close

Вход

Забыли?

вход по аккаунту

?

JP2008109279

код для вставкиСкачать
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 JP2008109279
The present invention provides an audio signal processing apparatus capable of reducing the
circuit size of an FIR filter that digitally filters audio signals having different sampling
frequencies. A sample rate converter down-converts an audio signal Xs based on a ratio β,
supplies the down-converted audio signal Xs to FIR filters 2a and 2b, and performs digital
filtering. FIR filters 2a and 2b are impulses generated when digitally filtering an audio signal
sampled at a sampling frequency equal to reference frequency fbase under an operating
frequency fa and fb equal to reference frequency fbase in a predetermined frequency range It is
formed by an FIR filter whose number of taps is determined according to the duration of
response. [Selected figure] Figure 2
Audio signal processing apparatus and audio signal processing method
[0001]
The present invention relates to an audio signal processing apparatus and an audio signal
processing method for processing an audio signal having different sampling frequencies by a
digital filter.
[0002]
In a digital audio device handling a discretized audio signal, a configuration for performing
digital signal processing such as frequency band division and sound quality adjustment
(equalizing) by an audio Finite Impulse Response Filter signal processing device provided with an
FIR (Finite Impulse Response) filter It has become.
09-05-2019
1
[0003]
FIG. 1A shows an example of the configuration of a conventional audio signal processing
apparatus.
In this audio signal processing apparatus, a digital signal processor (in which four systems of FIR
digital filters (hereinafter referred to as "FIR filters") F1, F2, F3, and F4, a DA converter, and an
amplifier for volume control are formed DSP: Digital Signal Processor) is provided.
In each system, for the audio signal X sampled at a predetermined sampling frequency fs
supplied from a signal source such as a CD player, the low-pass speaker SL, the low-mid-pass
speaker SLM, the middle-high pass speaker SMH, and the high-pass speaker SH Processing such
as frequency band division and sound quality adjustment shown in the characteristic chart of FIG.
1B is performed in accordance with each frequency band.
[0004]
Here, each of the FIR filters F1, F2, F3, and F4 is represented by a transfer function H (z) of the
following equation (1a), and is input series at time n for each sampling period T (that is, 1 / fs).
When an audio signal X (n) is supplied, the audio signal X (n) at that time n and the audio signal
X (n-1), X (n-2),. And the filter coefficients (tap coefficients) h0, h1, h2,..., HN of the number of
taps (N + 1) determined based on the duration time of the impulse response by the following
equation (1b) An output sequence Y (n) represented by a non-recursive difference equation is
generated and output.
[0005]
That is, the filter coefficients h0, h1, h2,..., HN in the FIR filters F1, F2, F3, F4 are respectively set
to specific coefficients, so that the FIR filter F1 is a low pass filter and the FIR filter F2 is low. The
middle pass filter, the FIR filter F3 is a middle high pass filter, and the FIR filter F4 is a high pass
filter.
[0006]
[0007]
The above-described conventional audio signal processing apparatus is configured to process
09-05-2019
2
only the audio signal X sampled at a sampling frequency fs of 44.1 kHz, for example, supplied
from a signal source such as a CD player.
[0008]
However, not only CD (Compact Disc) but also various media such as MD (Mini Disc) and DVD
(Digital Versatile Disc) have been developed, and with the progress of multi-media etc., only
sampling frequency of 44.1 kHz There is also a need for a compatible audio signal processor
capable of processing audio signals sampled at sampling frequencies such as 32 kHz, 48 kHz,
88.2 kHz, 96 kHz, 176.4 kHz, 192 kHz, etc.
[0009]
However, when the sampling frequency fs becomes high, generally, the number of taps of each of
the FIR filters F1, F2, F3, and F3 must be increased to lengthen the tap length, as is apparent
from the above formulas (1a) and (1b) The number of multipliers and delay elements for
multiply-accumulate operations increases, and the storage capacity of the buffer memory
required when performing signal processing also increases, so that the circuit scale becomes
large, and furthermore, digital signals The amount of processing of a processor (DSP) increases,
causing problems such as an increase in processing load.
[0010]
For example, if all the FIR filters F1, F2, F3 and F4 are designed in accordance with 192 kHz,
which is the highest sampling frequency described above, 32 kHz, 44.1 kHz, 48 kHz, 88. It is
theoretically possible to process audio signals such as 2 kHz, 96 kHz, 176.4 kHz, etc., but the
number of taps of each of the FIR filters F1, F2, F3 and F4 becomes enormous, and the circuit
scale increases, digital signal processor This causes problems such as an increase in the
processing load of (DSP).
In addition, if the tap length is increased, the convergence speed of the impulse response may be
degraded.
[0011]
In addition, as described above, when all the FIR filters F1, F2, F3, and F4 are designed in
accordance with the highest sampling frequency of 192 kHz, the FIR filters F1, F1 that process
low- and mid-range audio signals In F2, an echo (error component) is generated in a period of no
09-05-2019
3
signal (a period in which the input audio signal becomes almost 0 level), and the echo is output
as an output sequence Y. The sound quality of the reproduced sound reproduced by the low to
mid range speaker SLM is degraded.
[0012]
The present invention has been made in view of such conventional problems, and an audio signal
processing apparatus for performing processing such as frequency band division and sound
quality adjustment by an FIR filter on audio signals having different sampling frequencies. It is an
object of the present invention to provide an audio signal processing device and an audio signal
processing method capable of realizing reduction in circuit scale of a filter, reduction in
processing amount, and the like.
[0013]
The invention according to claim 1 is an audio signal processing apparatus having an FIR filter
that digitally filters a plurality of audio signals sampled at different sampling frequencies output
from a signal source, wherein the audio signal processing apparatus comprises: The sampling
frequency of the audio signal output from the signal source is compared with the reference
frequency using a predetermined sampling frequency as a reference frequency, and it is
determined that the sampling frequency of the audio signal output from the signal source is
higher than the reference frequency Then, control means for calculating a ratio between the
sampling frequency and the reference frequency as a resampling ratio, down-converting an audio
signal output from the signal source based on the resampling ratio, and supplying it to the FIR
filter The Pull rate conversion means, wherein the FIR filter digitally filters an audio signal
sampled at a sampling frequency equal to the reference frequency under an operating frequency
equal to the reference frequency in a predetermined frequency range It is characterized in that it
is an FIR filter whose number of taps is determined in accordance with the duration of the
impulse response that occurs in
[0014]
The invention according to claim 5 is an audio signal processing method for digitally filtering a
plurality of audio signals sampled at different sampling frequencies output from a signal source,
wherein a predetermined sampling frequency of the sampling frequencies is selected. The
sampling frequency of the audio signal output from the signal source is compared with the
reference frequency as the reference frequency, and the sampling frequency of the audio signal
output from the signal source is determined to be higher than the reference frequency. A control
step of calculating a ratio between a frequency and the reference frequency as a resampling
ratio, a sample rate conversion step of down converting an audio signal output from the signal
source based on the resampling ratio, the sample rate conversion step And FIR filtering for
09-05-2019
4
digitally filtering the downconverted downconverted audio signal, said FIR filtering being
sampled at a sampling frequency equal to said reference frequency under an operating frequency
equal to said reference frequency It is characterized in that it is an FIR filter process in which the
number of taps is determined in accordance with the duration of an impulse response which
occurs when digitally filtering an audio signal in a predetermined frequency range.
[0015]
A preferred embodiment of the present invention will be described with reference to FIGS. 2 and
3.
FIG. 2 is a block diagram showing the configuration of the audio signal processing device of the
present embodiment, and FIGS. 3A to 3D are block diagrams showing the basic configuration of
the FIR filter.
[0016]
In the following description, for convenience, an audio signal sampled at a sampling frequency
will be described as an "audio signal of sampling frequency", and a sampling frequency of an
audio signal sampled at a sampling frequency as "sampling frequency of an audio signal". I
decided to.
[0017]
In FIG. 2, the audio signal processing apparatus of the present embodiment has a control unit 20
and a digital signal processor 30, and the digital signal processor 30 controls the sample rate
converter 1 and four FIR filters 2a, 2b and 2c. , 2d and DA converters 3a, 3b, 3c, 3d and
amplifiers 4a, 4b, 4c, 4d for volume control.
[0018]
Then, the digital signal processor 30 is a signal source such as an MD player or a CD player or a
DVD player for reproducing media such as MD, CD or DVD, a receiver for receiving a digital
broadcast, a modem for receiving an audio source via the Internet Processing such as frequency
band division and sound quality adjustment according to each frequency band of low-pass
speaker SL, low-mid-pass speaker SLM, middle-high pass speaker SMH, and high-pass speaker SH
for audio signal Xs supplied from 10 I do.
09-05-2019
5
[0019]
Also, although the details will be described later, the audio signal processing apparatus of this
embodiment is 32 kHz, 44.1 kHz, 48 kHz, 88.2 kHz, 96 kHz, 176.4 kHz, 192 kHz specified in the
standards of CD, DVD, etc. The audio signal Xs sampled at any sampling frequency fs is
configured to exhibit compatibility for performing processing such as frequency band division
and sound quality adjustment.
[0020]
The control unit 20 detects the sampling frequency fs of the audio signal Xs output from the
signal source 10 by examining predetermined control data output from the signal source 10, and
the sample rate is detected according to the detected sampling frequency fs. A control signal
(code omitted) for setting a resampling ratio β at the time of downsampling to the converter 1 is
generated, and further, according to the detected sampling frequency fs, the FIR filters 2a to 2d
and the DA converters 3a to 2d Clock signals (symbols omitted) for setting the operating
frequencies fa, fb, fc and fd of the amplifiers 4a to 4d are generated.
[0021]
The resampling ratio β is the ratio of the sampling frequency fv to the sampling frequency fs
when the audio signal Xs sampled at the sampling frequency fs is downsampled to the audio
signal Xv at a lower sampling frequency fv (fv / defined as fs).
[0022]
Here, the control unit 20 performs the following processing on the control signal for specifying
the resampling ratio β, the clock signal for setting the operating frequency fa, the clock signal
for setting the operating frequency fb, and the operating frequency A clock signal for setting fc
and a clock signal for setting the operating frequency fd are generated.
[0023]
First, in the control unit 20, data indicating a predetermined reference frequency fbase is stored
in advance, and in the present embodiment, the reference frequency fbase is the lowest from the
lowest sampling frequency (32 kHz) defined in the above standard. Of the range up to the high
sampling frequency (192 kHz), the frequency equal to the sampling frequency fs of 44.1 kHz is
determined.
[0024]
09-05-2019
6
Then, the control unit 20 detects the sampling frequency fs of the audio signal Xs, compares it
with the reference frequency fbase, and sets a resampling ratio β according to the magnitude
relationship between the sampling frequency fs and the reference frequency fbase; The clock
signals for setting the operating frequencies fa, fb, fc and fd are generated.
[0025]
<< When fs and fbase are equal >> The sampling frequency fs and the reference frequency fbase
are compared, and when both frequencies are equal (when 44.1 kHz), the ratio (fbase / fs) is
calculated to calculate the resampling ratio β Set to 1.
As a result, the sample rate converter 1 is not downsampled, and the audio signal Xs sampled at
the sampling frequency fs of 44.1 kHz is output as the audio signal Xv as it is.
Further, the frequencies (operating frequencies) fa, fb, fc and fd of the clock signal supplied to
the FIR filters 2a to 2d are set to 44.1 kHz which is equal to the sampling frequency fs of the
audio signal Xs.
[0026]
Therefore, when the sampling frequency fs and the reference frequency fbase are equal, the
control unit 20 performs the condition setting represented by the following equations (2a), (2b)
and (2c).
[0027]
[0028]
<< When fs is lower than fbase >> When the sampling frequency fs is lower than the reference
frequency fbase, the resampling ratio β is set to 1.
As a result, the sample rate converter 1 is not downsampled, and the audio signal Xs sampled at
the sampling frequency fs of 32 kHz is output as the audio signal Xv as it is.
09-05-2019
7
Further, the frequencies (operating frequencies) fa, fb, fc and fd of the clock signal supplied to
the FIR filters 2a to 2d are set to 32 kHz which is equal to the sampling frequency fs of the audio
signal Xs.
[0029]
Therefore, when the sampling frequency fs is lower than the reference frequency fbase, the
control unit 20 performs the condition setting represented by the following equations (3a) (3b)
(3c).
[0030]
[0031]
<< When fs is higher than fbase >> When the sampling frequency fs is higher than the reference
frequency fbase, the ratio (fbase / fs) of the reference frequency fbase to the sampling frequency
fs is calculated, and the resampling ratio β is obtained.
This causes the sample rate converter 1 to downsample at the resampling ratio β, and generates
the audio signal Xv obtained by converting the audio signal Xs sampled at the sampling
frequency fs to the sampling frequency fv lower than that. Output.
Here, since the reference frequency fbase is determined to be 44.1 kHz, the resampling ratio β is
44.1 kHz / fs.
When the sample rate converter 1 downsamples the audio signal Xs based on the resampling
ratio β, the audio signal Xv of the sampling frequency fv equal to the reference frequency fbase
is generated and output.
[0032]
Furthermore, the frequency (operating frequency) fa and fb of the clock signal supplied to the
FIR filters 2a and 2b is set to 44.1 kHz equal to the reference frequency fbase, and the frequency
09-05-2019
8
(operating frequency) of the clock signal supplied to the FIR filters 2c and 2d ) Set fc and fd to a
frequency equal to the sampling frequency fs of the audio signal Xs.
[0033]
Therefore, when the sampling frequency fs is higher than the reference frequency fbase, that is,
when the sampling frequency fs is any of 48 kHz, 88.2 kHz, 96 kHz, 176.4 kHz, and 192 kHz, the
control unit 20 performs the following equation (4a) (4b) Perform the condition setting
represented by (4c).
[0034]
[0035]
Next, the sample rate converter 1 is formed by the down sampler, downsamples the audio signal
Xs at the resampling ratio β set conditionally by the control unit 20, and the sampling frequency
represented by the following equation (5) An audio signal Xv that has been subjected to sample
rate conversion to fv is generated and output.
[0036]
[0037]
Thereby, when the sampling frequency fs of the audio signal Xs is lower than the abovementioned reference frequency fbase (when 32 kHz) or when the sampling frequency fs of the
audio signal Xs is equal to the reference frequency fbase (when 44.1 kHz) The audio signal Xs is
output as it is as the audio signal Xv based on the relationship of the equations (2a), (3a) and (5),
and the sampling frequency fs of the audio signal Xs is higher 88.2 kHz, 96 kHz, 176.4 kHz, and
192 kHz), the sampling frequency equal to the reference frequency fbase with the audio signal
Xs at the sampling frequency fs, based on the relationship of the equations (4a) and (5) above.
The audio signal Xv of fv (that is, 44.1 kHz) is Convert the sampling rate and output.
[0038]
Next, the FIR filter 2a belonging to the first system is formed of an FIR filter whose tap number
Tp shown in FIG. 3A is determined to be P + 1, and P delay elements indicated by the delay time
.tau.a are used. , P + 1 multipliers that multiply the tap coefficients a0, a1, a2,..., AP-1, aP to the
input or output of each delay element, and addition that adds the outputs of all the multipliers It
has a basic configuration provided with the device ADa.
09-05-2019
9
[0039]
Here, the number of taps Tp is set at a delay time τa of each of the P delay elements to the
inverse number of the above-mentioned reference frequency fbase (that is, sampling period 1 /
fbase), and sampling 44.1 kHz equal to the reference frequency fbase It is designed to match the
duration of the impulse response that occurs when the audio signal sampled at the frequency fs
is low-pass filtered in the range of the frequency band of the low-pass speaker SL.
That is, by determining the number of taps Tp so that the tap length is equal to or longer than
the duration of the impulse response, the FIR filter 2a is designed so as not to be unnecessarily
large.
[0040]
Then, the FIR filter 2a exhibits the transfer function Ha (z) of z conversion notation shown in the
following equation (6a), receives the audio signal Xv output from the sample rate converter 1,
and receives the sampling period Tv (that is, When the audio signal Xv (m) which is an input
sequence is supplied at time m every 1 / fv), the audio signal Xv (m) at that time m and the audio
signal Xv (m-1) supplied in the past are supplied. By performing product-sum operation on Xv
(m−2),..., Xv (mP) and the filter coefficients (tap coefficients) a0, a1, a2,. The output sequence Ya
(m) is generated and output.
Then, the output sequence Ya (m) is converted from digital to analog by the D / A converter 3a,
and then the volume is adjusted by the amplifier 4a, and output to the high frequency speaker SL
side.
The D / A converter 3a and the amplifier 4a are also configured to perform digital-to-analog
conversion processing and amplification processing in synchronization with the clock signal of
the operating frequency fa.
[0041]
09-05-2019
10
[0042]
Furthermore, in the above equations (6a) and (6b), the resampling ratio β in the sample rate
converter 1 is set to 1, and the audio signal Xs sampled at the 32 kHz sampling frequency fs is
sampled as it is at the 32 kHz sampling frequency fv When supplied as the audio signal Xv
(hereinafter referred to as "first case"), the resampling ratio .beta. In the sample rate converter 1
is set to 1 and sampled at the sampling frequency fs of 44.1 kHz. When the audio signal Xs is
supplied as the audio signal Xv sampled at the sampling frequency fv of 44.1 kHz as it is
(hereinafter referred to as "second case"), the resampling ratio .beta. Set to (fbase / fs), 44.1kHz
Summarize the cases where the audio signal Xs sampled at the higher sampling frequency fs is
down-converted to the audio signal Xv at the sampling frequency fv of 44.1 kHz (hereinafter
referred to as the “third case”) In the “second case” and the “third case”, although the
same filter coefficients a0, a1, a2,. Filter coefficients a0, a1, a2,..., AP are adjusted to obtain the
same low frequency characteristics as the above-mentioned predetermined frequency
characteristics of the “second case” and the “third case”. Is supposed to be done.
Although not shown, adjustment means for adjusting the filter coefficients a 0, a 1, a 2,..., A P
based on the data stored in the database is formed in the control unit 20.
[0043]
Next, the FIR filter 2b belonging to the second system is formed of an FIR filter whose tap
number Tq shown in FIG. 3B is determined to be Q + 1, and Q delay elements shown by delay
time .tau.b are used. , Q + 1 multipliers that multiply tap coefficients b0, b1, b2,..., BQ-1, bQ to the
input or output of each delay element, and addition that adds the outputs of all the multipliers It
has a basic configuration provided with the device ADb.
[0044]
Here, the number of taps Tq is set at a delay time τb of each of the Q delay elements to the
reciprocal of the above-mentioned reference frequency fbase (that is, sampling period 1 / fbase),
and sampling at 44.1 kHz is equal to the reference frequency fbase. It is designed to match the
duration of impulse response that occurs when the audio signal being sampled at the frequency
fs is band pass filtered in the range of the frequency band of the low to mid speaker SLM.
09-05-2019
11
That is, by determining the number of taps Tq so that the tap length is equal to or longer than
the duration of the impulse response, the FIR filter 2b is designed so as not to be unnecessarily
large.
[0045]
Then, the FIR filter 2b exhibits the transfer function Hb (z) of z conversion notation shown in the
following equation (7a), receives the audio signal Xv output from the sample rate converter 1,
and receives the sampling period Tv (that is, When the audio signal Xv (m) which is an input
sequence is supplied at time m every 1 / fv), the audio signal Xv (m) at that time m and the audio
signal Xv (m-1) supplied in the past are supplied. By performing product-sum operation on Xv
(m−2),..., Xv (mQ) and the filter coefficients (tap coefficients) b0, b1, b2,. Generate and output
the output sequence Yb (m).
Then, the output sequence Yb (m) is converted from digital to analog by the DA converter 3b, and
then the volume is adjusted by the amplifier 4b and output to the high frequency speaker SLM
side.
The DA converter 3 b and the amplifier 4 b are also configured to perform digital analog
conversion processing and amplification processing in synchronization with the clock signal of
the operating frequency fb.
[0046]
[0047]
Furthermore, in the above formulas (7a) and (7b), the resampling ratio β in the sample rate
converter 1 is set to 1, and the audio signal Xs sampled at the sampling frequency fs of 32 kHz is
sampled as it is at the sampling frequency fv of 32 kHz. When supplied as the audio signal Xv
(hereinafter referred to as "the fourth case"), the resampling ratio .beta. In the sample rate
converter 1 is set to 1 and sampled at the sampling frequency fs of 44.1 kHz. When the audio
signal Xs is supplied as the audio signal Xv sampled at the sampling frequency fv of 44.1 kHz as
it is (hereinafter referred to as "the fifth case"), the resampling ratio .beta. It is set to (fv / fs) and
The case where the audio signal Xs sampled at a higher sampling frequency fs is down-converted
09-05-2019
12
to the audio signal Xv at the sampling frequency fv of 44.1 kHz (hereinafter referred to as “the
sixth case”) is summarized. In the “fifth case” and the “sixth case”, although the same filter
coefficients b0, b1, b2,. Filter coefficients b0, b1, b2,..., BQ are adjusted to obtain the same low
frequency characteristics as the above-mentioned predetermined frequency characteristics of the
“fifth case” and the “sixth case”. Is supposed to be done.
Although not shown, adjustment means for adjusting the filter coefficients b0, b1, b2,..., BQ based
on the data stored in the database is formed in the control unit 20.
[0048]
Next, the FIR filter 2c belonging to the third system is formed by an FIR filter in which the
number of taps Tr shown in FIG. 3C is determined to be R + 1, and R delay elements indicated by
delay time .tau.c are used. , R + 1 multipliers that multiply tap coefficients c0, c1, c2,..., CR−1, cR
by the input or output of each delay element, and addition that adds the outputs of all the
multipliers And a base unit ADc.
[0049]
Here, the number of taps Tr is set by setting the delay time τc of each of the R delay elements to
the reciprocal of the highest sampling frequency (192 kHz) defined by the above-mentioned
standard (that is, sampling period 1/192 kHz) The audio signal being sampled at the sampling
frequency fs of 192 kHz is designed in accordance with the duration of the impulse response that
occurs when band pass filtering is performed in the range of the frequency band of the middle
high frequency speaker SMH.
That is, by determining the number of taps Tr so that the tap length is equal to or longer than the
duration of the impulse response, the FIR filter 2c is designed so as not to be unnecessarily large.
[0050]
Then, the FIR filter 2c exhibits the transfer function Hc (z) of z conversion notation shown in the
following equation (8a), receives the audio signal Xs output from the signal source 10, and has a
sampling period Ts (that is, 1/1). When an audio signal Xs (n) which is an input sequence is
09-05-2019
13
supplied at time n every fs), the audio signal Xs (n) at that time n and the audio signals Xs (n-1),
Xs (n) supplied in the past are supplied. n-2), ..., Xs (nR) and the filter coefficients (tap
coefficients) c0, c1, c2, ..., cR of the number of taps Tr are expressed by the following equation
(8b) An output sequence Yc (n) is generated and output.
Then, the output sequence Yc (n) is converted from digital to analog by the D / A converter 3c,
and then the volume is adjusted by the amplifier 4c and output to the high frequency speaker
SMH side.
The DA converter 3c and the amplifier 4c are also configured to perform digital analog
conversion processing and amplification processing in synchronization with the clock signal of
the operating frequency fc.
[0051]
[0052]
Furthermore, the above equations (8a) and (8b) collectively indicate the case where the audio
signal Xs sampled at any sampling frequency fs of 48 kHz, 88.2 kHz, 96 kHz, 176.4 kHz, and
192 kHz is supplied. Filter coefficients for each sampling frequency fs of 48 kHz, 88.2 kHz, 96
kHz, 176.4 kHz, and 192 kHz so as to perform band pass filtering with the same frequency
characteristic to the audio signal Xs of any sampling frequency fs. c0, c1, c2,..., cR are to be
adjusted.
Although not shown, adjustment means for adjusting the filter coefficients c0, c1, c2,..., CR based
on the data stored in the database is formed in the control unit 20.
[0053]
Next, the FIR filter 2d belonging to the fourth system is formed of an FIR filter whose tap number
Tu shown in FIG. 3D is determined to be U + 1, and from U delay elements indicated by the delay
time .tau.d. , U + 1 multipliers that multiply the tap coefficients d0, d1, d2,..., DU-1 and dU by the
input or output of each delay element, and addition that adds the outputs of all the multipliers
And a device ADd.
09-05-2019
14
[0054]
Here, the number of taps Tu is set to the inverse of the highest sampling frequency (192 kHz)
defined by the above-mentioned standard (that is, sampling period 1/192 kHz) for the delay time
τd of U delay elements. The audio signal being sampled at the sampling frequency fs of 192 kHz
is designed in accordance with the duration of the impulse response that occurs when the highpass filtering is performed in the range of the frequency band of the high-pass speaker SH.
That is, by determining the number of taps Tu so that the tap length is equal to or longer than
the duration of the impulse response, the FIR filter 2 d is designed so as not to be unnecessarily
large.
[0055]
Then, the FIR filter 2d exhibits the transfer function Hd (z) of z conversion notation shown in the
following equation (9a), receives the audio signal Xs output from the signal source 10, and has a
sampling period Ts (that is, 1/1). When an audio signal Xs (n) which is an input sequence is
supplied at time n every fs), the audio signal Xs (n) at that time n and the audio signals Xs (n-1),
Xs (n) supplied in the past are supplied. n-2), ..., Xs (nU) and the filter coefficients (tap
coefficients) d0, d1, d2, ..., dU of the number of taps Tr are expressed by the following equation
(9b) An output sequence Yd (n) is generated and output.
Then, the output sequence Yd (n) is converted from digital to analog by the D / A converter 3 d,
and then the volume is adjusted by the amplifier 4 d and output to the high frequency speaker
SH side.
The DA converter 3d and the amplifier 4d are also configured to perform digital analog
conversion processing and amplification processing in synchronization with the clock signal of
the operating frequency fd.
[0056]
09-05-2019
15
[0057]
Furthermore, the above equations (9a) and (9b) collectively show the case where the audio signal
Xs sampled at any sampling frequency fs of 48 kHz, 88.2 kHz, 96 kHz, 176.4 kHz, and 192 kHz
is supplied. Filter coefficients for each sampling frequency fs of 48 kHz, 88.2 kHz, 96 kHz, 176.4
kHz, and 192 kHz so as to perform band pass filtering with the same frequency characteristic to
the audio signal Xs of any sampling frequency fs. d0, d1, d2,..., dU are to be adjusted.
Although not shown, adjustment means for adjusting the filter coefficients d0, d1, d2,..., DU based
on the data stored in the database is formed in the control unit 20.
[0058]
Next, the operation of the audio signal processing device of this embodiment will be described
with reference to FIG.
[0059]
When an audio signal Xs of 44.1 kHz equal to the reference frequency fbase is output from the
signal source 10, the control unit 20 checks predetermined control data output from the signal
source 10, and the sampling frequency fs of the audio signal Xs is It is determined that it is 44.1
kHz.
Then, the resampling ratio β is set to 1, and the operating frequencies fa, fb, fc and fd are set to
44.1 kHz, which is equal to the reference frequency fbase (in other words, equal to the sampling
frequency fs).
[0060]
As a result, the sample rate converter 1 outputs the audio signal Xs at the sampling frequency fv
equal to the sampling frequency fs without down-converting the audio signal Xs at the sampling
frequency fs.
[0061]
Further, the FIR filters 2a and 2b digitally filter the audio signal Xv at the operating frequencies
09-05-2019
16
fa and fb equal to the reference frequency fbase and the sampling frequency fs to generate
output sequences Ya and Yb, and the DA converters 3a and 3b and The audio signals Xs are
supplied to the speakers SL and SLM via the amplifiers 4a and 4b, and the FIR filters 2c and 2d
are similarly subjected to digital filtering of the audio signal Xs at the operating frequencies fc
and fd equal to the reference frequency fbase and the sampling frequency fs. An output series Yc,
Yd is generated and supplied to the speakers SMH, SH via the DA converters 3c, 3d and the
amplifiers 4c, 4d.
[0062]
Next, when an audio signal Xs having a sampling frequency fs (32 kHz) lower than the reference
frequency fbase is output from the signal source 10, the control unit 20 checks predetermined
control data output from the signal source 10, It is determined that the sampling frequency fs of
the audio signal Xs is 32 kHz.
Then, the resampling ratio β is set to 1, and the operating frequencies fa, fb, fc and fd are set to
32 kHz which is equal to the sampling frequency fs.
[0063]
As a result, the sample rate converter 1 outputs the audio signal Xs at the sampling frequency fv
equal to the sampling frequency fs without down-converting the audio signal Xs at the sampling
frequency fs.
[0064]
Further, the FIR filters 2a and 2b digitally filter the audio signal Xv at the operating frequencies
fa and fb equal to the sampling frequency fs to generate output sequences Ya and Yb, and the DA
converters 3a and 3b and the amplifiers 4a and 4b , And the FIR filters 2c and 2d similarly
generate the output series Yc and Yd by digitally filtering the audio signal Xs at the operating
frequencies fc and fd equal to the sampling frequency fs. , And to the speakers SMH and SH via
the DA converters 3c and 3d and the amplifiers 4c and 4d.
[0065]
Next, when an audio signal Xs having a sampling frequency fs (48 kHz, 88.2 kHz, 96 kHz, 176.4
09-05-2019
17
kHz, or 192 kHz sampling frequency) higher than the reference frequency fbase is output from
the signal source 10, control is performed. The unit 20 examines predetermined control data
output from the signal source 10 to determine the sampling frequency fs of the audio signal Xs.
Then, the resampling ratio β is set to a ratio (fbase / fs), and the operating frequencies fa and fb
are set to 44.1 kHz, which is equal to the reference frequency fbase, and the operating
frequencies fc and fd are set to a frequency equal to the sampling frequency fs. Do.
[0066]
Thereby, the sample rate converter 1 down-converts the audio signal Xs at the sampling
frequency fs with the resampling ratio β, and outputs the audio signal Xv which is sample rate
converted to the sampling frequency fv of 44.1 kHz equal to the reference frequency fbase. .
[0067]
Further, the FIR filters 2a and 2b digitally filter the audio signal Xv at the operating frequencies
fa and fb equal to the reference frequency fbase and the sampling frequency fv to generate
output sequences Ya and Yb, and the DA converters 3a and 3b and The output series Yc and Yd
are supplied to the speakers SL and SLM via the amplifiers 4a and 4b, and the FIR filters 2c and
2d digitally filter the audio signal Xs at the operating frequencies fc and fd equal to the sampling
frequency fs. It is generated and supplied to the speakers SMH and SH through the DA
converters 3c and 3d and the amplifiers 4c and 4d.
[0068]
As described above, according to the audio signal processing device of the present embodiment,
the following effects can be obtained.
[0069]
First, an FIR filter 2a, 2b that determines a specific sampling frequency (44.1 kHz) among a
plurality of sampling frequencies fs as a reference frequency fbase and performs digital filtering
of low and middle low frequencies is equal to the reference frequency fbase It is formed in
advance by an FIR filter of predetermined tap numbers Tp and Tq operating at frequencies fa and
fb.
09-05-2019
18
When the sampling frequency fs of the audio signal Xs output from the signal source 10 is
higher than the reference frequency fbase, the sample rate converter 1 downsamples the audio
signal Xs under the control of the control unit 20 to obtain the reference frequency. An audio
signal Xv at a sampling frequency fv equal to fbase is generated, and the audio signal Xv at the
sampling frequency fv is digitally filtered at operating frequencies fa and fb whose FIR filters 2a
and 2b are equal to a reference frequency fbase.
[0070]
Therefore, in order to process the audio signal Xs sampled at the highest sampling frequency
(192 kHz) among the plurality of sampling frequencies fs, the FIR filters 2a and 2b are preprocessed by an FIR filter with a long tap length (a large number of taps). There is no need to
form it, and it is possible to realize the reduction of the circuit scale of the audio signal
processing device and the reduction of the processing amount.
[0071]
Also, the FIR filters 2a and 2b for performing digital filtering in the low band and the low and
middle band are previously performed with FIR filters of predetermined tap numbers Tp and Tq
operating at operating frequencies fa and fb equal to the reference frequency fbase (44.1 kHz).
Since it is formed, when an audio signal Xs having a sampling frequency fs (32 kHz) lower than
the reference frequency fbase is output from the signal source 10, a sufficient tap length is
secured, and the audio signal Xs having the low sampling frequency fs It is possible to perform
digital filtering even if the sample rate conversion is not performed.
[0072]
In the embodiment described above, the reference frequency fbase is 44.1 kHz, but the reference
frequency fbase is another sampling frequency in the range from the lowest sampling frequency
(32 kHz) to the highest sampling frequency (192 kHz). For example, the frequency is set to 48
kHz, 88.2 kHz or 96 kHz, and the FIR filters 2a and 2b are formed in advance so as to operate at
the reference frequency fbase, and the control unit 20 uses the reference frequency fbase as a
reference. The resampling ratio β and the operating frequencies fa, fb, fc and fd may be set.
[0073]
However, the sampling frequency with the highest frequency of use is determined as the
reference frequency fbase, and the FIR filters 2a and 2b are formed in advance to operate at the
reference frequency fbase, and the control unit 20 uses the reference frequency fbase as a
reference. Setting the resampling ratio β and the operating frequencies fa, fb, fc and fd is
preferable in order to operate the audio signal processing apparatus efficiently.
09-05-2019
19
[0074]
Further, in the audio signal processing device according to the present embodiment described
above, as shown in FIG. 2, four systems of FIR filters 2a, 2b, 2c, and 4 are matched with the four
speakers SL, SLM, SMH, and SH. Although the audio signal processing apparatus according to
this embodiment includes one or a plurality of FIR filters that perform frequency division, sound
quality adjustment, and the like according to the frequency characteristics of one or more
speakers. The present invention can also be applied to an audio signal processing apparatus
provided.
[0075]
For example, as a modification, in the case of digitally filtering an audio signal supplied to one
speaker (full range speaker) having the entire audio band as a frequency band, for example, only
in the second system including the FIR filter 2b shown in FIG. It is assumed that digital filtering is
performed, and continuation of an impulse response that occurs when an audio signal sampled at
a sampling frequency fs equal to a predetermined reference frequency fbase is subjected to band
pass filtering over the entire audio band. Design according to the time.
When the audio signal Xs sampled at the sampling frequency fs higher than the reference
frequency fbase is output from the signal source 10, the audio signal Xs is downed to the audio
signal Xv at the reference frequency fbase by the sample rate converter 1. The conversion may
be performed, and the FIR filter 2b may be configured to digitally filter the audio signal Xv at the
operating frequency fb equal to the reference frequency fbase.
[0076]
It is a block diagram showing the composition of the conventional audio signal processing device.
It is a block diagram showing composition of an audio signal processing device concerning an
embodiment.
09-05-2019
20
It is a block diagram showing the basic composition of the FIR filter shown in FIG.
Explanation of sign
[0077]
DESCRIPTION OF SYMBOLS 10 ... Signal source 20 ... Control part 1 ... Sample rate converter 2a,
2b ... FIR filter SL, SLM ... Speaker
09-05-2019
21
Документ
Категория
Без категории
Просмотров
0
Размер файла
34 Кб
Теги
jp2008109279
1/--страниц
Пожаловаться на содержимое документа