close

Вход

Забыли?

вход по аккаунту

?

DESCRIPTION JP2006114998

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 JP2006114998
PROBLEM TO BE SOLVED: To provide a "digital signal processing system" capable of realizing a
soft clip excellent in arithmetic processing and small in distortion. A digital signal processing
method according to the present invention uses an input digital signal x up to a preset threshold
value P, where x is an input digital signal, g is a gain coefficient, and X is an output digital signal.
The linear processing of multiplying the gain coefficient g is performed, and in the region
exceeding the threshold value P, the nonlinear processing is performed by the order according to
the gain coefficient g. This produces a soft-clipped output digital signal X with low distortion.
[Selected figure] Figure 4
Digital signal processing system
[0001]
The present invention relates to digital signal processing systems, and more particularly to
processing such as gain adjustment of digital audio signals in digital amplifiers.
[0002]
2. Description of the Related Art With the development of semiconductor technology and
digitization technology, digital signal processing is widely performed in audio devices, navigation
devices, home appliance devices, and the like.
In the case of analog signal processing, although there is a problem of variations due to circuit
10-04-2019
1
characteristics and temperature characteristics, digital signal processing has excellent
reproducibility, but on the other hand, if high-precision processing is to be realized, Accordingly,
the amount of data to be processed becomes enormous. However, semiconductor processing
devices such as DSPs are capable of processing a large amount of data in real time, and it is
becoming possible to obtain an ideal analog signal waveform by digital signal processing.
[0003]
The applicant has disclosed a technique relating to a digital variable compressor of an input
audio signal as shown, for example, in Patent Document 1. This performs the following equation
(3) (hereinafter referred to as conventional method (3)) when x is an input audio signal, adds a
compression characteristic to the output audio signal X, and varies the n value Change the
compression characteristics. Here, the exponent value n is an integer that determines the
compression ratio, and | x | is the absolute value of the input signal x.
[0004]
The gain of the output digital signal X is adjusted by performing the operation shown in the
conventional method (3) on the input digital signal x, and a soft clip is realized without causing
overflow for all input signal levels. it can. Moreover, by executing the processing of the
conventional method (3) by the DSP, an output digital signal can be obtained in real time.
[0005]
Patent No. 3056815
[0006]
In the audio amplifier, when a large digital audio signal that causes an output exceeding the
maximum amplitude value is input, a portion exceeding the maximum amplitude of the output
signal is held (clipped) as it is, and distortion occurs in the output waveform. In addition, it
sounds like the noise appeared in hearing.
FIG. 8 shows an example in which the output signal is clipped in the portion exceeding the
10-04-2019
2
maximum amplitude value (maxlevel).
[0007]
In order to cope with such a problem, there has been proposed a process called soft clip, which is
a process of gently clipping the output signal so as not to overflow. Soft clipping smoothes the
waveform so that its output signal converges to the maximum amplitude value, for example, by
nonlinear processing of the input digital signal, but the calculation becomes very complicated
when the order of nonlinear processing becomes higher order It will In particular, it becomes
difficult to calculate a point (threshold value) at which non-linear processing should be started.
Further, there is a problem that when the threshold value is changed by gain, the sound quality
of the audio output is changed by gain.
[0008]
The above conventional method (3) does not propose a technique for solving the problems as
described above. Furthermore, when the signal processing of the conventional method (3) is
applied to a digital amplifier and the gain by the power supply voltage of the digital amplifier is G
[dB], the exponent value n is 1, 2, 3, 4, ... When variable, the gain of the output signal becomes G,
G + 6, G + 9, G + 12, ... [dB]. When the exponent value n is 2 or more, distortion is included in the
gain. For example, when the exponent value n is 2, the distortion component of x <2> is
superimposed on the output signal X, and when the exponent value n is 3, in addition to this, the
distortion component of x <3> is It is superimposed. This further degrades the distortion of the
output signal even in the reproduction state where the input signal level is low. The degree of
deterioration of the distortion rate changes depending on the set gain value, that is, the index
value n, but the distortion value is increased by about one digit.
[0009]
An object of the present invention is to solve the above-mentioned conventional problems, and to
provide a digital signal processing system which is excellent in arithmetic processing and can
realize a soft clip with small distortion. Another object of the present invention is to provide a
digital signal processing system capable of obtaining high-quality audio output. Another object of
the present invention is to provide a digital processing method capable of reducing distortion
more than the processing according to the conventional method (3).
10-04-2019
3
[0010]
In the digital signal processing method according to the present invention, when the input digital
signal is x, the gain coefficient is g, and the output digital signal is X, the calculation according to
equation (1) is performed up to a preset threshold value P. When the threshold value P is
exceeded, the calculation according to the equation (2) is performed. That is, if the input digital
signal x is less than or equal to the threshold value P, linear processing is performed by
multiplying the input digital signal x by the gain coefficient g, and in the region exceeding the
threshold value p, the order n corresponding to the gain coefficient g is Perform nonlinear
processing and soft clip to the set maximum amplitude value.
[0011]
According to the digital signal processing method of the present invention, the input digital
signal x is subjected to the linear processing according to the equation (1) up to the threshold P,
and the nonlinear processing according to the equation (2) when the input signal level is larger
than the threshold P Since the clip processing is performed, in the range where the input signal
level is low, good distortion performance can be obtained by complete linear processing, and in
the range where the input signal level is high, soft clipping can be performed without overflow. It
can be realized.
[0012]
Furthermore, since the threshold value P, which is the boundary at which linear processing and
non-linear processing should be performed on the input digital signal x, is set in advance,
arithmetic processing of the digital signal can be efficiently performed.
Also, by setting the threshold value P in advance, it is possible to make the point to be soft
clipped constant, and it is possible to stabilize the sound quality of the audio output against the
change in the gain. Furthermore, in the case of the conventional method (3), since the step width
is determined by the exponent value n where only positive real numbers are allowed, gain values
can also be obtained only discrete values such as G + 6, G + 9, G + 12. Although not present, any
gain value can be obtained in the present invention.
10-04-2019
4
[0013]
Hereinafter, embodiments of the present invention will be described in detail. The processing of
digital signals according to the present invention is preferably performed using a DSP (digital
signal processor) capable of high-speed operation including a multiplier or the like.
[0014]
FIG. 1 is a block diagram showing the configuration of a DSP that performs digital signal
processing according to the present embodiment. The DSP 10 has a data space and a program
space, and the data space is configured by an I / O port unit 22 connected to the data bus 20, a
multiplier / ALU unit 24, and a data memory 26. The program space is configured by the control
unit 32 and the program memory 34 connected to the program bus 30.
[0015]
The program memory 34 stores a program for executing arithmetic processing related to a soft
clip on an input digital signal. The control unit 32 controls the multiplier / ALU unit 24, the data
memory 26, the I / O port unit 22 and the like according to the program. The data memory 26
stores the gain coefficient g, the threshold value P, and the maximum amplitude value (maxlevel)
for the output signal. The gain coefficient g, the threshold value P, and the maximum amplitude
value can be set arbitrarily, and may be set by the user, for example.
[0016]
The gain factor g is a factor by which the input digital signal is multiplied. The threshold value P
is a boundary between linear processing and non-linear processing of the input digital signal. The
linear processing algorithm is executed for the input digital signal less than the threshold P and
the threshold value P is A non-linear algorithm is performed for larger input digital signals. Also,
the maximum amplitude value is the maximum amplitude level of the output signal, and the
output signal will be clipped at the maximum amplitude level. Preferably, the threshold value P is
set to the maximum output value (maxlevel) / 2 so that non-linear processing is performed from
half (-6 dB) of the maximum amplitude value.
10-04-2019
5
[0017]
The DSP 10 performs arithmetic processing on the digital signal x input to the I / O port 22 in
the multiplier / ALU unit 24 according to the following equations (1) and (2), and outputs digital
signal X from the I / O port 22 Output as
[0018]
FIG. 2 is a functional block for explaining the operations of the equations (1) and (2), and FIG. 3
is a diagram showing an operation flowchart thereof.
When the input digital signal x is input to the I / O port 22 of the DSP 10 (step S101), the
processing determination unit 50 performs gain coefficient g, threshold value P, and maximum
amplitude value, which are set in advance, as gain coefficients. Reading is performed from the
holding unit 52, the threshold value holding unit 54, and the maximum amplitude value holding
unit 56 (step S102). A predetermined address area of the data memory 26 is allocated to these
holding units.
[0019]
Next, the processing determination unit 50 compares the input digital signal x with the threshold
value P (step S103). If the input digital signal x is less than or equal to the threshold value P, the
linear processing unit 58 can be operated. Give an instruction 50a to indicate. Thereby, the linear
processing unit 58 performs linear processing of multiplying the input digital signal x by the gain
coefficient g according to the above equation (1) (step S104).
[0020]
When the input digital signal x is larger than the threshold value P, the processing determination
unit 50 gives the non-linear processing unit 60 an instruction 50b indicating that the operation
is possible. When the gain coefficient g is smaller than 2 (1 <g <2), the non-linear processing unit
60 performs non-linear processing with the exponent value n = 2 in equation (2). When the gain
coefficient g is 2 or more, Non-linear processing of the exponent value n according to the gain
coefficient is performed (step S105). The above-described calculation by linear and non-linear
10-04-2019
6
processing is executed by the DSP 10 in substantially real time, and the soft clipped digital signal
X is output from the I / O port 22 (step S106).
[0021]
The processing results when the maximum amplitude value (maxlevel) = 10 <−15/20> and the
threshold value P = maximum amplitude value / 2 by the above digital signal processing are
shown in FIG. 4 to FIG. 4 shows gain coefficient g = 10 <-14/20>, exponent value n = 1, FIG. 5
shows gain coefficient g = 10 <-11/20>, exponent value n = 2, FIG. 6 shows gain coefficient g =
10-5 / 20, and an index value n = 3. The vertical axis represents the amplitude level of the output
signal obtained when the output digital signal is converted into an analog signal, and the
horizontal axis represents the time axis.
[0022]
For example, referring to FIG. 4, a threshold P is a boundary, a threshold P or less is an output
waveform by linear processing, a portion exceeding the threshold P is an output waveform by
non-linear processing, and this waveform is It is soft clipped from the threshold value P to the
maximum amplitude value. The threshold P, which is the boundary between linear processing
and non-linear processing, is connected almost continuously. Further, as apparent from FIGS. 5
and 6, as the index value n increases, the connection of the boundary points of the threshold P
becomes smoother.
[0023]
As described above, when the level lower than the threshold P is low, the processing according to
the equation (1) is completely linear, so distortion of the output signal becomes extremely small.
By the non-linear processing according to (2), a smooth soft clip waveform with less distortion
can be realized. By setting the threshold value P and the maximum amplitude value to arbitrary
values, and thereby giving a constant to the equation (2), it is possible to extremely easily carry
out the calculation of the high-order non-linear processing. Also, by setting the threshold value P
constant in relation to the maximum amplitude value, it is possible to make the level at which
soft clipping is started constant regardless of the change in gain coefficient. Furthermore, in the
case of the processing operation according to the conventional method (3), only the positive
integer is allowed as the exponent value n, so the gain of the output signal is limited to the step
10-04-2019
7
width of G + 6, G + 9, G + 12. However, in this embodiment, it is also possible to obtain an
arbitrary gain in between.
[0024]
FIG. 7 is a block diagram showing an example in which digital signal processing according to this
embodiment is applied to a digital amplifier. The digital amplifier 100 includes a sound volume
adjustment unit 110, a gain adjustment unit 120, a gain coefficient setting unit 130, a PWM unit
140, and a drive unit 150. An m-bit digital audio signal x optically read from a recording medium
such as a CD, a DVD, etc. and A / D converted is input to the volume adjustment unit 110, where
the digital audio signal according to the volume instruction from the user Can be changed to
[0025]
The gain adjustment unit 120 is a part that executes digital signal processing of Equations (1)
and (2) according to the above-described embodiment. The coefficient setting unit 130 holds the
gain coefficient g, the threshold value P, and the maximum amplitude value. The gain coefficient
g can be changed according to an instruction from the user, or the gain coefficient g can be
automatically changed according to the external noise level. Although the threshold value P and
the maximum amplitude value can be set by the user, respectively, only one of them can be set if
a fixed relationship is established between the threshold value P and the maximum amplitude
value. May be set. The PWM unit 140 generates a PWM signal of a duty ratio according to the
digital signal, and the drive unit 150 performs a switching operation of selectively driving the
speaker according to the PWM signal.
[0026]
The gain adjustment unit 120 can adjust the gain in accordance with the level of the audio digital
signal input from the volume adjustment unit 110. In particular, when the volume level is a small
signal, the gain adjustment unit 120 can perform linear processing of the input signal, so the
distortion can be suppressed to a very small level. Furthermore, by setting the threshold value P,
even when the gain coefficient g is different, the soft clip can be operated from a certain level,
and it is possible to suppress such a defect that the timbre differs for each gain. . As described
above, by applying the digital signal processing method according to the present embodiment to
the gain adjustment unit 120 of the digital amplifier, it is possible to obtain an excellent audio
10-04-2019
8
output signal close to the analog method.
[0027]
Although the preferred embodiments of the present invention have been described above in
detail, the present invention is not limited to the specific embodiments, and various modifications
may be made within the scope of the present invention as set forth in the appended claims.
Modifications and changes are possible.
[0028]
The digital signal processing system according to the present invention is applicable to the
processing of any digital signal.
It can be used in digital devices such as audio devices, home electronics devices, navigation
devices, etc. that perform digital signal processing.
[0029]
FIG. 5 is a block diagram of a DSP for performing digital signal processing according to an
embodiment of the present invention. 3 is a functional block of digital signal processing
according to the present embodiment. It is a figure which shows the operation | movement flow
of the digital signal processing system which concerns on a present Example. It is a figure which
shows the example of an output signal waveform by the digital signal processing which concerns
on a present Example. It is a figure which shows the example of an output signal waveform by
the digital signal processing which concerns on a present Example. It is a figure which shows the
example of an output signal waveform by the digital signal processing which concerns on a
present Example. It is a block diagram of a digital amplifier to which this embodiment is applied.
It is a wave form diagram showing the soft clip in an audio amplifier.
Explanation of sign
[0030]
10: DSP 50: processing determination unit 52: gain coefficient holding unit 54: threshold value
10-04-2019
9
holding unit 56: maximum amplitude value holding unit 58: linear processing unit 60: non-linear
processing unit
10-04-2019
10
Документ
Категория
Без категории
Просмотров
0
Размер файла
19 Кб
Теги
jp2006114998, description
1/--страниц
Пожаловаться на содержимое документа