KR100513543B1 - BTSC encoder - Google Patents

Abstract

The BTSC encoder 200 receives a digital left channel audio signal L and receives a left digital high pass filter 212 and a digital right channel audio signal R for generating a digital left filter signal and generates a digital right signal. A right digital high pass filter 214, and an adder 216 for generating a digital sum signal L + R by adding the digital left and right filter signals, and converting one of the digital left and right filter signals from the other side of the digital left and right filter signals. A matrix including a subtractor 218 for subtracting the digital difference signal LR, a difference signal processor 230 for digitally processing the digital difference signal LR, and a digital sum signal L + R. A sum channel processor 220 for processing.

Description

BTC encoder

1 is a block diagram of a BTSC encoding system of the prior art, in particular BTSC encoder 100 as defined by the BTSC standard. The encoder 100 receives left and right channel audio input signals (indicated by "L" and "R" in FIG. 1, respectively) and generates a condition sum signal and an encoded differential signal therefrom. Although the prior art system and the system of the present invention are described as useful for encoding left and right audio signals of stereophonic signals transmitted as television signals, the BTSC system is also a separate audio signal, e.g. It should be understood that the present invention provides a means for encoding audio information of a language. Moreover, the noise reduction component of the BTSC encoding system can be used for purposes other than television broadcasting, such as to improve audio recording.

The system 100 includes an input terminal 110, a sum channel processing stage 120, and a difference channel processing stage 130. The input terminal 110 receives the left and right channel audio input signals and generates a sum signal (denoted by "L + R" in FIG. 1) and a difference signal (denoted by "L-R" in FIG. 1) therefrom. In stereophonic signals, the sum signal L + R can itself be used for monophonic audio reproduction, and it is well known that the signal is decoded by a conventional monophonic audio television set for sound reproduction. In the stereo sound set, the sum signal and the difference signal can be added to or subtracted from each other so as to recover the two original stereo sound signals (L) (R). Input stage 110 includes two signal adders 112 and 114. The adder 112 adds the left and right channel audio input signals to generate the sum signal, and the adder 114 subtracts the right channel audio input signal from the left channel audio input signal to generate the difference signal. As mentioned above, the sum signal L + R is transmitted through the transmission medium at the same signal-to-noise ratio as the conventional monophonic signal. However, since the difference signal L-R is transmitted through a very noisy channel, especially in the high frequency portion of the spectrum, the difference signal decoded has a poor signal-to-noise ratio due to the reduced dynamic range of the noisy medium and the medium. Dynamic range is defined as the signal range between the level of the noise floor and the maximum level at which signal saturation occurs. In the difference signal channel, the dynamic range decreases at high frequencies. In addition, since the difference signal is additionally processed rather than the sum signal, the dynamic range can be substantially preserved.

In particular, the sum channel processing stage 120 receives the sum signal and generates a condition sum signal therefrom. This processing stage 120 includes a 75 kHz preemphasis filter 122 and a band limiter 124. The sum signal is applied to the input of the filter 122, from which the output signal is applied to the input of the band limiter 124. The output signal generated by the band limiter is a condition sum signal.

The difference channel processing stage 130 receives the difference signal and generates an encoded difference signal therefrom. The processing stage 130 includes a fixed pre-emphasis filter 132 (shown as a cascade of two filters 132a and 132b), a variable gain amplifier 134 in the form of a voltage controlled amplifier, and a variable pre-emphasis / de-emphasis. Filter (referred to as " variable emphasis filter ") 136, overmodulation protector and band limiter 138, fixed gain amplifier 140, band pass filter 142, RMS level detector 144, fixed gain amplifier ( 146, a band pass filter 148, an RMS level detector 150, and a reversible generator 152.

The difference signal is applied to the input of the fixed pre-emphasis filter 132, which generates an output signal from which the output signal is applied to the input terminal of the amplifier 134 via line 132d. The output signal generated by the reversible generator 152 is applied to the gain control terminal of the amplifier 134 via line 152a. Amplifier 134 amplifies the signal on line 132d using a gain proportional to the signal value on line 152a to generate an output signal. The output signal generated by the amplifier 134 is applied to the input terminal of the variable emphasis filter 136 through the line 134a, and the output signal generated by the RMS detector 144 is filtered through the line 144a. Is applied to the control terminal of 136. The variable emphasis filter 136 generates an output signal by preemphasizing or deemphasing the high frequency portion of the signal in line 134a under the control of the signal in line 144a. The output signal generated by the filter 136 is applied to the inputs of the overmodulation protector and the band limiter 138 from which an encoding difference signal is generated.

The coded difference signal is applied to the inputs of the fixed gain amplifiers 140 and 146 via the feedback path 138a, and these amplifiers amplify the coded difference signal with Gain A and Gain B, respectively. The amplified signal generated by the amplifier 140 is applied to the input of the bandpass filter 142, from which the output signal is applied to the input of the RMS level detector 144. This detector generates an output signal as a function of the RMS value of the input signal level received from filter 142. The amplified signal generated by the amplifier 146 is applied to the input of the bandpass filter 148, which generates an output signal from which the output signal is applied to the input of the RMS level detector 150. This detector generates an output signal as a function of the RMS value of the input signal level received from filter 148. The output signal of detector 150 is applied to reversible detector 152 via line 150a, which generates a signal on line 152a representing the inverse of the signal value of line 150a. As mentioned above, the output signals generated by the RMS level detector 144 and the reversible generator 152 are applied to the filter 136 and the amplifier 134, respectively.

The present invention relates to a stereophonic audio encoder used for television broadcasting. In particular, the present invention relates to digital encoders for generating audio signals for use in broadcasting stereoscopic television signals in the United States or other countries.

In the 1980s, the Federal Communications Commission (FCC) proposed new regulations regulating the audio portion of television signals that allowed television programs to be broadcast and received in both channel audio, such as stereophonic sound. In these regulations, the FCC has been approved by the Electronic Industries Association and the National Association of Broadcasters and approved additional audio channel broadcasting methods called the Broadcast Television Systems Committee (BTSC) system. Special protection law was enacted. Such well-known criteria are sometimes called Multichannel Television Sounds (MTS), and the FCC document MULTICHANNEL TELEVISION SOUND TRANSMISSION AND AUDIO PROCESSING REQUIREMENTS FOR THE BTSC SYSTEM (OET Bulletin No. 60, Revision A, February 1986), MULTICHANNEL TELEVISION SOUND BTSC SYSTEM RECOMMENDED PRAETIECS (EIA Television Systems Bulletin No. 5, July 1985) published by the Electronics Industry Association. Television signals generated according to the BTSC regulations are called "BTSC signals".

The original monophonic television signal is carried only as a single audio channel. Because of the necessity of maintaining the composition of the monophonic television signal and the compatibility with the existing television set, the stereophonic sound system had to be placed in the high frequency region of the BTSC signal which constitutes the stereophonic sound channel with more noise than the monophonic audio channel. This is because the noise floor is essentially higher in the stereophonic signal than in the monophonic signal. The BTSC standard overcomes this problem by defining an encoding system that provides an additional signal for processing stereophonic audio signals. Before broadcasting the BTSC signal at the television station, the audio portion of the television program is encoded in a manner defined by the BTSC standard and upon reception of the BTSC signal the receiver (e.g. a television set) decodes the audio portion in a compensatory manner. This compensatory encoding and decoding ensures that the signal-to-noise ratio of the entire stereo audio signal is maintained at an acceptable level.

As shown in FIG. 1, the difference channel processing stage 130 is significantly more complicated than the sum channel processing stage 130. In addition to the compensatory processing performed by a decoder (not shown) receiving the BTSC signal, the additional processing performed by the difference channel processing stage 130 may be performed even in the presence of a high noise floor related to transmission and reception of the difference channel. Keep the signal-to-noise ratio of the difference channel at an acceptable level. The difference channel processing stage 130 dynamically generates a coded difference signal by dynamically compressing or reducing the dynamic range of the difference signal, so that the coded signal may be transmitted through a limited dynamic range transmission path associated with the BTSC signal, and receives a coded signal. By using the compensation method, the compressed difference signal can be extended to restore all dynamic ranges of the original difference signal. Difference channel processing stage 130 is a particular type of adaptive signal weighting system described in US Pat. No. 4,539,526, which is advantageous for transmitting signals having a relatively wide dynamic range through a transmission path having a relatively narrow frequency dependent dynamic range. It is known.

In summary, the difference channel processing stage may be considered to include a wideband compression unit 180 and a spectral compression unit 190. The wideband compression unit 180 is for generating a control signal for the variable gain amplifier 134 and the amplifier 134 in the form of a voltage controlled amplifier, and the amplifier 146, the bandpass filter 148, and the RMS level detector. A component of the feedback path, consisting of 150 and reversible generator 152. The bandpass filter 148 has a relatively wide passband weighted to a low audio frequency, so that in operation the output signal generated by the filter 148 and applied to the RMS level detector 150 substantially receives the encoded difference signal. Indicates. Accordingly, the RMS level detector 150 generates an output signal representing the weighted average of the energy levels of the encoded difference signal on the line 150a, and the reversible generator 152 represents the inverse of the weighted average on the line 152a. Occurs. In line 152a, this weighted average generates a signal representing the inverse. The signal on line 152a controls the gain of amplifier 134, which is inversely proportional to the weighted average of the energy levels of the encoded difference signal (i.e., weighted average weighted for low audio frequencies). ) Amplifies a signal with a relatively low amplitude and reduces the signal with a relatively wide amplitude, thereby "compressing" or reducing the dynamic range of the signal of line 132a.

The spectral compression unit 190 generates a variable emphasis filter 136, a control signal for the filter 136, and is composed of an amplifier 140, a band pass filter 142, and an RMS level detector 144. Contains components of the path. Unlike the filter 148, the bandpass filter 142 has a relatively narrow passband weighted to a high audio frequency. As is well known, the transmission medium associated with the differential portion of the BTSC transmission system has a frequency dependent dynamic range and the passband of the filter 142 corresponds to the spectral portion (ie, the high frequency portion) of this transmission path with the narrowest dynamic range. To be selected. In operation, the output signal generated by filter 142 and applied to RMS level detector 144 includes a high frequency portion of the encoding difference signal. Accordingly, the RMS level detector 144 generates an output signal indicating the energy level in the high frequency portion of the coded difference signal on the line 144a. The signal controls the pre-emphasis / de-emphasis applied by the variable emphasis filter 136 so that the spectral compression unit 190 is applied to the energy level of the high frequency portion of the encoding difference signal determined by the filter 142. The high frequency portion of the signal on the line 134a is dynamically compressed by the determined amount. By using the spectral compression unit 190 as described above, an additional signal compression is performed on the high frequency portion of the difference signal, which is a broadband provided by the variable gain amplifier 134 so that all compression is performed at a high frequency for compression at a low frequency. Combined with compression. This is done because the difference signal tends to be more noisy in the high frequencies of the spectrum. When the encoded difference signal is decoded by a spectral expander of a wideband expander and a decoder (not shown) in a compensating manner for the spectral compression unit 190 of the wideband compression unit 180 and the encoder, respectively, the difference processing stage 130 The signal-to-noise ratio of the LR signal applied to) is substantially preserved.

The BTSC standard strictly regulates the required operation of the 75 kHz pre-emphasis filter 122, the fixed pre-emphasis filter 132, the variable emphasis filter 136, and the band pass filters 142, 148 by the ideal analog filter. It is prescribed. In particular, the BTSC standard provides a transfer function for each of these components, which is described by the formula of an ideal analog filter. The BTSC standard also defines gains, gain A and gain B of amplifiers 140 and 146, and defines the operation of amplifier 134, RMS level detectors 144 and 150 and reversible generator 152. have. The BTSC standard also provides guidelines for the operation of overmodulation protectors and band limiters 138 and band limiters 124. In particular, the band limiter 124 and the overmodulation protector and the band limiter portion of the band limiter 138 are described as a low pass filter having a cutoff frequency of 15 kHz, and the overmodulation protection portion of the overmodulation protector and the band limiter 138. Is described as a limiting device that limits the amplitude of the encoded difference signal to 100% full modulation, where full modulation is the maximum allowable deviation level for modulating the audio subcarriers of the television signal.

Since the encoder 100 is defined by the formula of an ideal filter, it can be regarded as an ideal or theoretical encoder, and it is impossible for a person skilled in the art to physically realize a BTSC encoder that can accurately match the performance of the theoretical encoder 100. You will see that. Thus, the performance of all BTSC encoders deviates slightly from the theoretical ones and the BTSC criterion defines the maximum in the allowable amount of deviation. For example, the BTSC standard allows the BTSC encoder to provide at least 300dB of separation between 100Hz and 8,000Hz, where a signal applied to only one of the left or right channel inputs is separated from the left or right channel output of the other side. It is a measure of the amount of error.

The BTSC standard also defines a composite stereoacoustic baseband signal (hereinafter referred to as a "composite signal") used to generate the audio portion of the BTSC signal. This composite signal is generated using a conditional sum signal, an encoding difference signal, and a tone signal that is commonly referred to as a "pilot tone" or a simplified "pilot" and is a sinusoidal wave at a frequency fH (where fH is 15,734 Hz). The presence of a pilot in the received television signal indicates to the receiver that the television signal is a BTSC signal rather than a monophonic or other non-BTSC signal. The complex signal is multiplied by the oscillation waveform at twice the frequency of the frequency of the signal according to the cosine function cos (4πfHt) (where t is the time) in order to generate an amplitude modulated double sideband to suppressed carrier signal. Generated by adding the condition sum signal and the pilot tone.

2 is a graph of the spectrum of a composite signal. In FIG. 2, the main spectral band including the content of the condition sum signal is represented by "L + R", and the frequency shift coded difference signal (or the two spectral sidebands containing the content of "difference channel signal" is "LR"). The pilot tone is represented by an arrow at the frequency f H. As shown in Fig. 2, in the composite signal, the coded difference signal is used for 100% pre-modulation and the condition sum signal is used for 50% pre-modulation. Tons are used for 10% premodulation.

Stereo-acoustic televisions have been a great success and have performed well with conventional encoders, but practically all current BTSC encoders are constructed using analog circuitry technology. These analog BTSC encoders, especially analogue subchannel processing stages, are due to their increased complexity. It is relatively difficult and expensive to construct. Due to the variability of analog components, the selection of complex components and extensive calibration are required for the manufacture of an acceptable analog differential channel processing stage. Moreover, the propensity of analog components to deviate from their calibration operating points each time makes it difficult to manufacture analog differential channel processing stages that perform consistently and repeatedly within a given tolerance range. The digital channel processing stage can solve the problem of component selection, calibration, and performance change if it can be configured, and potentially increase performance.

Moreover, the analog characteristics of existing BTSC encoders are not convenient for digital devices that are newly developed and increasing in popularity. For example, current television programs can be stored using digital storage media such as hard disks or digital tapes, rather than conventional analog storage media, and the use of digital storage media will be further increased. Generating a BTSC signal from a digitally stored program requires converting the digital audio signal into an analog signal and applying the analog signal to the analog BTSC encoder. The digital BTSC encoder can be integrated with other digital equipment more easily by accepting the digital audio signal directly if it can be configured.

Although the digital BTSC encoder offers several advantages, it is not straightforward to construct the encoder using digital technology that is functionally identical to the ideal encoder 100 defined by the BTSC standard. One problem is that the BTSC standard defines all limiting components of the ideal encoder 100 by means of an analog filter transfer function. As is well known, it is possible to design a digital filter such that the size or phase response of the digital filter matches that of the analog filter, but without requiring a significant amount of processing capacity to process the sampled data at very high sampling rates or It is very difficult to match magnitude or phase response without significantly increasing complexity. Without increasing the sampling frequency or filter command, the amplitude response of the digital filter can be more closely matched to the analog filter at the expense of increased mismatch between the phase response of the two filters. However, even a small error in amplitude or phase reduces the amount of separation provided by the BTSC encoder, so that the digital BTSC encoder can closely match the amplitude and phase response of an ideal encoder of the type shown in FIG.

In digital BTSC encoders to provide acceptable performance, it is limited to preserve the characteristics of the analog filter of the ideal encoder 100. There are many techniques for designing digital filters to match the performance of analog filters, but generally these techniques produce digital filters (in the same order as analog filters) with amplitude and phase responses that exactly match the response of analog filters. Can not. The ideal encoder 100 is defined by an analog transmission function specified in the frequency domain, that is, the s-plane, and these transmission functions must be converted to the z-plane in order to design a digital BTSC encoder. This conversion can be performed in a "many-to-one" mapping manner from the s-plane to the z-plane, which is attempted to preserve time-domain characteristics. However, in this conversion the frequency domain response is aliased and can be changed significantly. This conversion can also be performed in an "one-to-one" mapping manner from the s-plane to the z-plane, which compresses the entire s-plane into a single circle z-plane. However, this compression can be a well known " frequency warping " between analog and digital frequencies. Although prewarping can be used to compensate for the effects of frequency warping, such prewarping can be used. It is not possible to completely eliminate the deviation from the required frequency response. These problems can be solved by manufacturing digital BTSC encoders that are effectively performed and are not overly complex or expensive.

Therefore, there is a need to overcome these problems and develop a digital BTSC encoder.

It is an object of the present invention to reduce or overcome the above mentioned problems of the prior art.

Another object of the present invention is to provide an adaptive digital weighting system.

It is another object of the present invention to encode an electrical information signal of a predetermined bandwidth so that the information signal is dynamically having a narrow portion that is dynamically limited in the first spectrum region rather than in at least one other spectral region of the predetermined bandwidth. An adaptive digital weighting system is recorded or transmitted through a limited frequency dependent channel.

Another object of the present invention is to provide a digital BTSC encoder.

It is still another object of the present invention to provide a digital BTSC encoder that prevents ticking, which can be a problem at substantially zero input signal levels.

Another object of the present invention is to provide a digital BTSC encoder using a sampling frequency that is a multiple of 15,734 Hz of the pilot tone signal frequency in order to prevent interference between the signal information of the encoded signal and the pilot tone signal.

It is still another object of the present invention to provide a digital BTSC encoder for generating a condition sum signal and an encoding difference signal that do not substantially contain signal energy at a pilot tone frequency of 15,734 Hz.

Another object of the present invention includes a sum channel processing stage for generating a condition sum signal and a difference channel processing stage for generating an encoded difference signal, wherein the sum channel processing stage is introduced into an encoded difference signal by the difference channel processing stage. SUMMARY To provide a digital BTSC encoder including a device for introducing a compensation phase error into a condition sum signal to compensate for a phase error.

Another object of the invention is a digital variable emphasis filter comprising a digital variable emphasis unit and the unit characterized by a variable coefficient transfer function, wherein the unit is a variable coefficient as a function of signal energy of the coded difference signal. It is to provide a digital BTSC encoder including a device for selecting the coefficient of the transmission function.

It is still another object of the present invention to provide a digital BTSC encoder including a complex modulator for generating a complex modulated signal from a condition sum signal and a coded difference signal.

Another object of the present invention is to provide a digital BTSC encoder that can be configured as a single integrated circuit.

Both these and other objects are achieved by an improved BTSC encoder which includes an input stage, a sum channel processing stage and a difference channel processing stage that can be realized with digital technology. In one aspect, the input includes a high pass filter that prevents the BTSC encoder from exhibiting "ticking". In another aspect, the BTSC encoder uses a sampling frequency equal to an integer multiple of the pilot frequency.

In another aspect, the sum channel processing stage generates a conditional sum signal, the difference channel processing stage generates an encoding difference signal, and the sum channel processing stage compensates for a phase error introduced into the encoded difference signal by the difference channel processing stage. And a device for introducing a phase error into the condition sum signal.

According to another aspect, the invention encodes an electrical information signal of a predetermined bandwidth such that the information signal has a narrow portion that is dynamically limited in the first spectrum region rather than in at least one other spectral region of the predetermined bandwidth. Provides an adaptive digital weighting system that is recorded or transmitted over a dynamically limited frequency dependent channel.

Other objects and advantages of the invention will be apparent to those skilled in the art from the following description in which various embodiments have shown the best mode of the invention. The present invention can be configured in other embodiments and can be modified in various respects without departing from the present invention. Accordingly, the drawings and the descriptions are for illustrative purposes only and are not intended to be limiting. The scope of the invention is defined in the claims.

Referring to the present invention in more detail based on the accompanying drawings as follows.

1 is a block diagram showing a conventional BTSC encoder of the prior art.

2 is a graph of the spectrum of the composite signal generated according to the BTSC reference.

3 is a block diagram of one embodiment of a digital BTSC encoder constructed in accordance with the present invention.

4A-C are block diagrams of a low pass filter used in the digital BTSC encoder shown in FIG.

FIG. 5 is a detailed block diagram of a broadband compression unit used in the digital BTSC encoder shown in FIG.

6 is a block diagram of a spectral compression unit used in the digital BTSC encoder shown in FIG.

7 is a flowchart used to calculate a filter coefficient of the variable emphasis filter used in the spectral compression unit shown in FIG.

8A-D are block diagrams illustrating signal scaling that can be used to preserve resolution and reduce saturation opportunities in fixed point implementations of digital BTSC encoders constructed in accordance with the present invention.

9 is a detailed block diagram of the composite modulator shown in FIGS. 8A-C.

Figure 10 is a block diagram of one preferred embodiment of the sum and difference channel processing stages that may be used in a digital BTSC encoder constructed in accordance with the present invention.

3 is a block diagram of one embodiment of a digital BTSC encoder 200 constructed in accordance with the present invention. The digital encoder 200 is configured to provide the same performance as that of the ideal encoder 100 (shown in FIG. 1). Like the encoder 100, the digital encoder 200 receives the left and right channel audio input signals and generates condition sum signals and encoded value signals therefrom, but in the digital encoder 200, these input and output signals are continuous analog signals. It is not a signal but a digital sampling signal.

The selection of sampling frequency fs for the left and right channel audio input signals has a significant effect on the design of the digital encoder 200. In a preferred embodiment, the sampling frequency fs is chosen to be a multiple of the pilot frequency f _H such that fs = Nf _H (where N is an integer), and in the best preferred embodiment N is chosen to be greater than or equal to three. In the encoder 200, it is important that the condition sum signal and the encoding difference signal do not have sufficient energy at the pilot frequency f _H so as not to interfere with the pilot tone included in the composite signal. Thus, as will be described in detail later, it is desirable to provide a particularly large attenuation at the pilot frequency f _H for at least some of the filters of the digital inode 200, and the selection of this sampling frequency f _S may be Simplify the design.

The digital encoder 200 includes an input terminal 210, a sum channel processing stage 220, and a difference channel processing stage 230. Rather than simply configuring the difference channel processing stage 23 using digital technology, all three processing stages 210, 220 and 230 may be entirely configured using digital technology. Most of the components of the digital encoder 200 coincide with the components of the ideal encoder 100. In general, the components of digital encoder 200 have been chosen such that their amplitude responses closely match each amplitude response of the corresponding component of encoder 100. In these, there is a relatively large difference between the phase responses of the components. According to one aspect of the invention, the digital encoder 200 is provided with means for compensating or nullifying these phase differences, i.e., phase errors. As is well known to those skilled in the art, a relatively small phase error in the difference channel processing stage 230 may be compensated by introducing a similar phase error to the sum channel processing stage 220, and the configuration of the sum channel processing stage using digital technology This simplifies the introduction of the required compensation phase error.

The input stage 210 of the encoder 200 includes two high pass filters 212 and 214 and two signal adders 216 and 218. The left channel audio input signal L is applied to the input of the high pass filter 212, from which the output signal is applied to the positive input terminal of the adders 216,218. The right channel audio input signal R is applied to the input of the high pass filter 214, from which the output signal is applied to the positive input terminal of the adder 216 and the negative input terminal of the trailing device 218. The adder 216 adds the output signals generated by the filters 212 and 214 to generate a sum signal (indicated by " L + R " in FIG. 3). The adder 218 is generated by the filter 212. The output signal generated by the filter 214 is viewed from the generated output signal to generate a difference signal (indicated by " LR " in Fig. 3). Thus, input stage 210 is similar to input stage 110 (shown in FIG. 1), but input stage 210 additionally includes two high pass filters 212 and 214 and generates a digital sum and difference signal.

The high pass filters 212 and 214 remove the D.C. component from the left and right channel audio input signals with the same response. As will be described in detail later, the D.C.component is removed so that the encoder 200 does not exhibit a propensity to be called " ticking ". Since the audio information content of the important left and right channel audio input signals is considered to be in the frequency band between 50 Hz and 15,000 Hz, eliminating the D.C. component is not an obstacle to transmitting the information content of the audio signal. Therefore, it is preferable that the filters 212 and 214 have a cutoff frequency of 50 Hz or less, and more preferably, to have a cutoff frequency of 10 Hz or less, so that they do not remove audio information included in the audio input signal. . Also, filters 212 and 214 preferably have balanced amplitude response in their passband. In one preferred embodiment, the filters 212 and 214 consist of a first order infinite impulse response (IIR) filter, each of which has a transmission function H (z) given by the formula shown in equation (1) below.

3 again, the sum channel processing stage 220 receives the sum signal and generates a condition sum signal therefrom. In particular, the sum signal is applied to the 75 kHz preemphasis filter 222. This filter 222 generates an output signal that is applied to the positive phase equalization filter 228. The filter 228 generates an output signal applied to the low pass filter 224 of the processing stage 220 generating the condition sum signal.

The 75 kHz preemphasis filter 222 processes the signal partially similarly to the filter 122 (shown in FIG. 1) of the ideal encoder 100. The amplitude response of filter 222 is preferably chosen to closely match the response of filter 122. As will be described in detail later, the difference channel processing stage 230 is provided with means for compensating for any differences in the phase response of the filters 222 and 122. In one preferred embodiment, the filter 222 consists of a first order IIR filter having a transmission function H (z) defined by the formula shown in equation (2) below.

The positive phase equalization filter 228 performs processing that is not directly similar to the components of the abnormal encoder 100 (shown in FIG. 1). As will be described in detail later, the positive phase equalization filter 228 is used to introduce a phase error that compensates for the phase error introduced by the difference processing stage 230. In summary, the phase-phase equalization filter 228 is an "all-pass" filter having a relatively balanced amplitude response and selected phase response. In one preferred embodiment, the filter 228 consists of a first order IIR filter having a transmission function H (z) defined by the formula shown in equation (3) below.

The low pass filter 224 performs a process similar in part to the band limiter 124 (shown in FIG. 1) of the encoder 100. Low pass filter 224 provides balanced amplitude response in the passband of 0-15 kHz and relatively sharp cutoff above 15 kHz. Filter 224 also provides particularly large attenuation at pilot tones of frequency f _H (i.e. 15,734 Hz). By providing such a particularly large degree of attenuation, the filter 224 ensures that the condition sum signal does not contain sufficient energy at the pilot frequency f _H that interferes with the pilot tone applied to the composite signal. As mentioned above, the sampling frequency f _S is selected to be equal to a multiple of the pilot frequency f _H , thereby simplifying the design of the filter providing a particularly large attenuation at the pilot frequency and thus simplifying the design of the filter 224. . Filter 224 is at least 70dB attenuation for all frequencies have to occur with respect to a null (null) at the pilot frequency f _H from the pilot frequency f _H to half the sample rate.

4 is a block diagram showing one preferred embodiment of the low pass filter 224. As shown in FIG. 4A, filter 224 consists of five continuous filter stages 310, 312, 314, 316 and 318 in series. In one preferred embodiment, all five filter stages 310, 312, 314, 316 and 318 have a second order IIR filter with a transmission function H (z) defined by the formula shown in equation (4) It consists of.

As such, in the embodiment shown in FIG. 4A, filter 224 is a tenth order IIR filter.

3 again, the difference channel processing stage 230 receives the difference signal and generates an encoded difference signal therefrom. The difference signal is applied to the low pass filter 238a, from which the output signal is applied to the fixed pre-emphasis filter 232a. The fixed pre-emphasis filter generates an output signal applied to the input terminal of the broadband compression unit 280 through the line 239 and the encoding difference signal is applied to the detector terminal of the broadband compression unit 280 through the feedback line 240. Is approved. The wideband compression unit generates an output signal applied to the input terminal of the spectral compression unit 290 through the line 281, and the encoding difference signal is applied to the detector terminal of the unit 290 through the feedback line 240. This unit generates an output signal that is applied to the fixed pre-emphasis filter 232b which in turn generates an output signal that is applied to the clipper 254. The clipper 254 generates an output signal applied to the low pass filter 238b, which generates an encoding difference signal.

The low pass filters 238a and 238b form a low pass filter 238 which performs processing similar in part to the overmodulation protector of the abnormal encoder 100 and the band limiter of the band limiter 138 (shown in FIG. 1). The filter 238 is configured in the same way as the low pass filter 224, which is used in the sum channel processing stage 220. Therefore, the phase error introduced into the coded difference signal by the filter 238 is compensated by the equilibrium phase error introduced by the filter 224. The filter 238 is preferably divided into two parts 238a and 238b, as shown for reasons to be described in detail later, and the filter 238a is sleepy with null at the pilot frequency f _H.

4B-C are block diagrams showing one preferred embodiment of each filter 238a and 238b. As shown in FIG. 4B, filter 238a is constructed by successively arranging three filter stages 310, 314 and 318 identical to the three filter stages used for filter 224 (shown in FIG. 4A), As shown in Figure 4C, filter 238b is constructed by successively arranging two filter stages 312 and 316 that are identical to the remaining two portions used in filter 224.

Fixed pre-emphasis filters 232a and 232b (shown in FIG. 3) are fixed pre-emphasis filters 232 that perform processing similar in part to filters 132 (shown in FIG. 1) of the ideal encoder 100. To form. The amplitude response of filter 232 is preferably chosen to closely match the amplitude response of filter 132. In one embodiment, the phase response of the filters 232 and 132 is significantly different, and as will be described in detail below, the resulting phase error is compensated by the filters 222 and 228 in the sum channel processing stage 220. do. The filter 232 is preferably divided into two portions 232a and 232b as shown for reasons to be described in detail later. In one preferred embodiment, the filters 232a and 232b are configured as first order IIR filters having a transmission function H (z) defined by the formula shown in equation (2). Thus, in this embodiment, the filter 232 is a secondary IIR filter.

In one preferred embodiment, the difference between the phase responses of the filters 232b and 132a closely matches the difference between the phase responses of the filters 222 and 122. Therefore, the phase error introduced into the encoding difference signal by the fixed preemphasis filter 232b is balanced by the phase error introduced into the condition sum signal by the 75 kHz preemphasis filter 222. Furthermore, in this embodiment, the phase response of the positive phase equalization filter 228 is selected to be close to the difference between the phase response of the fixed preemphasis filter 232a and the filter 132b and encoded by the filter 232a. The phase error introduced into the difference signal is balanced by the compensation phase error of the condition sum signal introduced by the positive phase equalization filter 228.

The clipper 254 performs processing similar in part to the overmodulation protector and band limiter 138 (shown in FIG. 1) of the abnormal encoder 100. In summary, the clipper 254 is configured as a limiter, but the operation of this clipper 254 will be described in detail later.

The broadband compression unit 280 and the spectral compression unit 290 perform processing functions that are partially similar to those of the units 180 and 190 of the ideal encoder 100 (shown in FIG. 1). In summary, wideband compression unit 280 dynamically compresses the signal on line 239 as a function of the overall energy level of the encoded difference signal and spectral compression unit 290 is a line 281 as a function of the high frequency energy of the encoded difference signal. Compresses the high frequency portion of the signal

5 shows a block diagram of a preferred embodiment of a digital broadband compression unit 280. The unit 280 includes a digital signal multiplier 434, a digital signal multiplier 446, a wideband digital band pass filter 448, a digital RMS level detector 450, and a digital reversible generator 458. These components are performed by each amplifier 134, amplifier 146, bandpass filter 148, RMS level detector 150, and reversible generator 152 of the ideal encoder 100 (shown in FIG. 1). It performs processing functions that are partially similar to the ones. The encoding difference signal is applied to the input of the wideband digital bandpass filter 448 through the feedback path 240, and generates an output signal from the RMS level detector 450. The detector generates an output signal representing the RMS value of the output signal generated by filter 448 and applies this output signal to reversible generator 458 via line 450a. Reversible generator 458 then generates an output signal representing the inverse of signal on line 450a and applies this output signal to multiplier 446 via line 458a. The digital signal multiplier 446 multiplies the signal on the line 458a by the gain value D to generate an output signal representing the reciprocal of the RMS value of D times and applies it to the input terminal of the multiplier 434 via the line 446a. do. The output signal generated by the fixed pre-emphasis filter 232a is applied to the other input terminal of the multiplier 434 via the line 239. Multiplier 434 multiplies the signal on line 239 by the signal of line 446a to generate the output of broadband compression unit 280 applied to the input of spectral compression unit 290 via line 281. The wideband digital bandpass filter 448 is designed to have an amplitude response that closely matches the amplitude response of the bandpass filter 148 (shown in FIG. 1). The filter 448 is preferably selected to minimize the mean square difference between its amplitude response and the amplitude response of the filter 148. In one embodiment, the phase responses of the filters 448 and 148 are significantly different, but these phase differences can be ignored since the output signal of the RMS level detector 450 is not affected by the phase of its input signal. In one preferred embodiment, the wideband digital bandpass filter 448 is configured as a secondary IRR filter having a transmission function H (z) defined by the formula shown in equation (4).

RMS level detector 450 is designed to approximate the performance of detector 150 used in ideal encoder 100 (shown in FIG. 1). The detector 450 includes a signal square device 452, a signal averaging device 454, and a square root device 456. Square unit 452 squares the signal generated by bandpass filter 448 and applies the squared signal to averaging unit 454 through line 452a. This average device calculates the time weighted average of the signal on line 452a and applies this average value to square root device 456 via line 454a. Square root device 456 calculates the square root of the signal on line 454a to generate a signal on line 450a that represents the RMS value of the output signal generated by wideband digital bandpass filter 448.

The averaging device 454 includes a digital signal multiplier 460, a digital signal adder 462, a digital signal multiplier 464, and a delay register 465. The output signal generated by the square device 452 is applied via line 452a to one input of a multiplier 460 which generates an output signal by scaling the signal on line 452a by a constant α. The scaling output signal generated by the multiplier 460 is applied to one input of the adder 462 and the output signal generated by the delay register 465 is applied to the other input of the adder 462. The adder 462 adds the signals of the two inputs to generate an output signal, which is the output signal of the average device 454 and is applied to the square root device 456 via the line 454a. This addition signal is also applied to one input of a multiplier 464 which generates an output signal by scaling the addition signal by a constant (1-α). The output signal generated by the multiplier 464 is applied to the input of the delay register 465. Those skilled in the art will appreciate that the averaging device is a circulating filter and performs the digital averaging function defined by the cyclic formula shown in the following equation (5).

Where y (n) represents the current digital sample of the signal output by the average device 454 on line 454a, and y (n-1) represents the signal output by the average device 454 on line 454a. Represents the preceding digital sample, χ (n) represents the current digital sample of the signal output by the square device 452 on line 452a. Those skilled in the art will appreciate that the averaging device 454 provides a digital approximation of the analog averaging function defined by the BTSC standard and executed by the RMS level detector 150 (shown in FIG. 1) of the abnormal corner 100. I will understand. The constant α is preferably selected so that the time constant of the RMS level detector 450 closely matches the time constant specified in the BTSC criterion of the RMS level detector 150.

Although the digital square root device 456 and the digital reversible generator 458 are shown in FIG. 5 as two separate components, the expert would have a single device that generates an output signal representing the inverse of the square root of the input signal. It will be appreciated that it can be realized using. Such a device may be configured, for example, with a memory lookup table (LUT) or using a processing component that calculates a Taylor series polynomial approximation of the inverse square root function.

6 shows a block diagram of the spectral compression unit 290 of the first embodiment. The unit 290 is a variable pre-emphasis / de-emphasis unit (hereinafter referred to as a "variable emphasis unit") 536, a signal multiplier 540, a spectral band pass filter 542, and an RMS level detector 544. These components are partially similar to each variable emphasis filter 136, amplifier 140, band pass filter, and RMS level detector 144 of the ideal encoder 100 (shown in FIG. 1). Perform the process. The encoded difference signal is applied through the feedback line 240 to the input of the signal multiplier 540 which generates the output signal by multiplying the encoded difference signal by the fixed gain value Gain C. The medium width output signal generated by the signal multiplier 540 is applied to the spectral band pass filter 542 which generates the output signal applied to the RMS level detector 544. The detector generates an output signal through which the variable emphasis unit 536 is applied to the control complex through the line 544a, and the output signal generated by the wideband compression unit 280 is connected to the unit 536 through the line 281. Is applied to the input terminal of). The variable emphasis unit changes the frequency response applied to the signal on line 281 as a function of the signal on line 544a and the signal on line 544a is within the frequency band passed by spectrum band pass filter 542. It is a function of the signal energy of the encoded difference signal. The output signal of the unit 290 generated by the unit 536 and applied to the input of the fixed pre-emphasis filter 232b is compressed in a large amount in the high frequency portion of the signal rather than in the rest of the main spectrum.

The spectral bandpass filter 542 is designed to have an amplitude response that closely matches the amplitude response of the bandpass filter 142 (shown in FIG. 1) of the encoder 100. As with filter 448 (shown in FIG. 5), this filter 542 is preferably chosen such that the difference between its RMS amplitude response and that of filter 142 is minimal. In one embodiment, the phase response of the filter 542 is significantly different but these phase differences can be neglected since the RMS output of the RMS level detector 544 is substantially unaffected by the phase of the input to the detector. In one preferred embodiment, the spectral band pass filter 542 comprises three consecutive secondary IIR filter stages 542a, 542b, 542c (with a transmission function H (z) described by the formula shown in equation (4) ( Shown in FIG. 6).

RMS level detector 544 is designed to approximate the performance of detector 144 used in ideal encoder 100 (shown in FIG. 1). The detector 544 includes a signal square device 522, a signal averaging device 554, and a square root device 556. Square device 552 squares the signal generated by spectrum band pass filter 542 and the squared signal is applied to averaging device 554 via line 552a. The averaging device has a function similar to the averaging device 454 (shown in FIG. 5) used in the broadband compression unit 280, but this averaging device 554 uses a constant β different from the constant α. The characteristic of the average device 554 is, of course, described by equation (5) in which β is substituted for α. The constant β of the device 554 is preferably chosen so that the time constant of the RMS level detector 544 is close to the corresponding time constant determined by the BTSC criterion for the RMS level detector 144 (shown in FIG. 1). The averaging device 554 calculates the time weighted average of corals on line 552a and applies this average value to square root device 556 via line 554a. Square root device 556 calculates the square root of the signal on line 554a, thereby generating a signal on line 544a in accordance with the RMS value of the output signal generated by spectral bandpass filter 542.

The signal of the line 544a is applied to the control terminal of the variable emphasis unit 536. The variable emphasis unit 536 performs a process similar in part to the filter 136 (shown in FIG. 1) of the abnormal encoder 100. As defined by the BTSC standard, the filter 136 has an amplitude and phase that change in accordance with the output signal generated by the RMS level detector 144. One preferred method in which the unit 536 is configured such that it has a similar variable response uses a digital filter having a variable coefficient to determine its transmission function and based on the value of the signal on line 544a or any given sample period To select the value of the coefficient in the sample period group.

6 shows one embodiment of a variable emphasis unit 536 that includes a logarithmic generator 558, a variable emphasis filter 560, and a lookup table (LUT) 562. The output signal generated by the RMS level detector 544 is applied to the logarithmic generator 558 via line 544a. This logarithmic generator generates a signal on line 558a indicating the logarithm of the signal on line 544a and applies this signal to LUT 562. LUT 562 generates an output signal that is selected from this LUT and indicates a filter coefficient to be used by variable emphasis filter 560. The coefficient generated in this way by the LUT 562 is applied to the coefficient selection terminal of the variable emphasis filter 560 through the line 562A. The output signal generated by the broadband compression unit 280 is applied to the input terminal of the variable emphasis filter 560 via the line 281. The variable emphasis filter 560 generates an output signal of the spectral compression unit 290 applied to the input of the fixed pre-emphasis filter 233b.

The variable emphasis filter 560 is designed to have a variable amplitude response that closely matches the variable amplitude response of the filter 136 (shown in FIG. 1) of the ideal encoder 100. The variable emphasis filter 560 uses a variable coefficient transfer function (i.e., the coefficient of the transfer function H (z) of the filter 560 changes) and the LUT 562 selects the value of the coefficient during the time based on the sample period. This provides a similar variable response. As will be described in detail later, LUT 562 stores the values of the filter coefficients used by filter 560, and LUT 562 is algebraic in line 558a during each sample period or during a selected group of sample periods. A series of filter coefficients is selected as a function of the output signal generated by generator 558. In one preferred embodiment, the variable emphasis filter 560 is configured as a first order IIR filter subtracting the transfer function H (z) described by the formula shown in equation (6) below.

Here, the filter coefficients b ₀ , b ₁ and a ₁ are variables selected by the LUT 562. The method of selecting the value of the filter coefficient used by the filter 560 and other filters of the encoder 200 will be described in detail later.

In FIG. 6, the logarithmic generator 558 and the square root device 556 are shown as two separate components for simplicity of explanation. However, those skilled in the art will understand that these two elements can be constructed using a single device, such as a LUT, or using a processing element that computes the Taylor series polynomial approximation of the logarithm of the signal on line 554A and divides this value by two. will be. Likewise, the algebraic generator 558, square root device 556, and functions performed by the LUT 562 can be combined in a single device as another method.

As mentioned above, the high pass filter 212, 214 (shown in FIG. 3) can be used to block the DC component so that the encoder 200 does not exhibit a propensity to be known as " ticking ". With respect to the stereophonic encoder, ticking refers to the relatively low frequency oscillation tendency of the encoder when there is no signal at the left and right channel audio inputs. When there is no signal at the audio input, the required propensity of the stereophonic system is to maintain no sound. However, an encoder that is connected to the loudspeaker through the decoder and tends to teak is an encoder that is connected to the loudspeaker through the coder and to the ticking tendency that the loudspeaker is somewhat related to the time constant of the RMS level detector in the broadband compressor. Make audible sounds called "teaks" for a specified time. In particular, in the encoder 200, when a very low level signal appears at the audio input and there is a DC component or offset in the signal of the line 239, the broadband compression unit 280 exhibits an unstable tendency to cause ticking. .

Assume that only a low level audio signal is present in line 239. In this case, the output of the RMS level detector 450 at line 450a will be very small so that they of the multiplier 434 will be very large. If the low level audio signal of the line 239 has a constant amplitude, the wideband compression unit 280 has a certain amount of time since the coding difference signal is fed back from the line 240 to the wideband compression unit 280. Will be stabilized after the time constant applied to Since the feedback is negative, when the audio signal of the line 239 increases in amplitude, the signal of the line 450a will increase and the gain of the multiplier 424 will decrease. When the audio signal of line 239 decreases in amplitude, the signal of line 450a decreases to increase the gain of multiplier 434.

However, if there is a significant dc signal along with a low level audio signal on line 239, the feedback process is hindered by the action of broadband bandpass filter 448 with zero response to the dc signal. . In particular, the dc component present in the encoded difference signal at line 240 is blocked by filter 448 and is not detected by RMS level detector 450. The dc signal present at line 239 will be amplified with the audio signal at line 239. The amplification factor or gain will only be determined by the audio signal amplitude as detected by the RMS level detector after filtering by the filter 448.

As mentioned above, if the amplitude of the audio signal of line 239 changes, the gain of multiplier 434 changes inversely. During this change in gain, the dc component on line 239 will be variable amplified and modulated with the dc signal resulting in an ac signal. Such a dc signal is not rejected by the filter 448 and is modulated to generate a signal of sufficient audio band sensed by the detector 450. When the audio signal of the line 239 is smaller than the dc of the line 239, the small change in the audio signal level that causes the gain change of the amplifier 434 is caused by the dc level (ac) in the line 281 through this modulation process. Equivalent to the signal). This generated ac signal results in the increase of all signals passing through the filter 448 regardless of whether the audio signal change that generates the ac signal increases or decreases in the signal level. In particular, negative feedback typically increases the gain of multiplier 434 when the audio signal level of line 239 is reduced. However, when there are enough dc signals in line 239, the reduction in the audio signal of line 239 causes the gain in multiplier 434 to be reduced in the signal sensed by detector 450. As such, the negative feedback process is reversed and the feedback becomes positive.

This positive feedback is coupled to any audio signal present on line 281 when the modulated dc signal at line 281 is weighted by the response of all filters and signal modifiers between line 281 and the output of field 448. If it is significantly larger in comparison, it will survive. When the gain of multiplier 434 decreases and the modulated dc signal of line 281 no longer provides sufficient input to detector 450, the feedback switches to its normal negative form. Depending on the time constant of the detector 450, the system will require an appropriate gain level based on the level of the audio signal of the line 239. However, if enough dc remains in the signal of line 239, the cycle will repeat itself when the gain of multiplier 434 is sufficiently increased. During each such period of positive feedback, a sharp change in the dc level of line 281 occurs. These changes are audible and sound somewhat similar to the "teak" of the clock. Since this dc change will occur to some degree regularly based on the time constant of the detector 450, this phenomenon is often referred to as " ticking ".

One way to prevent such ticking is to remove the dc component present in the input signal to the encoder 200. This is done by high pass filters 212 and 214. In addition, the high pass filters 212 and 214 may help to maximize the dynamic range of the encoder 200 by removing the dc component that consumes the variable dynamic range. As mentioned above, and as shown in FIG. 3, the low pass filter 238 is preferably composed of two filters 238a and 238b. Splitting filter 238 in this manner has several advantages. If the filter 238a is omitted and the entire filter 238 is placed on top of the clipper 254 (at the position of the filter 238b), a component of 15 dB or more in the audio input signal has a wide bandwidth similar to the above-mentioned ticking tendency. It will be unstable in compression unit 280. This is because more than 15 dB of signal components in line 239 are amplified by multiplier 434 (shown in FIG. 5) and these components are filtered by the low pass filter after clipper 254 (shown in FIG. 3). This is because it is not detected by the RMS level detector 450. The detector 450 increases the gain of the multiplier 434 when the detector detects the absence of a signal, so that the gain of the multiplier 434 shows that the signal of the line 239 has little or no audio signal (15 kHz or less). When configured with information (above 15 kHz) it can become relatively large. The multiplier 434 may amplify the high frequency information to generate a large signal that is easily clipped by the components of the processing stage 230. This clipping may generate harmonics called low frequencies to be sensed by the RMS level detector 450 which causes the system to tick as mentioned above. In addition, when the filter 238b is omitted and the entire filter 238 is disposed in front of the fixed pre-empty filter 232a (that is, at the position of the filter 238), the high frequency artifacts generated by the clipter 254 are encoded. It will be included in the signal and will interfere with the pilot tone in the composite signal. Thus, the split filter 238 as shown provides an optimal configuration while the filter 238a prevents ticking of the compression unit 280 and the filter 238b is generated by the clipper 254 Aftermath

The fixed preemphasis filter 232 is also preferably divided into two filters 232a and 232b as shown in FIG. The filter 232 typically requires relatively large ones at high frequencies, as defined by the BTSC standard, and increases the propensity of the filter 232 to be the cause of clipping by using only a single stage to configure the filter 232. It is advantageous to apply part of the gain of the filter 232 to the input side of the broadband compression unit 280 (to filter 232a) and part of the gain of the filter 232 to the output side of the broadband compression unit 280 (to filter 233b). Do. Unit 280 typically compresses its input signal, thereby dividing the gain of filter 232 relative to the compression made by unit 280, thereby reducing the propensity to bring the gain of filter 232 into an overflow state. .

In order to minimize size, power consumption and manufacturing cost, the encoder 200 may be configured using a single digital signal processing chip. The encoder 200 has been successfully constructed using one of the well known Motorola DSP 56002 digital signal processing chips (hereinafter referred to as "DSP devices"). The Mltorola DSP 56002 is a fixed-point 24-bit chip, or other forms of processing chips, such as floating-point chips or fixed-point chips with different word lengths. The DSP device of the encoder 200 uses a sampling frequency f _s that is three times the pilot frequency f _H (that is, f _S = 47202 Hz). The following table 1 lists all filter coefficients used in the DSP device of the encoder 200 except those used in the variable emphasis filter 560.

In the DSP device of the encoder 200, the value of the constant α used following the average device 454 (shown in FIG. 5) of the broadband compression unit 280 is set to 0.0006093973517 and the average device of the spectral compression unit 290 ( 553) (shown in FIG. 6), the value of the constant β is set to 0.001825967. In addition, the values of Gain C and Gain D used in the amplifier 540 (446) of each spectrum and broadband compression unit are set to 0.5011872 and 0.08984625 so that the DSP device of the encoder 200 performs a function similar to that of the encoder 100, respectively. do.

FIG. 7 shows a flowchart 700 illustrating a preferred method for precomputing all sets of filter coefficients used by the variable emphasis filter 560 (shown in FIG. 6) of the DSP device of the encoder 200. have. All sets of filter coefficients used by filter 560 prior to operation of encoder 200 are precomputed (eg, by a general purpose digital computer) and loaded into LUT 562. In the DSP device of the encoder 200, the filter 560 has a transmission function H (z) described by equation (6) so that the flowchart 70 describes the calculation of the coefficients b ₀ , b ₁ and a ₁ . As defined by the BTSC standard, the transmission function S (f, b) of the analog filter 136 (shown in FIG. 1), in which the filter 560 partially matches, is described by the formula shown by the following equation.

Here, F is 20.1 ms.

The first step of the flowchart 700 is an initialization step 710 in which several variables are initialized. In particular, the sampling frequency fs is set to 47202 Hz and the time T is set to 1 / fs. The variable W is the digital version of the variable F used in equation (7) and is set to π (20.1 ms) / fs. The variable dBRANG represents the required signal range of the RMS detector of the spectral compression unit. For DSP units the dBRANGE is set to 72.25 dB. The variable dBRES is related to the sensitivity of the filter 560 to changes in the energy level of the encoded difference signal. In the DSP device of encoder 200, dBRES is set to 0.094 dB so that filter 560 will use the coefficients based on the signal value of line 558a quantified to 0.094 dB close to home. The variable N is equal to the total number of filter coefficient sets used in the filter 560 and N is calculated by dividing the sensitivity (dBRES) by the range (dBRANGE) to be the nearest integer. In DSP devices, n is 768, but the expert will understand that this number can vary depending on sensitivity or range. In a DSP device, the LUT 562 stores 769 sets of systems for the filter 560, and of course a large LUT will be used to store an extra set of filter coefficients when N is increased. Those skilled in the art will also understand that the algebraic generator 558 can scale the signal of line 558a to reduce the number of filter coefficients stored by LUT 562 for a given minimum quantification of the signal value of line 558a. . However, in other devices, the logarithm generator 558 can be omitted and the LUT 562 can store multiple filter coefficient sets. Finally, the variables Scale and Address are set to 32 and zero, respectively. The variable Scale, which is used only for fixed point configuration, is chosen so that all filter coefficients are greater than or equal to -1 and less than 1 (when filter coefficient is two's complement).

Following the initialization step 710, the coefficient generation step 720 is executed. During the initial execution of step 720 the variables b ₀ (0), b ₁ (0) and a ₁ (0) are assigned to the values of the coefficients b ₀ , b ₁ and a ₁ to be stored at address location zero of the LUT 562. Calculated accordingly. Following this execution of step 720, an incremental step 730 is executed in which the value of the variable Address is incremented. Following step 730, a comparison step is executed in which the values of the variables Address and N are compared. If Address is less than or equal to N, steps 720, 730, and 740 are re-executed repeatedly so that the values of coefficients b ₀ , b _1, and a1 are calculated for each 769 addresses of LUT 562. When step 740 detects that the value of Address is greater than N, all 769 sets of coefficients are calculated and execution of flowchart 700 reaches end step 750.

In the coefficient generation step 720, the variable dBFS corresponds to the output of the logarithmic generator 558. Since the value of the variable address ranges from 0 to 769, the value of dBFS ranges from about -72.25 to zero dB corresponding to a single range of about 72.25 dB provided by the DSP device of the encoder 200 (where zero dB Corresponds to premodulation). The variable RMSd matches the output of the analog RMS level detector 144 (shown in FIG. 1), and since the value of the variable address ranges from 0 to 769, the value of RMSd is typically a single range of 72 dB provided by a prior art analog BTSC encoder. Is approximately -36 to 36 dB. The variable RMSb is the linear version of the variable RMSd and RMSb is equal to the variable b in the transfer function S (f, b) in equation (7). The variables K1 and K2 correspond to the terms (b + 51) / (b + 1) and (51b + 1) / (b + 1) in equation (7), respectively. The coefficients b ₀ , b ₁ and a ₁ are calculated as shown in step 720 using the variables K1, K2, W and Scale.

FIG. 8A shows a block diagram illustrating one method of using an DSP system of an analog system. In FIG. 8A, all the components of the 56002 integrated circuit are indicated by the reference numeral 200a. The analog system supplies analog left and right channel audio input signals (shown as " L " and " R ", respectively, in Fig. 8A) and these signals are applied to the inputs of the 16-bit analog-to-digital converters 810 and 812, respectively. Converters 810 and 812 sample their analog input signals using a sampling frequency of 47.202 Hz (i.e., 3f _H ), whereby converters 810 and 812 respectively represent 16-bit digital signals representing left and right channel audio input signals. Generate a sequence of samples. The signals generated by transducers 810 and 812 are applied to encoder 200a, where these signals are received by modules 292 and 294, respectively. Modules 292 and 294 are " divide by sixteen " modules (divide the amplitude of these inputs by a factor of 16), so they generate an output signal by dividing their signals by sixteen. Since dividing by powers of two using a shift register is easily performed in a digital system, modules 292 and 294 are configured as shift registers that shift their inputs by four binary positions.

As mentioned above, the 56002 chip is a fixed point 24-bit processor and the sample applied to the chip by converters 810 and 812 represents two's complement. Modules 292 and 294 divide the samples generated by converters 810 and 812 by 16, so that each sample is centered in a 24-bit word. In the sample, the four most significant bits are the least significant bits, the four least significant bits are zero and the 16 bits in the middle of the word correspond to one sample generated by one of the converters 810,812. In this way, accuracy is preserved by padding each 24-bit word from high-end to government bits and from low-end to zero, and the intermediate signal generated by encoder 200a is 16-bit without the occurrence of error conditions such as overflow. To exceed.

In encoder 200a, each bit of a 24-bit word is approximately 6 dB of signal range, such that module 292 and 294 correspond to a -24 dB (ie -6 x 4) attenuator. If the analog input signal applied to transducers 810 and 812 is considered a zero dB signal for reference purposes, the signal generated by modules 292 and 294 is attenuated by 24 dB.

The input terminal 210 receives a 24-bit word generated by the modules 292 and 294 and generates a sum signal applied to the sum channel processing stage 220 therefrom. The output signal generated by the sum channel processing stage 220 is applied to a " 16 times module " (which can be considered as a 24 dB amplifier) 296. This causes module 296 to compensate for -24 dB attenuators 292 and 294 so that the output of sum channel processing stage 220 is 100% modulated (ie, "full scale"). The output signal generated by module 296 is applied to a 16-bit digital-to-analog converter 814 that generates an analog condition sum signal.

The input terminal 210 also generates a difference signal applied to the difference channel processing stage 230. As mentioned above, as a result of the modules 292 and 294, the difference signal may be considered to be attenuated by 24 dB. In the DSP device of the encoder 200a, the clipper 254 (shown in FIG. 3) of the difference processing stage 230 includes an 18 dB amplifier (configured for 8 times). That is, the clipper 254 amplifies the signal generated by the fixed pre-emphasis filter 232b by 18 dB, and clips the amplified signal to output an output signal generated by the clipper 254 below 6 dB from the full modulation. Do not exceed the number of arguments. Therefore, the signal applied from the clipper 254 to the low pass filter 238b has a "headroom" of 1 bit (or 6 dB), so that the filter 238b is an output signal 6 dB larger than its input signal without being saturated. Occurs. It is desirable to leave this 1-bit headroom, because the headroom has some ringing that temporarily generates an instantaneous output signal whose transfer response is greater than the instantaneous input signal. This is because ringing of the filter 238b which causes the saturation condition is prevented. Again in FIG. 8A, the output signal generated by filter 238b is applied to a 16-bit digital-to-analog converter 816, which generates an output signal that is applied to a 6 dB analog amplifier 820. D / A converters 814 and 816 include analog anti-image filters that are well known as part of their functionality as full converters. An anti-image filter is an analog filter applied to analog signal-following digital-to-analog conversion that acts to attenuate any image of the desired signal that is symmetrical to the sample frequency and its multiples. The converters 814 and 816 are considered to be equal to each other, operate at the same sample rate, and include the same anti-image filtering. Typically such converters are commercially available, such as Crystal Semiconductor CS4328. The amplifier 820 amplifies the input signal by 6 dB so that the encoded difference signal is full scaled. Although FIG. 8A shows an encoder 200a coupled to analog-to-digital converters 810 and 812 to receive analog audio signals, converters 810 and 812 are, of course, omitted in the digital system so that the encoder 200a may be omitted. Direct digital audio signals can be received.

8B shows a block diagram of one preferred embodiment of a BTSC encoder 200b constructed in accordance with the present invention and configured as part of an analog system. Encoder 200b is similar to encoder 200a, but amplifies by 24dB at encoder 200a, while module 296 at encoder 200b amplifies its input signal by 18dB (8x). The output signal generated by module 296 is a scale version of the condition sum signal and is shown as S in FIG. 8B. In addition, the encoder 200b includes a module 298 for amplifying the output signal generated by the difference channel processing stage 230 by 6 dB (double magnification). The output signal generated by the module 298 is a scale version of the coded difference signal and is shown as D in FIG. 8B. Encoder 200b also includes a composite modulator 822 for receiving signals S and D and generating a digital version of the composite signal therefrom. The digital composite signal generated by the modulator 822 is applied to the digital-analog converter 818 and its output is an analog version of the composite signal. The D / A converter 818 is a full converter that includes the above mentioned analog anti-image filter as part of its function. Such converters are commercially available, such as Burr-Brown PCM1710. In the preferred embodiment, the modules 292, 294, the input terminal 210, the sum channel processing stage 220, the difference channel processing stage 230, the modules 296 and 298, and the composite modulator 822 are all It can be configured in a single digital signal processing chip.

Since the composite signal is generated as a digital signal at the encoder 200b, the module 298 waits after the digital-to-analog conversion and as shown in FIG. 8A, the sub-channel processing stage 230 rather than using an analog amplifier such as the amplifier 820. The output signal generated by) is included to be full scale. In addition, since the conditional signal is used for 50% modulation in the composite signal, the module 296 amplifies the input signal by only 18 dB so that the output signal generated by the module 296 is equal to the amplitude of the output signal generated by the module 298. Half.

9 shows a block diagram of one embodiment of a composite modulator 822. The composite modulator receives signals S and D from which a digital version of the composite signal is generated. The modulator 822 includes two interpolators 910 and 912, two digital low pass filters 914 and 916, a digital signal multiplier 918, and two digital signal adders 920 and 922. . S and D signals are applied to each input of interpolators 910 and 912. Interpolators 910 and 912, otherwise referred to as " up-samplers, " double the sampling frequency of input signals S and D by storing new samples between every two consecutive samples applied to their inputs. Generates an output signal with The output signal generated by the interpolators 910 and 912 is applied to each input of the low pass filter 914 and 916. These filters remove images introduced into the S and D signals by interpolators 910 and 912. The filter output signal generated by the filter 916 is applied to one input of the signal multiplier 918 and the digital oscillation signal is applied to the other input of the multiplier 918 as a function of cos [4π (f _H / f _S ) n]. Is approved. As a result, the multiplier 918 generates an amplitude modulated double sideband compressed carrier version of the difference signal used in the composite signal. The output signal generated by the multiplier 918 is applied to one input of the signal adder 920 and the filter output signal generated by the filter 914 is applied to the other input of the signal adder 920. The two signals present in the P are added to generate an output signal, and the signal is applied to the signal adder 922. A pilot tone signal oscillating as a function of Acos [2π (fH / f _S ) n], where "A" is a constant representing 10% of full-scale modulation, is applied to the other input of the signal adder 922 The adder adds two signals appearing at its input to generate a digital composite signal.

Complex modulator 822 includes iterators 910 and 912, as it is slightly smaller than 3f _H (shown in FIG. 2) of the maximum frequency component in the composite signal, thus signal multiplier 918 and signal adder 920. In order to satisfy the Nyquist criterion, the signal applied to the input of N) must have a sampling rate of at least 6f _{H in} order to satisfy the Nyquist criterion. The sample rate at the output of the complex modulator 822 is typically higher than the sample rate of the S or D signal, so if the input signals S and D applied to the D / A modulator 822 have a sample rate of 3f _H , Some form of interpolation (such as provided by interpolators 910 and 912) should be provided to double. Of course, interpolators 910 and 912 and low pass filters 917 and 916 may be omitted from modulator 822 when a sufficiently high sample rate is used through encoder 200b.

8C shows a block diagram of another embodiment of a BTSC encoder 200c constructed in accordance with the present invention. The encoder 200 is similar to the encoder 200b (shown in FIG. 8B), but the module 298 is omitted in the encoder 200c so that the signal generated by the difference channel processing stage 230 is the signal D and the direct complex modulator ( 822). Moreover, at encoder 200c, module 296 amplifies its input by 12dB (4x) rather than 18dB as done at encoder 200b. As such, in encoder 200b, signals S and D are 6 dB lower than the level of signal of encoder 200b. Thus, modulator 822 generates a version of the composite signal that is attenuated by these signals by 6 dB. The version of the composite signal thus attenuated is converted into an analog signal by the digital-to-analog converter 818 and is full-scaled by the 6 dB analog amplifier 820. Like the encoder 200b, the encoder 200c may be configured using a single digital signal processing chip.

The difference between the encoders 200b and 200c is design compatibility. As will be appreciated by an expert, digital-to-analog converters tend to minimize the loss of signal-to-noise ratio that a digital signal can produce as a result of conversion at full scale. Encoder 200b utilizes modules 296 and 298 to ensure that the digital version of the composite signal applied to converter 818 is full scale, thereby minimizing the loss of signal-to-noise ratio that may appear during operation of converter 818. However, even if the loss of signal-to-noise ratio that may occur as a result of the converter 200b and the converter 818 is minimized, the encoder 200b increases the tendency for clipping to appear in the composite signal. Since the difference channel processing stage 230 uses a relatively large gain provided by the fixed pre-emphasis filter 232 (shown in FIG. 3), some clipping may occur in the path of the coded difference signal. Encoder 200b uses module 298 so that the signal is full scale, which increases the chance that clipping will occur by removing headroom from the signal path of the D signal. As such, the encoder 200b minimizes the loss of signal-to-noise ratio that appears as a result of the converter 818 at the expense of increasing the clipping propensity in the path of the encoding difference signal. In contrast, encoder 200c reduces clipping propensity at the expense of preserving headroom in the path of the encoding difference signal, thereby increasing the loss of signal-to-noise ratio that occurs as a result of the operation of converter 818.

8D shows a block diagram of another embodiment of a BTSC encoder 200d constructed in accordance with the present invention. Encoder 200d is similar to encoder 200a (shown in FIG. 8A), but encoder 200d additionally includes part 822a of complex modulator. This portion 822a includes two interpolators 910 and 912, two low pass filters 914 and 916, a digital signal multiplier 918, and a digital signal adder 930. The S signal generated by the module 296 is applied to an interpolator 910 that "up-samples" the signal and applies the up-sampled signal to the low pass filter 914. The low pass filter filters this signal and applies the filtered signal to one input terminal of the adder 930. A digital pilot tone having twice the nominal amplitude (i.e., 2 Acos2π (f _H / f _S ) n) is applied to the other input terminal of the adder 930, which adds two signals of the input terminal to generate an output signal. do. The signal generated by the difference channel processing stage 230 is applied to an interpolator 912 that generates an up-sampled signal that is applied to the low pass filter 916. The low pass filter filters this signal and applies the filtered signal to one terminal of the multiplier 918. It is applied to one terminal of the multiplier 918 oscillating according to cos4? (f _H / f _S ) n. Like the encoders 200a-c, the encoder 200d may be configured using a single digital signal processing chip.

The encoder 200d is preferably used with two digital-to-analog converters 932 and 934, an analog -6 dB attenuator 936, an analog 6 dB amplifier 938, and an analog adder 940. The output signal generated by the adder 930 is applied to a converter 932 which generates an analog signal applied to the attenuator 936. The output signal generated by the multiplier 918 is applied to a converter 934 which generates an analog signal applied to the amplifier 938. The signals generated by the attenuator 936 and the amplifier 938 are applied to the input terminals of the signal adder 940, and the signal adder adds these signals to generate an analog composite signal. D / A converters 932 and 934 are full converters that include the analog anti-image filters mentioned above as part of their function. The converters 932 and 934 are assumed to be substantially identical and operate at the same sample rate to include the same anti-image filtering. Such converters are commercially available, such as Burr Brown PCM1710.

The interpolator 910 and the low pass filter 914 may be removed from FIG. 8D and the D / A converter 932 may be operated at the same sample rate as the sum channel processing stage 220. However, this is not practical, as low cost D / A converters, which are commonly available, are usually available in pairs swimming in a single direct circuit. These pairs of D / A converters operate at the same sample rate. On the other hand, the complexity of the DSP can be reduced by removing the interpolator 910 and low pass filter 914 from FIG. 8D, but in this case a simple stereo D / A converter is added to both D / A converters 932 and 934. It can't be used anymore, increasing the cost or complexity of the overall design.

Encoder 200d is a combination of the features of encoders 200b and 200c. Encoder 200d uses module 296 to make the signal full-scale to minimize the loss of signal-to-noise ratio that may occur as a result of the operation of converter 932. In addition, the encoder 200d preserves 6 dB of headroom in the signal path of the D signal, thereby reducing the loss of accuracy due to clipping. Although encoder 200d includes more components than encoders 200b and 200c, encoder 200d minimizes the loss of signal-to-noise ratio and the clipping tendency.

FIG. 10 shows a block diagram of a preferred embodiment of the sum channel processing stage 220a and the difference channel processing stage 230a used in the encoder 200 (these stages 220a and 230a as well as the encoder 200a-d). do). Process stages 220a and 230a are similar to stages 220 and 230 mentioned above, but stage 220a additionally includes an in-phase equalization filter 1010 and stage 230a additionally And a phase equalization filter 1012. In the illustrated embodiment, the output signal generated by the positive phase equalization filter 228 and the fixed pre-emphasis filter 232a is applied to the input terminals of the in-phase equalization filter 1010 and 1012, respectively, and the line 558a. The output signal generated by the majority generator 558 is applied to the control terminal of the filter (1010, 1012). The output signals generated by the filters 1010 and 1012 are applied to the low pass filter 224 and the broadband compression unit 280, respectively.

In-phase equalization filters 1010 and 1012 are used to compensate for phase errors introduced by the variable emphasis filter 560 used in the spectral compression unit 290. The phase response of the variable emphasis filter 560 is preferably matched as closely as possible to the variable emphasis filter 136 (shown in FIG. 1). However, by means of the variable emphasis filter 136 having a variable characteristic of signal dependency, the phase response is matched with the variable emphasis filter 136 for all the pre-emphasis / de-emphasis characteristics that change according to the signal level. It is very difficult to design the variable emphasis filter 560. Thus, in the encoder 200 of the exemplary embodiment, the phase response of the variable emphasis filter 560 and the variable emphasis filter 136 diverges according to the signal level. The in-phase equalization filters 1010 and 1012 introduce a compensation phase error into the sum and difference channel processing stages to compensate for divergence between the variable emphasis filter 560 and the variable emphasis filter 136.

Thus, in-phase equalization filter 1010 and 1012 perform a function similar to that performed by positive phase equalization filter 228. However, while filter 228 compensates for phase errors independent of the level of the encoding difference signal, filters 1010 and 1012 compensate for phase errors that depend on this signal level. Filters 1010 and 1012 are preferably configured as "full pass" filters having a relatively flat amplitude response and selected phase response. The in-phase equalization filter is included in both sum and difference processing stages because phase delay is required in the sum or difference channel to compensate for the phase error introduced by the variable emphasis filter 560. In a preferred embodiment, the filters 1010 and 1012 are configured in a manner similar to the variable emphasis unit 536 and include a filter having a variable coefficient transfer coefficient and a LUT to select a value of the filter coefficient at a specific time. . The signal generated by the logarithmic generator 558 in line 558a is applied to the control terminals of filters 1010 and 1012 and selects the filter coefficients used by these filters.

Although digital encoder 200 has been described in connection with certain specific embodiments, those skilled in the art will understand that variations of these embodiments are included within the present invention. For example, it has been described that the variable emphasis unit 536 (shown in FIG. 6) is configured using the variable emphasis filter 560 and the LUT 562. However, rather than precomputing all the coefficients for the filter 560 and storing them in the LUT 562, the variable emphasis unit 536 includes components for removing the LUT 562 and calculating the filter coefficients in real time. Can also be configured differently. Experts can consider compatibility between memory resources (such as those used by the LUT to store filter coefficients) and computational resources (such as those used by components to calculate filter coefficients in real time) and encoders ( It will be appreciated that in any particular form of 200) it may be solved differently. Similarly, the same considerations apply to computing square roots 456 and 556, reversible generators 458, and memory resources (e.g., LUTs for storing all values) or processing resources (e.g., Taylor series polynomial approximations. Is applied to algebraic generator 558 which alternately uses resources. In yet other embodiments, any or all of the components of encoder 200 may be configured using components such as each hardware component or software module running on a general purpose or special purpose computer.

Another variation of encoder 200 included in the present invention relates to scaling module 292 and 294 (shown in FIG. 8). These modules are particularly relevant to the fixed point configuration of encoder 200. In floating point configurations, each sample does not need to be padded with zero and positive bits to prevent overflow, and these modules can be removed from the floating point configuration. As another example, while the phase-phase equalization filter 228 (shown in FIG. 10) has been described for compensating for phase errors introduced by the filter 232A, the filter 228 is a different component of the difference channel processing stage 230A. It can be used to compensate for other phase errors introduced by the component. In addition, the filters 228 and 1010 may be configured as a single filter.

Since the present invention can be changed to some extent without departing from the scope of the present invention, it is to be understood that all matter contained in the above description and drawings is for the purpose of illustration and not by way of limitation.

Claims (86)

An adaptive signal weighted assembly comprising a signal path for conveying an information signal of a predetermined bandwidth, wherein the assembly is within a first selective spectrum region within a predetermined bandwidth by a first variable gain rate that varies with the first control signal. Filter means disposed in the signal path to change a gain imposed on a part of the information signal in the first path according to the signal energy of the information signal in the second selection spectrum region including at least a portion of the first selection spectrum region. A first control signal generating means, disposed in the signal path and coupled to the filter means to change the signal gain imposed on the information signal through a predetermined bandwidth by a second variable gain rate that varies in accordance with the second control signal. The second control signal according to the gain control means and the signal energy of the information signal in the third selected spectrum region in the predetermined wide band; An adaptive signal weighting assembly comprising a second control signal generating means for generating a signal, wherein the information signal is a digital information signal and the filter means is digital in a first selection spectrum region within a predetermined bandwidth by a first variable gain rate. A digital filter disposed in the signal path for changing a gain imposed on a part of the information signal, the control gain means being disposed in the signal path for changing the signal gain imposed on the digital information signal through a predetermined bandwidth Coupled to the means, the second control signal generating means digitally generating the second control signal in accordance with the signal energy of the digital information signal in the third selected spectrum region within the predetermined bandwidth, the digital filter means, the first control signal Generating means, digital gain control means and second control signal generating means to preserve the signal content of the information signal; Adaptive signal weighting assembly characterized in that it works in real time. 2. The apparatus of claim 1, wherein the assembly comprises input means for encoding an information signal, receiving the information signal, and output means coupled to the input means via a signal path to provide an information signal encoded by the assembly. The filter means comprising a variable coefficient digital filter means for filtering the information signal, the filter being specified by the variable coefficient transfer function, the filter being within a first selected spectrum region within a predetermined bandwidth by a first variable gain factor. An adaptive signal weighted assembly comprising changing a gain imposed on a part of an information signal at and changing a variable coefficient of the variable coefficient transfer function and a first variable gain rate according to the first control signal. The digital left and right channel audio signal according to claim 1, wherein the left and right channel audio signals encoded by encoding the two information signals consisting of the digital left and right channel audio signals in real time without distorting the signal contents of the digital left and right channel audio signals. Adaptive signal weight assembly, characterized in that it can be decoded to play. 4. The left high pass filter means for receiving a digital left channel audio signal and digitally high pass filtering the digital left channel audio signal to generate a digital left filter signal, the digital right channel audio signal being received and the digital right channel. Right high-pass filter means for digitally high-pass filtering an audio signal to generate a digital right filter signal, means for generating a digital sum signal by adding digital left and right filter signals, and other digital left and right filter signals from one digital left-right filter signal. Matrix means for receiving digital left and right filter signals including means for subtracting and generating a digital difference signal, difference channel processing means for digitally processing the digital difference signal, and digital processing of the digital sum signal; A sum channel processing means, the left and right high pass filter means, the matrix means, Null adaptive signal weighting assembly, characterized in that the processing means, and the sum channel processing means is operating in real time in order to preserve the signal information of the digital left and right channel audio signals. 5. An adaptive signal weighted assembly according to claim 4, wherein the left and right high pass filter means are characterized in a cutoff frequency of less than or equal to 50 Hz. 6. An adaptive signal weighted assembly according to claim 5, wherein the left and right high pass filter means are characterized by a passband and a flat response in the passband. An adaptive digital signal weighting system for using a digitally expressed difference signal representing a difference between two stereophonic audio signals, the system having a predetermined sampling rate through the system to preserve the signal content of the information signal. Comprising a signal path for delivering an electronic information signal containing information relating to a difference signal of a predetermined bandwidth, the system being arranged in the signal path and varying by a variable gain rate in accordance with a first control signal. A digital filter device configured to vary a gain imposed on a portion of the information signal within a first selected spectrum region within the predetermined bandwidth, the second selected spectrum region including at least a portion of the first selected spectrum region; Respond only to the signal energy of the digitally represented difference signal and accordingly A first control signal generator configured to digitally generate the first control signal, varying the signal gain imposed on the information signal through a predetermined bandwidth by a second variable gain rate that varies in accordance with a second control signal. A digital gain controller disposed in the signal path and coupled to the digital filter structure and digitally converting the second control signal in accordance with the signal energy of the digitally represented difference signal in a third selection spectrum region within a predetermined bandwidth. A second control signal generator configured to generate a digital filter structure, a first control signal generator, a digital gain controller, and a second control signal generator operating at a predetermined sample rate to preserve the signal content of the information signal; The frequency of the pilot tone, the sample rate may be added to the difference signal to identify the encoded signal for the receiver. Adaptive digital signal weighting system as claimed selected to be equal to an integer multiple of. Encoding the electrical information signal of a predetermined bandwidth at a predetermined sample rate to preserve the signal content of the electrical information signal so that the information signal is an encoded difference signal representing the difference between two stereophonic audio signals. A digital system for enabling recording or transmission over a channel having a dependent dynamic range, wherein the channel has a dynamic limiting portion narrower than at least one other spectral region of the predetermined bandwidth in a first spectrum region, wherein An input configured to receive a system information signal, a signal transmission path coupled to the input and configured to deliver the information signal received at the input, the signal to provide the information signal encoded by the system Through the transmission path An output coupled to an input, a digital gain controller imposed on the information signal through a predetermined bandwidth and coupled within the signal path to vary the signal gain that varies with the first control signal, the signal path and the digital gain controller A digital filter coupled to and configured to impose a signal gain varying according to a second control signal to a portion of the information signal in the first spectrum region to preemphasize a portion of the information signal with respect to the remaining portion of the information signal. A device, a first control signal generator configured to digitally generate the first control signal in accordance with the signal energy of the information signal in a second spectrum region of the predetermined bandwidth, and at least a portion of the first spectrum region The second agent according to the signal energy of the information signal within a predetermined bandwidth And a second control signal generator configured to digitally generate the fish signal, wherein the digital gain controller, the digital filter structure, the first control signal generator, and the second control signal generator are configured to output the signal content of the information signal. Operating at a predetermined sample rate for preservation, the sample rate being selected to be equal to an integer multiple of the frequency of the pilot tone that can be added to identify the encoded signal for the receiver. An adaptive digital signal weighting system for using a digitally expressed difference signal representing a difference between two stereophonic audio signals, the signal path for transmitting an electrical information signal of a predetermined bandwidth through the system, the information Configured to filter the signal and characterized by a variable coefficient transfer function so as to vary the gain imposed on a portion of the information signal within a first selected spectrum region within the predetermined bandwidth by a first variable gain rate. And a variable coefficient digital filter structure in which the variable coefficient transfer function and the variable efficiency of the first variable gain rate change in accordance with a first control signal, and in at least a portion of the first selected spectrum region. A first control signal configured to digitally generate the first control signal in accordance with signal energy Generator, digital gain disposed in the signal path and coupled to the variable efficiency digital filter structure to change the signal gain imposed on the information signal through a predetermined bandwidth by a second variable gain rate that varies in accordance with a second control signal. A controller and a second control signal generator configured to respond to signal energy of the information signal in a third selected spectrum region within a predetermined bandwidth and to digitally generate the second control signal accordingly, the digital filter structure The first control signal generator, the digital gain controller, and the second control signal generator operate at a predetermined sample rate to preserve the signal content of the information signal, the sample rate being used to provide a difference signal to the receiver. Selected to be equal to an integer multiple of the frequency of the pilot tone that can be added to the difference signal for identification. Adaptive digital signal weighting system. The adaptive signal weighting assembly according to claim 4, wherein the sum channel processing means and the difference channel processing means are configured in a single integrated circuit. 4. The adaptive signal weighted assembly of claim 3, wherein the digital left and right channel audio signal is a digital sampling signal sampled at a sampling frequency of N × 15,734 Hz (N is an integer greater than or equal to 3). 5. The apparatus of claim 4, wherein the difference channel processing means is coupled to difference input means for receiving the digital difference signal, difference output means for providing an encoded difference signal, and coupled to the difference input means and encoded from the digital difference signal. Adaptive signal weight assembly comprising a differential signal transmission path comprising means for generating a differential signal. The method of claim 4, wherein the difference channel processing means receives the spectral compression input signal from the difference signal transmission path and compresses the spectral compression input signal according to the energy level of the encoded difference signal to generate a spectral compression output signal. And means for applying the spectral output signal to the difference signal transmission path. 14. The apparatus of claim 13, wherein the spectral compression means comprises means for measuring a first energy level of the coded difference signal in a first selection spectrum portion and generating a first control signal indicative of the first energy level. Adaptive signal weight assembly. 15. The filter according to claim 14, wherein said spectral compression means comprises a variable emphasis filter for receiving said digital signal and digitally filtering said spectral output signal to generate said spectral output signal. And a means for selecting said plurality of coefficients in accordance with said first control signal, said spectral compression means being characterized by a transmission function comprising a plurality of coefficients. 16. The adaptive signal weighted assembly of claim 15, wherein said means for selecting the plurality of coefficients comprises a memory lookup table. 17. The apparatus of claim 16, wherein the means for selecting the plurality of coefficients comprises an algebraic generator means for receiving the control signal and compressing logarithmically to generate an algebraically compressed signal. Adaptive signal weighting assembly comprising means for applying a compressed signal. 16. The adaptive signal weighting assembly of claim 15, wherein said means for selecting said plurality of coefficients calculates said plurality of coefficients as a function of said first control signal. 16. The apparatus of claim 15, wherein the spectral compression means includes spectral bandpass filter means for receiving and filtering a signal representing the encoded difference signal, wherein the spectral bandpass filter means generates the spectral signal. Adaptive signal weight assembly characterized in that characterized by the pass band of the first selection spectrum portion. 20. The apparatus of claim 19, wherein the spectral compression means comprises first RMS level detector means for receiving the spectral signal and generating the first control signal therefrom, wherein the first control signal is an RMS value of the spectral signal. Adaptive signal weight assembly, characterized in that. 20. The adaptive signal weight assembly of claim 19, wherein the spectral compression means comprises amplifying means for receiving and amplifying the coded difference signal to generate the signal representing the coded difference signal. 5. The method of claim 4, wherein the difference channel processing means receives the wideband compression input signal from the difference signal transmission path, compresses the wideband compression input signal according to the energy level of the encoded difference signal, and generates a wideband compression output signal. And wideband compression means, and means for applying said wideband compression output signal to said differential signal transmission path. 16. The adaptive signal weighting assembly of claim 15, wherein the spectral compression input signal comprises the wideband compression output signal. 23. The apparatus of claim 22, wherein the wideband compression means includes means for measuring a second energy level of the coded difference signal and indicating the second energy level and generating a second control signal at a second selection spectrum portion. Adaptive signal weight assembly. 25. The adaptive signal weighted assembly of claim 24, wherein the wideband compression means receives and amplifies the wideband compression input signal using the gain controlled by the second control signal to generate the wideband compression input signal. . 26. The apparatus of claim 25, wherein the wideband compression means includes wideband bandpass filter means for receiving and filtering a signal representing the coded difference signal to generate a wideband signal. Adaptive signal weight assembly characterized in that it is characterized by the passage area of the selection spectrum portion. 27. The apparatus of claim 26, wherein said broadband compression means comprises second RMS level detector means for receiving said broadband signal and generating said second control signal therefrom, wherein said second control signal is adapted to determine the RMS value of said broadband signal. Adaptive signal weight assembly, characterized in that. 5. The method of claim 4, wherein the difference channel processing means includes first low pass filter means for receiving a first low pass filter input signal from the difference signal transmission path, and the first low pass by low pass filtering the first low pass filter input signal. Means for generating a filter output signal, said means for applying said first low pass filter output signal to said difference signal transmission path. 5. The apparatus according to claim 4, wherein said difference channel processing means includes second low pass filter means for receiving a second low pass filter input signal from said difference signal transmission path, and means for low filtering said second low pass filter input signal. And means for applying said second low pass filter output signal to said difference signal transmission path. 5. The apparatus of claim 4, wherein the sum channel processing means is coupled to a sum input means for receiving the digital sum signal, a sum output means for providing an adjusted sum signal, the sum input means and the sum output means, and And a sum signal transmission path comprising means for generating the adjusted sum signal from a digital sum signal. 5. The apparatus according to claim 4, wherein said sum channel processing means comprises sum channel low pass filter means for receiving a sum channel low pass filter input signal from said sum signal transmission path and comprises means for low filtering said sum channel low pass filter input signal. And means for generating a sum channel low pass filter output signal, the means for applying the sum channel low pass filter output signal to the sum signal transmission path. 30. The adaptive signal weighted assembly according to claim 28 or 29, wherein the filter by the continuous first and second low pass filter means is similar to the filter by the sum channel low filter means. 33. The adaptive signal weighted assembly of claim 32, wherein the filtering by the sum channel low pass means is zero at 15,734 Hz. 33. The adaptive signal weighted assembly of claim 32, wherein the filter by the sum channel low pass means has a pass band between 0-15 kHz. 35. The adaptive signal weighted assembly according to claim 34, wherein the filter by the sum channel low pass filter cuts off at 15 kHz or more. 5. The method of claim 4, wherein the difference channel processing means includes a fixed first pre-emphasis filter means for receiving a first pre-emphasis input signal from the difference signal transmission path and receives the first pre-emphasis input signal. Means for generating a first pre-emphasis output signal by filtration and means for applying said first pre-emphasis output signal to said difference signal transmission path. 5. The method of claim 4, wherein the difference channel processing means filters the second pre-emphasis filter means and the second pre-emphasis input signal to receive a second pre-emphasis input signal from the difference signal transmission path. Means for generating a second pre-emphasis output signal and means for applying said second pre-emphasis output signal to said difference signal transmission path. 5. The 75 kHz preemphasis filter means for receiving a 75 kHz preemphasis input signal from the sum signal transmission path, and the 75 kHz preemphasis input signal to filter the 75 kHz preemphasis input signal. Means for generating a preemphasis output signal and means for applying said 75 kHz preemphasis output signal to said sum signal transmission path. 5. The apparatus of claim 4, further comprising: equalization filter means for receiving the equalization input signal from the sum signal transmission path by the sum channel processing means, means for filtering the equalization input signal to generate a level output signal; And means for applying said equalized output signal to said sum signal transmission path. 39. The filter according to claim 37 or 38, wherein the filter by the second pre-emphasis filter means is characterized by a second pre-emphasis phase response, and the filter by the 75-kHz preemphasis filter means is 75-kHz free. And the difference between the second pre-emphasis phase response and the first reference phase response is similar to the difference between the 75 ms pre-emphasis phase response and the second reference phase response. Adaptive signal weight assembly. 40. The filter according to claim 36 or 39, wherein the aftershock by the first pre-emphasis filter means is characterized by a first pre-emphasis phase response, and the after-effect by the equalization filter means is characterized by the equalization phase response. And the first pre-emphasis phase response is similar to the normalized phase response. 37. The adaptive signal weighted assembly of claim 36, wherein said first preemphasis input signal comprises said first low pass filter output signal. 43. The adaptive signal weighted assembly of claim 42, wherein said wideband compression input signal comprises said first preemphasis output signal. 38. The adaptive signal weight assembly of claim 37, wherein said second preemphasis input signal comprises said spectral compression output signal. 5. The apparatus of claim 4, further comprising: differential equalization filter means for receiving the differential equalization input signal from the difference signal transmission path, means for generating a differential equalization output signal by filtering the differential equalization input signal; And means for applying said differential equalization output signal to said differential signal transmission path. 5. The apparatus according to claim 4, wherein said sum channel processing means comprises: a joint equalization filter means for receiving a joint equalization input signal from said sum signal transmission path, means for filtering said joint equalization input signal to generate a joint equalization output signal; And means for applying said joint equalization output signal to said sum signal transmission path. 47. The filter according to claim 45 or 46, wherein the filter by said differential equalization filter means is characterized by a differential equalization phase response, and the filter by said variable emphasis filter is characterized by a variable emphasis phase response. The filter by means of a joint equalization filter means is characterized by a joint equalization phase response, wherein the differential equalization, variable emphasis and joint equalization phase response all change in accordance with the second control signal. 31. The adaptive signal weighting assembly according to claim 30, further comprising a complex modulator means for receiving said encoded difference signal and said sum signal and generating a complex modulated signal therefrom. 8. The adaptive digital signal weighting system of claim 7, wherein the electrical information signal comprises a pilot signal of pilot frequency and the pilot frequency is 15,734 Hz. 8. The adaptive digital signal weighting system according to claim 7, wherein the DC signal energy is removed from the signal content of the information signal to prevent the ticking phenomenon. 9. The digital system of claim 8 wherein the electrical information signal comprises a pilot signal of pilot frequency and the pilot frequency is 15,734 Hz. The digital system according to claim 8, wherein the DC signal energy is removed from the signal content of the information signal to prevent the ticking phenomenon. 10. The adaptive digital signal weighting system of claim 9, wherein the electrical information signal comprises a pilot signal of pilot frequency and the pilot frequency is 15,734 Hz. 10. The adaptive digital signal weighting system according to claim 9, wherein the DC signal energy is removed from the signal content of the information signal to prevent the ticking phenomenon. An adaptive digital signal weighting system for using a digitally expressed difference signal representing a difference between two stereophonic audio signals and a digitally represented sum signal representing a sum between two stereophonic audio signals. Delivering a second electrical information signal comprising information relating to a difference signal path for transmitting a first electrical information signal including information related to a difference signal of a predetermined bandwidth and a sum signal of a predetermined bandwidth through the system; Wherein said system is disposed in said difference signal path and said first within said first selected spectrum region within said predetermined bandwidth by a first variable gain rate that varies in accordance with a first control signal. And configured to vary the gain imposed on a portion of the information signal and functioning as a function of frequency. A first digital filter structure for introducing a phase error into the second digital filter structure, the signal energy of the digitally represented difference signal in the second selected spectrum region including at least a portion of the first selected spectrum region and thus the first control. A first control signal generator configured to digitally generate a signal, the difference signal to change the signal gain imposed on the first information signal through a predetermined bandwidth by a second variable gain rate that varies in accordance with the second control signal A digital gain controller disposed in the path and coupled to the digital filter structure, responsive to the signal energy of the digitally represented difference signal in a third selective spectrum region within a predetermined bandwidth, thereby digitally converting the second control signal A second control signal generator configured to generate and introduced into said difference signal path by said digital filter structure. And a signal compensator disposed in the sum signal path to compensate for the phase error, wherein the digital filter structure, the first control signal generator, the digital gain controller, and the second control signal generator preserve the signal content of the information signal. Adaptive digital signal weighting system characterized by operating at a predetermined sample rate. 56. The adaptive digital signal weighting system of claim 55, wherein the signal compensator comprises a static phase equalization filter. 56. The adaptive digital signal weighting system of claim 55, wherein the signal compensator comprises a dynamic phase equalization filter. 56. The adaptive digital signal weighting system of claim 55, wherein the signal compensator includes all pass filters having a relatively flat amplitude response characteristic and a preselected phase response characteristic. 56. The adaptive digital signal weighting system of claim 55, wherein the signal compensator comprises an IIR filter. A digital device of an analog adaptive digital signal weighting system for using a digitally expressed difference signal representing a difference between two stereophonic audio signals, the system having a predetermined sampling rate for preserving the signal content of the information signal. A signal path for delivering an electronic information signal containing information related to a difference signal of a predetermined bandwidth through the system, wherein the analog adaptive digital signal weighting system comprises a first analog filter element; In a form comprising a second analog control signal generator comprising an analog control signal generator and a second analog filter component, wherein the digital device is arranged in the signal path and varies by a variable gain rate that varies in accordance with a first control signal. In the first selection spectrum region within the predetermined bandwidth A digital filter structure configured to vary a gain imposed on a portion of the complementary signal, responsive only to the signal energy of the digitally represented difference signal in a second selected spectrum region comprising at least a portion of the first selected spectrum region Thus, the first control signal is configured to generate digitally and the difference between the amplitude response characteristic of the first digital filter element and the amplitude response characteristic of the first analog digital filter element is minimal regardless of their respective phase response characteristics. A first digital control signal generator comprising a first digital filter element having an amplitude response characteristic, wherein the information is imposed on the information signal through a predetermined bandwidth by a second variable gain rate varying in accordance with a second control signal. Digital gain disposed in the signal path and coupled to the digital filter structure to change the signal gain A controller and configured to respond to the signal energy of the digitally represented difference signal in a third selected spectrum region within a predetermined bandwidth and thereby to digitally generate the second control signal and to amplitude of the second digital filter component. A second digital control signal comprising a second digital filter component having an amplitude response characteristic such that the difference between the response characteristic and the amplitude response characteristic of the second analog digital filter component is minimal regardless of their respective phase response characteristics. An analog adaptive digital, comprising a generator, wherein the digital filter structure, the first control signal generator, the digital gain controller, and the second control signal generator operate at a predetermined sample rate to preserve the signal content of the information signal. Digital device of signal weighting system. 61. The digital device of claim 60 wherein the first digital filter component is comprised of a spectral bandpass filter. 61. The digital device of claim 60 wherein the first digital filter component is comprised of an IIR filter. 61. The digital device of claim 60, wherein the first digital filter component is comprised of a cascade of IIR filters. 61. The digital apparatus of claim 60 wherein the second digital filter component is comprised of an IIR filter. A system for generating a broadcast television stereo sound signal from a left channel signal and a right channel signal, the system configured to convert a right channel signal into a right digital signal and a left channel signal into a left digital signal. A signal combiner device coupled to a converter device and an analog-to-digital converter device to generate a sum signal consisting of a sum of a right digital signal and a left digital signal and to generate a difference signal consisting of a difference between the right digital signal and the left digital signal. A sum and difference signal generator device configured to generate a first pre-emphasis digital signal in accordance with the sum signal and a second pre-emphasis digital signal in accordance with the difference signal, and convert the first pre-emphasis digital signal into a digital BTSC compliant signal. Converts to an L + R signal and converts a second preemphasis digital signal into a digital BTSC compliant LR Signal converter connected and configured to convert the signal, to convert the digital BTSC compliant L + R signal to an analog BTSC compliant L + R signal and to convert the digital BTSC compliant LR signal to an analog BTSC compliant LR signal And a composite signal generator configured to generate a composite signal according to a combination of a modulated version of an analog BTSC compliant L + R signal and an analog BTSC compliant LR signal. Stereo sound signal generation system. 66. The system of claim 65, wherein the sum and difference signal generator devices are configured with a digital signal processor device programmed to digitally add preemphasis to each sum and difference signal. 66. The system of claim 65, wherein the signal conversion device is composed of an L-R data path and an L + R data path, and each path has a preselected sampling rate. A method for generating a broadcast television stereo sound signal from a left channel signal and a right channel signal, the method converting a right channel signal into a right digital signal and converting a left channel signal into a left digital signal; Generating a sum signal consisting of a sum of left digital signals and generating a difference signal consisting of a difference between a right digital signal and a left digital signal, generating a first pre-emphasis digital signal in accordance with the sum signal, Generating a second preemphasis digital signal, converting the first preemphasis digital signal into a digital BTSC compliant L + R signal and converting the second preemphasis digital signal into a digital BTSC compliant LR signal, Converts a digital BTSC compliant L + R signal to an analog BTSC compliant L + R signal and converts the digital BTSC compliant L + R signal. Converting the signal into an analog BTSC compliant LR signal and generating a complex signal according to a combination of an analog BTSC compliant L + R signal and a modulation version of the analog BTSC compliant LR signal. Method of generating an acoustic signal. 69. The method of claim 68, comprising converting a digital BTSC L-R signal to an analog BTSC compliant L-R signal and generating a modulated version of the analog BTSC compliant L-R signal. 69. The apparatus of claim 68, wherein generating a first pre-emphasis digital signal and generating a second pre-emphasis digital signal comprise a digital signal processor device programmed to digitally add pre-emphasis to each sum and difference signal. Method for generating a broadcast television stereo sound signal, characterized in that consisting of steps. 69. The method of claim 68, wherein converting the first preemphasis signal and converting the second preemphasis signal sample the first preemphasis signal at a preselected first sample rate and preselect a second sample rate. And sampling the second pre-emphasis signal in the broadcast television stereo sound signal generation method. A method for generating a broadcast television stereo sound signal from a left channel signal and a right channel signal, the method comprising converting a right channel signal into a right digital signal, converting a left channel signal into a left digital signal, right digital Generating a sum signal consisting of a sum of a signal and a left digital signal; generating a difference signal consisting of a difference between a right digital signal and a left digital signal; generating a first pre-emphasis digital signal matching the sum signal; Generating a second pre-emphasis digital signal corresponding to the difference signal, converting the first pre-emphasis digital signal into a digital BTSC compliant L + R signal, and converting the second pre-emphasis digital signal into a digital signal. Converting the BTSC compliant LR signal, converting the digital BTSC compliant L + R signal to an analog BTSC compliant L + R signal. Converting the digital BTSC compliant LR signal into an analog BTSC compliant LR signal and generating a composite signal according to a combination of an analog BTSC compliant L + R signal and a modulated version of the analog BTSC compliant LR signal. Method for generating a broadcast television stereo sound signal, characterized in that. 73. The method of claim 72, comprising converting a digital BTSC L-R signal into an analog BTSC compliant L-R signal and generating a modulated version of the analog BTSC compliant L-R signal. A system for generating a broadcast television stereo sound signal from a left channel signal and a right channel signal, the system configured to convert a right channel signal into a right digital signal and a left channel signal into a left digital signal. A signal combiner device coupled to a converter device and an analog-to-digital converter device to generate a sum signal consisting of a sum of a right digital signal and a left digital signal and to generate a difference signal consisting of a difference between the right digital signal and the left digital signal. A sum and difference signal generator device configured to generate a first pre-emphasis digital signal in accordance with the sum signal and a second pre-emphasis digital signal in accordance with the difference signal, and convert the first pre-emphasis digital signal into a digital BTSC compliant signal. Converts to an L + R signal and converts a second preemphasis digital signal into a digital BTSC compliant LR A signal converter connected and configured to convert into a signal, a composite signal generator device configured to generate a composite signal according to a combination of an analog BTSC compliant L + R signal and a modulation version of the analog BTSC compliant LR signal, and a digital composite signal And a digital-to-analog converter connected and configured to convert the analog signal into an analog composite signal. 75. The system of claim 74, wherein the sum and difference signal generator devices are configured with a digital signal processor device programmed to digitally add preemphasis to each sum and difference signal. 75. The system of claim 74, wherein the signal conversion device is composed of an L-R data path and an L + R data path, and each path has a preselected sampling rate. A method for generating a broadcast television stereo sound signal from a left channel signal and a right channel signal, the method converting a right channel signal into a right digital signal and converting a left channel signal into a left digital signal; Generating a sum signal consisting of a sum of left digital signals and generating a difference signal consisting of a difference between a right digital signal and a left digital signal, generating a first pre-emphasis digital signal in accordance with the sum signal, Generating a second preemphasis digital signal, converting the first preemphasis digital signal into a digital BTSC compliant L + R signal and converting the second preemphasis digital signal into a digital BTSC compliant LR signal, Generates a composite signal based on a combination of an analog BTSC compliant L + R signal and a modulated version of the analog BTSC compliant LR signal Method for generating stereo sound broadcast television signal for the stage, a digital composite signal, characterized by consisting of a step of conversion into an analog composite signal. 78. The method of claim 77, comprising generating a modulated version of the digital BTSC compliant L-R signal before converting the digital composite signal into an analog composite signal. 78. The apparatus of claim 77, wherein generating a first preemphasis digital signal and generating a second preemphasis digital signal comprise a digital signal processor device programmed to digitally add preemphasis to each sum and difference signal. Method for generating a broadcast television stereo sound signal, characterized in that consisting of steps. 78. The method of claim 77, wherein transforming the first preemphasis signal and converting the second preemphasis signal sample the first preemphasis signal at a preselected first sample rate and preselect a second sample rate. And sampling the second pre-emphasis signal in the broadcast television stereo sound signal generation method. A method for generating a broadcast television stereo sound signal from a left channel signal and a right channel signal, the method comprising converting a right channel signal into a right digital signal, converting a left channel signal into a left digital signal, right digital Generating a sum signal consisting of a sum of a signal and a left digital signal; generating a difference signal consisting of a difference between a right digital signal and a left digital signal; generating a first pre-emphasis digital signal matching the sum signal; Generating a second pre-emphasis digital signal corresponding to the difference signal, converting the first pre-emphasis digital signal into a digital BTSC compliant L + R signal, and converting the second pre-emphasis digital signal into a digital signal. Converting the BTSC compliant LR signal to a modulated version of the analog BTSC compliant L + R signal and the analog BTSC compliant LR signal. Generating a composite signal according to the sum, a method of generating a digital composite signal to broadcast television stereo sound signal, characterized by consisting of a step of conversion into an analog composite signal. 84. The method of claim 81, comprising generating a modulated version of the digital BTSC compliant L-R signal before generating the digital composite signal. A digital signal processor device used to generate a broadcast television stereo sound signal from a left channel signal and a right channel signal, the device comprising a digital sum signal corresponding to a sum of a left channel and a right channel, and a difference between a left channel and a right channel. A signal generator device configured to generate a digital difference signal according to the present invention, a sum signal signal processing device including a filter device configured to filter a digital sum signal to generate an adjusted digital sum signal, and to generate an adjusted digital difference signal. A difference signal signal processing device including a pre-emphasis filter device and a signal compressor device configured to adjust and compress a digital difference signal, converts the adjusted digital sum signal into an analog sum signal, and converts the adjusted digital difference signal into an analog difference signal. To analog converter device, analog sum signal and analog car Digital signal processor device for broadcasting TV stereo sound signals generated, characterized by consisting of a signal combiner unit configured to combine the modulated version. 84. The digital signal processor of claim 83, wherein the filter device of the sum signal signal processing apparatus is configured to filter the digital sum signal with 75 kHz pre-emphasis to generate an adjusted digital sum signal. Device. A digital signal processor device used to generate a broadcast television stereo sound signal from a left channel signal and a right channel signal, the device comprising a digital sum signal corresponding to a sum of a left channel and a right channel, and a difference between a left channel and a right channel. A signal generator device configured to generate a digital difference signal according to the present invention, a sum signal signal processing device including a filter device configured to filter a digital sum signal to generate an adjusted digital sum signal, and to generate an adjusted digital difference signal. Difference signal signal processing device including pre-emphasis filter device and signal compressor device configured to adjust and compress digital difference signal, and a modified version of digital sum signal and digital signal adjusted to generate complex modulated signal A signal combiner device configured to combine and convert a complex modulated signal into an analog output signal. Digital signal processor device for broadcasting TV stereo sound signals generated, characterized by consisting of a to-analog converter device for digital. 86. The digital signal processor device according to claim 85, wherein the filter device of the sum signal signal processing device is configured to filter the digital sum signal with 75 kHz pre-emphasis to generate an adjusted digital signal. .