JPS5947848A - Reducing device of impulsive noise - Google Patents

Abstract

PURPOSE:To interpolate the missing of a signal produced during the period of impulsive noise, by connecting the 1st and the 2nd de-emphasis circuits in cascade, and providing a signal correcting circuit between the both. CONSTITUTION:The 1st and the 2nd de-emphasis circuits DE1, DE2 having a characteristic as shown in the fig. are provided in cascade in a transmission line of an audio signal, and the overall characteristic of the transmission line is the addition of both characteristics. Thus, the audio signal having a flat frequency characteristic is obtained as to the frequency band being the objective of reproduction, by passing the audio signal through the 1st de-emphasis circuit DE1. Then, the linear interpolation to the missing part of a desired signal in a signal correcting circuit 5 or the like is performed in excellent way, since the signal component in the frequency band being the objective of reproduction has a flat frequency characteristic.

Description

[Detailed description of the invention]

(Industrial Application Field) The present invention is directed to preventing external interference from entering the audio signal system in audio equipment, audio receivers, television receivers, video disc players, etc. The present invention relates to an apparatus for reducing pulsed noise (a) in which the reduction of pulsed noise (σ) is performed perceptibly well. (Prior art) Telephone equipment or electronic equipment having an audio signal system
C If pulse noise, such as the ignition noise of a car or other electrical equipment, is mixed into the audio signal system of any type of equipment, the 5-o-1 audio signal will be affected. It is well known that quality deteriorates. Conventionally, as a means to reduce the deterioration in audio signal quality caused by the contamination of pulse noise, there are two methods: (a) reducing the gain of the signal transmission system during the period in which pulse noise occurs; , or a method that attempts to reduce pulse noise by cutting off the signal transmission system (lower the gain by one generation to zero...employing a squelch circuit). The signal level of the Kero signal is maintained at the signal level immediately before the period of puffy noise σ) to reduce pulse noise. Means and practical use are the first, but these σ)(Il, (
In the 1,100th period of (C)), there is a drawback that there is a signal dropout during the period of (k). Yo, everywhere is wholesale Masa υ)
One problem is that a sufficient one-area effect cannot be obtained. By the way, the period of Zonmen Shin's V Kogyushi=I・I1−, to interpolate the missing month, the blog conversion'rl to Ajdal J%
” After converting p and f, the signal corresponding to the missing part of the signal is converted to 1'1 by applying the linear prediction method, and the complementary IL signal is used to interpolate the signal during the noise period. Although it may not be possible to use some digital/equipment/[i], when implementing it, it may be necessary to
The use of a 1'b circuit is required.
:The U/C solution has not been applied to general audio equipment. Now, as mentioned above, 1-te Z)/
・In the case of −y, x when the luscious noise is reduced,... a loss of reliability θ) occurs corresponding to the period of existence of the lusious '+ WG sound. The problem is that audio quality of good quality cannot be obtained depending on the region.
The first part of solving the problem mentioned above is B.
The fact that a complex and expensive circuit is required to interpolate the missing portion of the bar makes it difficult to apply to general audio equipment. In order to solve the two conventional problems, the applicant's company first uses a differentiating circuit, a Zamplepoort circuit, and information on the slope of the desired signal during a period in which pulse noise occurs in the input audio signal. By supplying a control signal with a certain degree of reliability, an operation for removing pulse noise in the input audio signal and a linear interpolation operation for the desired signal during the period in which pulse noise is occurring will be performed. An anamig circuit with a simple TC circuit configuration consisting of one signal correction circuit, etc., is used to create a correction signal that can interpolate the missing part of the signal during the period when pulse noise is occurring, thereby producing a signal with good quality. We proposed a device for reducing pulse noise so that audio signals can be obtained. Figure 1 shows one proposal already made at the previous station θ) B/Lノ, +/1 scrap 1γ
? θ)1 This is a block diagram of the internal organs apparatus. is a delay circuit, C
The SG is based on the Lusty Noise Detection Circuit 3 and the Pulse Shaping Circuit 4. 5. A controllable generation circuit configured with 5. 5 This control signal generation circuit C8G outputs an input 4-dio signal S.
, a control No. 12 S2 of pulse rlJ corresponding to the period (a) and (b) of the pulse noise mixed in is generated. Control (S Rock Generation Circuit C8G i'Coage 2) As the pulse miscellaneous H6 detection circuit 3 and the pulse shaping circuit 4, an appropriate one may be selected from well-known configurations. By the way, the control signal S2 generated from the control signal generation circuit C8G correctly corresponds to the position of the pulse noise mixed in the input audio signal during the 11.1 axis. Necessity I'1.7. , but in the control 1 Jingei generation circuit C8G, human audio (V in L8%)
The pulse noise mixed in is detected, and the pulse l corresponding to the period in which the pulse noise exists is detected accordingly.
] Before the control signal S2 is generated, there is a predetermined time delay determined depending on the operating characteristics of the pulse noise detection circuit 3 used. delaying the human-powered audio signal supplied to the input terminal tK by a delay circuit 2 having a delay time approximately equal to the time difference between the pulsed noise and the one-tone signal generated corresponding to the pulsed noise; The above control (Signal 52VCJ) ensures that various signal processing to be performed is performed correctly at the position where pulse noise exists in the input audio signal.The input audio signal S shown by a in FIG. is the input audio signal Sl which has been given the required time delay by the delay circuit 2, and is the pulse noise mixed in the human audio conversion II shown in a of FIG. Existence location and b in Figure 2
It correctly matches the position on the time axis of the control signal -1-%S2 shown in 2). In addition, in Fig. 2, for the input audio Shingei, time 1
．． → t2, time t3 → t, time t, → t, pulse noise N, , N2. The case where N3 is mixed is also exemplified as θ). In FIG. 1, an input audio signal outputted from the delay circuit 2 is supplied to an input terminal 5a of a signal correction circuit 5. The signal correction circuit 5 has a relatively heavy configuration as the third example.
The circuit in 7U block 5 in the figure is as shown in FIG. s5 (e in Figure 2).
) is connected to the terminal 5d&, so that the output terminal 5b receives the signals S12 and Z/C as shown in C in FIG.
In other words, human power audio signal - Pulse M in @S1 -
B't is removed, and the desired signal during the period in which the pulse noise occurs is linearly interpolated.2. A state output signal S, is sent out. Details of the signal correction circuit 5 will be described later with reference to FIG. Output signal from the above-mentioned @month correction circuit 5-p:4 S...
is outputted to the output terminal 8 of the device and supplied to the differentiating circuit 6 along with 2). The differentiating circuit 6 differentiates the signal 3'+483 shown in σ)C in FIG. 2 and outputs a differential signal S as shown in d in FIG. The above-mentioned [-1 difference signal S4 is -90 with respect to the original signal (desired (6)), the output signal 83 from the confidence correction circuit 5, etc.
In addition to showing the phase difference of degrees, the signal 83 is also linearly interpolated with -C.
In issue 7. During this period) when there is noise, J! It is said that a signal section indicating the slope of ≦rock corresponds to a certain value of 1-j, , -, -, and is said to be 5/:c. Then, the signal f in the differential signal S4 is constant.
The monk's bell in the signal section indicating Rehel becomes a positive signal level depending on the direction of the gradient l in the original signal, and it becomes 1 -
Alternatively, the polarity can be changed depending on the direction of the slope of the original signal, such that the signal level becomes negative, and the polarity is changed depending on the direction of the slope of the original signal.
A certain level of correspondence in 4th grade. The magnitude of the difference between the reliability I/hell and the sea level of the signal section indicating 1 is changing.
. The differential circuit 6 outputs f and the differential is 1°13 chi, , +z
8. +t, sample pole 1 circuit 't II <-assembly 7j, phosphor/' 7L from Holt circuit 7, Figure 2 υ
) e VIE /J, output Mooi1□S÷S5. 0) Information-""r s5 is device 11 power bow L'
r:'; When operating in state 1α, it is written as rfD (U k; same as S i). Zanonolpo-1L
1. The circuit 7 enters operation in the device power bow J-frame state 2)
Waiting time 2, 1... Gei) operation θ) f komu (zi - undecided. Said 1. Munirino Norvo) The/Nori/Gharno for the 111 circuit, and (-te(゛1- Control) Generated by the trust bone generation circuit C8G and controlled by T (Ilt: signal s
2 can stay 4 times a month. The above-mentioned sample port circuit 7 outputs the differential signal S, and the differential signal S, &,1.・It shows a certain signal rail.B) It has a signal section that shows a group of monks and a bell that corresponds to the signal section, and has already been mentioned σ)
As shown above, since the signal section showing a constant confidence level in the differential signal 84M shown above indicates the slope information of the original confidence (hope confidence), -υ nonorhole [・
The output signal S,,4+,n'l from the circuit 7 contains the slope information of the desired signal depending on the signal section indicated by (14-"l/・-l). -Signal s5 output from Sononorhole 1 circuit 7,-
J-1'ch, signal S having slope information of desired temperature generation is supplied to terminal 5d Ic9 (supplied) of correction circuit 5;
In the signal correction circuit 5, No. 45 8. Based on the slope information of the desired signal, the missing part of the signal is linearly interpolated to compensate for the input audio. Make art,
Then, linear interpolation of the missing portion of the signal θ) is performed by Masanobu, and a signal S as shown in FIG. 2 is sent to the output terminal 8. Next, referring to FIG.
14 configuration and operation will be explained. In Figure 3,
No. III circuit 6 is a differentiating circuit 6, and Fundenza C (
It is composed of I, a resistive surface, an amplifier A, and an IJ circuit 7.
It consists of switch SWs, condenser C8, and amplifier A.2) 7a is Sanono°
Control (i, 8” s2) given as a ring pulse
Human power immunity cA Barry → Non-pulling pole [・Circuit 7
is controlled" 52(1)/, switch 5Wsi+r, 1no)-1ch-C-II during the period of IL-\le:. Denoza Cs [switch SWs /+; (,/,'Musa J1 ], 1 white front θ) 13 Maintain the rail 72).
l'<I Ill &im, + 5a is human power audio 4,;+Hθ) human power is low 11.1, 5)) is the output terminal, 5c is control iD'+I signal j b2υ) 1''I, 1
f (j end; r, 5d +: at the supply terminal of L C'i No. S), S, i muni, A, is the [υ)th amplification 2: A2 is the second amplifier, - c-1st. )) J! ,'Q
°Width 1 Xi; A, is IJ (output impedance, 1st 2nd θ) amplifier A2 is sieved Kainobi-9
';/' 7. (The name of the story). The output side of the first amplifier A and the second amplifier) Amplification 2:;
σ) The control transmission line between the input end and the control tll signal V,
When s2 is in the state of ・rerail, the state of ``pregnancy'' is h.
A 1-f notch SW is provided, ;F-5 second σ
) There is a charge accumulation hole between the input side of A2 in amplifier 2 and the ground.
,・Denault 1) C is installed, and the second amplification 2 is equipped with a variable constant current circuit vcσ) on the human power side of A2.
The output side is connected. In the example shown in FIG. 3, the variable constant current circuit VC includes a phase inverting amplifier 8 with a gain of -], a transistor X whose emitter is connected to the positive power supply VDC via a resistor R1, , a transistor X whose collector is connected to the collectors of the above-mentioned transistors, a resistor I connected between the emitter of the above-mentioned transistor
Series connection of resistor R3, variable resistor VR, and resistor R4 connected between C and negative power supply -VDC l! 11
The base of the transistor X is connected to the connection point between the resistor and the variable resistor VR;
fy, :l - The helix of lance X2 is connected to the connection point between resistor R4 and variable resistor (2). The variable resistor■ is a single-use device that corrects imbalances in the circuit due to variations in the characteristics of the circuit components, and if the circuit can be balanced correctly, it may be possible to replace it with two fixed resistors. can. The variable constant current circuit VC has a voltage of zero σ at its terminal 5d.
), when the Z point σ) voltage is BIU and BUI), the operation is performed under the standard operating condition of C: 11;
i'f1, Jl- or 1 [When polarity, the voltage at the two points becomes the same positive polarity voltage as the voltage at terminal 5d, and
i'5dpi)''i11.When the polarity is negative, the voltage at the two points is the same as the voltage at the terminal 5d, and the voltage is negative.
. Therefore, the variable constant current circuit Cσ) Z point VC is given to the terminal 5d.
16 No. 1-.-, a voltage having a polarity and a voltage value of T corresponds to the polarity of the signal σ in the interval (;Ii) and the magnitude of σ) appears, and from 4), the above 1 If you connect VCJ node/the C between two points and ground, that point/the C
is 4r; ” S5 vc (-) is a signal section that shows a certain reliability level): has the same polarity as the negative of the signal S, and is a signal section that shows a signal level of 1. θ) It is charged with a constant charging current determined according to the 1J rising rail and the 1J rising rail. In the state where 4 is not mixed in, there are four main states of control supplied to the terminal 5c: 8□ or 1, and the switch SW is turned off. The first amplifier A is supplied to the input terminal 5a.
,→switch SW→second amplifier A2→output terminal 5b, and passes through the signal transmission path from input terminal 5a to output terminal 5b.
transmitted to. At this time, the charge storage cable/te/ther C connected between the signal transmission path and the ground is

[71] The terminal voltage value is shown according to the voltage value of the signal transmitted to the signal transmission line. In addition, in the above state in which the input audio signal S is not mixed with Parno noise, the output terminal of the variable constant current circuit VC is connected to the output terminal of approximately zea ohms via the switch SW which is in the OFF state. The first one has extremely low output impedance as
The amplifier A1σ) is connected to the output side of the amplifier A1σ) and is connected to the variable constant current circuit VC from A) to the variable constant current circuit VC via the terminal 5d.
If a current is generated at the output side of the first amplifier A1 having a low output impedance of about 7C, and is output from a variable constant current circuit with a high output impedance, the voltage generated is not much. ;C is smaller than σ) and 1ε, so the current generated by the variable constant current circuit VC or C- is from the first amplifier A1 to the second σ1'1 width 3: 4 A,,'--(li16.0) 4'4, there is no way to give a hindrance to the desired signal to be transmitted. It's a signal section σ indicating the signal 1 novel.゛na1l)1
There is no such thing as 1). (2) The output 4.tH8゜ can be supplied as is to the constant current circuit VC. . Next, the input audio from the other station S1 is mixed with russive noise, and in one case, it is pulsed 'AV? , ; The control signal S2 is issued in correspondence with the period in which NI-N3 is occurring, and the 1, 1 and 2 SWs are turned off during the period in which the control signal S2 is at a high level. Said 1-f 7, Inona S
When W is set to A, the terminal electric terminal J is turned off and the above-mentioned switch SW is turned off. It is held as 0+ and ・le when the belief is (when). In addition, in this state, the terminal 5d of the variable constant current circuit VC is connected to the signal S. The variable constant current circuit VC has a polarity C of a signal S given to a terminal 5d.
Accordingly, a 11L current is output from the IX pole (/1, and a constant current value σ corresponding to the signal at the output terminal), thereby charging the charge storage 1-1 node → node C. . And A
The charging operation for charge storage C/denosa is performed during the period when pulse noise is occurring, and the voltage at the terminal C increases linearly. However, at the moment when the pulse noise is no longer mixed in, the control signal S2 becomes low level and the switch sw is turned on. It is instantly discharged by the output impedance. The variable current circuit VC is supplied with a signal to the terminal 5d,
- That is, it is driven by a signal S that has the slope information of the desired signal at a polarity and a constant confidence level, so that T
A current of this polarity and a constant current value is caused to flow into the two-inch capacitor C for charge storage, and the terminal voltage of the capacitor C is changed to the signal four.
S, ni(1i, -sho(ha1.1si1shiruro small signal section iX,; station pole +11 i1J"r-!l/,
LE f'c When the voltage at the terminals of C is increased, the inflow of current from the current circuit VC increases. / Denori Cθ) 4A voltage is when the switch SW is in the 8 position.
F41. / =71. Therefore, human power λ(i
; -Contamination with F48H (, Muni,.../LS noise period and Z) (-VC'-l in the issue: U fko1,1
It is clear that the missing lip θ) is well linearly interpolated by the signal correction circuit 5θ) above, and the signal 81. Haharai 1
No. 4 & C near (IJ, those with l-r waveforms). In Fig. 2, f is a correction signal for linear interpolation in the compensation circuit 5, which outputs the correction signal for linear interpolation as a solid line, and produces a state in which there is no linear noise. ) The desired +4 waveform is shown in the dotted line in Figure 2. θ)f is an explanatory diagram that makes it easier to understand the operation. In the actual θ) operation, the signal is corrected. The circuit 5 outputs a signal S3 such that θ" c f/C is smaller than that shown in FIG. In the KL= reduction device, the signal S3 is passed through the differentiating circuit 0 (
From the differential circuit 6, the output signal S, etc. At the same time,
Generate a differential (F signal S,
According to the slope jFt information ic of the desired emotion included in the above differential differentiation of f-j8s4, the desired E number σ' + i;
M-line interpolation is performed] 1-a) 2) This already proposed θ) pulse noise reduction device is mixed into the human-powered audio system, and 1-a) pulse noise is t and c are also set to σ) so as to indicate a significantly short existence period compared to the period of the desired signal. Linear interpolation for the missing part of the cow-jiro signal in 111 for existence le is done well/:
However, it is not possible to effectively perform the pulsed 4L sound in the 6/C condition, which makes the auditory sense unnatural. 4 -J-'a) If the period of pulse noise is of IC length that is negligible with respect to the period of the audio signal,...Luno, 1/1 The problem is that the signal loss 'fil' between the sound A and I and the target line drawing is performed well and becomes tc, and the problem is that the above The problem is that Calus 1'1-%
(Kat θ) Even if the 4 periods are the same, the pulse has 1
The frequency of Kibo Shinga in the part where one sound is mixed is lil!
1 (iI:f here &C, pulse noise σ) existence period Ib desired A1; +4 period jυ1/11- is relatively long A) That means C, desired signal V) A) For the frequency components of the crystal region, make it so that it goes to the line where 9f7-, +- linear complement 111 is done.
4). Figure 4 is Figure 1 (Figure 3) UC small tl, tii)
The previously proposed pulse-like noise θ) surrounding equipment (j) does not explain the problem 1). A, N force, residence (;-j It is a waveform diagram of a signal showing the operation of the proposed device,
A in Figure 4 is in the human power audio trust, so! 1σ)1.
A balloon showing a period of existence close to /4 times jul&c, (
This is an example of the waveform of the human audio signal SIo'N in the state where L: I input signal is mixed in. Also, b in Fig. 4 shows the waveform of the input audio signal S, and the pulse noise N mixed in it.
The rut control signal S generated from the control signal generation circuit C8G in the circuit arrangement shown in FIG. 1 (FIG. 3) corresponds to the period of
2, and C in FIG. 4 is the differential signal S4 output from the differentiating circuit 6 in the circuit arrangement shown in FIG.・This is a waveform diagram that commonly shows the signal S output from the circuit 7, and d in FIG. 4 shows the signal S having the waveform as shown in C in FIG. FIG. 1 is a waveform diagram of the output signal S3 sent from the terminal 5b of the signal correction circuit 5 to the output terminal 8vc when supplied to the terminal 5d of the signal correction circuit 5 in the circuit arrangement shown in FIG. As is clear from the waveform diagram shown at a = d in FIG. , the period of pulse noise mixed into the input audio signal may be nearly 1/4 of the period of the audio signal.
A) and... 74 dimensions of the lack of monks in the period of Rusu matter 'Ar' M, interpolation 0) state? Please refer to Figure 2 for the fact that 11 shown for reference in d in Figure 4 will not be as shown in the dotted line, but will be as shown in the solid line in d in Figure 4. This is clear from the description of the circuit arrangement shown in FIG. 1 (FIG. 3) and the circuit operation of the previously proposed device. In other words, in a circuit layout pattern (i"<0.) in FIG. 1 (FIG. 3), the L input audio (u S, If the desired symbol is mixed with pulse noise N having an existence period of 4 to 4 periods V (from is the peak position of (6 or C outputs from the differential circuit at the Kth point where the slope at the time l extraction of the H number is 0)) 1. The period σ) of the pulsed noise in the output signal S is defined by cVC in FIG.
[0, therefore, the signal correction circuit 5
If slope information is not given to G, interpolation G for signal loss in the part of the input audio signal S1 mixed with noculus noise N, as shown by the solid line in d of FIG. The input audio (SO8°σ) signal level at the beginning of the signal is maintained for 1/4 period of the desired signal. Input side - be mixed into the audio signal (even if the start of the period of existence of the pulse noise is at the peak position of the desired signal, the period of existence of the pulse noise is compared to the v4 period of the desired other layer) - If the signal level of the desired signal during that period is kept at the peak value of the desired signal, this state will be difficult for the missing part of the signal during the period of response noise. The effective tc linear interpolation is σ), but if the existence period of the 5-pair noise N is long as shown in FIG. 4a, as described above, the first The previously proposed system has the configuration shown in the figure (Fig. 3).
The pulseless noise reduction device cannot perform good linear interpolation for the missing portion of the signal σ) during the pulse noise σ) period. 1. In explaining the problem with reference to FIG. 4, it is assumed that pulsed kl white t is mixed into the L portion from the peak position of the desired signal to approximately 1/4 period of the desired 15 months. Taking as an example the case where L is mixed with the desired signal 1, 4 t, c, the mixing position of the pulse noise mixed in with the desired signal 1, 4 t, c is different from the above on the waveform of the desired signal, 7. However, hope 4. If the interval is long, the slope information contained in the differential signal output from the differentiating circuit 6 - station S may be insufficient 7ic or 7c. , the presence of pulse noise in the desired signal 4411 VC (a) In the missing portion of the signal, the linear interpolation of the signal is in an inappropriate tc state. The present applicant's association U has proposed a device for reducing pulse characteristics that can solve the above-mentioned problems in the previously proposed device for reducing pulse nature M￥f shown in FIG. In FIG. 5, the same components as those in the pulse noise reduction device shown in FIG.
The same drawing symbols as those used in the figures are attached [, 2). In Figure 5, ■ indicates that pulse noise is mixed in (b)
Input terminal of Im Rock S1, 2 is delay circuit,
A'
-], from this control signal generation circuit C8G, the human audio signal ``13Slvc'' is mixed.
82 will be generated. In FIG. 5, the delay circuit 2, the signal correction circuit 5, the first-order differentiator circuit 6, and the normal halt circuit 7t (which will be explained with reference to FIG. 1 (FIG. 3)) are constant pulse noise reduction devices. Delay circuit 2 - signal correction circuit 5, differentiation circuit 6
, sample and hold circuit 7, etc., and each of these components will be described in the description of the pulse noise reduction device shown in FIG.
The structure and operation are explained in detail. In FIG. 5, the signal S, which is sent from the terminal 5b of the signal correction circuit 5 to the output terminal 8, is set as follows: ) ray, the differential circuit (](1st order 1
−Ij is input as a differentiated signal to the next subpath 6), and the VC− signal level is set to a predetermined value.
Installed f' fi,! tf is also given as a differentiated signal to the second-order differentiator circuit 11 via the second level setter 10. ．． . As mentioned above 1. Output from the 211th order fine circuit 6: l'L 7.
, Z lθζ differential signal S41 and the second-order differential signal S42 outputted from the second-order differentiator circuit port are added in the adder 12 by −4 and 2, resulting in an addition 18 S4
So 4I becomes 5. -〇〇Addition 1. j C] S is 1 in state 90 degrees advanced with respect to the differentiated signal S3.
The second differential signal M S, , and the differentiated signal S, ilc
The state in which the phase is advanced by 180 degrees with kJ l- θ) By adding the second-order differential A 5 columns 842 and σ), the relationship between frequency and phase shift angle changes to the required change characteristic between 90 degrees and 180 degrees. It is said to be ``to show'', U/C. Figure 6 shows that one pulse mixed in the desired signal train I/C and the human input signal S has a period of IOf]-v microseconds. At some point, the signal correction circuit 5 outputs a signal S, in which the missing part of the desired signal that occurred during the existence of the Halsian miscellaneous person has been linearly interpolated well by the signal correction circuit 5. 2) Addition signal S4
It shows the frequency vs. phase shift angle characteristic of 1. What kind of frequency vs. phase shift angle characteristic (eE) (frequency vs. phase shift characteristic) should be added to the summed signal S is determined by the desired frequency that occurs during the existence period of pulse noise. I would like to have good linear interpolation for missing parts2)
It is determined depending on how long the period of existence of the pulse noise is, or the frequency range of the desired signal for which linear interpolation is to be performed well. 7. Similar to the input audio signal S1 shown in the waveform diagram a of FIG. 4 described above, FIG. ) The period of existence of the pulse noise is 1 to 4 of the desired signal.
Even if the signals are close to the period, the device of the present invention adds each output signal from the first-order differentiator circuit 6 and the second-order differentiator circuit 11 to n,'A:4+z and adds t9[-t-Kabao Monk S , the phase is advanced by 90 degrees from the desired signal by 1, +/)i'5i, and a) is obtained, the warpage becomes, and it is given to the Nordholt circuit 7, and the result is Rinonor,1, -Nosi,1. The output from circuit 7 @S will be described later. A) Benefit () Limiting circuit 13 I+
"(Ju K'X") to the terminal 5d of the signal correction circuit 5 through
- In particular, , J: From the signal correction circuit 5's l'ilA completion 5b (-1,...signal loss during the existence period of nine-dimensional noise i, the difference between linear interpolation and signal is obtained 2
) without further explanation A) f, knee θ) waveform diagram, and Figure 7 σ) a is the waveform of No. 16 of ``Manual audio with +/l external RN mixed in'' ( Yes, MaTko, 7th
b in the diagram is control! ) The waveform 1 of S2, C in FIG. 7, is 140 degrees ahead of the desired signal.7. , L: The waveform of the added signal S which has been added and the output signal S5 from the hole 1 circuit 7. 8Vc output ba) 1a)i'tA ”i
b3θ) waveform diagram. Figure 5 shows the waveform diagram of a = d in Figure 7, which was used to explain the operation of the pulse top 1 shoulder 1.14θ) surrounding area device island, and the waveform diagram of pulse nature maintenance shown in Figure 1. As is clear from a comparison with Figures a to d of Figures 9-1 used to explain the operation of the area device, the device shown in Figure 5 can perform linear interpolation better than the device shown in Figure 1. Among them, 1.1 Even when pulse noise N having such a long existence period is mixed into the input audio signal, linear interpolation for the missing part of (No. 8) is performed well during the existence period of the pulse noise N. Therefore, according to the device shown in Figure 5, the first
The above-mentioned problems of the illustrated device are successfully solved. By the way, it is well known that the amplitude or peak value of the output signal from a differentiating circuit increases as the frequency of the signal to be differentiated becomes uglier.Also, even if the signals to be differentiated are the same, the differentiating circuit is quadratic. It is also well known that in the case of an image dividing circuit, a larger differential signal is applied to the differentiating circuit than in the case of a primary image dividing circuit. By the way, in the apparatus shown in FIG.
Contains the slope information required for linear interpolation so as to obtain: A, -: >÷8. is supplied to the signal correction circuit 5 (IiCl). , therefore, differentiated-2)8. of,! 5 area θ) Input: The component is differentiated and the output is 41.
= s has a large peak value, and therefore has the same waveform as Jinto 84. The signal Rs, which is given to the terminal 5dK of the signal compensation circuit 5, is sent to the signal correction circuit. 5 in l
When dynamic interpolation is performed between the l'j line and J1 A, there is an 'IF-3\ for linear interpolation in the part VCZ that exceeds the linear interpolation part, i.e. , Supplement 1
4- An output signal S3 with an error appears, which causes noise in the world.
Especially lI:I. 2) Pc in the device (jroli#!j
The lr+1 limiter circuit 13 is a solution to the above problem. 4. Wave 1 of river S: Even if the value of i increases and the output signal S5 increases, f' is determined and fixed so that it does not exceed the magnitude of 1. The magnitude of the output signal S5 of the output signal S5 from the output circuit 7 is limited so that it is supplied to the terminal 5d of the signal complement circuit 52). The lines CI and -CL shown at C in FIG. 71 are from the gain limiting circuit 13 described above (the limit values of S and Bell are exemplified).As the gain limiting circuit 13 having such characteristics, For example, a slicer of known configuration can be used. By providing the gain limiting circuit 13 as described above between the Sansolpoort circuit 7 and the signal correction circuit 5, noise due to excessive linear interpolation that may occur in high-frequency, large-amplitude signal components is eliminated. The occurrence of this can be effectively suppressed. (Problems to be Solved by the Invention) As mentioned above, according to the pulse noise reduction device having the circuit layout shown in FIG. Even if it is rumored to have a long existence period close to 1/4 period of the input audio signal, the missing portion of the desired signal during the existence period of pulse noise can be linearly interpolated well.
The human-powered audio signal supplied to the input terminal 1 is in the state where it is being turned on and off, and the human-powered audio signal is input to the human-powered audio signal.
++ range θ) A5 component and... Lusty noise etc. are emphasized compared to the signal component in the 1 range due to the predetermined peripheral gauze characteristic. , pulse noise θ) Missing part of the signal of the wishing monk between Inno and 4101, 4! JVc existence (2) Absolute level of linear interpolation error for the missing part of the signal in the Ij region in month I - Therefore, the hope (,; 'r, ;'s Shinkyu θ) Missing 111! The problem is that it is difficult to perform linear interpolation for the length e. In other words, it is obtained by FM demodulating a signal in which causal noise is mixed into the FM$ information signal created using the Prienonoγ/Susano'shi1koji-dio signal as a modulation signal. 2-Zoio signal +:) Naturally, its high-frequency signal components and pulse noise are compared to low-frequency signal components. It is emphasized that 1. However, an audio signal in which the frequency is emphasized so that the amplitude of VCC10 of r6 (f, In this case, when the level of the high-frequency signal component σ) is high, the wave crystal value of the addition station S also increases, and the signal component σ) is supplied to the device as a human-powered audio signal. If the absolute level of the linear interpolation error is large at 1C, noise that causes problems with hearing will occur in the raw sound. In order to solve the above two problems, the human audio signal to be supplied to the input terminal 1 is, for example, a digital audio signal that has been passed through a digital interface circuit. In this case, the pulse noise mixed into the human-powered audio signal is 1117
The de-emphasis in the 7771 circuit causes the waveform to become dull and distorted, so the period of existence of the pulse noise is longer than that before the de-emphasis. Also, it is difficult to properly perform linear interpolation for the missing portion of the desired signal during the pulse noise presence period. Then, if the audio meter is in a state where it is nosed, then σ) For example, in an FM broadcast receiver, 71 How to output from the FM demodulator Trust - 11-voice demodulation to the television receiver, etc. - Audio signals are transmitted as frequency modulated waves,
b) In the case of recording, the output signal from the FM demodulator can be cited as a), but this θ)J:
The 2-dio signal in the preennonosed state is transferred to the pulse reduction device of A (I) as described above.
Feed it as an input signal! This is required, for example, in various in-vehicle electronic devices, and therefore, there is a need for a solution to the above-mentioned problems. child. (Means for Solving the Problems) The present invention provides a method for detecting a desired signal in a pre-ennounced period when pulse noise exists in an input signal. The audio frequency to be applied to the circuit section that performs the linear interpolation operation is approximately equal to the pre-enhanced audio frequency (σ) relative to the frequency band of interest. - By passing through the first 11177771 circuit configured to be able to carry out the Roughly speaking, the frequency band that is higher than the frequency band targeted for birth was designated as 7C O-1, so that it still has frequency characteristics in accordance with the Norienno f/s characteristics. teii) to do it:,! 2) A linear interpolation operation is performed for the missing part of the signal. Also, when the absolute level of the interpolation error is small,
The desired signal processing is carried out in a state where the waveform of the cursive noise is rounded.
The above-mentioned problems 1-r and 1 are to be solved, and furthermore, the above-mentioned problems 1-r and 1 are all aimed at beef cattle. The output signals sent to the output terminals via a second de-emphasis circuit configured to perform an emphasis operation are the same as those of the first de-emphasis circuit described above. 2nd θ) de f1 horn f/7. Di/F/S of the circuit ``P1θ) Comprehensive #!J+''1 is made into a complementary tI:Der・-7 Noah Mess special property with Nori 1-7 phasis characteristic f vs. kc, 1: Su, Tree 7 Tsume/
Using one di-7no, f/s circuit with complementary T17no/person characteristics to vc, we can create a tc audio signal as in the case of f, i'+, 7+ and so on. −
2). (Example) Next, with reference to the attached drawings, specific examples of the present invention σ) and the 41c noise reduction device will be described.

[Contents 1) Levi C liver + r111
Explain. Figure 8 shows the barno of the present invention, case 4VC3y)
8 is a schematic diagram of one embodiment of the enclosure device. In this FIG. ．． A first de-emphasis circuit DE, and a second de-emphasis circuit DE connected between the output terminal 5b and the output terminal 8 of the (a)
(The 1q component is the same as the pulse-assisted, blind 1-component device shown in FIG. 5 previously described. 4
・Nori:・- It is a curve diagram showing an example of the characteristic of the 7th phase.
A signal used to de-amplify an audio signal that has been pre-amplified according to a pre-amplification characteristic, such as a pre-amplifier that has been pre-amplified according to a pre-amplifier characteristic. This is a curve diagram showing the de-emphasis/s'14 of a general output/switching circuit, in which the de-emphasis characteristics shown in FIG. 9 and the de-emphasis characteristics shown in FIG. 10 are complementary characteristics. being done. FIG. 11 shows a first de-emphasis circuit DE and a second de-emphasis circuit DE2 used in the pulse noise reduction circuit of the present invention shown in FIG.
The curve DE in FIG. 11 shows the frequency response characteristic curve of the first de-emphasis circuit DE, and the curve DE2 in FIG. 11 shows the frequency response characteristic curve of the second de-emphasis circuit DE. 10 shows the frequency response characteristic curve of the /s circuit DE2, and the overall characteristics of the two characteristic curves DE, DE2 in FIG.
This is shown a) σ which matches the characteristic of
). That is, the first def[-7 noance circuit D in FIG.
E is Jd(I'Z) shown in curve DE in Figure 11.
As such, in the frequency band lower than the frequency 1ife, the frequency is flat! /subo7s% 111・ir fJ'
",・su, the first one, frequency fc to frequency fel
t:L at υ) exhibits a characteristic 11 that descends with the same slope as the -u-1 (Teienono f/S' Samurai 1'U curve shown in H Nu 1), and furthermore, the frequency fch It is constructed so that it has a flat frequency-responsive characteristic in a frequency band higher than 1. The second di1/ins γ/s circuit DE2 in FIG.
is shown by the curve DE2 in FIG.
It has a flat frequency response characteristic,
In the frequency band V higher than the frequency fch, the second
The curve is constructed so that it descends at the same slope as the characteristic curve shown in the figure. ]-1 Previously, 1st! 7) De-emphasis circuit DE
. In the pulse noise reduction device shown in FIG. 8, which is provided cascaded in the audio signal transmission path, the second de-emphasis circuit DE2 and the amplifier are connected to the de-emphasis circuit DE2, which is applied to the audio signal. is obtained by integrating the characteristics shown by curves DE and DE2 in FIG. c-ro. 1st. In FIG. 11 showing the characteristic curve LIE, DE2 of the second de-emphasis circuit DE, DE, DE2,
The frequency fch is selected to be a frequency value close to the maximum frequency i (=J) in the frequency band targeted for birth.
Therefore, the audio signal which is supplied to the input terminal (1) of the pulse noise reduction device of the present invention shown in Figure 8 and which is being ampliconed is - after being given the required time delay by the delay circuit 2. By passing through the Vc-first dee'-7 noance circuit DE, an audio signal with a flat frequency characteristic for the frequency band targeted for birth (in the frequency band targeted for birth-
) is the audio in the state of being +7 + 7 / s, t is -pa ) and / [, and the frequency above the 1st frequency band 9th band is considered to be the object of birth. Band If? : Qi-C is the input terminal 5 of the Jinjisode+E circuit 5 as the signal +1 in the state where Nori 1 is being applied.
a is supplied to Z), also known as ItC'/:c. Then, the signal correction circuit 5, the first σ1/... - le setting device 9, the second level setting device 105 and the Sa device 12.
The/Pull Bold Circuit 7, 4) Limiting Circuit 1: Due to the above-described operations in each component such as 3, 1. In the missing part of the signal of line tC, please trust
Linear interpolation is performed using (i
If the H rock acid component has a flat frequency characteristic at -4-R, θ), then the ζ-Russian miscellaneous imitation reduction device shown in Figure 5 should be used.
/c will be explained and it will be done well as described above.It is obvious from the above mentioned L1. In addition, in the output signal output from the layer correction circuit 5σ) output signal 5b, natural θ)
Since the frequency band above the branch band does not really exist 2) the component of Hals 1/2 is still in the state of No. 1 and 2), so the pulse noise is Waveform Ganasho 2. , l: 1 (<, then the pulse noise σ) presence period 1B5 expands), and the linear interpolation for the missing part of the desired signal is performed in good condition. be. A second de-emphasis circuit DE2 provided between the output terminal 5b and the output terminal 8 of the signal correction circuit 5 is connected to the L
It is output from the output terminal 5b of the public support IJ twice 1 heat 5 and 2] output Jf4v Koogero birth f) performs D L nonoa/s for the frequency band above the target frequency band/[
Therefore, the original audio signal is output to the output terminal 8. (Effects) As is clear from the detailed explanation above, the pulse noise reduction device of the present invention simply attenuates the gain of the transmission system during the period when pulse noise is mixed. Or, try to reduce the pulse noise by keeping the signal level during the ItJJ of the pulse noise at the level just before the pulse noise. As already mentioned, pulse noise θ)1 by the conventional method differs from the shallow area device/C, pulse noise σ) period/-1, same, (,-ith θ) missing C Interpolation is also performed.It is possible to effectively reduce the amount of pulsed noise by 1i(-'(18) without causing any auditory artifacts, and to reduce the amount of missing information. 1θ) Interpolation (/1f, the second circuit configuration is also easy)
In addition, it is low cost and has excellent performance.
2K can be easily provided. Yo1ko, Barsuze 1'41roθ) reduction device of the present invention □i
I,! Is there a sufficient correction effect not only when the time I11 is narrow, but also when the pulse noise is wide? VF ``-+re, furthermore, A-dioiki slip is pre-e-no-r) 2. In the case where good linear interpolation can be performed, the problem in the already proposed device is also good. VC cannot be solved, and the device of the present invention can solve problems such as automobiles and O-1-by-TC.
Yon noise, pulsed 4Mf noise generated from a device with a built-in electric motor, dust and dirt adhering to the audio disc (4-a), 1° knob noise, video disc 0) 1・u occurs in the Kl voice signal when the signal is missing
7 no γ udo υ) miscellaneous ~ so σ) 1 ya σ)・(pulse characteristic be f
It is used ineffectively to reduce the sound 11-.
0

[Brief explanation of the drawing]

Figures 1 and O-5' are already proposed θ)... Nosif, (pull)
5KI='row σ) (Uji government attire (neck) 1372-2
11 J'll and] ζ, ■-7 is an explanation of the operation, for clarity 0) Waveform diagram - 31st ("L,; - it has 1'+ f
An example configuration of the circuit and its related circuits σ)4, θ)θ)
+1'' 1 route map, Figures 6 and 9 to 111 Bacteria 1. The dynamic 1'1- theory 1)) 1σ)! Figure 8 is a diagram of one embodiment of the present invention. 110
．． Input terminal, 2... delay circuit ~ C8G...: 11 j
Start-up speech development II1. Circuit-DE,... 1st σ) Di' Tsuno 7 no, II! 1st road, DIE,... 2nd end face/su 1111th road -: 3 ノ; Rusue noise detection circuit, 4... ζ 11 circuits, 5 - A word correction circuit, 6 - Primary image division circuit, 7...-' note +t; - 1 - circuit, 8 Output terminal -9, 1, 0, .20th) Level setter-11...Secondary image component l111
Currency '1-1) L・Book 1j iQ 1lill l hot water circuit~
VC: Variable constant current circuit, C: 44'J/
Denza, AI + A2...11th 2nd σ] lump width instrument,
A3. A4 amplifier, SW, SWs...Swinopt
el I tz, t5, t2tz, j6 Ka 2 Figure 7

Claims (1)

[Claims] 1. Input audio conversion percentage that is pre-en-no-anosed
means for detecting the pulsed noise mixed in the VC and generating a control signal having a pulse period corresponding to the period during which the pulsed noise occurs; Above IL m! I Is the Idl signal generated by the means for generating it? Between the 1111th signal and 5 its control (between the 8th signal and the corresponding pulse noise θ'l L'
means for delaying the input audio signal 42 containing 5-pulse noise by a delay circuit having a delay time approximately equal to the difference between means for applying the control signal to a first de-emphasis circuit configured to perform a de-emphasis operation on a signal in the band, the control signal being supplied as a timing signal for the operation; 7) During the period when pulse noise is generated in the input audio signal, a signal having the ei multiplier signal slope '1I4i signal is supplied. The output of the 1-coil de-energization circuit is constructed as follows: Iro-go na-r-
5e2° means and the output 4i from the signal correction circuit
4= in the first deennor/s circuit described above,
・2) A second de-emphasis circuit is configured to perform a di-noise operation on a frequency band higher than the frequency band in which the -7 phasis operation is being performed. means for transmitting to the output terminal through the
The output from the signal correction circuit 73"LA is passed through the first level setting circuit to the first-order differential circuit vcJ-iE2)
At the same time, the output signal from the signal correction circuit is
After adding the output signal of the first-order differentiating circuit and the output signal of the second-order differentiating circuit to the second-order differentiating circuit through the general constant circuit, the above-mentioned control signal becomes a sampling pulse. means for supplying the output signal to the sample hold circuit supplied as a sample and hold circuit; Pulse noise reduction device 2 comprising means for outputting the slope information of the desired signal between 1411 and giving it to the above-mentioned confidence correction circuit. A structure is used in which the charge storage capacitor is charged with the output current of the variable constant current circuit only during the period when pulse noise is occurring, and the discharge operation is performed instantaneously at the end of the period. 1 Pulse noise reduction circuit 3 according to claim 1 As a gain limiting circuit, instantaneously performs a gain limiting operation based on σ) for a level equal to or higher than a preset reference input level. σ) Pulse noise reduction device 4 as described in claim 1. 8. As a correction circuit, a first amplifier with a low output impedance and a first amplifier with a low input impedance are used. 7) Intermediate transmission between EL1 and EL1 is provided in the entertainment transmission line from the first amplifier to the second amplifier, and pulse noise is generated. The output current value is determined by Takumi Shinzuki, who has the slope information of 2) Desired A^H. The output flllll of the 13J variable current circuit that operates smoothly and the charge storage capacitor are the same as those described above.
) Connected to the input side and using a TC/C filter as described in claim 1, the Milno property 4-[one θ) reducing device straight 5, variable constant 1L flow circuit is used for ~.il. The current value is set according to the signal level of Zui/Human Power Ia, and the input signal σ) for it is 1 pole (configured so that a constant current output of 1 can be obtained) 1 depending on the polar region. (i) Hals noise 0) reduction device