RU2165127C2 - Method and device for automatic setting of peak electrical broadcast signals to desired level in controlling relative mean power - Google Patents

Abstract

FIELD: radio engineering. SUBSTANCE: peak values are automatically set in channels carrying high- and low-frequency components of output-signal envelope using preset threshold level and desired level of relative mean power. Proposed method and device provide for regulation of desired level of relative mean power which ensures equal reproduction volume of different sound programs, increase in relative mean power of signal being converted by 1.5 to 4 times while maintaining its peak level, reduction of broadcast transmitter power by 1.5-2 times at same coverage zone while maintaining high sound quality responded by different persons, and reduction of cost of sound-recording system for public address systems by two times while enhancing intelligibility of voice messages by 10-15% under high noise conditions. EFFECT: provision for non-smearing limitation and companding of broadcast signals without enriching their spectrum, enlarged functional capabilities. 2 cl, 10 dwg

Description

 The invention relates to techniques for transmitting signals and can be used when recording and reproducing audio signals.

 A known method [1] of automatically adjusting the peak values of electric broadcast signals to a given level while stabilizing the relative average power, which includes converting the input analog signal to digital and the inverse conversion of the digital signal to analog output, as well as the operation of controlling the transmission coefficient of the link in the broadcast signal channel wherein the control signal is generated from the input and output signal, for which the envelope of the signal is isolated, it is filtered, and the regulation of the transmission coefficient according to a given law is performed when the signal exceeds a certain threshold level. As a rule [1], the envelope obtained by rectifying the broadcast signal and subsequent low-pass filtering is used for the control signal.

 Due to the inertia of this method of regulation, distortion of broadcast signals and, in particular, large volume differences at the junction of programs with different spectra (for example, speech - music) that are identical in peak level occur. In addition, the regulation in this way does not allow to sufficiently increase the relative average power (OSM) of the audio signal without making any noticeable changes in it.

 The closest variation to the proposed one is a method of inertialess conversion of the dynamic range of an audio signal [2], in accordance with which regulation is carried out using a nonlinear transformation of the analytical (Hilbert) envelope of the audio signal. However, the strict regulation characteristic, which is not adaptable to the properties of the broadcast signal, leads to violations of the artistic correspondence of the volume transitions and does not provide a sufficient increase in the OSM signal. In addition, this method of regulation is characterized by a significant level of non-linear distortion and, as a consequence, the expansion of the signal spectrum.

 It is also known a device [3] for implementing known methods for converting the dynamic range of audio signals, comprising an orthogonal signal generating unit, an envelope extraction unit, a control signal unit containing a correction circuit and a threshold input of the device, and transmission coefficient control units.

 The disadvantages of the known device include a rigid control characteristic that is not adaptable to the properties of the broadcast signal (which leads to violations of the artistic correspondence of the volume transitions); insufficient increase in the OSM signal; significant level of nonlinear distortion and the expansion of the spectrum of the input signal.

The known method and device can be considered as a prototype. The task of the invention is to eliminate the above disadvantages, namely:
- inertia-free limitation, compression and companding of broadcast signals in the absence of enrichment of their spectrum;
- a significant increase in the OSM signal;
- equalization of the volume of diverse programs (speech - music), that is, adaptive volume control;
- expansion of the loud dynamic range.

The problem is solved in that in the method for automatically adjusting the peak values of the electrical signals of audio broadcasting to a predetermined level while stabilizing the OSM, which includes the following operations: analog-to-digital conversion of the audio broadcasting signal to the input digital sequence Z _in , functional processing and digital-to-analog conversion of the output digital sequence Z _o in the output analog signal, while the control signal of the transmission coefficient in the processing path is generated from the input and the output digital sequence, which is isolated envelope signal is performed its filtering and regulating coefficient by a given law transmission is performed in excess of broadcasting a certain threshold level signal, according to the invention in the sequence Z _Rin using N-point transform Hilbert formed orthogonal signal, calculating an envelope and the cosine of the instantaneous phase of the input signal, the envelope is divided into low and high frequency components, which are separately regulated under the influence control signal it, after which the envelope is reduced by summing the transformed components to obtain an output digital sequence Z _O by multiplying the restored envelope of the cosine of the instantaneous phase, wherein for generating a control signal sequentially on each segment containing kN samples, determine the peak value of the input | Z _Bx | _max and output | Z _o | _max sequences and relative average power P _OS.out. the output sequence, and the control signal is formed, on the one hand, when comparing the digital peak values | Z _I | _{the maximum} input signal with a digital threshold voltage level Z _pores , and on the other hand, when comparing P _OSM.output with a threshold level of relative average power P _OSM.pore so that when the inequality | Z _I | _max > Z _then the control signal changes the transmission coefficient in the processing path, implementing the envelope expansion operation, if P _OSM.output > P _OSM.pore , or its compression operation if P _OSM.output <P _OSM.pore , and the degree of expansion or compression determined in accordance with the magnitude of the difference P _OSM.out - P _OSM.pore , and when the inequality | Z _I | _max ≤ Z _{then the} control signal sets the transmission coefficient equal to unity, while the threshold values of Z _then and P _OSM.pore , as well as the parameters k and N are set in accordance with the requirements for the quality characteristics of the output analog signal.

Thus, the proposed method automatically controls the peak values of broadcast signals in the low-frequency (LF) and high-frequency (HF) paths of the envelope of the input signal, which ensures inertia-free control and the absence of “spurious” enrichment of the spectrum of the broadcast signal. The introduction of two threshold levels: Z _pore - according to the level of peak values of the input signal and P _OSM . _Pore - according to the level of the OSM of the output signal provides a significant increase in OSM, equalization of the volume of various programs and allows you to expand the volume dynamic range.

 The problem is also solved in that the device for automatically controlling the peak values of electric broadcast signals to a predetermined level while stabilizing the relative average power, comprising an orthogonal signal generating unit, an envelope extraction unit, a control signal unit containing a correction circuit and a threshold device input, and coefficient adjustment units transmission, according to the invention, additionally included blocks of analog-to-digital and digital-to-analog conversion, block separation the low-frequency and high-frequency components, the gear ratio meter, the envelope recovery unit, the signal recovery unit, and the transmission coefficient control units are not included in the paths of the original and orthogonal signals, but in the paths of the low and high-frequency components of the envelope, and the control signal unit is additionally equipped inputs for connecting the input and output signals of the device and for setting the threshold level of the relative average power of the output signal.

 The proposed device allows to implement the proposed method for automatically adjusting the peak values of broadcast signals to a given level while stabilizing the OSM.

фиг. 2 изображает гильбертовскую огибающую

сформированную из исходного Z_вх и ортогонального

сигналов;
фиг. 3 изображает мгновенную фазу

сформированную из исходного Z_вх и ортогонального

сигналов;
фиг. 4 изображает низкочастотную S_нч(t) и высокочастотную S_вч(t) составляющие огибающей S(t) = S_нч(t) + S_вч(t);
фиг. 5 изображает компрессирующую 1, экспандирующую 2 и линейную 3 характеристики регулирования составляющей S_нч(t) огибающей;
фиг. 6 - структурная схема предлагаемого устройства;
фиг. 7 - структурная схема известного устройства [3], в котором:
1 - блок формирования ортогонального сигнала, 2 - блок выделения огибающей, 3 - блок сигнала управления, содержащий схему коррекции и пороговый вход 4 устройства, 5 - блок регулирования коэффициента передачи;
фиг. 8 изображает совмещенные осциллограммы сигналов до и после регулирования с увеличенной ОСМ музыкального сигнала;
фиг. 9 изображает совмещенные осциллограммы сигналов до и после регулирования при улучшении соотношения "речь - музыка";
фиг. 10 иллюстрирует увеличение ОСМ речевого сигнала для систем массового оповещения.BRIEF DESCRIPTION OF THE DRAWINGS
The invention is further illustrated by oscillograms and drawings, in which:
FIG. 1 shows a combined on a single graph the original signal Z _Rin and an output signal Z _O, and formed by the Hilbert transform, orthogonal to the original signal

FIG. 2 shows a Hilbert envelope

formed from the original Z _in and orthogonal

signals;
FIG. 3 depicts an instant phase

formed from the original Z _in and orthogonal

signals;
FIG. 4 shows the low-frequency S _woofer (t) and the high-frequency S _hf (t) envelope components S (t) = S _woofer (t) + S _treble (t);
FIG. 5 shows a compression 1, expanding 2, and linear 3 characteristics of the regulation of the S _{component of the low} (t) envelope;
FIG. 6 is a structural diagram of the proposed device;
FIG. 7 is a structural diagram of a known device [3], in which:
1 - an orthogonal signal generating unit, 2 - an envelope extraction unit, 3 - a control signal unit containing a correction circuit and a threshold input 4 of the device, 5 - a transmission coefficient control unit;
FIG. 8 depicts combined waveforms of signals before and after adjustment with an enlarged OSM of a music signal;
FIG. 9 depicts combined waveforms of signals before and after adjustment while improving the speech-music ratio;
FIG. 10 illustrates an increase in OSM speech signal for public address systems.

ортогонального исходному Z_вх, т.е. сопряженного с ним по Гильберту [1, разд. 4.3] (иллюстрируется осциллограммой фиг. 1).The proposed method is implemented by sequentially performing the following operations:
1. The formation of a sequence of Z _I using the N-point Hilbert transform [1, 4] signal

orthogonal to the original Z _in , i.e. Hilbert conjugate to it [1, Sec. 4.3] (illustrated by the waveform of Fig. 1).

(фиг. 2) и косинуса мгновенной фазы cos φ (t) = Z_вх/S(t) из исходного Z_вх и ортогонального

сигналов.2. Formation of the Hilbert envelope

(Fig. 2) and the cosine of the instantaneous phase cos φ (t) = Z _in / S (t) from the original Z _in and orthogonal

signals.

3. Separation of the envelope S (t) into low-frequency (LF) S _LF (t) and high-frequency (HF) S _HF (t) components (Fig. 4a, b).

4. Regulation of the LF component S _LF (t) in the control link, the amplitude characteristic of which, depending on the polarity and magnitude of the control signal, is compression or expander. If the control signal is zero, then the amplitude characteristic is linear (Fig. 5).

5. Regulation HF component _HF S (t) in the control unit of the linear amplitude response which is determined by the transmission link coefficient LF-regulating component, and a scale factor γ is not determined by the requirement to the quality characteristics of the output analog signal.

6. Restoring the envelope by summing the generated LF and HF components: S _stand (t) = S _woof.wd (t) + S _wf.wd (t).

7. Preparation output digital sequence by multiplying the operation Z = S _{O adt} (t) cosφ (t).

8. Adaptive generation of a control signal taking into account a given input signal level (set by a digital threshold voltage Z _pores ) and a given OSM value of an output signal (set by a digital threshold level P _OS.por ).

In the process of regulation carry out the measurement of peak values of the input signal | Z _I | _max and OSM output signal P _OSM.out . Measurements are taken sequentially at each time interval, the duration of which is set equal to kNIF ≈ 40 ms (F is the sampling frequency) for information broadcasting and approximately 160 ms for art broadcasting. Parameters k, N and F are determined by the requirements for the quality characteristics of the output analog signal. So, for example, to control the broadcast signal with the upper cutoff frequency of the spectrum equal to 10 kHz, k = 7, N = 512, F = 22 kHz are chosen.

When the peak value of the signal | Z _I | _{the maximum of the} given threshold Z _pores and in the case of deviation of P _OSM.output from the set value of P _OSM.pore as a result of regulation, the LF component of the envelope is multiplied either by the compression component (if P _OSM.output <P _OSM.pore ) or by the expander (if P _OSM.output > P _OSM.pore ) amplitude response of the control link. In this case, the RF component of the envelope varies linearly in proportion to the change in the RF component in the control interval and taking into account the magnitude of the scale factor γ, which is chosen equal to 0.5 for compander systems of long distance audio channels, 1.0 for art broadcast systems and 1,5 - for information broadcasting and mass warning systems.

 A device for automatically adjusting the peak values of broadcast signals to a given level during OSM stabilization is shown in FIG. 6. It contains an analog-to-digital conversion unit 1, an orthogonal signal generating unit 2, an envelope and cosine extraction unit of the instantaneous phase of the signal 3, an envelope separation unit for low and high frequency components 4, a transmission coefficient adjustment unit of components 5, a transmission coefficient meter 6, envelope recovery unit 7, signal recovery unit 8, digital-to-analog conversion unit 9, control signal unit 10.

ортогональный исходному (фиг. 1). Формирование ортогонального сигнала осуществляют с помощью специальной, вычислительно достаточно простой процедуры синтеза такого сигнала в частотной области на основе БПФ [4]. Этим обеспечивают высокую точность формирования в блоке 3 гильбертовской огибающей

(фиг. 2) и косинуса мгновенной фазы cosφ(t) = Z_вх/S(t) вещательного сигнала. В блоке 4 огибающую разделяют на низкочастотную S_нч(t) и высокочастотную S_вч(t) составляющие (фиг. 4) и по отдельности регулируют в блоках 5 под воздействием сигнала управления, после чего огибающую восстанавливают, суммируя составляющие в блоке 7. Выходную цифровую последовательность Z_вых получают в блоке 8 путем перемножения восстановленной огибающей и косинуса мгновенной фазы Z_вых = S_вст(t)cosφ(t). Цифроаналоговым преобразованием в блоке 9 получают выходной аналоговый сигнал.The device in accordance with the proposed method works as follows. The input analog broadcast signal is converted by an analog-to-digital converter 1 into a digital Z input _circuit , which is fed to the inputs of the blocks for generating an orthogonal signal 2, for extracting the envelope and cosine of the instantaneous phase 3 and for the control signal 10. In block 2, a signal is generated using the N-point Hilbert transform

orthogonal to the original (Fig. 1). The formation of the orthogonal signal is carried out using a special, computationally simple enough procedure for the synthesis of such a signal in the frequency domain based on the FFT [4]. This ensures high accuracy of formation in block 3 of the Hilbert envelope

(Fig. 2) and the cosine of the instantaneous phase cosφ (t) = Z _in / S (t) of the broadcast signal. In block 4, the envelope is separated into a low-frequency _LF S (t) and high-frequency S _RF (t) components (Fig. 4) and individually adjusted in blocks 5 under the influence of the control signal, after which the envelope is reduced by summing the components in the block 7. The output digital the Z _o sequence is obtained in block 8 by multiplying the reconstructed envelope and the cosine of the instantaneous phase Z _o = S _cn (t) cosφ (t). Digital-to-analog conversion in block 9 receive the output analog signal.

The transmission coefficients K _low (V _y ) and K _high (V _y ) blocks 5 determine the nature of the regulation, respectively, the LF and HF components of the envelope:
_Nch.r S _LF = K (V _y) S S _LF and _{HF vch.r} = K (V _y) S _HF,
where the indices correspond to the envelope components at the input and output of the control blocks 5.

Compressing or expanding the LF component of the envelope, that is, the function K _LF (V _y ) (Fig. 5), is determined by the magnitude and sign of the control signal V _y produced by block 10. The change in the amplitude of the RF component is proportional to the output signal of the gear ratio meter 6, then there is K _hf (V _y ) = γ K _woof (V _y ).

The low-frequency component (Fig. 4a) contains infrasonic frequencies, the spectrum of which is located below the spectrum of the broadcast signal. Therefore, a slow change in the transmission coefficient K _LF (V _y ) does not lead to a “spurious" enrichment of the signal spectrum. A linear (proportional) change in the RF component in the process of regulation is also not accompanied by distortion of the signal spectrum.

In block 10, adaptive control signal generation is performed at a given level of peak values of the input signal | Z _in | _max and set value of the OSM output signal. In this case, the specified level of peak values of the input signal is determined by the digital threshold Z _pores , and the level of the OSM output signal is determined by the digital threshold P _OSM.pore .

The control signal V _y form, on the one hand, when comparing the digital peak signal levels | Z _I | _max with a threshold of Z _pores , and on the other hand, when comparing the digital values of the OSM of the output signal P _OSM.out with another threshold level P _OS.pore so that when the inequality | Z _I | _max > Z _then the control signal changes the transmission coefficient, including the envelope expansion operation, if P _OSM.output > P _OSM.pore , or the operation of its compression if P _OSM.output <P _OSM.pore , and the degree of expansion or compression is determined in accordance with with the difference value P _OSM.out - P _OSM.pore , and when the inequality | Z _I | _max ≤ Z _{then the} control signal sets the transmission coefficient equal to one. The threshold values of Z _pores and P _OSM.pore , as well as the parameters k, N, and γ, are set in accordance with the requirements for the quality characteristics of the output analog signal.

 In FIG. 7 is a structural diagram of a known device [3] for dynamic compression of a broadcast signal. A comparison of the circuits in FIG. 6 and 7 illustrate the fundamental difference between the devices.

 The authors performed experimental modeling of the claimed method and device based on it on real broadcast signals. The simulation was carried out on a personal computer of the latest generation, supplemented by specialized boards. Simulation illustrations are presented on combined waveforms of signals before and after regulation: FIG. 8 is an increase in the OSM of the music signal; 9 is an example of improving the speech-music ratio; FIG. 10 - a large increase in the OSM speech signal for public address systems.

According to the authors, the proposed method and device provide:
- inertia-free restriction of broadcast signals;
- lack of spectrum enrichment, which makes them applicable to a wide class of signals;
- processing broadcast signals with studio quality, which is a consequence of the high accuracy of the formation of the orthogonal signal;
- the ability to use in compander systems of long-distance channels of sound broadcasting;
- stabilization of a given OSM, which guarantees equal playback volume of diverse sound programs;
- a significant (3 ... 4 times) increase in the OSM of the converted signal while maintaining its peak level, which makes it possible to increase the intelligibility of voice messages in mass notification systems (by 10 ... 15% compared with signals processed by regulation devices known to the authors ) in conditions of a high level of noise and interference;
- increase the volume of broadcasts, equivalent to an increase in the power of radio transmitting devices and acoustic emitters by 1.5 ... 2 times.

 The proposed method and device can be widely used on all channels of sound broadcasting in our country and abroad. This will significantly improve the quality of sound broadcasting services.

Literature
1. Sound broadcasting. / A.V. Coming, P.M. Zhmurin, I.F. Zorin et al .: Ed. Yu.A. Kovalgin: Reference. - M .: Radio and communications, 1993, 464 p.

 2. Ishutkin Yu.M., Plyushchev V.M. Transformation of the envelope of sound signals // Technique of cinema and television. - 1983, N 4, p. 3-6.

 3. Copyright certificate of the USSR N 1058077, cl. H 04 B 1/64. Ishutkin Yu.M. , Plyushchev V.M., Uvarov V.K. Device for converting the dynamic range of audio signals.

 4. Marple. ml S.L. Digital spectral analysis and its applications: Per. from English - M .: Mir, 1990, 584 p.

Claims (2)

1. A method of automatically adjusting the peak values of the electrical signals of audio broadcasting to a predetermined level while stabilizing the relative average power, including the operation of generating a control signal and adjusting the transmission coefficient according to a given law when a predetermined threshold level is exceeded, characterized in that the analog-to-digital conversion of the input the sound broadcasting signal into a digital sequence z _Bx , from which using the N-point Hilbert transform form the orthogonal signal is z _{inx ort} , then z _inx and z _{inx ort} are used to form the envelope and cosine of the instantaneous phase of the signal, then the envelope is divided into low-frequency and high-frequency components, which separately regulate, while generating a control signal on the control interval containing k · N samples carry out determining the peak values of the input | z _I | _max and output | z _o | _max sequences and relative average power P _{OSM. of the} output sequence, and the control signal is formed taking into account the comparison of digital peak values | z _in | _max with a digital threshold voltage level z and P _{OSM.vyh pores} with a threshold level relative average power P _OSM.por so that when the inequality | z _Bx | _max > z _then the control signal changes the transmission coefficient of the low-frequency component of the envelope, realizing the operation of its expansion if P _OSM.output > P _OSM.pore , or the operation of its compression if P _OSM.output <P _OSM.pore , and the degree of expansion or compression determined by the magnitude of the difference P _OSM.out - P _OSM.pore , and when the inequality | z _I | _max ≤ z _pores, the control signal sets the transmission coefficient of the low-frequency component equal to unity, and the transmission coefficient of the high-frequency component of the envelope is changed proportionally to the change in the coefficient of the low-frequency component taking into account the scale factor γ, after which the envelope is restored by summing the converted components and obtaining the output digital sequence z _o by multiplying the restored envelope with the cosine of the instantaneous phase, and then carry out digital conversion of the z _output sequence into an analog output signal, and the threshold values of z _pores and P _{OSM output} , as well as the parameters k, N, and γ, are set in accordance with the requirements for the quality characteristics of the output analog signal. 2. A device for automatically adjusting the peak values of electric broadcast signals to a predetermined level during stabilization of the relative average power, comprising an orthogonal signal generating unit, a control signal unit with a threshold input, transmission coefficient control units, characterized in that an analog-to-digital conversion (ADC) unit is additionally included ), which converts the input analog broadcast signal into a digital sequence, an envelope and cosine extraction block of the instant phase, cat ing forms the envelope

and the cosine of the instantaneous phase cosφ (t) = z _in / S (t), where z _in and

accordingly, the digital sequence generated by the ADC and the signal generated in the block for generating the orthogonal signal, the block for dividing the envelope into low (LF) and high-frequency (HF) components, and the transmission coefficient control units separately regulate the LF and HF components and provide the nature of regulation, respectively LF and HF components of the envelope S _{LF R} = K _LF (V _y ) · S _LF and S _{HF R} = K _HF (V _y ) · S _HF , where K _LF (V _y ) and K _HF (V _y ), respectively transmission coefficients of the LF and HF components, V _y - control signal, forming which is represented by the control signal block, S and S _p are the envelopes at the input and output of the blocks for controlling the transmission coefficients, where K _hf (V _y ) = γ · K _lf (V _y ), where γ is the scale factor, which is set in accordance with the requirements of the qualitative characteristics of the output analog signal, the envelope recovery unit, summing the bass and treble components after adjustment, the signal recovery unit, in which the output sequence is obtained by multiplying the restored envelope and the cosine of the instantaneous phase, digital-to-analog eobrazovatel, wherein the receive analog output signal, wherein the digital sequence z _Rin is input to the block formation of the orthogonal signal and a block selection envelope and cosine of instantaneous phase signal, and the block control signal generates a control signal by comparing the digital peak input levels | z _Bx | _max with a threshold of z _pores and when comparing the digital values of the relative average power of the output signal P _OSM.out with the threshold level P _OSM.poor in such a way that when the inequality | z _in | _max > z _then the control signal changes the transmission coefficient K _low (V _у ), including the envelope expansion operation, if P _OSM.output > P _OSM.por , or the operation of its compression if P _OSM.output <P _OSM.pore , and when the inequality | z _i | _max ≤ z _{then the} control signal sets the transmission coefficient equal to one.

Images