KR840001997B1 - The circuit device for signal compression and expansion in variable amplitude - Google Patents

Abstract

A system has a linear dynamic range with regard to an input signal, when it acts as a signal compressor, or with regard to an output signal, when it acts as a signal expander. A frequency from the filter is automatically adjusted according to an input level by restricting the low level below the predetermined threshold of the high signal level, so that the second output component when compared to the first one is less than unity.

Description

Circuit device for compressing or extending the amplitude fluctuation range of the information signal

1 is a block diagram of a compressor for an audio signal.

2 is a block diagram of a complementary expander.

3 and 4 are circuit diagrams of two filter and limiter circuits used in a compressor and an expander.

FIG. 5 is a diagram showing an example of modification of FIG.

6 is a circuit diagram of a modified filter and limiter circuit used in the color television system.

7 is a circuit diagram of a filter and limiter circuit for an audio compressor.

The present invention is used in a noise reduction method of a type that compresses the amplitude of a signal before it is transmitted through a transmission line or a recording / reproducing apparatus that introduces noise and complementarily extends the signal after transmission. A circuit arrangement for compressing or stretching. For convenience, a device that introduces noise is called an "information channel". Noise in an audio system is usually hissing, hum, rumble, click, or crosstalk. The main noise in the video system is the high frequency noise.

Many noise reduction schemes are known, including preemphasis in the low and high frequencies of the audio domain, but their use is limited. This is because overloading must be avoided when the audio signal contains a significant amount of energy in the high and low frequencies of the frequency. As another noise reduction method, recording is sometimes made up of two channels, where one channel is recorded at a higher level up to 30 decibels than the other. During playback, the low level part of the recording content automatically selects the high level channel to achieve a good SN ratio, and the high level part of the recording content automatically selects the low level channel. In other words, the output of the low level channel is amplified with amplification greater than that of the high level channel, so that a correct balance is achieved between the low level and the high level parts. However, this technique is limited in scope and inevitably uses two channels instead of one. It is known to use a nonlinear circuit as another known noise reduction method, but this introduces intermodulation distortion in an audio system. It is also known to automatically control the low pass filter in response to the signal level so that the hissing is removed by the filter only when the low level passes. It is also known to use a compressor and a decompressor to automatically change the degree of decompression when reproducing according to the signal level.

Their prior art is limited in scope and efficiency because of unwanted side effects such as hissing.

The object of the present invention is to eliminate these drawbacks and generally have a wide range of applications.

SUMMARY OF THE INVENTION The present invention is a circuit arrangement for compressing or stretching an amplitude fluctuation range of an information signal, comprising: a first channel for conveying a signal having a linear transmission characteristic over the entire bandwidth of the signal, parallel to and affecting the amplitude; In addition, the first and the second channels including at least one frequency selective filter and the output of the second channel increase the output of the first channel for compression and reduce the output of the first channel for expansion. A coupling device for coupling the outputs of the two channels, the electrical filter being automatically varied to narrow its bandwidth in response to the signal level of the circuit arrangement, thereby outputting the second channel at a high signal level. The signal is suppressed so that it is in an area smaller by one unit or more than the output of the first channel or exceeds a low level amplitude. All.

This technique of forming compression or stretching characteristics by combining signals in two channels has several advantages.

The contribution of the limiter is negligible at high signal levels, so the distortion is significantly reduced. The problem of giving a precise correspondence between compression and elongation is practical.

Thus, a good feature of this method is that it is relatively insensitive to the difference in level between compression and stretching.

This method is also advantageous under fluctuating signal and transient signal conditions. In particular, the overshoot can be kept at a negligible value, thereby avoiding overload of the information channel. In this way, in addition to the selective filter, the second channel may comprise a non-linear limiter continuous to the filter, which is a transient current passed by the filter when the filter operates as a linear limiter as in the embodiment of FIG. It serves to suppress.

Firstly, the second channel has a bandpass characteristic (which is obtained by the highpass or the lowpass characteristic of each of the high and low frequencies of the entire frequency range), so that the main system of the video system, for example, hissing, power, high frequency noise, etc. Noise reduction is selectively performed in certain parts of the parse spectrum. In general applications, a plurality of second channels that selectively process separate bands are used so that a signal having a large amplitude in one band does not interfere with noise reduction in the other band in which a low level signal exists. need.

In the case of a color television signal, the drastic reduction of the moire pattern and the noise reduction of the broadband are such that the secondary channel, which is an auxiliary channel, is the color subcarrier and preferably the second channel.

In phased systems such as video systems, the output of the auxiliary signal path is directly connected to the filter and automatically shifted to the cutoff frequency of the filter so that the output of the filter is higher than any level. The passband is limited by a simple diode circuit that operates narrowly. In audio systems, a linear limiter is used and its cutoff frequency is controlled by a smoothed signal to prevent the introduction of distortion. However, at this time, after the linear limiter, the time interval is short without introducing an audible distortion for a sustained period of time of the generated smoothed signal, and the amplitude of this limiter output is equal to the amplitude of the output of the main channel. For the reason of being small in comparison, a nonlinear limiter may be connected which acts so as not to overload during the time interval. Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

H이다. 이에 따라, 고 입력 신호 레벨에 있어서 신호의 오우버슈트에 기여하는 고주파 성분은 작으므로, 도입되는 오우버슈트는 무시할 수 있다.Figure 1 shows a black and white television video signal compressor. The signals of the first channel 18 are combined at the circuit 20 with the outputs of the two channels 9 and 11, which are installed to reduce the noise (with extension) from 100KC to 1MC or 1MC to 4MC, respectively. The frequency band from 100KC to 1MC and the frequency band from 1MC to 4MC approximately match the range required for a normal television video signal. The outputs of channels 9 and 11 are added in a second adder 19 before being fed to the first adder 20. To achieve noise reduction of 10 decibels, amplifier 13 amplifies the low level input signal to 6.17 decibels (2.16 times), and reducer 16 adjusts the gain of the two channels 9 and 11 as needed. Each channel contains 10, 12 circuits that combine filters and limiters (selecting corresponding bands), and these circuits

H. As a result, the high frequency component contributing to the overshoot of the signal at the high input signal level is small, so that the overshoot introduced can be ignored.

The expander shown in FIG. 2 restores the signal to its original size since 9 and 11 are connected between the output and the input of the subtractor 21 as two signals. The circuit configurations of 9 and 11 as signals are the same as those in FIG. Alternatively, only one unit may be made, and switching means or switching connecting lines for changing the unit to either the compressor or the expander as necessary may be provided.

If the output of the expander is indicated by Z, a signal having an amplitude of 3.16L is added to the input of the expander at a low signal level, and the following equation can be obtained.

3.16L-2.16Z = Z

Thus L = Z.

H)의 진폭을 가진 신호가 신장기의 증폭기 19의 입력에 가해지고, 이때 이 증폭기의 출력의 진폭은 2.16H이다. 그러나, 필터 및 리미터가 결합된 회로장치 10, 12는 이값을 상술한 값 H에 비교해서 대단히 작은 값인 0.01HM보다 크지않은 값으로 제한시키므로, 감산장치 21의 출력 H-0.01HM은 H와 거의같다.At high signal levels, H + 0.01HM (

A signal with an amplitude of H) is applied to the input of amplifier 19 of the expander, where the amplitude of the output of this amplifier is 2.16H. However, the circuit arrangements 10 and 12 in combination with the filter and the limiter limit this value to a value no greater than 0.01 HM, which is a very small value compared to the above-described value H, so the output H-0.01HM of the subtractor 21 is almost equal to H. .

As can be seen from the foregoing, at low signal levels, the amplitude of the input to the expander is reduced to one-third of 3.16 regardless of whether the low level signal is of high or low frequency. Therefore, the amplitude of the noise (this is a signal that is input to the expander and has a higher frequency than the signal transmitted from the compressor) is also reduced to one third of 3.16. This means that noise reduction of 10 decibels has been achieved in the stretcher. In this way, the noise of the high frequency of the low signal level which produces a granular image is attenuated. When the high frequency noise supplied to the input of the expander is at a fairly high signal level, this noise will not be attenuated. However, since the amplitude of such noise is usually less than the amplitude of the high frequency signal transmitted from the compressor, this noise is masked by the details of the image represented by the high frequency signal component. Because it becomes, it is not detected by eyes.

In the audio system to be described later, the same masking effect is obtained satisfactorily, and the detected noise is removed. Two second channels 9 and 11 are shown as examples in the expanders and compressors of FIGS. 1 and 2. For example, assuming that only one signal is used for the entire band from 100KC to 4MC, high amplitude signals up to 1MC are sometimes higher than 1MC.

Figure 3 shows a very suitable circuit that can be used as filter and limiter circuit 12. A simple RC network consisting of a capacitor 24 and a resistor 26 (for example)

Since diode 28 is a largely conducting germanium diode, the voltage drop between both ends is minimal (for example, about 0.1 or 0.2 volts). Even in this case, the input signal voltage must be at a high level when the limit is operated at 1% of the maximum input (eg). For diodes with a bias of 1 volt of moisture, a peak to peak value of the input voltage is 40 volts. 5 is a modified embodiment of FIG. 3, which alleviates the need for driving, which is convenient for transistor circuits and the like. The diode 28 is connected via the capacitor 35 to the output point 34 of the filter and to the output of the feedback amplifier circuit, which is composed of an amplifier 36, a gain setting reducer 38 and an emitter follower. The swing of the voltage required at the input is reduced by the gain of the feedback loop, and the effect of this feedback loop will shift the limit threshold in the direction opposite to the direction of change in the input signal level. It should be noted that the limiting diode is connected directly to the filter at point 34, with no full-band amplifier inserted. If the filter's capacitor charge current is a high amplitude transient, it is usually supplied by this diode. Thus, the recovery time of this circuit is less than if the capacitor charging current is supplied by resistor 26 using a buffer amplifier. The circuit elements of the filter have a transient and pass through

This extender can be used to remove noise from video signals that contain noise that has not been previously acted upon by the compressor. This results in a loss of low level detail, but this loss is almost undetectable by ordinary observation. For example, when transmitting video signals from recording, the operator may increase the gain of each channel 9 and 11 to sufficiently reduce the noise introduced by the defective tape head to an acceptable level. It can be seen that when the compressor and the expander are used together, it is necessary to combine the characteristics of the second channel into their auxiliary signals in a strictly consistent state.

Since the present invention can also be applied to color video signals, it is particularly advantageous not only for reducing the noise of the wide band but also for reducing the wave pattern generated from the interaction between the color carrier and the FM carrier of the video recorder.

This compression and stretching can be applied individually to the R, B and G signals, or more preferably to the Y, I and Q signals or sometimes to the composite color signal when the component signals are not independently obtained as such. have. In all cases, it is necessary to subtract the color subcarriers (at least 50 decibels) from each secondary signal, i.e., if the subcarriers are not removed from the second channel, this subcarrier renders the compressor and the extender inoperable. Noise reduction is impossible when the composite signal is processed. Therefore, a lowpass filter (maintaining a cutoff frequency of about 2MC) is needed for each channel,

When the cutoff frequency is increased (eg, within the 400KC range of the subcarrier frequency) to allow the sidebands of the subcarrier to pass, the total range of noise reduction is increased, but every time there is a drastic change in the amplitude or phase of the subcarrier. You will have to accept the introduction of "dead time" (which can be calculated in ms). "Floating time" refers to the time when the signal level drops below the limit level but as an auxiliary signal, i.e., the output of the second channel stays at the limit level. According to the MTSC standard method using 525 scan lines, the noise reduction of the Y channel can be obtained in the frequency band from 100KC to 3.58MC-400KC = 3.18MC, and from 400KC to 1.5MC in the I channel and 400KC in the Q channel. It can be obtained in the frequency range of up to 500KC. This dead time hinders the noise reduction that occurs after a large, sudden color signal transition, but the burst of noise is incidentally masked by the transition portion of the color signal and usually the resulting luminance change. I will throw it away.

The choking effect of this sideband is compared with the operation described above for the black and white method using a buffer amplifier between the filter circuit and the limiter circuit. In no case does the diode enter the fabrication of the filter network in the new state given by the generation of the transient current. As the diode interacts with the filter, it was first described that the filter becomes nonlinear when the amplitude of the transient is large and automatically quickly becomes suitable for the new state. This state of operation implies that any similar arrangement in response to the carrier and its sidebands is possible.

The idealized operation can be obtained by a subcarrier resonant circuit having a high Q (narrow bandwidth) when the color side band and the high frequency Y component are at a low level, and a noise reduction action is possible under these conditions. When either or both of the color sideband and the high frequency Y component exceed a predetermined level, the Q of this known circuit drops momentarily and dissipates the energy introduced by the sideband so that it is not stored in the reactance element. . As soon as the energy from the sideband disappears, the Q of the circuit increases and the noise reduction action recovers again.

6 shows a suitable circuit by the circuit operation described so far. In addition to the capacitor 60 and the resistor 62, which define the lower frequency of the noise reduction band as shown in FIG. A series resonant circuit forms a shunt for this signal path. Either of these two resonant circuits can eliminate one circuit, but in general, both circuits are used simultaneously. The low level sideband component is transmitted to the output stage without dropping the level, and the resonant circuit

Noise reduction then covers the Y channel from 100 KC up to the frequency except for the sharp notch at the subcarrier frequency. The I and Q channels are covered by a frequency defined by the bandwidth of the resonant arc (in some cases a plurality of resonant circuits), for example, 50 KC to the frequency range of the upper limit of the frequency.

The above description mainly relates to the NTSC and PAL schemes, but of course it is also applicable to the SECAM scheme. However, when applied to the SECAM method, some circuit changes are necessary because the subcarriers are frequency modulated. For example, the resonant circuits 54A and 56A of FIG. 6 can be replaced with a plurality of such resonant circuits (or a plurality of parallel resonant circuits in series) in parallel and they can work together to cover all subcarrier deviation bands. Alternatively, the FM subcarrier may be taken out of the tap, limited and tracked to cancel the subcarrier of the synthesized signal transmitted to the auxiliary signal of the compressor or expander.

Techniques such as those described above for black and white video signals can also be used to handle applications to broadband metrology, where the frequency range to the auxiliary signal must be changed as appropriate. When the carrier frequency method is used, it is good to divide the signal into a plurality of independent bands and provide their own noise reduction device. Various carriers

In an audio compressor or expander, the limiting action in the low frequency and high frequency channels is achieved by moving the upper and lower limits of the cutoff frequency of each band downward and upward, respectively, when the channel's output puts a certain level. The following description is limited to the high frequency channel, but the opposite treatment in the low frequency channel is apparent in this description. In this way, the second channel of the compressor (corresponding to channels 9 or 11 in FIG. 1) at the high signal low signal level amplifies the signal of 2 KC or more. When the amplitude signal is generated at a frequency higher than this frequency, the cutoff frequency of the filter shifts upwards. Figure 7 shows a suitable filter and limiter circuit to achieve this.

Capacitor C ₁₀ , inductor L ₁ and low profile R ₁₁ form a common highpass filter. The voltage control resistor 108, which may consist of a balanced diode circuit with means for changing the bias for the diode, is configured in parallel with the resistor R ₁₁ and biased to show high impedance during normal operation. However, the output signal is amplified by the amplifier 110, rectified and filtered by the circuit 112 to generate a control signal for the voltage control resistor 108 to lower the resistance of the voltage control resistor 108 when the amplitude of the signal is added. this

This temporary cut-off frequency filter can take a variety of other forms, and the controlled elements themselves include hall effect devices, reactance tubes, voltage control capacitors, light control resistors, transistors or field effect transistors.

It can be configured to give full isolation at high signal levels. One method is to monitor the alternating current through the voltage controller 108. When this current exceeds a certain value, the value obtained from this current is sent to the control circuit, which causes the control circuit to lower the resistance. At this time, the output signal is interrupted and stays intact until the input voltage or frequency distribution changes to reduce the sampled current.

Alternatively, a simple frequency discrimination network can be inserted at either side of the output of the circuit shown in control amplifier 110 or FIG. 7 for the purpose of lowering the threshold for very high or very low frequencies. This lowering of the frequency threshold is necessary to prevent overloading of some type of highly equalized recording device. The audio compressor and the extender using the second channel described in connection with FIG. 7 have a tape recording,

Claims (1)

A first channel section for generating a signal having a linear amplitude fluctuation range with respect to an input signal of the circuit device over the entire bandwidth of the signal, at least one second circuit section affecting the amplitude and including a frequency selective filter; Consisting of a coupling device for coupling the output of the first and second circuit parts to increase the output from the first time in the case of signal compression by the output of the two circuit parts and to reduce the output of the first circuit part in the case of signal extension, A circuit device for compressing or stretching an amplitude variation range of an information signal, wherein the input signal of the second circuit portion has a linear amplitude variation range with respect to the input signal of the circuit apparatus when the circuit apparatus is operated as a compressor. When operating as a linear amplitude fluctuation range with respect to the output signal of the circuit arrangement, the frequency selective filter has a signal level in the circuit arrangement. And the passband is automatically varied so as to narrow, and at a high signal level, the output signal of the second circuit means exceeds an amplitude of any lower level such that the output signal of the second circuit means is in one unit or less than the output signal of the first circuit means. A circuit device for compressing or extending the amplitude fluctuation range of an information signal, wherein the signal is suppressed.

Images