JPS6336200B2 - - Google Patents

Description

[Detailed description of the invention]

[Industrial Application] The present invention reproduces the information on four separate channels after being transmitted or recorded on a medium with only two tracks, provided by four loudspeakers, and reproduces the information on an equal number of separate The present invention relates to a device that provides a listener with sound that appears to come from a sound source. In particular, it relates to the directional enhancement system described in pending British Patent Application No. 13047/74 (corresponding Japanese application, JP 50-153602), and forms part of the processing unit of that system. Here, the directional enhancement system receives the output signals of a 4-channel matrix decoder, continuously monitors the directional information encoded in those signals, and changes the degree of separation between those channels. This refers to a system that increases the perceptual directivity of reproduced sound in the listening area. [Prior art and its problems] A directional enhancement system is connected between the output terminal of a 4-channel matrix decoder and the input terminal of a power amplifier that operates the speakers of a 4-channel audio reproduction system. By performing functions that increase the degree of separation between channels, while maintaining all audio being played at the same total sound output power level as would be obtained in the absence of a record or This is a system that enhances or emphasizes directional information encoded in broadcasts. A matrix decoder uses loudspeakers placed around a listener with an acoustic field intended to reproduce the original sound signal recorded from all directions around the microphone in the same apparent direction around the listener. A pair of stereophonic signals is combined by phase shifting and amplitude matrixing to generate a plurality of speaker supply signals for driving. Typical such decoders are SQ decoders and QS decoders. The directional enhancement system consists of three important subsystems: a direction sensing system;
It includes a control signal processor and a matrix multiplier. These belong to the prior art. The direction detector system is shown in FIGS. 5 and 6 of the accompanying drawings of the present application, which correspond to FIGS. 2 and 4 of the previously referenced Japanese patent (JP-A-50-153602).
The system includes a plurality of signal rectifiers forming a gain control loop that together provide a set of six or more output signals corresponding to different coded audio directions and control subsequent circuitry. used for The control signal processor adds appropriate attack and decay time constants to the control signals and limits the signals to a certain maximum level having a value logically equal to one. In integrated circuits with directional enhancement systems, this level is approximately 1.9V, but may be any convenient voltage. Reference Japanese Patent No. 7
After passing through this circuit, which is shown in FIG. 9 of the accompanying drawing corresponding to the figure, the control signal is used to generate voltages corresponding to the coefficients of the matrix defined in the Japanese patent and called coefficient voltages. A matrix multiplier is a combination of an analog multiplier and a summing node that performs matrix multiplication of a set of input signals by a matrix whose coefficients are defined by the coefficient voltages extracted from the directional information obtained in the input signals. and generating a set of output signals. The output signal is then amplified and provided to a speaker. The attack and decay time constants added to the control signals in directional enhancement systems pose some problems while providing desirable characteristics to audio reproduction. Specifically, if two or more control signals are active at the same time,
The combined effect of the control signals acting on the coefficient generator is
This results in the formation of a matrix that has coefficients that exceed the level at which crosstalk is properly canceled, thereby generating excessive crosstalk. This can only occur if the sum of the control signals exceeds a predetermined value, a logic value of one. OBJECTS OF THE INVENTION The present invention is a subcircuit that can be incorporated between the attack and attenuation control circuitry of a control signal processor and the coefficient signal generation circuitry. The aim is to prevent the occurrence of the above-mentioned undesirable effects, and the invention therefore also improves the degree of dynamic separation that can be achieved, which makes it possible to change the audio reproduction from the original sound recorded by the microphone. This is advantageous in allowing more accurate correspondence to events. Its operating principle is to limit the control signals such that the sum of the control signals can never exceed a logic value of one. Additionally, it has been found desirable for the user to be able to adjust the degree of control over the maximum dynamic resolution achieved using a manually operated potentiometer. Such control is referred to as dimensional control, since crosstalk has the effect of reducing the apparent dimensionality of the spatial effects reproduced by a surround sound system. Because the present invention maintains the degree of separation at or near a level set by user control, it is referred to as automatic dimension control, following automatic frequency control and similar terms. In the processing unit of the directional enhancement system, the control signal issued by the direction detection device is subjected to appropriate level limiting characteristics and attack-attenuation (attack-attenuation).
decay) characteristics on the control signals prior to generating the coefficients of the modified matrix for provision to the matrix multiplier of the system. The attack attenuation characteristics herein refer to the attack and attenuation time constants applied to the control voltage by the attack and attenuation control circuits of the control signal processor of the directional enhancement system. The attack and attenuation control circuit is a circuit similar to the circuit shown in FIG. Refers to the circuit for generating electricity. Furthermore, a matrix multiplier is a matrix multiplication of a vector signal consisting of several time-varying signals by a matrix consisting of several coefficients expressed as voltages to generate an output vector signal consisting of several individual output signals. A combination of an analog multiplier and a summing amplifier configured to perform the numerical operations of It is an object of the invention that such control signal events
The purpose is to dynamically change the level limiting characteristics so that the optimal arrangement and separation of the corresponding sound sources can be obtained when the above-mentioned problems occur simultaneously. If you do this,
A single sound source located in an intermediate direction between the sound sources providing the control signal can be more accurately located and better separated. SUMMARY OF THE INVENTION It is therefore an object of the present invention to improve the performance of directional enhancement systems by providing the maximum possible separation and the best possible placement accuracy for a plurality of localized sound sources. A control signal event here refers to an attack on one or more control signals, ie, a sudden increase in voltage. The term "accurate placement" originally refers to the fact that a sound source at a certain azimuth angle in the recording environment should be reproduced so that it is perceived by the listener to be localized in the original direction as accurately as possible. It means something. The term corresponding sound source refers to the "control signal event" that occurs in the directional enhancement system when encoded and subsequently decoded.
represents the sound source that caused the sound. For example, the onset of sound generation from an instrument located in the left rear location causes the left rear control signal to rise, which is considered to constitute a "control signal event." At the same time, if another instrument, for example located halfway between center front and right front, were to emit a note, both the center front control signal and the right front control signal would rise.
If the invention is operating correctly, a logical value equal to the sum of the left rear control signal, the center front control signal, and the right front control signal (if the invention is not employed,
(which will probably exceed 1) is limited to 1. This prevents over-cancellation and the sound from the left rear sound source appears primarily on the left rear speaker (crosstalk to other speakers is minimized). The sound from the front instrument is transmitted to the front left speaker,
It appears at a different level to the front right speaker, and crosstalk to the two rear speakers is minimized. The front speakers should be panned at a ratio such that the apparent sound source direction is exactly halfway between the center front location and the right front location. Here, the "front-stage matrix decoding device" in the claims refers to, for example, several (usually four) output signals to be reproduced through speakers placed at (four) corners of a listening room. This is a device that matrixes stereo sound source signals from a phonograph to generate phonographs. The number of speakers may be 6 to 8 depending on the system. Also, "enhance the directional content of information"
This means that the reproduced signal appears in a more clearly localized state. The directional content of the information contained in an audio signal is such that the signal can be distinguished as a correlated signal in a specific phase-amplitude relationship defined by a phase-amplitude matrix coding system for each sound source direction. It is an ingredient. A directional detector system that is part of the directional enhancement system generates an output signal that varies according to this directional information content. Since the composite signal from the matrix decoder contains crosstalk, enhancing the directional content means increasing the separation between the (four) output channels by reducing crosstalk. This process is performed dynamically by the directional enhancement system and is called directional enhancement. This is why the expression "enhancing the directional content" is used. Additionally, "variable control device" refers to a manually controlled potentiometer or similar device operated by the user. DESCRIPTION OF THE EMBODIMENTS The present invention provides a method and apparatus for processing concurrent control signals in a directional enhancement system. This device is designed to obtain the optimal placement and separation characteristics of the corresponding sound source when reproduced by the output device, and in particular, the most effective effect is obtained by applying a directional enhancement system to the 4-channel sound device. demonstrate. This device is currently being applied for in UK Patent Application No. 13047/74.
(Corresponding Japanese Application, Application Publication No. 50-153602), which forms part of the processing unit of the directional enhancement system described in US Pat. Characterized by means of limiting the signal. It can also be used in conjunction with a limited type of manual dimensional adjustment. Here, in order to facilitate understanding of the present invention, the contents of British Patent Application No. 13047/74 (corresponding Japanese application, application publication number 50-153602) are explained based on the attached drawings and FIGS. 4 to 9. First, I will explain in detail. FIG. 4 shows a block diagram of a four channel stereo sound system. As shown in FIG.
Added to 00,101. Left (L) signal and right (R)
These pair of input signals consisting of signals are, for example, 2
It is obtained from track records and contains directional information. For example, the L and R signals are processed by the matrix 4-channel stereo decoder 10 of the SQ system.
Added to 2. Four output signals L′ _p (front left), R′ _p
(front right), R′ _B (rear right), and L′ _B (rear left) are the decoders 10
2. These four signals are processed by the circuit in the dotted block 114 to produce four emphasis signals L″ _p (front left), R″ _p (front right), R″ _B (rear _right ,
left rear). 4 speakers 110, 111, 112, 11
3 is shown in a rectangle 115 surrounded by a dotted line, and this rectangle 115 represents, for example, a room in which a four-channel stereo system is arranged. The speakers 110, 111, 112, and 113 are front left, front right, rear left, and rear right speakers, respectively. In this way, the signal L" _p is applied to the speaker 110 via the amplifier 106, and the signal R" _p is applied to the speaker 110 via the amplifier 107.
The signal L″ _B is applied to the speaker 112 via the amplifier 109, and the signal R″ _B is applied to the speaker 113 via the amplifier 108. Detector 103, processor 104, and matrix multiplier 105 generate emphasis signals L″ _p , R″ _p ,
Make L″ _B and R″ _B . L′ _p from decoder 102,
The R' _p , R' _B , and L' _B signals are applied to both detector 103 and matrix multiplier 105 . Detector 10
3 is activated by the presence of a dominant sound source in the corresponding direction θ according to the L′ _p , R′ _p , R′ _B , L′ _B signals applied thereto, using a suitable technique of amplitude comparison. A large number of control signals Cθ are generated. These control signals are applied to processor 104 as its input. Processor 104 adjusts the attack attenuation characteristics of each signal to produce optimal results, with circuitry controlling the charging and discharging of capacitors according to each control signal present. Based on this adjustment of the attack attenuation characteristics, processor 104 amplitude limits each signal and combines the signals in various ratios to produce signals corresponding to the coefficients of the matrix. Similarly, any other four-channel stereo phase matrix required to modify the output signal of a BMX, for example, may be determined. Additionally, the system generally described here matrix transforms the information for a given number of channels into a smaller number of channels, and then matrix transforms the information back to the given number of channels. All previous multichannel matrices are converted so that each output channel of the system contains both the desired signal and other crosstalk components, which may include directional information, in a given phase and amplitude relationship. Used to enhance the output signal from the system. These matrix coefficient signals are applied to matrix multiplier 105 _{shown in FIG .} The matrix multiplier 105 has vector components L' _p , R' _p , R' _B , L' _B
The process of matrix multiplication of the incoming signal vector d consisting of
The two audio signals L″ _p , R″ _p , R″ _B , and L″ _B are generated. These output signals are then sent to matrix decoder 1
02 is virtually identical acoustically to the original signal applied to the matrix encoder used for recording or transmission of the signals L, R applied to the 02. Regarding the detector 103, this detector 10
3 can produce any number of control signals as desired, but typically produces between 5 and 10 signals. Figures 5 and 6 detail two embodiments of the decoder 103 added to the SQ system. In FIG. 5 ten control signals are produced and in FIG. 6 six control signals are produced. In the embodiment of the detector 103 shown in FIG.
The four signals L' _p , R' _p , R' _B , L' _B are applied to four variable gain amplifiers 116, 117, 118, 119, respectively. Variable gain amplifier 116, 11
7, 118, and 119 form part of an automatic gain control device. The outputs of these amplifiers L′ _pp , R′ _pp ,
R' _B0 and L' _B0 are variable gain amplifiers 116, 117,
are essentially the same as the input signals L' _p , R' _p , R' _B , L' _B except that they are normalized to a predetermined level by signals 118 and 119 . Amplifiers 116-119
These output signals from
8,129. For example, amplifier 11
The output signal L' _pp from 6 is applied to an attenuator 120 and signal combiners 124, 127, and also to an amplifier 1
The signal R' _pp from 17 is applied to attenuator 121 and signal combiners 124, 125, 128, 129. Attenuators 120-123 attenuate these signals by the coefficients shown, and signal combiners 124-
129 is 10 signals S ₄₅ , S ₁₈₀ , S ₃₁₅ , S ₂₇₀ , S ₂₂₅ ,
The output signals are combined in the ratios shown to yield S ₀ , S ₁₃₅ , S ₉₀ , S' ₂₇₀ , and S' ₉₀ . Each of the above 10 signals becomes 0 when the signal arrives from the angular direction indicated by the subscript. Two signals S′ ₂₇₀ with dashes,
S′ ₉₀ is a signal that becomes 0 for the position encoded sound sources at the center left and center right, respectively. These signals also reach a maximum level for some other direction of the sound source, and the attenuation and coupling amounts of the attenuator and signal combiner have values that bring these maximum values all to the same level. The above direction sensing signals S ₄₅ , S ₁₈₀ , S ₃₁₅ , S ₂₇₀ ,
S ₂₂₅ , S ₀ , S ₁₃₅ , S ₉₀ , S' ₂₇₀ , S' ₉₀ are rectifier 13
0,131,132,133,134,135,
136, 137, 138, and 139, respectively. The transfer characteristics of each rectifier 130-139 are shown diagrammatically on the block representing each rectifier. There is no smoothing time constant at the output of the rectifier, and the rectifiers 130, 131, 132, 133,
134, 135, 136, 137, 138, 13
The unsmoothed outputs from 9 are connected to resistors 14, respectively.
0,141,142,143,144,145,
Added to 146, 147, 148, 149.
All these resistors 140-149 are connected to amplifier 1
50 inputs. Therefore, resistance 14
0-149 combine the signals from rectifiers 130-139 at the inputs of amplifier 150. resistance 14
If the resistance value of 3,144,145,147,148,149 is R, then resistor 140,141,14
The resistance value of 2,146 is 2R. Each of these resistance values corresponds to the corner signal, front center and rear center signals,
Equal proportions of each of the average left center signal and right center signal are applied to amplifier 150. resistance 1
51 serves as a feedback resistor, the value of which is chosen to produce an output DC level equal to a fraction of the maximum DC level occurring at any rectifier output. Capacitor 152 is a smoothing capacitor. amplifier 150
The output from is applied to a unity gain inverter 153. This inverter 153 is connected to variable gain amplifiers 116~
The output of amplifier 150 is inverted to obtain the correct polarity to drive the gain control input of 119.
In this manner, the signal from inverter 153 provides a gain control voltage to variable gain amplifiers 116-119. Amplifiers 116-119 may typically have gain control characteristics similar to those shown in FIG. The characteristic curve shown in FIG. 8 has a rather sharp turning point in the control voltage and a rather narrow operating range. The control action is essentially logarithmic, so that the amplifier gain decreases, for example, at a rate of 1 dB per mV above the knee voltage. The logarithmic characteristic improves the overall stability of the automatic gain controller, and the steepness of its slope ensures that the normalized signal remains very close to a predetermined level over a wide range of input levels. In addition to providing drive voltages to amplifiers 116-119, inverter 153 provides reference level voltages to comparison amplifiers 164-173. Comparison amplifier 164-1
73 each have two inputs, the first of which
is coupled to an input of an inverter 153, and a second input is coupled to a smoothing filter 15 as shown in FIG.
It is connected to the outputs of 4 to 163 different filters. The smoothing filters 154 to 163 are the rectifier 1
30-139 and comparator amplifiers 164-173 such that each rectifier output after smoothing by an associated smoothing filter is applied as a second input to its associated comparator amplifier. That is, the output of rectifier 130 is smoothed by filter 154 and then applied to the second input of comparison amplifier 164, and the output of rectifier 131 is smoothed by filter 155 and then applied to the second input of comparison amplifier 165.
is added to the input of , and the following connections and operations are similar. The output of each smoothing filter 154-163 is compared to a reference level voltage at the output of inverter 153 by its associated comparison amplifier 164-173. Comparison amplifiers 164-173 have inputs from their associated smoothing filters 154-163 connected to inverter 153.
produces an output only when the voltage is lower than the reference level voltage from The reference level voltage is the comparator amplifier 1
Since each direction signal applied to comparators 164 to 173 may have a minimum value other than the minimum value in a particular direction, only signals applied to comparators 164 to 173 that fall within a limited range of characteristic directions are below the reference level. be chosen to descend to comparison amplifiers 164, 165, 166, 167,
168, 169, 170, 171, 172, 17
The output signals from 3 include C ₄₅ , C ₁₈₀ , C ₃₁₅ , C ₂₇₀ ,
C ₂₂₅ , C ₀ , C ₁₃₅ , C ₉₀ , C′ ₂₇₀ , and C′ ₉₀ are added. These signals are the raw directional control signals applied to processor 104. FIG. 7 shows a typical interface between the smoothing filter of FIG. 5 and its associated rectifier, and shows typical filters that can be used as smoothing filters 154-163. Figure 7 shows resistance 1
A two-stage ladder filter consisting of capacitors 80 and 181 and capacitors 182 and 183 is shown. This filter is designed to satisfactorily attenuate ripples at the lowest signal frequencies and to have as fast a transient response as possible. Since the purpose of this filter is to detect the absence of a signal at the detection point, the PNP transistor 184 is configured to provide a full-wave rectifier with a signal applied to the base of the transistor 184 from the associated rectifier, typically with a non-zero average level. It is used to pull down the input whenever the audio signal drops to zero. Current is supplied to the emitter of transistor 184 by current source 185. The power supply current 185 has a value such that the positive peak voltage of the input signal can be faithfully reproduced at the emitter. In this way, the response to sudden cessation of the signal is as quick as possible. FIG. 6 shows a simplified embodiment of the detector. In other words, FIG. 6 is a simplified version of the circuit of FIG. 5. In FIG. 5, ten control signals are generated, but in FIG. 6, only six control signals are generated, and the control signals for the left and right centers are omitted. Of course, this saves the cost of the detector and subsequent circuitry, but the cost savings may be accompanied by a degradation in the quality of the final signal to the loudspeaker. Except for the reduction in the number of control signals produced, the circuit of FIG. 6 is the same as the circuit of FIG. _A comparison _between FIG. 6 and _{FIG .} 5 shows that in FIG. The resulting comparison amplifier, smoothing filter, rectifier, signal combiner and resistor 140,1 of FIG.
41, 142, and 146 are omitted in the circuit of FIG. 6, and in other respects both circuits are the same, and it is clear that the operation of the same circuit is the same. Therefore, the numbers used in FIG. 6 are the same as those used in FIG. 5 for corresponding elements. To discuss the circuit of FIG. 6 in detail would merely be a repetition of the description of FIG. 5, and thus would not require further explanation. However, since the center left, center right signals and the alternative center left, center right signals do not occur in FIG. 6, the resistor in FIG. 5 that is omitted from FIG. Resistors 140, 141, 142, 146 having
Please note that. Regarding the detector 103, the detailed circuits shown in FIGS. 5 and 6 are given as examples of preferred circuit configurations of the detector 103, but it is possible for those skilled in the art to make various changes and modifications to these circuits. It will be obvious to those who have done so. but,
It is of particular significance that these circuits provide a reference voltage that is virtually independent of the direction of the sound source, and it is also possible to determine the reference voltage solely from the rectified levels of these signals without combining the four input signals. This results in an improvement over devices of this type. for example,
The reference level signal obtained by the ten directional control signals in the system of FIG. 5 varies by less than 0.5% in the direction of the sound source. However, in other systems that determine the reference level voltage only from the rectified levels of the four signals, the reference voltage is 14% in the direction of the sound source.
The degree may also vary. Furthermore, by comparing the reference voltage obtained in FIGS. 5 and 6 with the rectified signal, the detection efficiency can be made virtually independent of the signal level. This ensures that the directional enhancement contained in each signal remains effective over a wider range of input signal levels than that obtained in other systems. Figure 9 shows a typical circuit and processor 1 that can be used in place of the comparison amplifiers in Figures 5 and 6.
A representative circuit of a portion of 04 is shown. Therefore, the comparison amplifier shown in detail in FIG. 9 may be the comparison amplifier 164 of FIGS. 5 and 6. All other comparator amplifier circuits in FIGS. 5 and 6 are the same as the circuit shown in FIG. Furthermore, other suitable comparison circuits known in the art may also be used as the comparison amplifiers of FIGS. 5, 6, and 10. Referring now specifically to FIG. 9, comparison amplifier 1
64 includes a first transistor 300 and a second transistor 302. transistor 300
The emitter of is connected to the transistor 30 via the resistor 301.
It is connected to the second emitter. The input signal from the associated rectifier, the comparison amplifier shown in detail is 164
Therefore, in the case described here, the input signal from regulator 154 of FIG. 5 or 6 is applied to the base of transistor 300. A reference level voltage is applied to the base of transistor 302. Current source 312 is connected to transistor 30.
The emitter of 2 is connected to the connection point of the resistor 301,
The collector of transistor 302 is connected to ground. The circuit connection diagram in FIG. 9 is a portion of the entire circuit for processor 104. As shown in FIG. 9, this circuit includes a pair of Darlington-connected transistors 310, 311, a second pair of base-connected transistors 306, 307, and a plurality of transistors 315a-315x. The number of transistors 315 provided is one less than the number of comparison amplifiers. Thus, in the 10-signal system shown in FIG.
5 is provided at each stage. In the case of 10 signals, the 9th
The same circuit as shown in the figure is provided in 10 stages. Three diodes 316, 31 connected in series
7, 318 are connected between the collector of transistor 311 and ground, and current source 314 is connected to the collector of transistor 311. Any output signal present at the collector of transistor 311 is applied to the coefficient generator. Each transistor 315 is all connected in parallel, and a diode 308 and a resistor 309 are connected between ground and the emitter of each transistor 315. The base of each transistor 315 is connected to one of the other stages.
It is connected to point 313 of two different stages. For example, the base of transistor 315a is connected to point 313 at the stage of processor 104 associated with comparison amplifier 165 in FIG.
is connected to point 313 at the stage of processor 104 associated with comparator amplifier 166. Both emitters of transistors 307 and 306 are connected to ground, and their bases are coupled together and connected to the common junction of series connected resistor 309 and diode 308. The collector of transistor 307 is connected to the base of Darlington transistor 310 and the common junction of resistor 303 and the collector of comparison transistor 300. A resistor 303 is connected in series with a capacitor 304 between ground and comparator transistor 300. Diode 305 is capacitor 3
04, and the collector of the transistor 306 is connected to the diode 305, the capacitor 304,
It is connected to the common connection point of resistor 303. As discussed above and shown in FIG. 9, a signal from an associated rectifier is applied to the base of comparator transistor 300 and a reference level voltage is applied to the base of comparator transistor 302. Under normal conditions, the current provided by current source 312 flows through the collector of transistor 302. However, if the input signal voltage applied to the base of comparator transistor 300 falls below the reference level voltage applied to the base of comparator transistor 302, a portion of the current will flow through the collector of comparator transistor 300. be distracted by Resistor 301 effectively transfers the entire current from transistor 3 well before the output from the associated rectifier reaches zero.
It is chosen to have a resistance value that guarantees a transition to 0.00. This property aids in detecting the dominant signal when there are signals of significant magnitude from other directions. Current from current source 312 flows through transistor 300
When deflected, a voltage drop occurs across resistor 303 and capacitor 304 begins to charge. At the same time, the voltage at the emitter of transistor 311, that is, the Darlington pair, is
immediately rises to a maximum value determined by . A PNP Darlington pair of transistors 310 and 311 acts as a buffer and level shifter. When the signal to the base of comparator transistor 300 is a narrow pulse, capacitor 3
04 does not hold enough charge and transistor 31
The current in the emitter 1 decreases rapidly. On the other hand, when the signal to the base of comparator transistor 300 is a wide pulse, capacitor 304 charges to a level limited by diode 305, and when the input signal to comparator transistor 300 is removed, capacitor 304 slowly charges. is discharged, thereby slowly attenuating the output current at the emitter of transistor 311. If, after some time, another signal from a different direction occurs, one of the other stages will operate and the output voltage from that stage will be applied to the base of transistor 315 connected to point 313 of that stage, and Therefore, current flows through the resistor 309. resistance 309
The current flowing through resistor 3 creates a forward bias in the bases of transistors 306 and 307.
The current of 09 causes the same current to flow through transistors 306 and 307. Its forward bias is limited by diode 308. transistor 3
This current at 06,307 immediately pulls down the base voltage of transistor 310 and begins to discharge capacitor 304 fairly rapidly. When the signals in the other stages disappear, currents in transistors 306 and 307 stop flowing. Transistors 306, 307
The time that the current flows is very short and the capacitor 3
04 is not completely discharged, the voltage at the base of transistor 310 will be lower than transistor 30.
It rises again when the short-term current of 6,307 is gone. This feature allows the directional control signal for the dominant signal to be momentarily snatched by a short signal from a different direction. Here, the output signals C ₀ , C ₄₅ , C ₁₃₅ , C ₁₈₀ , C ₂₂₅ , and C ₃₁₅ output from the circuit shown in FIG. 6 described above are supplied to the circuit shown in FIG. 1. On the other hand, the present invention includes an automatic control device in a device that enhances the directional content of information contained in a plurality of composite signals output from a decoding device. This automatic control device consists of the following three things. 1. A summing device having a number of inputs equal to the number of control signals output by the directional enhancement system and one output. It serves to output a signal that is proportional to the sum of the signals received at its inputs. 2. A plurality of dividing devices each having a first input, a second input, and an output - divides a first signal entering the first input by a second signal entering the second input. , serves to output a signal proportional to the quotient. One dividing device is respectively applied to one control signal, and one of the different control signals is
is applied to a first input of each divider, and the sum signal output from the adder is applied to a second input of each divider. 3 A plurality of limiting devices each having one signal terminal and one control terminal, which prevent the signal level at the signal terminal from exceeding the level at the control terminal. The control terminal of each limiter is applied with one of the output signals of a different divider, the signal terminal of each limiter is connected to the first input terminal of the divider, and the output of this divider is connected to the same limiter. is supplied to the control terminal of The working of this entire system is to limit the maximum value of the sum of the control signals entering the input of the system to a constant value, without changing the ratio between these control signals, and to limit the maximum value of the sum of the control parameters Cθ represented by the control signals. This limits the value to 1. One of the effects of the directional enhancement system specified in pending British patent application No. 13047/74 (corresponding Japanese application, JP 50-153602) is that it follows a predetermined equation and that the detection device of the system The second effect is to generate the coefficients of a correction matrix M according to the determined direction of the most dominant sound source. generate a vector m of decomposed signals such that the dominant sound source appears only in the corresponding loudspeaker.
Change the overall transmission matrix T ₀ . The third effect is to maintain a constant total output power at the loudspeaker for each source. This process is expressed by the following equation. m = M d ... (1) = T s ... (2) This T is the 4 × 4 vector from the original signal vector s to the converted and decoded signal vector m .
transformation matrix, and T=MT ₀ …(3) where T ₀ is from the original signal vector s ,
The total transformation matrix up to the decoded signal vector d provided to the input of the directional enhancement system. The correction matrix is divided into the following two parts. M=B+I (4) where I is its identity transformation and B is the time-varying difference between M and I required to make this required correction. If the control signal C〓 is given several different control directions θ, then the matrix B can be written as a linear combination B〓 of a given matrix for each direction, where C〓 varies with time. . B(t)=Σ〓C〓(t)B〓...(5) When a sound source occurs between the two directions for which control signals are prepared, the control parameter becomes 0.
If we take an intermediate value between A modification matrix results. For the left side or right side signal,
It was specifically shown that complete isolation and correct alignment can be achieved, although the total output is reduced by 1.8 dB. An advantage of the present invention is that by dynamically limiting the sum of the control signals closer to unity, all single sound sources occurring anywhere can be optimally separated. A precise state has been established in which sound sources in any direction are completely separated, and this is as shown below. When applied to the CBS SQ4 channel sound system, a directional enhancement system with six control parameters and six directional correction matrices was used, given the control signals and correction matrices. The direction angle θ is 0° (straight ahead), 45° (right front), 135° (right rear), 180° (straight back), 225° (left rear), and 315° (
front left). The modification matrix was designated by a subscript equal to the direction angle, and the coefficients corresponding to the B matrix were: Find the necessary combination of these matrices in order to completely separate sound sources located at corners, directly in front of you, and directly behind you. We need to know how these sound sources are encoded into symbols and decoded by the SQ device. The sound sources for the front and rear quadrants are SQ
Although encoded directly by the matrix equation, the sine and cosine functions are usually given by a sine and cosine rotating voltage divider. Therefore, 0° and
For signals located between 45° and 45°, the original signal vector takes the form: s = cos (θ + 45°) sin (θ + 45°) 0 0 ...(12) Between 45° and 135°, the signal vector has the following form. There is a symmetrical relationship that creates a similar original signal vector in other directions as well. Therefore, in order to completely separate the signals in the above-mentioned formats, and only to ensure that the symmetry of the system is correct, it is necessary to consider the appropriate values of the control coefficients, and the reason for this will be explained in the following. This is to completely isolate any sound sources located around the person. Let us consider the necessary conditions for these coefficients. 0° and
For directional angles located between 45°, we need:
The only non-zero control parameter is C ₀
and C ₄₅ , and these parameters should take numerical values as a function of the direction angle. Let's apply it to the equation to simplify it. C ₀ =f(θ)...(14A) and _C45 =g(θ)...(14B) Then, the modification matrix M(θ) is: M(θ)=f(θ)B ₀ +g(θ)B ₄₅ +I (15) The decoded signal vector is generated by applying the transformation matrix T ₀ to the signal vector s of the sound source, and a= 45°−θ …(16) With the triangle identity cosX=sin(90°−X),
The signal vector of the sound source is S=six a cos a 0 0 (17), and the decoded signal vector is as follows. So, the modified signal vector is m = M d (1), where M is as shown below. The necessary condition for erasing the signal of the later channel is that if two lines under M are multiplied by d , a zero product is obtained. This applies to both real numbers and imaginary numbers in the product part, and at first glance there seem to be four equations to determine the two unknowns, but two of these equations are equal to the other two,
The result is two separate equations. 0.644ft+0.356f+g-1=0...(20) 0.644f+0.356ft-0.155gt-t=0...(21) Here, t=tan a...(22) If we eliminate g here, f=t/0.5574+0 .3564t+0.0862t ² …(23) Or, by eliminating f, g=1−t ² /1+0.6392t+0.1547t ² …(24) where the real and imaginary components of the two channels before the modified signal is found for each direction angle and the resulting level of output is also calculated. or,
Since the regenerated signal appears only in the previous two channels, its placement can be confirmed, and in terms of ratio, t′=tan a′=m ₁ /m ₂ …(25) Here, the regenerated direction is The angle θ' is expressed by the following formula. θ′=45°−a′ …(26) The placement error is the difference between the original direction angle and the reproduced direction angle, and as seen in Table 1, the maximum placement error within this range of sound source angles is is 1.36°. Such errors are only slightly noticeable when directly comparing the reproduced signal to the original signal. Perfect channel separation can be defined by a ratio expressed in decibels: the ratio of the signal output of the two channels of the signal requiring directional correct reproduction to that of the other channel. In an ideal decoder, this would always be infinite for all source directions, but in practice this is not possible due to component tolerances and detector performance. Since the total channel separation (TQS) of a simple matrix decoding device made under the above conditions is always 0 dB, the TQS of a simple matrix decoding device is based on the above, and the numerical value obtained with an actual decoding device is It is a reliable indication of improvement. The above theory for source angles of 0° and 45° gives the same equations for f and g at 135° and 180°, 180° and 225°, and 315° and 360°, and their control The relationship between the parameters and these functions is expressed by the method below. C ₁₃₅ = g, C ₁₈₀ = f, a=θ−135° 135°θ180° C _{225 = g, C 180 =} f, a=225°−θ 180°θ225° C ₃₁₅ = g, C ₀ = f, a=θ−315° 315°θ360° (27) In the horizontal quadrant, a necessary condition for complete separation is that the component of the reproduced signal appearing on the opposite side is always zero. According to the above-mentioned theory, the numerical value of the control signal required for the sound source signal configuration expressed by equation (13) is as follows. d=(1-t)(0.5774+0.2391t)/0.5774+0.478
3t+0.0991t ² …(28) e=t(0.8165+0.3382t)/0.5774+0.4783t+0.09
91t ² ...(29) Here, as before, t=tan a, or the relationship between a and θ in these two quadrants is as follows. C ₄₅ = d, C ₁₃₅ = e, a = 1/2 (θ-45°) 45° θ135° C ₂₂₅ = d, C ₃₁₅ = e, a = 1/2 (θ-225°) 225° θ315° …(30) Another useful feature of the numerical values of e and d is that their sum is always 1; moreover, the sum of the numerical values of f and g is also approximately 1, so that a simple restriction The circuit provides approximately the correct properties for the summation of the control signals. Before describing the circuit that accomplishes this, I would like to draw your attention to Table 1. d,
The methods of e and fg are calculated at 5° intervals by a computer, and the resulting decoded signals are arranged and integrated into 4
It was found that the separation of the channels and the total effective power output from the four channels. As seen in this table, the placement error in the front and rear quadrants is always less than 1.4°, and in the horizontal quadrants is always less than 2.25°,
The difference in total output is less than 0.25 dB in the front and rear quadrants, and less than 1.9 dB in the horizontal quadrant. The overall performance of such a system is thus virtually perfect in these respects. The effect of this control system, as seen in the listening experiments, is that when there are two or more signals, they are all placed correctly. This is because, in many cases, the control signal is below the maximum possible value, but when there are two equal signals on opposite sides of the room, the control circuit will not dominate either of the Serves to increase signal separation. This idea is
The resulting correction matrix and the influence on the sound source in the main direction of the control signal are confirmed by mathematical considerations. The separation is also increased when there are three equal signals at the same time. This system works instantly, so
It is not perceptible to the ear that multiple sources are not separated; this is because when one source is sufficiently dominant over another, the separation is increased to the maximum extent variable by the tolerances of the components of the system. Here, the maximum variable range will be explained. Ideally, the degree of separation of a single signal applied to a directional enhancement system, such as a signal intended only for the left front speaker, would be infinite; This also suggests that no crosstalk occurs. However, component tolerances limit the maximum degree of separation that can be achieved. In practice, separations of 35 dB to 55 dB are achieved. This is the "maximum possible range",
Directional enhancement systems can increase the degree of separation to this range. Without the use of a directional enhancement system, the total four channel separation (as defined herein) is 0 dB. When only one sound source is primarily active, the control signal or signals generated by directional information about that sound source reach the highest achievable level, and the degree of separation is as high as physically possible. It becomes Satoshi.
In mixed situations, where only a single sound source is not dominant, the separation for up to three sound sources is somewhat improved, and the correct position can be almost fully determined by ear. As the power of any one of these sound sources increases, its separation increases and its location can be accurately heard. The brain assumes that the sound source remains stationary unless there is evidence that it has moved. Therefore, when a previously dominant sound source loses its power, the ear perceives the source's directionality as unchanged, even though other sounds present may cause the decoder to place the source in a different location. do.

〔Effect of the invention〕

An advantage of the present invention is that by dynamically controlling the sum of control signals closer to unity, all single sound sources occurring anywhere can be optimally separated.

[Brief explanation of the drawing]

FIG. 1 is a detailed diagram of an attack/attenuation control circuit of a processing device of a directional enhancement system according to the prior art. FIG. 2 is a detailed circuit diagram of the automatic dimensional control part according to the present invention, and FIG. 3 shows another embodiment of the present invention capable of automatic dimensional control of a processing device. Figure 4 is a block diagram showing a four-channel stereo sound system; Figure 5 is a diagram of a detector applied to an SQ sound system. Detailed configuration diagram, No. 6
The figure is a configuration diagram showing a modification of the detector shown in Fig. 5.
Figure 8 is a schematic diagram of a preferred interface between the rectifier output and a typical smoothing filter of the detector.
The figure is an amplifier gain control characteristic diagram, and FIG. 9 is a circuit diagram of a comparator used in the processor. Explanation of symbols of main parts, addition device (signal addition device)...201, division device (output device)...202~
207, restriction device...137-142.

Claims (1)

[Claims] 1. In a device for enhancing the directional content of information included in a plurality of control signals output from a matrix decoding device in the previous stage, the device is configured to a summing device having an equal number of inputs and one output and generating at its output a signal proportional to the sum of the signals received at its inputs; a summing device provided for each of said control signals and receiving one of said different control signals; 1 input terminal, a second input terminal for inputting the sum signal output from the adding device, and dividing the control signal entering the first input terminal by the sum signal entering the second input terminal; and an output terminal for outputting an output signal proportional to the quotient; and a plurality of dividing devices each having an output terminal for outputting an output signal proportional to the quotient; one signal terminal connected to a flowing line and a control terminal connected to the output terminal of the dividing device;
and a plurality of limiting devices such as diodes that prevent the voltage at the signal terminal from exceeding the voltage at the control terminal, and by limiting the maximum value of the sum of the control signals to a constant value, An automatic dimension control device characterized in that the maximum value of the sum of control parameters C represented by a signal is limited to 1. 2. In the automatic dimensional control device according to claim 1, another connecting device connected to the variable control device operates to limit the maximum value output from the adding device, thereby causing the control signal to An automatic dimension control device that limits the maximum value of the sum of control coefficients to be less than 1.