of a practical embodiment wherein the drawback
that the frequency-gain characteristic shown in Fig. 2 has
large gain within an unnecessary frequency range is
compensated;
Fig. 5 is a view showing the frequency-phase
characteristic of the audio signal waveform emphasis
processing device shown in Fig. 1 when the frequency-gain
characteristic shown in Fig. 4 is realized;
Fig. 6 is a view showing an example of the frequency-
gain characteristic obtained with the audio signal waveform
emphasis processing device shown in Fig. 1;
Fig. 7 is a view showing the frequency-phase
characteristic of the audio signal waveform emphasis
processing circuit shown in Fig. 1 when the frequency-gain
characteristic shown in Fig. 6 is realized;
Fig. 8 is a view showing the frequency-gain
characteristic of a practical embodiment when the drawback
that the frequency-gain characteristic shown in Fig. 6 has
large gain in an unnecessary frequency range is compensated;
Fig. 9 is a view showing the frequency-phase
characteristic of the audio signal waveform emphasis
processing device shown in Fig. 1 when the frequency-gain
characteristic shown in Fig. 8 is realized;
Fig. 10 is a circuit diagram illustrating a specific
construction of an audio signal waveform emphasis processing
device to which the audio signal waveform emphasis processing
8
CA 02241000 1998-06-19
device and method according to the present invention have
been applied;
Fig. 11 is a circuit diagram showing a specific example
of the double integrating circuit of Fig. 10;
Fig. 12 is a circuit diagram showing a specific example
of the double integrating circuit of Fig. 10;
Fig. 13 is a circuit diagram illustrating yet another
embodiment of an audio signal waveform emphasis processing
device according to the present invention;
Fig. 14 is a circuit diagram illustrating a specific
example of the waveform emphasis circuit shown in Fig. 13;
Fig. 15 is a circuit diagram showing another specific
example of the waveform emphasis circuit shown in Fig. 13;
Fig. 16 is a view showing the frequency-gain
characteristic of the waveform emphasis circuit in the
circuit shown in Fig. 13 and the frequency-gain
characteristic with respect to the original signal of an
addition circuit;
Fig. 17 is a view showing the frequency-gain
characteristic of the circuit shown in Fig. 13;
Fig. 18 is a circuit diagram of yet another embodiment
of an audio signal waveform emphasis processing device
according to the present invention wherein the degree of
emphasis of the low frequency range and the high frequency
range can be adjusted by a single variable resistor, by
adding a variable resistor to the circuit shown in Fig. 13;
and
9
CA 02241000 1998-06-19
Fig. 19 is a view showing the frequency-gain
characteristic when the degree of emphasis of the low
frequency range and high frequency range of the circuit shown
in Fig. 18 is varied.
BEST MODE FOR CARRYING OUT THE INVENTION
An embodiment of an audio signal waveform emphasis
processing device and method according to the present
invention is described in detail below with reference to the
appended drawings.
Fig. 1 is a block diagram showing a diagrammatic
configuration of an embodiment of an audio signal waveform
emphasis processing device which is arranged to be able to
obtain a frequency-gain characteristic in which the high
frequency range and low frequency range are emphasized as
desired without destroying the mutual phase relationship
between the frequency components of the middle frequency
range and the frequency components of the low frequency range
and high frequency range close to the middle frequency range
which constitute the audio signal.
Referring to Fig. 1, the audio signal waveform emphasis
processing device comprises a low frequency range emphasis
processing section 100 that performs emphasis processing on
the low frequency range of original signal f(t), which is a
function of time t and applied to input terminal INPUT; high
frequency range emphasis processing section 200 that performs
emphasis processing on the high frequency range of original
signal f(t) applied to input terminal INPUT; a coefficient
CA 02241000 1998-06-19
generator 300 that determines the degree of emphasis of the
low frequency range by multiplying by an overall coefficient
B the low frequency range emphasized signal g(t) that was
subjected to emphasis processing on the low frequency range
by means of the low frequency range emphasis processing
section 100 so as to output Bg(t); a coefficient generator
400 that determines the degree of emphasis of the high
frequency range by multiplying by an overall coefficient C
the high frequency range emphasized signal h(t) that was
subjected to high frequency range emphasis processing by
means of high frequency range emphasis processing section 200
so as to outputting Ch(t); and an adder 500 that adds the
signal Bg(t) whose low frequency range was emphasized that is
output from coefficient generator 300, the original signal
f(t) applied to input terminal INPUT and the signal Ch(t)
whose high frequency range has been emphasized that is output
from coefficient generator 400 so as to output an output
signal F(t) through output terminal OUTPUT.
Low frequency range emphasis processing section 100
comprises a plurality of double integrators II1, II2, ...,
Iip that output intermediate signals gl(t), g2(t), ..., gp(t)
by successively double-integrating original signal f(t) that
is applied at input terminal INPUT; a plurality of
coefficient generators bl, b2, ..., by that respectively
multiply by coefficients -H1, H2, ..., (-1)PBp that are set
beforehand intermediate signals gl(t), g2(t), ..., gp(t) that
are output from double integrators II1, II2, ..., Iip; and an
adder SUMI that outputs a low frequency range emphasized
11
CA 02241000 1998-06-19
signal g(t) obtained by adding the outputs of coefficient
generators bl, b2, ..., bp.
The group of coefficients B1, B2, ..., Bp that are used
as multipliers by coefficient generators bl, b2, ..., by are
determined in accordance with the desired frequency-phase
characteristic and may have negative as well as positive
values.
High frequency range emphasis processing section 200
comprises a plurality of double differentiators DD1, DD2,
..., Ddq that output intermediate signals hl(t), h2(t), ...,
hq(t) obtained by successive double differentiation of the
original signal f(t) that is applied to input terminal INPUT;
a plurality of coefficient generators cl, c2, ..., cq that
respectively multiply by coefficients -C1, C2, ..., (-1)qCq
that are set beforehand the intermediate signals hl(t),
h2(t), ..., hq(t) that are output from double differentiators
DD1, DD2, ..., Ddq; and adder SUMD that outputs the high
frequency range emphasized signal h(t) obtained by adding the
outputs of coefficient generators cl, c2, ..., cq.
It should be noted that the group of coefficients C1,
C2, ..., Cq that are applied as multipliers by coefficient
generators cl, c2, ..., cq are determined in accordance with
the desired frequency-phase characteristic, as will be
described, and may have negative as well as positive values.
The principles of the low frequency emphasis processing
action of low frequency emphasis processing section 100 will
now be described.
12
CA 02241000 1998-06-19
Since the original signal f(t) that is applied to input
terminal INPUT is an audio signal, in general it contains a
plurality of frequency components. The principles of the
present invention may therefore be explained by expressing
the original signal f(t) as:
f ~t~ = AlSin(tr~lt + B1) + A2Sin(c~2t + 92)+---+AnSin(t~nt + Bn) - ~ - ( 1 )
where A1, A2, ..., An respectively represent the amplitudes
of the frequency components constituting the original signal
f ( t ) ; cul , ua2, . . . , cam respectively represent the angular
frequencies of the frequency components constituting the
original signal f(t), and 81, A2, ..., An respectively
represent the phase angles of the frequency components
constituting the original signal f(t).
The equation (1) may be expressed in matrix notation:
A1
f ~t) = A2 ~Sin(wlt + 81), Sin(tr~2t + 62), ~ ~ -, Sin(~nt + 6n)~ - ~ - ( 2 )
An
Original signal f(t) represented by equation (1) that is
applied at input terminal INPUT is first of all subjected to
double integration in double integrator II1. Consequently,
the intermediate signal gl(t) that is output from double
integrator II1 is:
A1~1-Z
A2~2-2
gl(t) _ ~Sin(~ It + Bl), Sin(~2t + 92), ~ ~ -, Sin(~nt + 8n)~ - ~ - ( 3 )
Ant<m-2
13
CA 02241000 1998-06-19
As can be seen by comparing equation (2) with equation
(3), the term that is represented by the frequency component
and phase component of intermediate signal gl(t) that is
output from double integrator II1 is identical with the term
that is expressed by the frequency component and phase
component of original signal f(t). Consequently, it can be
seen that intermediate signal gl(t) that is output from
double integrator II1 maintains unchanged the mutual phase
relationship between the frequency components of original
signal f(t).
The amplitudes of the respective frequency components of
intermediate signal gl(t) that is output from double
integrator IIl have values that are inversely proportional to
the square of the angular frequency, namely, a value obtained
by dividing A1 by the square of w, a value obtained by
dividing A2 by the square of c~, and so forth. That is, the
amplitude of each frequency component of intermediate signal
gl(t) that is output from double integrator II1 increases in
inverse proportion to the square of the angular frequency as
the angular frequency becomes lower.
The intermediate signal gl(t) that is output from double
integrator II1 is again double-integrated by double
integrator II2, so that the intermediate signal g2(t) that is
output from the double integrator II2 is:
A1~1~'
A2~2~°
g2~t~= Sin(~lt+81),Sin(~2t+92),~~~,Sin(wnt+6n)~ ~~~ (4)
Ancvn~
14
CA 02241000 1998-06-19
As can be seen by comparing equation (2) with equation
(4), the term that is expressed by the frequency component
and phase component of intermediate signal g2(t) that is
output from double integrator II2 and the term that is
expressed by the frequency component and the phase component
of original signal f(t) are identical. It can therefore be
seen that the intermediate signal g2(t) that is output from
double integrator II2 maintains unaltered the mutual phase
relationships of the frequency components of original signal
f(t).
The amplitudes of the frequency components of the
intermediate signal g2(t) that is output from double
integrator II2 have values inversely proportional to the 4th
power of the angular frequency, namely, a value obtained by
dividing A1 by the 4th power of c~, a value obtained by
dividing A2 by the 4th power of w, and so forth. In other
words, the amplitude of each frequency component of
intermediate signal g2(t) that is output from double
integrator II2 increases in inverse proportion to the 4th
power of the angular frequency as the angular frequency
becomes lower.
Thus, the intermediate signal gp(t) that is output from
double integrator Iip is:
A1~1-2°
~~t~=(_1)P A2~2-Zp ~S,in(t~lt+61),Sin(~2t+B2),~~~,Sin(t~nt+6n)~ -~~ (5)
Antvn-Z''
CA 02241000 1998-06-19
As can be seen by comparing equation (2) with equation
(5), the term which is represented by the frequency component
and the phase component of intermediate signal gp(t) that is
output from double integrator Iip is identical with the term
that is expressed by the frequency component and phase
component of original signal f(t). It can therefore be seen
that, in the intermediate signal gp(t) that is output from
double integrator Iip, the mutual phase relationship of the
frequency components of original signal f(t) is maintained
unaltered.
Also, the amplitudes of the frequency components of
intermediate signal gp(t) that is output from double
integrator Iip have values inversely proportional to the
power 2p of the angular frequency, namely, a value obtained
by dividing A1 by c~ raised to the power 2p, a value obtained
by dividing A2 by w raised to the power 2p, and so forth. In
other words, the amplitudes of the frequency components of
intermediate signal gp(t) that is output from double
integrator Iip increase in inverse proportion to the power 2p
of the angular frequency as the angular frequency becomes
lower.
The intermediate signals gl(t), g2(t), ..., gp(t) shown
in the equations (3) to (5) that are respectively output from
double integrators II1, II2, ..., Iip are respectively
multiplied by pre-set coefficients -B1, B2, ..., (-1)pBp by
coefficient generators bl, b2, ..., by and added by adder
SUM1.
16
CA 02241000 1998-06-19
Consequently, the low frequency range emphasized signal
g(t) that is output from adder SUM1 is:
g(t~ _ -Blgl(t) + B2g2(t)-~ ~ ~+(-1)~ Bpgp(t) ~ ~ ~ ( 6 )
In other words, the low frequency range emphasized
signal g(t) is:
A1~ Bjr~l-zj
j=I
g~t) = A2~ Bjr,~2-z' ~Sin(~lt + B1) Sin(r,~2t + B2)
j=I , , ,Sin rant + Bn ( 7 )
An ~ Bj wn-z'
j=1
Consequently, as can be seen by comparing equation (2)
with equation (7), the term of the low frequency range
emphasized signal g(t) expressed by the frequency component
and the phase component is identical with the term of the
original signal f(t) expressed by the frequency component and
phase component, so in the low frequency range emphasized
signal g(t) the mutual phase relationship of the frequency
components of the original signal f(t) is preserved
unchanged.
The term that expresses the frequency components of the
low frequency range emphasized signal g(t) contains no
resonance pole. This is a reason why, as a result of the
sound quality improvement of the present invention, a gentle
sound quality is obtained with no specific resonance in the
low frequency range.
Next, the principles of operation of high frequency
range emphasis processing by high frequency range emphasis
17
CA 02241000 1998-06-19
processing section 200 will be described below. The original
signal f(t) indicated by equation (1) and applied to input
terminal INPUT is first of all subjected to double
differentiation by double differentiator DD1. The
intermediate signal hl(t) that is output from double
differentiator DD1 in this case is therefore:
A1~12
A2~2z
hl~t~ _ ~Sin(~lt + 81), Sin(~2t + B2), ~ ~ ~, Sin(cont + 6n)~ ~ ~ ~ ( 8 )
An~n2
As is clear by comparing equation (2) with equation (8),
the term expressed by the frequency component and phase
component of intermediate signal hl(t) that is output from
double differentiator DD1 is identical with the term
expressed by the frequency component and phase component of
original signal f(t). In the intermediate signal hl(t) that
is output from double differentiator DD1, the mutual phase
relationship of the frequency components of original signal
f(t) is therefore preserved unchanged.
The amplitudes of the frequency components of
intermediate signal hl(t) that is output from double
differentiator DD1 have values proportional to the square of
the angular frequency, namely, a value obtained by
multiplying A1 by the square of aa, a value obtained by
multiplying A2 by the square of a.~, and so forth. In other
words, the amplitude of the frequency components of
intermediate signal hl(t) that is output from double
differentiator DD1 increases in proportion to the square of
18
CA 02241000 1998-06-19
the angular frequency as the angular frequency becomes
higher.
The intermediate signal hl(t) that is output from double
differentiator DD1 is again subjected to double
differentiation by double differentiator DD2, so the
intermediate signal h2(t) that is output from the double
differentiator DD1 is:
A1~1°
A2~2°
h2~t~ _ ~Sin(tolt + Bl), Sin(tv2t + 92), ~ ~ ~, Sin(tont + 6n)~ ~ ~ ~ ( 9 )
Antvn°
As can be seen by comparing equation (2) with equation
(9), the term that is expressed by the frequency component
and phase component of intermediate signals h2(t) that is
output from double differentiator DD2 is identical with the
term expressed by the frequency component and phase component
of original signal f(t). It can therefore be seen that the
intermediate signal h2(t) that is output from double
differentiator DD2 likewise preserves unchanged the mutual
phase relationship of the frequency components of original
signal f(t).
The amplitudes of the frequency components of the
intermediate signal h2(t) that is output from double
differentiator DD2 have values proportional to the 4th power
of the angular frequency, namely, a value obtained by
multiplying A1 by the 4th power of c~, a value obtained by
multiplying A2 by the 4th power of a~, and so forth. In other
words, the amplitude of the frequency components of the
19
CA 02241000 1998-06-19
intermediate signal h2(t) that is output from double
differentiator DD2 increases in proportion to the 4th power
of the angular frequency as the angular frequency becomes
higher.
Likewise, intermediate signal hq(t) that is output from
double differentiator Ddq is:
Al~lZq
A2~2Zq
hq~t~ _ (-1)q ~Sin(~lt + 81),Sin(to2t + 62),~ ~ ~,Sin(~nt + 6n)~ ~ ~ ~ ( 1 0 )
An cm zy
As will be clear from comparison of equation (2) with
equation (10), the term that is expressed by the frequency
component and phase component of intermediate signal hgq(t)
that is output from double differentiator DDq is identical
with the term expressed by the frequency component and phase
component of original signal f(t). It can therefore be seen
that the intermediate signal hq(t) that is output from double
differentiator DDq preserves unchanged the mutual phase
relationship of the frequency components of original signal
f(t).
The amplitudes of the frequency components of
intermediate signal hq(t) that is output from double
differentiator DDq have values proportional to the power 2q of
the angular frequency, namely, a value obtained by
multiplying A1 by cu raised to the power 2q, a value obtained
by multiplying A2 by cu raised to the power 2q and so forth.
In other words, the amplitude of the frequency components of
intermediate signal hq(t) that is output from double
CA 02241000 1998-06-19
differentiator IIq increases in proportion with the angular
frequency raised to the power 2q as the angular frequency
becomes higher.
The intermediate signals hl(t), h2(t), ..., hq(t)
indicated by the equations (8) to (10) and output from double
differentiators DDl, DD2, ..., Ddq are respectively
multiplied by coefficients -C1, C2, ..., (-1) qCq that were
set beforehand, in coefficient generators cl, c2, ..., cq,
and then added by adder SUMD.
The high frequency range emphasized signal h(t) that is
output from adder SUMD is therefore:
h(t)=-Clhl(t)+C2h2(t)----+(-1)qCqhq(t) ~-- ( 1 1 )
That is, the high frequency range emphasized signal h(t)
is:
Al~ Cj~lz'
;_'
A2~Cj~2z'
h~t~= ;-, Sin(~lt+91),Sin(w2t+B2),---,Sin(tr~nt+9n)~ --- ( 1 2 )
An~Cj~n2'
''
Consequently, as will be clear by comparing equation (2)
with equation (12), the term that is expressed by the
frequency component and phase component of high frequency
range emphasized signal h(t) is identical with the term that
is expressed by the frequency component and phase component
of original signal f(t), so the high frequency range
emphasized signal h(t) preserves unaltered the mutual phase
21
CA 02241000 1998-06-19
relationships of the frequency components of the original
signal f(t).
The term that expresses the frequency components of high
frequency range emphasized signal h(t) contains no resonance
pole. This fact is the reason why a gentle sound quality
with no specific resonance is obtained in the high frequency
range as a result of the sound quality improvement of the
present invention.
After the low frequency range emphasized signal g(t)
represented by equation (7) has been subjected to low
frequency range emphasis processing as described above in low
frequency range emphasis processing section 100, it is
multiplied by overall coefficient B that determines the
degree of low frequency range emphasis in coefficient
generator 300 and supplied as low frequency range signal
Bg(t) to adder 500.
High frequency range emphasized signal h(t) represented
by equation (12) that has been subjected to high frequency
range emphasis processing as described above in high
frequency range emphasis processing section 200 is multiplied
by overall coefficient C that determines the degree of high
frequency range emphasis in coefficient generator 600 and is
applied to adder 500 as high frequency range signal Ch(t).
The original signal f(t) represented by equation (1),
which was applied at input terminal INPUT, is applied
directly to adder 500. Adder 500 adds the low frequency
range signal Bg(t) and high frequency range signal Ch(t) and
original signal f(t) referred to above, and outputs the
22
CA 02241000 1998-06-19
result to output terminal OUTPUT as output signal F(t). The
output signal F(t) that is output to output terminal OUTPUT
is therefore:
F(t~ = Bg(t) + f (t) + Ch(t) ~ ~ ~ ( 1 3 )
The output signal F(t) that is output at output terminal
OUTPUT is therefore:
A1C1 + B~ Bjwl-Zj + C~ Ck~l2k~
j=I k= JI
h~t) = A2C1 + BJ~ Bjc~2-zj + CkE Ck~22k) ~Sin(~lt + 81), Sin(~2t + B2), ~ ~ ~,
Sin(~nt + 8n)}
AnCI+B~Bj~n-Zj +C~Ckr~nZk~
j=I k=I
It can be seen by comparing the equation (14) with
equation (2) that the term expressed by the frequency
component and phase component of the output signal F(t) is
identical with the term expressed by the frequency component
and phase component of original signal f(t), so the output
signal F(t) preserves unaltered the mutual phase
relationships of the frequency components of original signal
f(t).
In the equation (14), if we assume that B = 0 and C = 0,
output signal F(t) is the same as the original signal f(t).
By extracting from equation (14) only the amplitude
term, we have:
23
CA 02241000 1998-06-19
A1C1+B~Bj~l-Zj +C~Ckwlzk~ _
j=1 k-I
A2~1+B~Bj~2-z'+C~Ck~22k~ [
k_1 1~1~...~1 ... ( 1 5 )
An(1 + B~ Bjton-Zj + C~ Ck~n2k~
j=i k= JI
If the respective coefficients associated with values
A1, A2, ..., An representing the amplitude of the frequency
components of the original signal are plotted on respective
frequency axes, the frequency-gain characteristic of the
waveform emphasis processing circuit can be obtained.
Fig. 2 shows an example of the frequency-gain
characteristic of the audio signal waveform emphasis
processing circuit shown in Fig. 1, which is obtained by
suitably selecting coefficients B1, B2, ..., Bp, B, C1, C2,
..., Cq, and C in accordance with equation (15).
As is clear from Fig. 2, the frequency-gain
characteristic of the audio signal waveform emphasis
processing device is flat in the middle frequency range. In
the low frequency range, the gain increases as the frequency
becomes lower, while, in the high frequency range, the gain
increases as the frequency becomes higher.
Coefficients H1, B2, ..., Hp and C1, C2, ..., Cq can be
found by solving the following linear simultaneous equations
with number of unknowns n=p+q indicated below.
V1= 1+B~Bjr.~l-zj +C~Ckc~lZk
j=1 k=I
V2 = 1+B~Bj~2-zj +C~Ck~2Zk ", ( 1 6 )
j-I k=1
Vn = 1 + B~ Bjr.~rt-Zj + C~ Cktr~n2k
j=I k=I
24
CA 02241000 1998-06-19
where V1, V2, ..., Vn are expected values of the gain
corresponding to angular frequencies wl, cal, ..., can on the
desired frequency-gain characteristic.
The simultaneous equations (16) have no general
solution, but solution can be obtained under certain
conditions.
In the simplest case, by putting p+q = n in the
simultaneous equations (16), coefficients H1, H2, ..., Hp and
C1, C2, ..., Cq can be uniquely found.
For example, in the configuration of Fig. 1, if p = q =
l, in other words, in the case where double integrators IIl,
II2, ..., Iip and double differentiators DD1, DD2, ..., Ddq,
are respectively of single-stage construction, by setting two
expected points V1, V2, coefficients H1 and C1 can be
uniquely obtained by solving the linear simultaneous
equations with two unknowns:
V1=1 + B1~1-Z + Clc~lz ". ( 1 7 )
V2 =1 + Blw2-2 + Cl~2z
Also, in for example the configuration of Fig. 1, if p =
q = 4, i.e., in the case where double integrators II1, II2,
..., Iip and double differentiators DD1, DD2, ..., Ddq are
respectively of 4-stage construction, the coefficients B1,
H2, B3, B4 and C1, C2, C3, C4 can be uniquely found by
solving the linear simultaneous equations with 8 unknowns if
the eight expected points V1, V2, ..., V8 are set.
CA 02241000 1998-06-19
V1= 1+B~Bj~I-Zj +C~Ck~lzk
j=I k=I
V2 =1 + B~ Bj~2-Zj + C~ Ckr.~2zk _
j=I k=I
V8=1+B~B~~8 2' +C~Cktv82k
j=I k=I
In other cases, a general solution cannot be obtained,
but approximate solutions and special solutions can be
obtained.
In the configuration of Fig. 1, if p = q = 1, in other
words, in the case where double integrators II1, II2, ...,
Iip and double differentiators DD1, DD2, ..., Ddq are
respectively of single-stage construction, output signal F(t)
is:
A1~1 + Bc~l-2 + C~lz
h~t~ = A2~1 + Br.~2-2 + Cw2z ~ ~Sin(tr~lt + Bl), Sin(r,~2t + 92), ~ ~ ~,
Sin(rvnt + 8n)~ ~ ~ ~ ( 1 9 )
An~l + Br.~n-z + Cwnz
The audio signal waveform emphasis processing device
with double integrators and double differentiators of this
construction is an example which exhibits highest cost-
performance.
Incidentally, in equation (14), if coefficients B1, B2,
..., Bp and coefficients C1, C2, ..., Cq are all positive,
output signal F(t) diverges as the angular frequency cap
approaches 0. Likewise, in equation (14), the output signal
F(t) diverges when the angular frequency w becomes large.
Therefore, in the configuration of Fig. 1, there is an
unwanted increase in gain in the very low frequency range and
26
CA 02241000 1998-06-19
very high frequency range: this is not appropriate in
practical use.
Accordingly, in a practical circuit configuration, in
order to ensure that the gain does not diverge in frequency
ranges that are entirely unnecessary in respect of the audio
signal, a configuration may be adopted in which the signs of
coefficients H1, B2, ..., Bp and coefficients C1, C2, ..., Cq
are appropriately selected or a cut-off filter is employed.
The location where such a cut-off filter is to be inserted is
determined taking into account the dynamic range and/or S-N
ratio of the actual device configuration and/or during
calculation process.
In the audio signal waveform emphasis method of the
present invention, coefficients B1, B2, ..., Bp and C1, C2,
..., Cq that can accurately satisfy a frequency-gain
characteristic of any arbitrary form do not in general exist.
However, this fact does not impair the effectiveness of the
present invention.
The present invention lies in ensuring mutually in-phase
characteristics between frequency components in a practically
important frequency range using a very straightforward
principle; its essence does not lie in obtaining a general
solution for arbitrary requirements.
Although the present invention is inappropriate for
satisfying fine emphasis characteristics, it makes it
possible to cope with characteristic compensation of electro-
mechanical systems such as microphones and/or speakers,
compensation for sound wave propagation characteristics in
27
CA 02241000 1998-06-19
air, compensation for sensitivity of auditory perception, and
environmental effects etc. over a wide range, by a simple
adjustment (for example adjustment of overall coefficients B
and C). This is one of the main features of the present
invention.
Also, with the sound quality emphasis of the present
invention, since no resonance occurs in the entire frequency
range, a natural sound quality is obtained with no unusual
characteristics even though emphasis is performed in the low
frequency range and/or high frequency range.
It is generally said that, in audio engineering, the
characteristics of the time-axis waveform of the audio signal
are not considered to be important in respect of sound
quality, but, in practice, it is known that they affect tone
characteristics.
Whether the waveform characteristic on the time axis
affects sound quality or not is not of the essence of the
present invention. The essence of the present invention lies
in emphasizing the characteristic of the waveform on the
sound axis without destroying the waveform characteristics of
the audio signal.
With regard to the specific techniques of integration,
differentiation and addition etc. of the audio signal, any
technique, whether of the analogue type or digital type, may
be employed.
An arbitrary audio signal in a given time zone can be
represented by a Fourier series without loss of generality in
practice. The frequency components and phase components of a
28
CA 02241000 1998-06-19
series obtained by applying integration of even-number degree
and differentiation of even-number degree on such a Fourier
series are in-phase and reverse-phase with respect to the
signal components of the original signal. Since it is known
whether any term of this series is in-phase or reverse-phase,
it is therefore possible to obtain a target frequency-gain
characteristic by adding the frequency components obtained by
multiplying by a group of appropriate coefficients, the
signal obtained by applying even-order differentiation and
integration in the necessary frequency region.
Fig. 3 shows the frequency-phase characteristic of an
audio signal waveform emphasis processing device shown in
Fig. 1 when the frequency-gain characteristic shown in Fig. 2
is realized. It can be seen that, with the frequency-phase
characteristic shown in Fig. 3, the phase distortion of the
output signal F(t) with respect to original signal f(t) is 0°
in the effective frequency range.
Fig. 4 shows the frequency-gain characteristic of a
practical embodiment in which the drawback that the
frequency-gain characteristic shown in Fig. 2 has large gain
in unnecessary frequency ranges is compensated. Such a
frequency-gain characteristic can be implemented by employing
a low frequency cut-off filter such that gain is not
increased in a frequency range below a frequency L3 and
employing a high frequency cut-off filter such that the gain
does not increase in a frequency range above a frequency H3.
Fig. 5 shows the frequency-phase characteristic of the
audio signal waveform emphasis processing device shown in
29
CA 02241000 1998-06-19
Fig. 1 in which the frequency-gain characteristic shown in
Fig. 4 is realized.
In general, if the low frequency range is cut off by a
low frequency cut-off filter, the phase in the low frequency
range is advanced and, if the high frequency range is cut off
by a high frequency cut-off filter, the phase in the high
frequency range is delayed.
If therefore a frequency-gain characteristic as shown in
Fig. 4 is to be achieved, the constants of the low frequency
cut-off filter and high frequency cut-off filter are
determined such that the frequency ranges where phase
distortion occurs are beyond L3 and H3, in order to ensure
that phase distortion does not occur in the effective audio
frequency range.
Fig. 6 shows an example of the frequency-gain
characteristic obtained by an audio signal waveform emphasis
processing device as shown in Fig. 1 in which the frequency-
gain characteristic shown in Fig. 8 is realized.
The frequency-gain characteristic shown in Fig. 6 can be
obtained by setting p = q = 5 or so in the configuration
shown in Fig. 1. The frequency-gain characteristic shown in
Fig. 6 is characterized in that the gain drops to the bottom
in the vicinity of frequencies L3 and H3.
Fig. 7 shows the frequency-phase characteristic of the
audio signal waveform emphasis processing device shown in
Fig. 1 in which the frequency-gain characteristic shown in
Fig. 6 is realized. The frequency-phase characteristic in
Fig. 7 shows that the phase distortion of the output signal
CA 02241000 1998-06-19
F(t) with respect to the original signal f(t) is 0° in the
effective frequency range.
However, in this case also, as shown in Fig. 6, the gain
increases rapidly below frequency L4 and above frequency H4,
so this is still unsuitable for practical application. A low
frequency cut-off filter and a high frequency cut-off filter
are therefore provided to respectively cut off the
unnecessary low frequency range below frequency L4 and the
unnecessary high frequency range above frequency H4.
Fig. 8 shows the frequency-gain characteristic of a
practical embodiment wherein the drawback that the frequency-
gain characteristic shown in Fig. 6 has large gain in
unnecessary frequency ranges is compensated. This frequency-
gain characteristic can be implemented by employing a low
frequency cut-off filter such that the gain does not increase
in the range below frequency L3 and by employing a high
frequency cut-off filter such that the gain does not increase
in the range above frequency H3.
Fig. 9 shows the frequency-phase characteristic of the
audio signal waveform emphasis processing circuit shown in
Fig. 1 in which the frequency-gain characteristic shown in
Fig. 8 is realized.
In this case also, if the low frequency range is cut off
by a low frequency cut-off filter, the phase in the low
frequency range is advanced and, if the high frequency range
is cut off by a high frequency cut-off filter, the phase in
the high frequency range is delayed.
31
CA 02241000 1998-06-19
Therefore, when a frequency-gain characteristic as shown
in Fig. 8 is to be realized, the constants of the low
frequency cut-off filter and high frequency cut-off filter
are determined such that the range in which phase distortion
is generated lies outside L3 and H3, in order that phase
distortion is not generated in the effective audio frequency
range.
The audio signal waveform emphasis processing device may
be realized by:
(1) an analogue device constituted by analog passive
elements and active elements;
(2) a digital device that digitally calculates a
digital signal;
(3) a digital signal processor (DSP) that performs
software calculation;
(4) a device constituted by package software and
soundboard control software.
Fig. 10 is a circuit diagram showing a specific circuit
configuration wherein an audio signal waveform emphasis
processing device according to the present invention is
realized by an analogue circuit constituted by passive
elements and active elements.
In Fig. 10, low frequency range emphasis circuit 10
corresponds to low frequency range emphasis processing
section 100 shown in Fig. 1; high frequency range emphasis
circuit 20 corresponds to high frequency range emphasis
processing section 200 shown in Fig. 1; low frequency range
adjustment circuit 30 corresponds to coefficient generator
32
CA 02241000 1998-06-19
300 shown in Fig. 1; high frequency range adjustment circuit
40 corresponds to coefficient generator 400 shown in Fig. 1;
and addition circuit 50 corresponds to adder 500 shown in
Fig. 1.
Low frequency range emphasis circuit 10 comprises N
double integrating circuits 11-1; 11-2, ..., 11-N; N
coefficient generators 12-1, 12-2, ..., 12-N respectively
connected to the outputs of the N double integrating circuits
11-1, 11-2, ..., 11-N; a first inverting amplification
circuit constituted of an operational amplifier 13 and
resistor 14, for adding the outputs of coefficient generators
12-2, 12-4, ..., 12-N; and a second inverting amplification
circuit constituted of a resistor 15 and operational
amplifier 16 and a resistor 17, for adding the output of the
first inverting amplification circuit and the outputs of
coefficient generators 12-1, 12-3, ..., 12-(N-1).
The N double integrating circuits 11-1, 11-2, ..., 11-N
correspond to the double integrators II1, II2, ..., Iip shown
in Fig. 1; the N coefficient generators 12-1, 12-2, ..., 12-N
correspond to the coefficient generators bl, b2, ..., by
shown in Fig. 1; and the first inverting amplification
circuit and second inverting amplification circuit correspond
to adder SUMI shown in Fig. 1.
The N double integrating circuits 11-1, 11-2, ..., 11-N
may respectively be constituted by the circuit shown in Fig.
11.
In Fig. 11, the double integration circuit (II H) 11
comprises a first integrating circuit consisting of resistor
33
CA 02241000 1998-06-19
II R1; capacitor II C1, resistor II R3, and operational
amplifier II OP1 and a second integrating circuit consisting
of resistor II R2, capacitor II C2, resistor II R4, and
operational amplifier II OP2. A signal that is input from
input terminal II_in is double-integrated and is output from
output terminal II out.
High frequency range emphasis circuit 20 comprises a
first inverting amplification circuit constituted of N double
differentiating circuits 21-1, 21-2, ..., 21-N; N coefficient
generators 21-1, 22-2, ..., 22-N respectively connected to
the outputs of the N double differentiating circuits 21-1,
21-2, ..., 21-N; a first inverting amplification circuit
constituted by operational amplifier 23 and resistor 24, for
adding the outputs of coefficient generators 22-2, 22-4, ...,
22-N; and a second inverting amplification circuit
constituted by resistor 25 and operational amplifier 26 and
resistor 27, for adding the output of the first inverting
amplification circuit and the outputs of coefficient
generators 22-1, 22-3, ..., 22-(N-1).
The N double differentiating circuits 21-1, 21-2, ...,
21-N of high frequency range emphasis circuit 20 correspond
to the double differentiators DD1, DD2, ..., Ddq shown in
Fig. 1; the N coefficient generators 22-1, 22-2, ..., 22-N
correspond to coefficient generators cl, c2, ..., cq shown in
Fig. 1, and the first inverting amplification circuit and
second inverting amplification circuit correspond to adder
SUMD shown in Fig. 1.
34
CA 02241000 1998-06-19
The N double integrating circuits 21-1, 21-2, ..., 21-N
may be respectively constituted by the circuit shown in Fig.
12.
In Fig. 12, the double differentiating circuit (DD B)12
comprises a first differentiating circuit consisting of
resistor DD R3, capacitor DD C1, resistor DD R1, and
operational amplifier DD OP1; and a second differentiating
circuit consisting of resistor DD R4, capacitor DD C2,
resistor DD R2, and operational amplifier DD OP2. A signal
that is input from input terminal DD in is double-
differentiated and is output from output terminal DD out.
The basic operation of low frequency range emphasis
circuit 10 and high frequency range emphasis circuit 20 shown
in Fig. 10 is the same as the processing of low frequency
range emphasis processing section 100 and high frequency
range emphasis processing section 200 shown in Fig. 1.
Low frequency range adjustment circuit 30 comprises a
variable resistor 31 and fixed resistor 32 and performs
adjustment of the output level of low frequency range
emphasis circuit 10; the output of the low frequency range
adjustment circuit 30 is input to addition circuit 50.
High frequency range adjustment circuit 40 is
constituted of variable resistor 41 and fixed resistor 42 and
performs adjustment of the output level of high frequency
range emphasis circuit 20~ the output of the high frequency
range adjustment circuit 40 is input to addition circuit 50.
Addition circuit 50 comprises a first inverting
amplifier consisting of resistor 51, operational amplifier
CA 02241000 1998-06-19
52, and resistor 53 and that adds the original signal applied
to input terminal Input and the output of low frequency range
adjustment circuit 30 and the output of high frequency range
adjustment circuit 40; and a second inverting amplifier
consisting of resistor 54, operational amplifier 55, and
resistor 56 which inverts the output of the first inverting
amplifier and outputs this to output terminal Output.
The basic operation of low frequency range adjustment
circuit 30 and high frequency range adjustment circuit 40 and
addition circuit 50 shown in Fig. 10 is the same as the
processing of coefficient generator 300 and coefficient
generator 400 and addition circuit 500 shown in Fig. 1.
With the configuration of Fig. 10, just as in the audio
signal waveform emphasis processing device shown in Fig. l,
an output signal having the desired frequency-gain
characteristic in which the high frequency range and low
frequency range are emphasized can be obtained from the
output terminal Output, without destroying the mutual phase
relationship between the frequency components constituting
the audio signal that is input from input terminal Input.
Fig. 13 is a circuit diagram showing yet a further
embodiment of an audio signal waveform emphasis processing
device according to the present invention, wherein the
configuration shown in Fig. 10 is simplified.
In Fig. 13, this circuit is constituted of a waveform
emphasis circuit (FL1) 600 that performs waveform emphasis in
respect of the high frequency range and low frequency range
on an original signal applied between input terminal Input
36
CA 02241000 1998-06-19
and ground terminal GND, and an addition circuit 700 that
adds the original signal applied between input terminal Input
and ground terminal GND and the waveform-emphasized signal
that has been subjected to waveform emphasis by waveform
emphasis circuit 600.
Fig. 14 shows a specific example of the waveform
emphasis circuit 600.
The waveform emphasis circuit 600 shown in Fig. 14 is
constituted by combining a double differentiating circuit
(high pass filter) of maximum gain about 30 times consisting
of resistors R3, R4, and capacitors C3, C4 with a double
integrating circuit (low pass filter) of maximum gain about
30 times consisting of resistors R1, R2 and capacitors C1,
C2.
In the waveform emphasis circuit 600 shown in Fig. 14,
the absolute value of the impedance in the low frequency
range of capacitors C3, C4 constituting the double
differentiating circuit is a value that is negligible in
comparison with the absolute value of the impedance of
resistors R1, R2 and capacitors C1, C2 constituting the
double integrating circuit.
Consequently, the effect of the double differentiating
circuit consisting of resistors R3, R4 and capacitors C3, C4
on the frequency-gain characteristic in the low frequency
range of the double integrating circuit consisting of
resistors R1, R2 and capacitors C1, C2 is negligible.
Also, the absolute value of the impedance in the high
frequency range of capacitors C1, C2 constituting the double
37
CA 02241000 1998-06-19
differentiating circuit is a value that is negligible in
comparison with the absolute value of the impedance of
resistors R3, R4, and capacitors C3, C4 constituting the
double differentiating circuit.
The effect of the double integrating circuit consisting
of resistors R1, R2 and capacitors C1, C2 on the frequency-
gain characteristic in the high frequency range of the double
differentiating circuit consisting of resistors R3, R4 and
capacitors C3, C4 can therefore be neglected.
The waveform emphasis circuit 600 shown in Fig. 4
therefore acts as a circuit that performs waveform emphasis
for the high frequency range and low frequency range.
The waveform emphasis circuit 600 shown in Fig. 13 could
be constructed of the circuit shown in Fig. 15 instead of the
circuit shown in Fig. 14.
The circuit shown in Fig. 15 is constituted by coupling
a high pass filter consisting of resistors R4, R5, R6 and
capacitors C4, C5, C6 with a low pass filter consisting of
resistors R1, R2, R3, and capacitors C1, C2, C3.
Waveform emphasis in the high frequency range and low
frequency range can be obtained with the circuit shown in
Fig. 15 just as in the case of the circuit shown in Fig. 14.
The circuit shown in Fig. 15 is in principle third-order
in respect of both the low frequency range and the high
frequency range, but, due to the mutual interaction of the
respective constants of differentiation and integration in
the real constant circuit, a characteristic close to second-
38
CA 02241000 1998-06-19
order is obtained, so a circuit is obtained which is
effective in practice.
In Fig. 13, addition circuit 700 is constituted of
resistors Ri, Rf, Rg, capacitor Cg and operational amplifier
OP1.
Since Ri is set to be equal to Rf (Ri=Rf), operational
amplifier OP1 functions as an inverting amplifier of gain 1.
In other words, the output of the operational amplifier OP1
is of the same magnitude as the input but inverted in phase.
Resistance Rg has scarcely any effect on the input/output
gain of this circuit.
Also, the degree of emphasis in the low frequency range
and high frequency range in the addition circuit 700 is
determined by the relationship of resistor Rg and resistor
Rf, since the effect of resistor Ri is negligible.
Capacitor Cg is a capacitor that is necessary in an
actual operating circuit: it has the function of reducing the
DC offset of operational amplifier OP1 and the function of
removing unnecessary signal components of the low frequency
range.
Fig. 16 shows the frequency-gain characteristic CLH of
waveform emphasis circuit 600 in the circuit shown in Fig. 13
and the frequency-gain characteristic CM with respect to the
original signal of addition circuit 700.
As is clear from the frequency-gain characteristic CLM
shown in Fig. 16, the low frequency range and high frequency
range of the original signal are emphasized by waveform
emphasis circuit 600, and, as is clear from the frequency-
39
CA 02241000 1998-06-19
gain characteristic CM, the frequency-gain characteristic of
the original signal in addition circuit 700 becomes flat.
With the circuit shown in Fig. 13, a waveform-emphasized
signal whose low frequency range and high frequency range
have been emphasized by wave band emphasis circuit 600
together with the original signal that is input from input
terminal Input are output from output terminal Output, so the
overall frequency-gain characteristic of the circuit shown in
Fig. 13 is a frequency characteristic obtained by adding the
frequency-gain characteristic CLH shown in Fig. 16 and the
frequency-gain characteristic CM, in other words, the
frequency-gain characteristic CA shown in Fig. 17.
Fig. 18 is a circuit diagram showing yet another
embodiment of an audio signal waveform emphasis processing
device according to the present invention wherein the degree
of emphasis of the low frequency range and high frequency
range can be adjusted by addition of a variable resistor VRg
to the circuit shown in Fig. 13.
In Fig. 18, variable resistor Vrg is connected between
resistor Rg and the ground. The rest of the configuration in
Fig. 16 is the same as the circuit shown in Fig. 13.
Specifically, in the circuit shown in Fig. 18, addition
circuit 800 is constituted of resistances Ri, Rf, Rg, RVg,
capacitor Cg and operational amplifier OP1.
In the configuration shown in Fig. 18, when variable
resistor Vrg is adjusted, the relationship of resistance (Vrg
+ Rg) and resistance Rf is altered and the degree of emphasis
CA 02241000 1998-06-19
of the low frequency range and high frequency range with
respect to the original signal can thereby be adjusted.
Fig. 19 shows the frequency-gain characteristic realized
with the circuit shown in Fig. 18.
As shown in Fig. 19, in the circuit shown in Fig. 18,
when variable resistance Vrg is adjusted, its frequency-gain
characteristic can be altered as shown by CA1, CA2, or CA3.
It should be noted that, in the above configuration,
integration, differentiation, multiplication and addition are
not strictly mathematical but include allowable error in the
realization of a practical device. They could also include
expression by calculation using a stored program system in
addition to functions based on physical phenomena.
Calculation processing of the signal is to be performed
in the frequency range of the audio signal or in a frequency
range somewhat wider than this.
The double integration and double differentiation do not
need to be pure integration and differentiation so long as
they are sufficient to enable a practical audio device to be
realized.
Hy "phase" in the description of the present invention
is meant the phase dependent on the complex impedance of the
lumped parameter circuit. "Phase" delay of an element having
a delay time and "phase" dependent on the complex impedance
of a linear circuit should be distinguished from "phase" in
the description of the present invention.
"Phase distortion" means the phenomenon that the mutual
phase relationship between the frequency components of an
41
CA 02241000 1998-06-19
output of a function that inputs a signal comprising a
plurality of frequency components in a specific phase
relationship of the input is different from the phase
relationship of the input.
In the present invention "low frequency range" means for
example the frequency range below 300 Hz.
"High frequency range" means for example the frequency
range above 1000 Hz.
"Middle frequency range" means for example the frequency
range from 300 Hz to 1000 Hz.
"Unnecessary low frequency range" means the low
frequency range which is clearly unnecessary in auditory
perception, for example a frequency range of below 20 Hz; the
"unnecessary high frequency range" means the high frequency
range that is clearly unnecessary for auditory perception,
for example a frequency range of more than 20 Khz.
INDUSTRIAL APPLICABILITY
The present invention provides an audio signal waveform
emphasis processing device and method wherein the sound
quality of audio devices of various types is improved by
emphasis processing of the audio signal waveform, and is
applicable to the various types of audio devices. According
to the present invention, it is possible to obtain a desired
frequency-gain characteristic without destroying the phase
relationship of the frequency components constituting the
audio signal, in particular the phase relationship between
the frequency components of the middle frequency range and
42
' CA 02241000 1998-06-19
the frequency components of the low frequency range and high
frequency range close to the middle frequency range. Thus,
the function not to destroy the phase relationship in the
middle frequency range is realized even though a great
emphasis is performed on the low and high frequency ranges,
as a result of which sound quality of audio signals in
various audio devises is enormously improved.
43
