KR850001908B1 - The system of noise reduction - Google Patents

Abstract

A noise reduction curcuit has a first signal path called the main- path whose gain is greater than a unit with linear chracteristics, and has a second path called the auxiliary path, which lowers the high amplitude signal to 0.1 times and amplifies the low level amplitude signal of the main-path output signal. The compressor, using lower unit gain positive feedback, has high Nyquist stability. The expander, employing an open loop, has complementary characteristics to the compressor. The auxillary input of the compressor is the main-path output of the compressor and its output is added to the main-path.

Description

Noise reduction

1 and 2 are basic diagrams of the noise reduction method of life 1 and life 2.

3 is an amplification specialty curve.

4 shows an auxiliary signal path for a video system.

5 is a circuit diagram of a limiter for an imaging device.

6, 7a and 7b are curve diagrams relating to property shaping.

The present invention relates in particular to a noise reduction method consisting of a compression device arranged in front of a noisy device and a complementary stretching device arranged behind it.

The present invention is an improvement on the noise reduction method constituted by the compression device and the stretching device described in the specification of Patent Application No. 76-729, and the invention of the above application will be referred to as a prior invention in the future. According to the principle of the compression or decompression device according to the prior invention, the first signal path, which is a straight signal path, and a gain of, for example, about 10 decibels, can be obtained, and the maximum output is obtained through the first signal (the main signal). Using a second signal (supplementary signal) that limits to a small portion of the amplitude, adds the output to the first signal if the device is a compression device, and to the output to the first signal if the device is an expansion device. To subtract the output of the second signal.

와 같은 형식으로 표시되며, 방식 전체에 대한 법칙은

이다. 단, X는 압축장치의 입력 Y는 압축장치의 출력, 즉 신장장치의 입력, Z는 신장장치의 출력, F₁및 F₂는 각 압축장치및 신장장치에 있어서 제 2 신호로의 특성을 나타내는 연산자이다. F₁과 F₂도 저신호 레벨에서는 승수의 형식을 가지며, 레벨이 높아짐에 따라 저하해서 제 2 신호로의 출력이 대략)일정하게 되던지 또는 경우에 따라서 X 혹은 Y의 감소함수로 된다. 상보적 압축장치및 신장장치에 대해서는 F₂=F₁이므로 Z=X 이다.In this configuration, the gain from the low input signal level to the second signal can significantly increase (compress) or decrease (extend) the signal level. At the high input signal level, the output to the second signal is not as important as the output to the straight signal. This is because the output to the second signal is limited as described above. The advantages of such a configuration are as described in detail in the specification of the preceding invention, the most important of which is that the result can be remarkably distortion free. As described in the specification of the prior invention, the input to the second signal is the same as the output of the compression device. So the law of compression is Y = (1 + F ₁ ) X and the law of extension is Z = YF ₂ Z,

Is written in the form

to be. Where X is the input of the compression device, Y is the output of the compression device, that is, the input of the decompression device, Z is the output of the decompression device, and F ₁ and F ₂ are the characteristics of the second signal in each of the compression device and the decompression device. Operator. F ₁ and F ₂ also have a form of multiplier at the low signal level, which decreases as the level rises, so that the output to the second signal is approximately constant or, depending on the case, decreases to X or Y. For complementary compression and extension devices, F ₂ = F _1, so Z = X.

A compression device and an expansion device operated by the above rule will be referred to as type 1 for convenience.

의 형식을 갖도록 되어 있으므로, 방식 전체에 대한 법칙은

된다. 여기서도 F₁=F₂이면 당연히 Z=X 이다. 본 발명의 구성에서는 뒤에 설명하는 바와같이 완전히 F₁=F₂로 할 수가 있다.The present invention has been invented from the recognition that the input to the second signal in the compression device and the decompression device, which are components of the noise reduction method, can be connected in a format different from that described in the specification of the preceding invention. In the decompression device, the second signal path receives the same input as the straight signal path. The law of decompression has the form of Z = (1-F ₂ ) Y, and the compression device compresses the input with the second signal. From the output of the device the compression law is equal to Y = X + F ₁ Y or

Since it is in the form of, the law for the whole

do. Again, if F ₁ = F _2, then of course Z = X. In the configuration of the present invention can be entirely F ₁ = F ₂ as will be described later.

Compression and extension devices operating according to this new law will be referred to as Type 2.

In the compression device and the decompression device, the second signal can process the entire signal band, but in practice, it is common to pass a signal in a limited portion of the signal band. In addition, as described in the specification of the present invention, by setting two or more second signals, that is, an auxiliary signal path by a known method, the upper frequency band can be selectively processed.

The main feature of the Type 2 unit compared to the Type 1 unit is that: a) the compression unit uses a square loop with a gain less than 1, so that the circuitry inherent in this unit has a unique Nyquist stability. B) Even in the stretching apparatus, the failure of stability can be avoided by not using the closed loop. Therefore, each of the compression device and the decompression device constituting the noise reduction method has high stability over a wide band and, in the case of the noise reduction method as a whole, provides a noise reduction method in which complete complementarity is maintained between the compression device and the decompression device. I can do it. We will now discuss the compression and stretching devices used in the Type 2 noise reduction scheme, which is simplified for installation on a home TV set.

In the specification of the prior invention, the second signal path is described as using a single limiting channel, but in the compression and extension devices which are components of the present invention, such limiters are different from each other by using a plurality of limiters connected in parallel. Alternatively, it is possible to have the applied signal level or the like. Combining the output of the limiter totally or partially subtractively before combining the signal into the straight signal can provide compression and stretching characteristics with various shapes as required. This can be achieved with a high degree of precision and reliability. In addition, overshoot can be ignored by setting a non-linear limiter with a limit after the linear limiter, as described in the specification of the preceding invention.

In one form of curve shaping, the input signal level initially increases, but the output can be obtained from the second signal so as to decrease later. Such a characteristic is referred to as falling or downward bending. The downward curvature characteristic is an important characteristic. This is because the output to the second signal can be disregarded at a high signal level by this characteristic, so that the output from the compression device, the decompression device, or the decompression device is almost a straight signal only. The downward curvature characteristic corresponds to an audio signal system and a video signal system.

It will be described in detail according to the drawings.

Before describing the type 2 compression and stretching device, FIG. 1 illustrates the type 1 complete noise reduction scheme.

In the compression apparatus, the second signal path 10 as an auxiliary signal extracts a specific signal portion from the input signal, and the signal portion to the second signal is added to the straight signal to the main signal by the adder 12. . Therefore, at low levels, the output signal is increased in amplitude, that is, compressed. The transmission characteristic d of the compression device of type 1 is represented by the logarithm, or dashbell, in FIG. The signal b, which rises from the limiter circuit and curves downward, forms a compressed signal d by going to the straight signal component a, and supplies the signal to a recording device or a transmission device. Depending on the demand, the signal from the second signal is supplied to a separate recording or transmission channel and then combined with the straight signal component at the receiving end.

In the decompression device, the second signal path 14 obtains a signal portion from the output signal of the device with the same second signal, and subtracts it from the signal recorded or transmitted by the subtraction device 16. As shown in FIG. 3, the transmission characteristic e of the decompression device is almost constant at a high level, and at a low level, the gain decreases to achieve noise reduction.

The type 2 device in FIG. 2 is a basic diagram showing a complete noise reduction scheme. Similarities to the shape of Type 1 can be seen from the drawing. However, in the type 2 decompression device, the input to the second signal path 14 is obtained from the direct recording or transmission channel and the subtraction device 16 reduces the output from the straight signal to the second signal. The compression device has a complementary configuration and the input to the second signal is obtained from the output of the compression device and the network output is supplied to the adder 12. Therefore, the second signal is included in the positive feedback loop. However, since the open loop gain is less than 1, there is no tendency to generate oscillation. In addition, it is not necessary to adjust the amplitude and phase response characteristics of the loop outside the pass band as in the case of the type 1 stretching device. If a signal path that sends a straight signal is compared with a signal path including the second signal path 14 and is expressed in more general terms, this may be referred to as a first circuit, and the signal path including the second signal path 14 may be referred to as a signal. May be referred to as a second signal.

에 있어 완전히 F₁=F₂로 할수가 있다. 이것은 타이프 1의 방식에 있어서와 완전히 똑같다. 단 타이프 2의 경우에는 방식 전체의 각 구성요소인 압축장치 및 신장장치에 전술한 바와같이 안정성이 있다는 특징이 더 추가된다.It can be seen from the configuration shown in FIG. 2 that it is very easy to make the second signal path 10 of the compression device and the second signal path 14 of the decompression device have the same transmission characteristics. In other words, it can be seen that the adder 12 and the subtractor 16 apply the same functionally, except that the difference is that the subtractor adds by reversing the sign of one input. When the input of the subtractor 16 is reversed from the second signal 14 so that the nationality is lowered (for example, by inserting an inverter), the subtractor 16 has the same configuration as that of the adder 12. In the device after such substitution, the compression device may also have one adder (the compressor and the expander may have the same configuration) and one second signal (the compressor and the expander). For the network comprising this one adder and one second signal path, as there is no characteristic limit to be taken between that of the device, When operating as a compression device, install a suitable switch to make the connection as shown in the lower left of FIG. 2, and to switch as shown in the lower right of FIG. 2 when operating as an extension device. Here, one common signal path is used for the second signal path 10 of the compression device and the second signal path 14 of the decompression device, and also as the compression device and the expansion device as the addition device. An adder can be used. As described earlier

Can be completely F ₁ = F ₂ . This is exactly the same as in type one. However, in the case of Type 2, the characteristics of stability as described above are further added to the compression apparatus and the stretching apparatus, which are components of the entire system.

As can be seen from this, it is possible to achieve complete complementarity between the compression device and the decompression device as a whole, but in addition to how the compression device and the decompression device can be configured as components of the method, and Understanding what characteristics are given by it (ie what characteristics can be given as F ₁ F ₂ ) is important for understanding the whole approach. This will be explained as follows.

As in the type 1 method, in order to achieve the compression or decompression device transmission characteristics by the type 2 method, it is necessary to use a slightly conspicuous shape as the transmission characteristics to the second signal. In Fig. 3, the type device is not a curve b but a curve c. In other words, it uses a relatively high one and slightly flat characteristics. Such curves are specific to static characteristics and the same adjustments must be made to any curves in these characteristics. In addition, a relatively long time constant is used for the control signal smoothing circuit of the type 2 device.

Regarding the Principle of the Type 2 Device To construct a composite device to which the principle of the type 1 device is applied, the input to the second signal may be maintained at an arbitrary ratio with respect to the input and output signals of the complex device. Thus, a generalized compression device or decompression device includes a first adder in a first signal path, a second adder configured to add at an input ratio to the first adder and an output ratio from the first adder, and It is conceivable that it consists of an auxiliary signal which obtains an input from the first adder. The output to the auxiliary signal is applied to the first adding device without inversion in the case of the compression device, and inverted so as to subtract in the case of the stretching device.

For type 2 devices or type 1-2 devices, when the output of the amplitude fluctuation range and the output to the auxiliary signal have to be sent to the stretching device through the respective channel, it is compressed as possible with the type 1 device. The adder of the device cannot simply be removed and placed in the stretching device. In the apparatus of type 2 and the composite apparatus of type 1-2, an adder is necessary to form an input to the second signal. However, the output to the auxiliary signal and the signal whose amplitude fluctuation range cannot be changed are also transmitted or recorded respectively, and in this case, the adder in the stretching device forms a compressed signal so that it is installed in the stretching device in duplicate.

When the compression device and the decompression device are formed complementarily, an addition device and a subtraction device are installed in the main signal path, and the adding device expands from the compression mode by selectively supplying the output of the auxiliary signal to the subtraction device at an appropriate ratio. A smooth transition to the mode can be achieved.

In the complete noise reduction scheme, the signal component of the second signal in all devices is first added to the signal as the main signal and then subtracted again so that the output signal of the device becomes the same as the input. Whatever the specific characteristics of the two identical second signals, the device system cannot have an overall effect on the dynamic or frequency response characteristics of the signal, but an appropriate characteristic must be selected to achieve an excellent noise reduction effect.

As for the noise reduction device, four auxiliary signal paths as shown in FIG. Is raising. Next to the filter is a linear limiter 20 of the same configuration having a transmission characteristic as shown in Fig. 3B. Each limiter is controlled by a control signal obtained by rectifying and smoothing one of the input and output signals of the limiter from one direction or two directions. This combination imparts the necessary rising and downward bending characteristics. Next to the linear limiter is a nonlinear limiter 22 diode clipper circuit, whose function can limit the amplitude of the transient signal passing over the limiter's start-up time (about 1 ms). The four band outputs are coupled at the adder 24, followed by a separate nonlinear limiter 26 for minimizing the overshoot at the clip for the combined transient signal. As shown in the figure, the network imparts a signal component to the masonry signal which can reduce rumble, hum, crosstalk and hiss.

In the video signal noise reduction apparatus, the linear phase frequency response characteristics of the video signal recording or transmission apparatus should be considered. Under these conditions, a relatively simple nonlinear limiter can be designed using a diode. As an example of such a filter and limiter circuit, diode 30 is not conducting for a low level signal from a low impedance signal source so that the signal passes through the full amplitude despite the high pass characteristics of resistor R and capacitor C. However, for large transient signals, the diode conducts and rapidly charges the capacitor C, so the low level signal passes outside of the shortest non-conducting time. The output of the limiter is combined with the output of another limiter covering various frequency bands as in the device for the above mentioned voice signal.

An extended interpretation of the compression and extension curve shaping methods using a single limiter described so far, several parallel limiters, one or both of which have different limits or applied signal levels, either in the same frequency band or overlapping bands. It can include the operation of. In this way, most of the advantages of the configuration using the basic single limiter, i.e., high accuracy, low noise, noise level and low oversuit, etc. can be obtained without losing the transmission characteristics of various shapes.

In order to easily explain the principle, FIG. 6 shows a characteristic curve of the compression height of type 1 in which three limiters are connected in parallel in only one auxiliary signal path.

The characteristics of the straight signal path are (a) and the characteristics of the three limiters are indicated as (b), (d) and (e). Since (b), (d) and (e) have different limiting limits, there are differences in their positions. Therefore, as in the case of the characteristic (b) and a single limiter, a comprehensive characteristic (c) is not obtained but a comprehensive characteristic (f) is obtained. The overshoot can be made extremely low if a non-limiting limiter (clipper) with a suitable limit is installed after each linear limiter and another limiter, other than the fourth one, to handle the signal as an auxiliary signal.

The shape of the overall characteristic can be precisely controlled by the selection of the shape, position, and number of the limiter characteristics, and if necessary, the output of one or more limiters can be controlled by subtracting rather than adding up. Thus, a combination of a so-called type 1 compression device and a type 2 extension device can be obtained.

In a typical device using multiple limiters or multiple auxiliary signals, it may be possible to stand alone to achieve compression or stretching of Type 1 or Type 2.

In a video signal system, downward curvature limiter characteristics are useful mainly for reducing or eliminating overshoot at the output of a compression apparatus. When a synthesized video signal is sent through a compression device using a limiter having downward curvature characteristics, for example, the output synchronous pulse has a type without overshoot as in the input. Therefore, the compressed video signal becomes the same as a normal signal, and processing and transmission are simple. The shape of the compressed video signal itself is also almost identical to that of a typical video signal. If the overshoot is limited to a low level signal, an extremely sharp and clear image is obtained, and in many other respects there is no significant difference from the normal signal.

Next, a case in which the downward curvature characteristic applied to the audio signal is applied to the video signal will be described.

In the above description, a characteristic curve shaping technique using a large number of limiters has been studied. For application in the field of video signals, it is easy to understand with reference to the limiter circuit of FIG. 5 and the transmission characteristic curve of FIG. In FIG. 7, a specific curve marked by engagement is indicated for easy understanding of the actual state of subtraction. If two limiters, like the fifth degree, are operated in parallel under the same conditions, and the outputs are subtractively combined, offsetting occurs. However, by appropriately changing the operating conditions of one writer (limit limit, type of diode, signal gain), an effective combined output can be obtained that achieves the required downward bend transmission characteristics.

FIG. 7A shows two output circuits with different levels of output as shown. The output (a) is multiplied by the appropriate coefficient K and the next output (b) is reduced. At a high input level, this calculation result is zero, and at a low level, a normal output is obtained.

7B is a diagram showing a complementary method in which one limiter is driven with low gain although the limiter of both limiters is the same. For the high amplitude input, the output levels of the two limiters are parallel or almost the same as in the figure, but at low levels, the output difference is as much as the difference of the applied signal levels. By subtracting the two outputs, the necessary rising and downward bending characteristics can be obtained.

In the specification of the prior invention, an embodiment which does not directly operate on a signal but operates only on a carrier signal modulated by the signal is described. Although application to color televisions and measurement devices has been described, the principle applies to the entire carrier signal.

Linear limiters should be used for applications such as voice signals, but non-linear limiters can be used in other cases. In other words, if the specific value of the limiter is kept at an extremely low value or must be completely zero as a whole, the limit must be non-linear.

It is necessary to use a carrier blocking filter for input to the auxiliary signal except when processing the suppressed carrier signal. In the case of using a plurality of auxiliary signal paths, a carrier blocking filter may be disposed at a common input contact to all auxiliary signals.

The auxiliary signals used in the carrier compression apparatus and the captain apparatus are similar to the auxiliary signals described above. Once the carrier is blocked, the sidebands are processed in various ways. If only one auxiliary signal is used and no filters are used, then a wideband compression or extension is obtained, and as a separate method, using multiple auxiliary signal paths, each with its own tuned filter over a given sideband frequency range, Compression or stretching is performed in the modulated signal frequency range corresponding to each frequency of the filter used.

In the bilateral carrier system, it is preferable to treat the sideband symmetrically in each auxiliary signal path in order to prevent large distortion. A combination of real band filters and band cancellation filters can achieve this purpose. In this case, center both sides of the filter on the carrier frequency and make a reasonable difference in bandwidth.

In a simple carrier noise reduction device suitable for an audio signal, a carrier blocking filter may be selected to attenuate low and medium voice frequency sidebands as well as carriers. Therefore, noise reduction is mainly performed at high audio frequencies, and hiss reduction is also achieved. When using a separate auxiliary signal with a narrow stop filter, the noise reduction extends to lower voice frequencies. The acting as an auxiliary signal imparts a signal of sideband frequency corresponding to a high speech frequency but the first auxiliary signal when the second auxiliary signal is blocked by a high level sideband corresponding to a medium speech frequency. Only works with.

In the limiter circuit for audio and video signals as described in the specification of the prior invention, when the filter output reaches a certain level or more, the filter automatically narrows this band so that a limiting effect is achieved. Such a method is also possible in the compression device and the decompression device of the carrier signal. For a channel whose phase is maintained, a technique by a nonlinear limiter used in connection with a color subcarrier may be used. For audio signals, the filter should be operated linearly and a general configuration as shown in FIG. 5 for practical circuits may be used. However, in this case, a parallel resonant carrier blocking circuit may be used instead of a capacitor, such as a field effect transistor. The controlled resistive element is connected in parallel to the resistor R and operated with a rectified and smoothed control voltage.

The circuit operates linearly, except for a period of rapid paste that corresponds to a high amplitude negative transient signal and conducts a diode during the limiter start-up time. In a practical circuit, a clipper diode is arranged to a relatively end of the circuit, and signal processing is performed after amplifying the limited signal. A series resonant circuit may be arranged in parallel with such a diode.

In another form of the FIG. 5 circuit suitable for use with carrier signals, capacitor C is replaced as a parallel tuned carrier blocking circuit to remove diode 30 and a symmetrical nonlinear device or circuit in series with the previous parallel tuned circuit is employed. Insert it. This nonlinear device or circuit has a relatively low impedance up to a constant applied voltage and maintains a substantially constant current beyond that. In this connection, for example, the collector characteristics of a constant current diode or a transistor may be used. A series tuned carrier resonant circuit may be arranged in parallel with the resistor R.

In addition, a compression device and an expansion device can be used independently for the purpose of various signal changes. For example, in order to reduce the amplitude fluctuation range of the audio signal, a compression device having an auxiliary signal path covering the low band can be used. In addition, the use of a number of auxiliary signal paths covering several frequency ranges yields special dynamic balance characteristics.

Full-band stretching devices are effective for reducing signal noise, particularly when no signal or low-level signals pass. An expansion device in which a filter is provided in one or more auxiliary signal paths can effectively reduce the noise in the presence of a signal. For example, by using such an expansion device, the hissing can be reduced in the old 78-turn record, and no signal disturbance is caused more than when the conventional filter method is adopted.

In the case of using an extension device to reduce noise by frequency, band filters are provided in each auxiliary signal path corresponding to each disturbing frequency, and the limit of each limiter is appropriately set. Heaths in audio signals, harmonics in power lines, disturbed radio frequencies in video signals, and the like have little effect on the signals themselves, and can be reduced by the above method. When a signal arrives at any of the noise frequencies, the noise reduction action is immediately prevented, but then the disturbance is effectively masked by the signal.

In the above description, little attention was paid to the placement of the amplifier, but this is only a minor design problem. However, in principle, an amplifier is unnecessary for the type 2 device of the present invention.

As can be seen from the above description, according to the present invention, perfect complementarity is maintained between the compression apparatus and the stretching apparatus, and each component of the compression apparatus and the stretching apparatus, which is each component, can obtain high stability over a wide band. Noise reduction methods are described.

Claims (1)

Before transmitting or recording a signal, compress the linear width range of the signal and extend the amplitude variation range of the signal after reception or playback, and the amplitude changer is linearly related to the amplitude ward range of the input signal. A first circuit portion for outputting a first signal component having a range and a linear limiter such that a significant level increase occurs at a low input signal level and at a level less than about 0.1 times the first signal component level at a high input level. A second circuit portion for outputting a second signal component, the expander outputting a third signal component having an amplitude variation range linearly related to the amplitude variation range of the input signal, and only at the low input signal level Significant level reduction occurs, and at high input signal levels, the linearity is about 0.1 times less than the third signal component level. In a noise reduction scheme comprising the output of a fourth signal component limited by a meter and having a second circuit portion identical to the compressor or a second circuit portion having substantially the same configuration, the input of the second circuit portion in the compressor is connected to the output of the compressor. Coupled and the input of the second circuit portion in the expander is coupled to the input of the expander.

Images