US3786193A

US3786193A - Four channel decoder with variable mixing of the output channels

Info

Publication number: US3786193A
Application number: US00272439A
Authority: US
Inventors: K Tsurushima
Original assignee: Sony Corp
Current assignee: Sony Corp
Priority date: 1971-07-19
Filing date: 1972-07-17
Publication date: 1974-01-15
Anticipated expiration: 1991-01-15
Also published as: DE2235238A1; IT963000B; GB1400061A; CA991086A; FR2146394B1; FR2146394A1; JPS5213082B1; CA1001957A; NL7209930A

Abstract

A crosstalk eliminating circuit for use in a four channel stereo decoder of the type which converts two composite signals LT and RT into four output signals containing dominant signal components Lf, Rf, Lb and Rb, respectively, with each of the output signals further including subdominant signal components as crosstalk. The elimination circuit includes a crosstalk detecting circuit having a plurality of all wave rectifying circuits and subtraction circuits for producing separate signals for controlling a plurality of mixing circuits connected between separate pairs of output channels for transmitting the output signals, the mixing circuits being controlled so as to mix two output signals in accordance with the respective control signal in such a manner that the crosstalk signals contained therein are cancelled.

Description

Tsurushima Jan. 15, 1974 1 FOUR CHANNEL DECODER WITH VARIABLE MIXING OF THE OUTPUT CHANNELS Katsuaki Tsurushima, Yokohama, Japan [75] Inventor:

Assignee:

Filed:

Appl. No.:

Sony Corporation, Tokyo, Japan July 17, 1972 Foreign Application Priority Data July 19, 1971 Japan 46/53771 179/100.1 TD, 15 HT References Cited UNITED STATES PATENTS 1/1972 561161136! 179/1 GO 1/1973 Bauer 179/1 GQ OTHER PUBLICATIONS The New Sansui 20db Matrix, Audio Magazine, July 1972. t

Primary Examiner-Kathleen H. Claify Assistant Examiner-Thomas DArnico At i0 rneyLewis 1-1. Eslinger, Esq. and Alvin Sinderbrand, Esq.

1 s7 ABSTRACT A crosstalk eliminating circuit for use in a four channel stereo decoder of the type which converts two composite signals L and R into four output signals containing dominant signal components L], R,, L,, and R respectively, with each of the output signals further including subdominant signal components as crosstalk. The elimination circuit includes a crosstalk detecting circuit having a plurality of all wave rectifying circuits and subtraction circuits for producing tseparate signals for controlling a plurality of mixing circuits connected between separate pairs of output channels for transmitting the output signals, the mixing circuits being controlled so as to mix two output signals in accordance with the respective control signal in such a manner that the crosstalk signals contained therein are cancelled.

9 Claims, 11 Drawing Figures PATENTED JAN 15 I974 saws 0F 6 PATENTEDJAK 15 1974 SHEET 5 BF 6 5 @QQQQ is w Sm PATENTEDJAN 15 m4 Ta w P Q @T QQNL W 0% M x FOUR CHANNEL DECODER WITH VARIABLE MIXING OF THE OUTPUT CHANNELS BACKGROUND OF THE INVENTION 1. Field of the Invention This invention relates to a multisignal transmission apparatus and more particularly to improved decoding and reproduction by a plurality of loudspeakers to give a listener a highly realistic multichannel sound program.

2. Description of the Prior Art A so-called matrix four channel stereo system has been heretofore proposed in which four original sound signals (which, for convenience, are identified as L,, L,,, R, and R,, for left front, left back, right front and right back, respectively) are converted into signals of only two channels by matrix networks called encoders for transmission by, or recording on conventional two channel media such as FM multiplex transmission or magnetic tape recording. In order to reproduce the encoded signals from the two channel media they are decoded to four signals by matrix networks which are called decoders.

It is preferred that the corresponding original sound signals L L,,, R, and R,, be reproduced only from separate loudspeakers. With such matrix four channel stereo systems, however, in addition to the corresponding original sound signal, another sound signal which must be reproduced through another loudspeaker is atthe same time, reproduced as a crosstalk, which is an obviously undesirable result.

In order to avoid such undesired crosstalk, it has been suggested to use a gain control amplifier with a special logic circuit, for example, as described in U.S. Pat. No. 3,632,886. In such a system the gain of channels in which the crosstalk exists is reduced with a control signal. Accordingly, if only the crosstalk signal is taken into account, no crosstalk exists with the result that the separation between the channels is improved.

' However, if a main signal of small level exists in the channel together with the crosstalk signal, the mainsignal is also reduced, thereby preventing correct reproduction. Further, accompanying the decrease or increase of the gain of the gain control amplifier are noise components peculiar to the sound components included in the channel which are also decreased or increased. These noise components are reproduced at the same time from the loudspeaker to give a listener an unrealistic sense.

SUMMARY OF THE INVENTION The above and other disadvantages are overcome by the present invention of a multisignal decoding apparatus comprising, a first channel for transmitting a first composite signal containing a first dominant signal and at least one subdominant signal, a second channel for transmitting a second composite signal containing a second dominant signal and at least one subdominant signal which is of the same kind and in phase opposition to the subdominant signal in the first composite signal, a third channel for transmitting a third composite signal containing a third dominant signal and at least one subdominant signal, and a fourth channel for transmitting a fourth composite signal containing a fourth dominant signal and at least one subdominant signal which is of the same kind and in phase opposition to the subdominant signal in the third composite signal. Circuit means are provided for producing a first and a second control signal by comparison with at least two of the first through four composite signals. The first control signal is supplied to a first variable mixing means connected between the first and second channels for mixing the first and second composite signals with each other so as to cancel the out of phase subdominant signals contained thereinQThe second control signal is supplied to a second variable mixing means connected between the third and fourth channels for mixing the third and fourth composite signals with each other so as to cancel the out of phase subdominant signals contained therein.

Accordingly, an object of the invention is to provide a multisignal transmission apparatus in which separation between channels is improved.

Another object of the invention is to provide a multisignal transmission apparatus in which a crosstalk signal can be cancelled without losing a main signal.

Another object of the invention is to provide a multisignal transmission apparatus which can eliminate a crosstalk signal between a pair of channels with a variable mixing circuit of simple construction.

A further object of the invention is to provide a multisignal transmission apparatus having a novel crosstalk canceller which applies encoded composite signals of different types to a matrix circuit for decoding the respective composite signals to thereby improve separation between channels.

Yet another object of the invention is to provide a multisignal transmission apparatus which cancels a crosstalk between channels without varying the gain of a channel transmitting a signal and consequently keeps a noise signal peculiar to sound elements substantially constant.

The foregoing and other objects, features and advantages of the invention will be more readily understood upon consideration of the following detailed description of certain preferred embodiments, taken in conjunction with the accompanying drawings.

BRIEF DESCRIPTION OF 'THE DRAWING FIG. I is a schematic diagram illustrating an encoder for a better understanding of the invention;

FIG. 2 is a schematic diagram of one embodimentof a decoder in accordance with the invention;

FIG. 3 is a schematic diagram showing a crosstalk signal detecting circuit to be employed in the apparatus illustrated in FIG. 2;

FIG. 4 is a circuit diagram showing a mixing circuit to be employed in the apparatus illustrated in FIG. 2;

FIG. 5 is a schematic diagram of another encoder for explaining the invention;

FIGS. 6 and 7 are phasor diagrams for use in explaining the operation and advantages of the encoder shown in FIG. 5;

FIG. 8 is a schematic diagram of a second embodiment of a decoder used in connection with the invention;

FIG. 9 is a schematic diagram of another crosstalk detecting circuit to be employed in the decodershown in FIG. 8; I

FIG. 10 is a schematic diagram of a third embodiment of a decoder used in connection with the invention, and

FIG. 11 is a schematic diagram of another crosstalk detecting circuit to be employed in the decoder shown in FIG. 10.

DESCRIPTION OF THE PREFERRED EMBODIMENTS For a better understanding of the invention an encoder will be described which encodes four original sound signals into two composite signals. An encoder, as illustrated in FIG. 1, has four

input terminals

10, 12, 14 and 16 to which four original sound signals L,, L,, R, and R,, depicted as in-phase signals of equal amplitude in the figure, are respectively applied. The total L, signal applied to the input terminal is added in a summing junction 18 to 0.707 of the R signal applied to the input terminal 14. The output of the summing junction 18 is applied to a I -network (all-pass phaseshifting network) 20 which provides a phase-shift of I 90 (where I is a function of frequency). The full R, signal applied to the input terminal 16 is added in a summing network 22 to 0.707 of the L, signal appearing at the input terminal 12 after it is passed through a phase inverter 12a. The output of the summing network 22 is passed through a I -network 24, which also provides a phase-shift of I 90. The phase-reversed L,, signal from the inverter 12a and the R signal are also applied to respective I -

networks

26 and 28 which each provide a phase-shift of I.

The full signal appearing at the output side of the I -network 20 is added in a summing circuit 30 to 0.707 of the signal appearing at the output side of the network 26 to produce at its output terminal 32 a composite signal designated as L Similarly, the full signal from network 24 is added in a summing circuit 34 to 0.707 of the signal from the network 28 to produce at its output terminal 36 a composite signal designated at R,. It will be observed that a phasor group 38 of the composite signal L consists of the L, signal, a 0.707R signal in phase with respect to the L, signal and a 0.707L,, signal lagging the L, signal by 90. The phasor group of the composite signal R consists of the R, signal, a 0.707R signal leading the R, signal by 90 and a 0.707L,, signal leading the 0.707R, signal by 90.

Further, it will be understood that the composite signals L and R may be transmitted by an FM multiplex radio, or they may be recorded on any two-channel medium such as a two-track tape or stereophonic record for later reproduction.

The two composite signals L, and R, which are encoded by the encoder illustrated in FIG. 1 can be decoded by a decoder illustrated in FIG. 2 to four output signals containing dominant signals corresponding to the respective original signals. The composite signals L, and R represented by the

phasor groups

38 and 40 are applied to

input terminals

50 and 52, respectively, of the decoder shown in FIG. 2. From the

terminals

50 and 52 the signals are applied to respective I -networks 54 and 56. In this manner, the composite signal L passes without relative phase-shift through the network 54 and is thereafter applied to an amplifier 58. The other composite signal R passes with a relative phaseshift of 90 through the network 56 and is next applied to an amplifier 60. A 0.707 portion of the signal appearing at the output side of the network 54 is added in a summing junction 60 to -0.707 of the signal appearing at the output side of the network 56, the summed signal thus formed being applied to an amplitier 62. A 0.707 portion of the signal appearing at the output side of the network 54 is also added in a summing junction 64 to 0.707 of the output signal of the network 56, and the summed signal thus formed is applied to an amplifier 66.

Phasor groups

68, 70, 72 and 74 indicate the output of the

amplifiers

58, 62, 66 and 60, respectively. At this point the phasor group 68 consists of the L, signal, a 0.707 R, signal in phase with the L, signal and a 0.707 L,, signal lagging with respect to the R, and 0.707 R, signals by 90. The phasor group 70 consists of the L,, signal, a 0.707 L, signal leading with respect to the L,, signal by 90 and a 0.707 R, signal leading with respect to the 0.707 L, signal by 90. The phasor group 72 consists of the R signal, a 0.707 L, signal in phase with the R signal and a 0.707 R, signal lagging with respect to the R and 0.707 L, signals by 90. The phasor group 74 consists of the R, signal, a 0.707 R, signal leading with respect to the R, signal by 90 and a 0.707 L,, signal leading with respect to the 0.707 R, signal by 90. It should be noted that the 0.707 L,, signal of the phasor group 68 and the 0.707 L,, signal of the phasor group 74 are out of phase with each other, and also that the 0.707 R, signal of the phasor group 70 and the 0.707 R, signal of the phasor group 72 are out of phase with each other.

The output signals of the

amplifiers

60 and 66 are supplied to phase-inverters and 82, respectively, so that the polarities of the output signals of the phaseinverters 80 and 82 are reversed as shown by phasor groups 84 and 86. It will be noted that the 0.707 R, signal of the phasor group 68 and the 0.707 R, signal of the phasor group 84 are out of phase with each other,-and also that the 0.707 L, signal of the phasor group 70 and the 0.707 L, signal of the phasor group 86 are out of phase with each other. The output signals of the

amplifiers

58 and 62 and the output signals of the phase-inverters 82 and 80 are respectively applied to I -

networks

88, 90, 92 and 94. The P-

networks

88 and 92 provide a phase-shift of I while the P-networks 90 and 94 provide a phase-shift of I 0.

Output signals appearing at the output sides of I -

networks

88 and 90 are transmitted to

output terminals

112 and 114 through phase-inverters 96 and 98, respectively, and output signals appearing at the output sides of I -networks 92 and 94 are directly transmitted to

output terminals

116 and 118, respectively. Phasor groups of the output signals derived at the respective output terminals 112-118 are designated as 104, 106, 108 and 110, respectively. It will be noted that respective dominant signals, namely the L,, L,,, R, and R, signals in the respective phasor groups are in phase with one another.

With the apparatus described above the original sound signals L,, L,,, R, and R contained in the composite signals L, and R appear in the signals derived at the output terminals 112 118 as the dominant signals, respectively. As shown in the phasor group 104, however, the output signal appearing at the output terminal 112 contains undesired subdominant signals L,, and R, which will cause crosstalk. Likewise, the output signal appearing at the output terminal 114 contains undesired subdominant signals L, and R, which will also cause crosstalk.

In the present invention means, such as, for example,

components appear in reverse phase relationship. The mixing circuits are controlled in operation with the crosstalk components applied thereto in order to cancel or reduce the crosstalk components.

Since the 0.707 L signal of the phasor group 68 and the 0.707 L, signal of the phasor group 74 are 180 out of phase with each other as described above, a first crosstalk canceller or a first variable mixing circuit means 120, in the embodiment of FIG. 2, is connected between the output sides of

amplifiers

58 and 60. Similarly a second variable mixing circuit means 122 is connected between the output sides of

amplifiers

62 and 66,. the outputs of which contain the out of phase subdominant signals R,. A third variable mixing circuit means 124 is connected between the output side of the amplifier 58' and the phase-inverter 80, in which the subdominant signals R, are out of phase. A fourth variable mixing circuit 126 is connected between the output sides of the amplifier 62 and the phase-inverter 82 in which the subdominant signals L, are out of phase.

The amount of signal mixing produced by the circuits 120 126 is controlled with signals obtained from a crosstalk signal detecting circuit described hereinafter.

.FIG. 3 shows an example of the crosstalk, signal detecting circuit employed in the apparatus shown in FIG. 2. In the figure, reference numeral 130 generally des'ig nates the crosstalk signal detecting circuit.

Input terminals

132, 134, 136 and 137 of the circuit 130 are respectively connected between the output terminals 132a, 134a, 136a and 137a of the

amplifiers

58, 62, 66 and 60 (referring to FIG. 2) and the all-wave rectifier circuits I38, 140, 142 and 144. The signals represented by the

phasor groups

68, 70, 72 and .74 are then allwave-re'ctified. The output of the rectifier circuit 138, which is shown in the figure as a phasor group 146, includes the 0.707 L signal reversed in phase withrespect to the input signal component. The output of the rectifier circuit 144, which is shown in the figure as a phasor group l48,includes the R, signal reversed in phase with; respect to that applied thereto previously. Similarly, the outputs of the rectifier circuitsl40 and 142, are shown'in the figure as phasor groups 150 and 152, respectively and: the phasor group 150 includes the L, signal in phase with the 0.707 R,signal thereof while the phasor group 152includesthe'0.707 R, signal leading the R signal by 90".

The outputs of the rectifier circuits 138 and 144 are supplied to a subtracting circuit 154 to produce a subtraction signal of L, and R, signals, namely |L,| |R,| 09 because the 0.707 L, and 0.707 R, signals of both the

phasor groups

146 and 148 are in phase with each other and accordingly are cancelled by the subtraction. Similarly, the output of the

rectifier circuits

140 and 142 are supplied to a subtracting circuit 156 to produce a subtraction signal of L and R, signals, namely |L,| lR,,| because the 0.707 R, and 0.707 L, signals of both the phasor groups 150 and 152 are the same in phase and consequently are cancelled by the subtraction. Accordingly, the outputs of the subtracting circuits I54 and 156 can be referred to as absolute comparison outputs between the dominant signals in the front and back channels.

Referring to the output signals of the subtracting circuits I54 and 156 as a and B respectively, the output signal a is supplied to an output terminal 160a through a rectifier 160 and also to an output terminal 164a through a phase-inverter 162 and a rectifier 164. The

' output signal B is delivered to an output terminal 166a through a rectifier 166 and also to an output terminal 170a through a phase-inverter 168 and a rectifier 170. The

output terminals

160a, 164a, 166a and 170a of the crosstalk signal detecting circuit 130 are connected to the

input tenninals

126a, 122a, 120a and 124a of the mixing

circuits

126, 122, 120 and 124 respectively (referring to FIG. 2).

An example of the mixing circuits will be described with reference to FIG. 4 which shows the mixing circuit 120, by way of example. The mixing circuit 120 comprises a transistor Tr, the collector of which is connected to the output side of the amplifier 58 through a resistor R and the emitter of which is connected to the output side of the amplifier through a resistor R The base of the transistor Tr serves as the input terminal a of the mixing circuit 120 to which the control signal is supplied. The other mixing circuits may be constructed similarly. It will be, however, apparent that circuits of other types can be employed as the mixing circuits of this invention.

Examples of the elimination of crosstalk throughthe operation of the decoder shown in FIG. 2 will-now be described. If only the L, signal exists in the composite signal L the L, signal appears at the output terminal 112 as the dominant signal L, but at the sametime appears at the output terminals ll4'and 116 as the subdominant signals or the crosstalk signals respectively. The D.C. component of the L, signal or the control signal or, however, is obtained at the output terminal 160a of the crosstalk signal detecting circuit by the subtractingcircuit 154 and is then supplied from the output terminal a to theinput terminal 126a of the mixing circuit 126 as the control signal. As a result, the transistor Tr of the mixing circuitl26 becomes conductive between its collector and emitter and the respective signals of the phasor groups 70 and 86 are mixed with each other in thernixing circuit 126 to can: tie] the 0.707 L, signals thereof which are out of phase. Accordingly, no L, signals are derived from the output terminals 114, and 1.16. The inverted a signal, -a, which is supplied at the same time by the inverter 162 to the anode of diodel64 is not passed by the diode be! cause it is of the wrongpolarity.

Simultaneously with the above operation, if thecomposite signal R contains the R signal only, the R signal appears at the output terminal 116 as the dominant signal. R and also appears at the

outputterrninals

112 and 118 as the subdominant signals or crosstalk signals. By means of the subtracting circuit 156, however, the D.C. component of the IR, signal, or the B signal is obtained as the control signal. That the ,8 signal is negative is apparent when the phasor group 152 is subtracted from the phasor group 150, taking into consideration that no L, or R, signal components are present. The control signal B is delivered to the output terminal a through the phase-inverter 168 which converts it to +13 so that is passed through the anode to cathode terminals of the rectifier 170 and is then supplied to the input terminal 124a of the mixing circuit v 124. The respective signals of the phasor groups 68 and 84 are mixed with one another in the mixing circuit 124 with the result that the 0.707 R, signals contained therein which are out of phase are cancelled. Accordingly, the 0.707 R, signals which will cause the crosstalk are prevented from being delivered to the output terminals 112 and l18.'The -/3 control obtained from the subtracting circuit 156 is also supplied to the anode of the rectifier 166 at the same time but it is not transmitted to the output terminal 166a of the circuit 130 due to the fact that it is a negative signal component and is therefore prevented from being passed through the anode to cathode terminals of the rectifier 166.

In a similar manner the 0.707 R, signals, which will cause crosstalk, contained in the

phasor groups

70 and 72 are cancelled in the mixing circuit 122 and the 0.707 L signals, which will cause crosstalk, contained in the phasor groups 68 and 74 are cancelled in the mixing circuit 120, so that crosstalk signals are prevented from being delivered to the output terminals whereby separation between the respective channels is easily and positively achieved.

FIG. shows an encoder of modified form which has four

input terminals

210, 212, 214 and 216 to which four original sound signals I..,, L,,, R,, and R, depicted as in phase signals of equal amplitude are respectively applied. The total L signal is added in a summing junction 218 to 0.707 of the R signal, with the output of the summing junction 218 being applied to a phase-shifting network 220 which introduces a reference phase-shift I, which is a function of frequency. The full R, signal at the terminal 216 is added in a summing network 222 to 0.707 of the L signal appearing at the input terminal 212, and the output is passed through a I -network 224, which also provides the reference phase-shift I. The L, and R signals are also applied to respective P-

networks

226 and 228, each of which provides a phase-shift of I 90. The full signal appearing at the output of the network 220 is added in a summing circuit 230 to 0.707 of the signal appearing at the output of the network 226 to produce at its output terminal 232 a composite signal designated L Similarly, the full signal from the network 224 is added in a summing junction 234 to 0.707 of the signal from the network 228, the latter in this case being in the positive sense. The signal appearing at an output terminal 236 of the network 234 is the composite signal designated R The significance of the modifications to the encoder of FIG. 1 to provide the encoder of FIG. 5 (namely, the reversal of the phase of the 0.707 terminals of the two summing circuits in the upper half of the diagram) will be appreciated from an analysis of the phasor relationship of the L and R composite signals portrayed as

phasor groups

238 and 240, respectively. It will be observed that the phasor group 238 consists of the signal L, (which although shown in the same phase relationship as the input signal L, has a I'as-a-function-offrequency angle difference between them), a signal 0.707 R, in a negative sense with respect to its corresponding input phasor, and a 0.707 L signal which lags the phasor 0.707 R, by 90 because of the action of the netowrk 226. The phasor group 240 consists of the original signal R, in the same relative phase position as its corresponding input signal, a signal 0.707 L in phase with the R, signal, and a 0.707 R, signal lagging the 0.707 L signal by 90 due to the action of the I -network 228.

Referring now to FIG. 6, the effect of the panning" is to divide the signal (as by means of two coupled attenuators) between two channel inputs. At the midpoint of the panning operation, the signal becomes precisely divided between the front channels L, and R,, or between the back channels L,, or R,,; this condition will now be examined. The

phasor groups

238 and 240 from FIG. 5 are repeated here as phasor groups 250 and 252, respectively, and the panned center signals have been added. The front center signal C, is placed in the proportion 0.707 C, and in-phase in the phasor groups 250 and 252, appearing as

phasors

254 and 256.

Further, it will be noted that the center-back channel C is divided in the proportion 0.707 in the left back channel and 0.707 in the right back channel, and since these two phasors appear as a 0.707 fraction, the corresponding fraction of the C,, signal which is in phase with the 0.707 L, phasor is 0.5 and the fraction which is in phase with the 0.707 R phasor is 0.5, in both the phasor groups. With this convention in mind it is seen that the two phasors in each group add to the larger phasors 0.707 C, in each of the phasor groups 250 and 252; however, it should also be observed that the phasor 0.707 C,, in the phasor group 250 is out-of-phase with the corresponding phasor in group 252.

Another significant feature of the encoder is illustrated by the

phasor groups

256 and 258 in FIG. 7, the former depicting the situation which results when the phasor groups 250 and 252 of FIG. 6 are added and the latter depicting the situation when the composite signal R (phasor group 252) is subtracted from the signal L (phasor group 250). It will be noted that when the L and R signals are added the phasors L,, L R and R, all have an intensity equal to unity, whereas the front center signal C, is augmented by a factor 1.414, which is exactly what happens when a stereophonic record is played over a monophonic player. The back center signal C is cancelled, however, because of the aforementioned out-of-phase relationship. When the phasor groups are subtracted, the phasors L,, L,,, R,, and R, again all appear with unity amplitude, but this time the center back signal C, is augmented by the factor 1.414 while the center front signal C, is cancelled. The relationship portrayed by the

phasor groups

256 and 258 are extremely important since they indicate that if only a center front signal is present (i.e., no center back signal) the phasor group 256 will be greater than the group 258 and, conversely, if there is only a center back signal but no center front signal, the phasor group 258 will be the larger. This interesting property is used to advantage to enhance the operation of a decoder which will now be described to be utilized with the encoder of FIG. 5.

The decoder, illustrated in FIG. 8, is in many respects similar to the decoder of FIG. 2. The composite signals L and R represented by the

phasor groups

238 and 240 in FIG. 5, respectively, are applied to

respective input terminals

300 and 302, from whence they are applied in parallel to respective pairs of P-

networks

304, 306, 308 and 310. In this manner, each of the signals L and R passes with only I -function of frequency phase-shift through the

networks

304 and 308, respectively, and also passes with I -function of frequency a phase-shift through the

networks

306 and 310.

The outputs of the

networks

304 and 308 are applied directly to the input terminals of

respective amplifiers

313 and 315, the outputs of which are applied to

respective loudspeakers

316 and 318 which are positioned at the front left and front right comers in a listening room M. The signals applied to the

loudspeakers

316 and 318 contain the dominant original signals L, and R respectively, and the two subdominant contaminating" signals 0.707 L, and 0.707 R,,. Equal proportions, namely, 0.707 of the output of the

networks

306 and 308 are summed at a summing junction 320 to produce a composite signal consisting of a dominant signal L,,, which is applied to an amplifier 322 and thence to a loudspeaker 324 which is positioned at the left back corner of the room M. Equal negative portions, namely, 0.707 of the outputs of the

networks

304 and 310 are summed at a second summing network 326 to produce a composite signal composed of a dominant signal R together with 0.707 R, and 0.7071 signal after amplification, and this composite signal is applied to an amplifier 328 whose output is fed to a loudspeaker 330 which is positioned at the right back corner of the room M. It will be observed from the phasor groups 332, 334, 336 and 338 which respectively portray the composite signals appearing at

loudspeakers

316, 324, 330 and 318, that the dominant phasors at all four loudspeakers are in-phase.

In this embodiment, if the composite signals L and R contain the center front signal C,, the 0.707 portion of the center front signal C, appears in phase in the L, and R, signals of the decoded phasor groups 332 and 338, respectively. The total center front signal 1.4 C, of both the phasor groups 332 and 338 can be represented by a phasor 340 in FIG. 8. The 0.5 of the center front signal components C; also appears in both the L, and R, signals of the phasor groups 334 and 336 in phase. Accordingly, the center front signal components C, of the phasor group 334 can be shown in total by a phasor 342 and the center front signal components of the phasor group 336 can be shown in total by a phasor 344. As mentioned above, if a center front sound exists in an original sound field, the center front sound is reproduced from the

loudspeakers

324 and 330 for the back channels in opposite phase relationship, respectively. Similarly, if a center back sound exists in an original sound field, in a reproduced sound field the center back signal C appears in both the phasor gorups 334 and 336 and also in the phasor groups 332 and 338. Similarly, these center back signals in the phasor groups are respectively represented by

phasors

346, 348 and 350 in the figure. Accordingly, if a center back sound exists in an original sound field, in a reproduced sound field a center back sound is respectively repro-- duced, from the

loudspeakers

316 and 318 for the front channels in opposite phase relationship, respectively. For this reason, a listener in the reproduced sound field can not spearate the center front and center back sounds and is subjected to an uncomfortable experience.

It is, however, noted that the center back signals leaked to the front channels andthe front center signals leaked to the back channels are reversed in phase but same in magnitude. Accordingly, with the present in vention the crosstalk signals of reverse phase can be cancelled in the manner similar to that described in connection with the first embodiment. For this purpose, in the embodiment shown in FIG. 8 a mixing circuit 360 is connected between the input sides of the

amplifiers

313 and 315 as a first crosstalk canceller and a second mixing circuit 362 is connected between the input sides of the

amplifiers

322 and 328 as a second crosstalk canceller. The degree of mixing in the

circuits

360 and 362 is changed or controlled as a function of the crosstalk signals. Mixing circuits similar to the circuit 120 described in reference to FIG. 4 may also be used as the variable mixing means of this embodiment.

FIG. 9 shows a circuit generally designated 370, which detects the center front or center back signal (crosstalk signals). Input terminals 372 and 374 of the circuit 370 are respectively connected to terminals 300a and 302a shown in FIG. 8, so that the input signals L and R applied to the

input terminals

300 and 302 are also applied to the input terminals 372 and 374 and then are applied to a summing; junction 376 and to a subtracting junction 378 respectively. The sum signal appearing at the output of the summing junction 376 is the sum of L and R and is depicted by the phasor group 256 in FIG. 7. The output of the subtracting junction 378 is a composite signal having the properties portrayed by the phasor gorup 258 in FIG. 7. It will be evident from reconsideration of FIG. 7 that if the front center signal dominates, the output of the summing circuit 376 will exceed the output of the subtraction circuit 378, and, conversely, if the back center signal dominates, the output of the subtracting junction will exceed the output of the summing junction.

In order to obtain the relative ratio between these two quantities, the outputs of the

junctions

376 and 378 are amplified in respective

quasi-logarithmic amplifiers

380 and 382, respectively, and are then rectified by

respective rectifiers

384 and 386, which are preferably full-wave rectifiers. The rectified signals from

rectifiers

384 and 386 are integrated by

leaky integrators

388 and 390, respectively. The output signals of the integrators 388 and'390 are applied to a subtracting circuit 392 the output of which is applied to an output terminal 394a through a rectifier 394 and also to an output terminal 398a through a phase-inverter 396 and arectifier 398. The

output terminals

394a and 398a of the circuit 370 are connected to control terminals 362a and 360a of the mixing

circuits

362 and 360, respectively.

If the two composite signals L and R contain a center front signal, the center front signal is detected by the summing junction 376 and is then applied to the subtracting circuit 392 through the logarithmic amplifier 380 and the rectifier 384. Meanwhile, since the output of the subtracting junction 378 contains no cen ter front signal, a positive output appears at theoutput terminal of the subtracting circuit 392 as a control signal, which is then applied to the mixing circuit 362 through the rectifier 394. As a result, the outputs of the summing

junctions

320 and 326 are mixed in the mixing circuit 362 to cancel the out of phase center front signal components C,. The output of the subtracting circuit 392 is also applied to the phase-inverter 396 and is thereby made into a negative control signal which is blocked at the anode terminal of rectifier 398 so that no control signal is delivered to the output terminal 398a. Accordingly, the mixing circuit 360 does not operate and hence thecenter front signal is reproduced from the

front loudspeakers

316 and 318.

Conversely, if the two composite signals L and R include the center back signal'res pectively, a negative control signal is derived from the output terminal of the subtracting circuit 392, since the output of the subtracting junction 378 becomes greater than that of the subtracting junction 376. The negative control signal is inverted to a positive control signal by I the phaseinverter 396 and is then applied to the control terminal 360a of the mixing circuit 360 through the rectifier 398. Accordingly, the outputs of the

networks

304 and 308 are mixed in the mixing circuit 360 and the out of phase center back signal components C,, are cancelled. The negative control signal from the subtracting circuit 392 is blocked from delivery to the output terminal 394a due to the function of the rectifier 394.

As is apparent from the foregoing description, with the invention it is possible to cancel crosstalk caused by the center front or center back signal by means of the mixing circuits. The above described circuit 370 may also be applied to the decoder shown in FIG. 2 with the same effect. Further, it is also possible to connect the input terminals 372 and 374 to the output terminals of the

networks

304 and 308 for the same purpose. It should be clear that if the subdominant signals of the same kind are reversed in phase in the circuit 130 (referring to FIG. 3) then the subtracted output becomes substantially zero, since they pass through full-wave rectifiers and are subtracted from each other.

FIG. shows another decoder according to the invention. The decoder of FIG. 10 has a pair of

input terminals

402 and 404 which are supplied with the composite signals L, and R The composite signals L and R, are produced by an encoder (not shown) which is known as a Regular Matrix Encoder" in Japan. The composite signal L, consists of an L, signal, a 0.4 R, signal in phase with the L, signal, an L, signal leading by 90 with respect to the L, signal and a 0.4 R,, signal in phase with the L, signal, while the composite signal R, consists of an R, signal, a 0.4 L, signal in phase with the R, signal, an R, signal lagging with respect to the R, signal by 90 and a 0.4 L, signal in phase with the R signal.

A basic matrix circuit 406 is connected to the

input terminals

402 and 404 and provides four

output terminals

408, 410, 412 and 414 for delivering four output signals therefrom, shown in the figure as

phasor groups

420, 422, 424 and 426, respectively. The output from terminal 408 is transmitted through a I 90 -network 440 to an output terminal 460. The output from terminal 410 is transmitted through a phase-inverter 442 and a I +0 -network 444 to an output terminal 462. The output from terminal 412 is transmitted through a phase-inverter 446 and a IN-0 -network 448 to an output terminal 464 and the output from terminal 414 is transmitted through a I 90 -network 450 to an output terminal 466.

The phasor group 420 consists of the L, signal, a 0.707 R, signal in phase with the L, signal and a 0.707 L signal leading with respect to the L, signal by 90. The phasor group 422 consists of an L, signal, a 0.707 R, signal in phase with the L, signal and a 0.707 L, signal lagging with respect to the L, signal by 90. The phasor group 424 consists of an R, signal, a 0.707 L signal in phase with the R signal and a 0.707 R, signal leading with respect to the R signal by 90. The phasor group 426 consists of an R, signal, a 0.707 L, signal in phase with the R, signal and a 0.707 R, signal lagging with respect to the R, signal by 90. In this case, the L,, L,,, R and R, signals in the respective phasor groups 420 426 are dominant signals, and the other signals are subdominant signals which produce crosstalk. However, it will be noted that the signal components of R, in the

phasor groups

420 and 424 are of the same amplitude but are out of phase with each other. Similarly the signal components of R in the

phasor groups

422 and 426 are of the same amplitude but are out of phase with each other.

The output signals from the

output terminals

410 and 412 are inverted in phase by the phase-

inverters

442 and 446, respectively, and are shown in the figure as

phasor groups

454 and 456, respectively. As seen from the

phasor groups

454 and 456, the signal components of L, in the

phasor groups

420 and 456 are the same in amplitude but are out of phase, and similarly, the signal components of L,in the

phasor groups

426 and 454 are the same in amplitude but are out of phase. The P-

networks

440, 444, 448 and 450 act to make the dominant signals, contained in the output signals delivered to the

output terminals

460, 462, 464 and 466, the same in phase. The outputs obtained at the output terminals 460 466 are shown in the figure as phasor groups 470 476, respectively.

In this embodiment, in order to cancel the oppositely phased crosstalk signals variable mixing circuits 480 486 are employed as crosstalk cancellers. The mixing circuit 480 is inserted between the

output terminals

408 and 412 of the basic matrix circuit 406, the mixing circuit 482 is inserted between the

output terminals

410 and 414, the mixing circuit 484 is inserted between the output terminal 408 and the output side of the phase-inverter 446, and the mixing circuit 486 is inserted between the output side of the phase-inverter 442 and the output terminal 414. Circuits similar to that described in reference to FIG. 4 may be used as the mixing circuits of this example.

FIG. 11 shows a circuit diagram of a crosstalk signal detecting circuit 488. The circuit 488 provides four

input terminals

490, 492, 494 and 496 which are connected to the

output terminals

408, 412, 410 and 414, respectively, of the basic matrix circuit 406. The

input terminals

490 and 492 are connected to full-

wave rectifiers

498 and 500, respectively. The outputs from the

rectifiers

498 and 500 are connected to a first subtracting junction 502. The

input terminals

494 and 496 are connected to full-wave rectifiers 504 and 506, respectively, and the rectifier outputs are connected to a second subtracting junction 508. A first signal representative of the subtraction of the L, and R, signals is produced in the junction 502 and a second signal representative of the subtraction of the L, and R, signals is produced in the junction 508 as first and second control signals, respectively.

The first control signal is then applied through the anode to cathode terminals of a rectifier 512 to a first output terminal 510 which is connected to a control terminal 486a of the mixing circuit 486. The first control signal is also applied through a phase-inverter 516 to the anode of a rectifier 518 whose cathode is connected to a second output terminal 514. The terminal 514 is connected to a control terminal 482a of the mixing circuit 482. The second control signal is applied through a rectifier 522 to a third output terminal 520 which is connected to a control terminal 480a of the mixing circuit 480 and is also applied through a phaseinverter 526 and through the anode to cathode terminals of a rectifier 528 to a fourth output terminal 524 which is connected to a control terminal 484a of the mixing circuit 484.

The circuits shown in FIGS. 10 and 11 operate to cancel the crosstalk signals in a manner similar to the first and second examples already described. The operation of the circuits of FIGS. 10 and 11, however, will now be described by way of example. The R, signal component contained in the composite L, and R,

signals appears at the

output terminals

408 and 412 of the basic matrix circuit 406 as crosstalk signals. Since the output signals of the

output terminals

408 and 412, however, are mixed by the mixing circuit 480, the R, signal components contained therein which are the same in amplitude, but which are out of phase with each other, are cancelled. In this embodiment the mixing circuit 480 is controlled in its mixing condition by the control signal derived from the output terminal 520 of the detecting circuit 488.

In the foregoing description, the detecting

circuits

130 and 488 are supplied with input signals from the output terminals of the decoders, but in other embodiments they may be supplied with input signals obtained from the composite signals which are passed through a matrix circuit which is separate from the decoder.

The terms and expressions which have been employed here are used as terms of description and not of limitation, and there is no intention in the use of such terms and expressions, of excluding equivalents of the features shown and described, or portions thereof, it being recognized that various modifications are possible within the scope of the invention claimed.

What is claimed is: i

l. A multisignal transmission apparatus comprising;

a. means for converting two composite signals into four output signals, each of said output signals containing a separate dominant signal and at least two subdominant signals,

b. four separate transmitting channels for respectively transmitting each of said output signals,

c. means for arranging the phase of at least one subdominant signal in each of said output signals to be one hundred and eighty degrees out-of-phase with respect to the same subdominant signal contained in another of said output signals,

d. means for producing at leastone control signal by comparing the absolute magnitudes of said dominant signals in at least two of said transmitting channels, and

e. variable mixing means connected between the other two of said transmitting channels which con tain said out of phase subdominant signals, said variable mixing means being controlled with said control signal so as to mix said out of phase subdominant signals and thereby substantially cancel them.

2. A multisignal transmission apparatus as recited in claim 1, wherein said converting means comprises allpass phase-shifting networks for giving said two composite signals a relative phase difference of 90, a summing circuit for summing said phase-shifted two com posite signals derived from said all-pass phase-shifting networks, and a subtracting circuit for differencing said phase-shifted two composite signals derived from said all-pass phase-shifting networks.

3. A multisignal transmission apparatus as recited in claim 1, wherein said variable mixing means comprises a variable impedance device to which said control signal is applied and the impedance of which is varied with the control signal.

4. A multisignal transmission apparatus as recited in claim 1, wherein said control signal producing means comprises detector circuit means for receiving a first pair of said output signals which contain the same subdominant signals and a seocnd pair of signals corresponding to the other pair of said output signals which contain the same subdominant signals and comparing circuit means for separately comparing said first pair of signals and for comparing said second pair of signals, respectively, and to produce separate control signals representative of whether said pairs of signals contain subdominant signals which are in phase coincidence or in phase opposition.

5. A multisignal transmission apparatus as recited in claim 4, wherein said detector circuit means includes four input terminals for separately receiving said four output signals, each of said input terminals being connected to a separate full-wave rectifier circuit, and said comparing circuit includes a pair of subtracting circuits for subtracting the outputs from selected pairs of said full-wave rectifier circuits.

6. A multisignal transmission apparatus adapted to receive first and second composite signals L and R respectively containing dominant signals L, and R; in phase with each other and each including at least two subdominant signal components L and R,,, the apparatus comprising;

a. circuit means including all-pass phase-shifting networks and combining networks for deriving first, second, third and fourth output signals containing in phase dominant signal components L,, R,, L,, and R respectively, said first and fourth output signals also containing subdominant L, and R, signals with said L' subdominant signal in each one of said first and fourth output signals being; in phase opposition to said R subdominant signal in the other of said first and fourth output signals, said second and third output signals also containing subdominant L, andR, signals with said L, subdominant signal in each one of said second and third output signals being in phase opposition to said R, subdominant signal in the other of said second and third output signals,

b. first, second, third and fourth channels for transmitting said respective first through fourth output signals,

c. first and second variable mixing means connected between said first and fourth channels and between said second and third channels, respectively, for mixing said first and fourth and second and third output signals, respectively,

d. control circuit means for comparing the sum and difference of said L and R composite signals to produce a first control signal when the sum of said L and R composite signals exceeds the difference of said L and R composite signals and a second control signal if the difference of said L and R composite signals exceeds the sum of said L and R composite signals, and

e. means for supplying said first control signal to said first variable mixing means and. said second control signal to said second variable mixing means so as to control the mixing conditions thereof in response to said first and second control signals, respectively.

7. A multisignal transmission apparatus comprising;

a. first, second, third and fourth. channels for transmitting, respectively, first, second, third and fourth composite signals each of which contains separate dominant signals a, b, c and d, respectively, said first composite signal also containing two subdominant signal components corresponding to portions of said c and d signals, said second composite signal also containing two subdominant signal components corresponding to portions of said c and d signals, said third composite signal also containing two subdominant signal components corresponding to portions of said a and b signals, and said fourth composite signal also containing two subdominant signal components corresponding to portions of said a and b signals,

b. means for arranging the phase of said c and d subdominant signal components in said first composite signal so as to be in phase opposition with respect to said c and d subdominant signal components, respectively, contained in said second composite signal,

means for arranging the phase of said a and b subdominant signal components in said third composite signal so as to be in phase opposition with respect to said a and b subdominant signal components contained in said fourth composite signal,

(1. first and second variable mixing networks connected between said first and second channels,

e; third and fourth variable mixing networks connected between said third and fourth channels,

f. comparing control means including means for comparing the magnitudes of said dominant a and b signals and for producing a first control signal when IaI-lbl 0 and a second control signal when lal Ibl 0; means for comparing the magnitudes of said dominant c and d signals and for producing a third control signal when Icl |d| 0 and a fourth control signal when lcl |d| 0; means for applying said first and second control signals to said third and fourth variable mixing networks, respectively, to thereby cancel said out of phase 0 and d subdominant signal components, respectively, and means for applying said third and fourth control signals to said first and second variable mixing means, respectively, to thereby cancel said out of phase a and b subdominant components, respectively.

8. A multisignal transmission apparatus as recited in claim 6 wherein said control circuit means further comprises means for summing said L and R composite signals, means for subtracting said L and R composite signals, means for rectifying the outputs of said summing and subtracting means, and means for subtracting the rectified outputs to produce said first and second control signals.

9. A multisignal decoding apparatus comprising, first, second, third and fourth channels for transmitting first, second, third and fourth signals, respectively, as dominant signals, said first and fourth channels also containing oppositely phased subdominant signals corresponding to said second signal, said second and third channels also containing oppositely phased subdominant signals corresponding to said fourth signal, first comparing circuit means for producing at least a first control signal by comparisoon of the absolute magnitudes of at least said first and fourth dominant signals, first variable mixing means connected between said second and third channels for mixing said signals contained therein with each other so as to cancel said oppositely phased subdominant signals corresponding to at least one of said first and fourth dominant signals in response to said first control signal, second comparing circuit means for producing at least a second control signal by comparison of the absolute magnitudes of said second and third signals, and second variable mixing means connected between said first and fourth channels for mixing said signals contained therein with each other so as to cancel said oppositely phased subdominant signals corresponding to at least one of said second and third dominant signals in response to said second control signal.

Claims

1. A multisignal transmission apparatus comprising; a. means for converting two composite signals into four output signals, each of said output signals containing a separate dominant signal and at least two subdominant signals, b. four separate transmitting channels for respectively transmitting each of said output signals, c. means for arranging the phase of at least one subdominant signal in each of said output signals to be one hundred and eighty degrees out-of-phase with respect to the same subdominant signal contained in another of said output signals, d. means for producing at least one control signal by comparing the absolute magnitudes of said dominant signals in at least two of said transmitting channels, and e. variable mixing means connected between the other two of said transmitting channels which contain said out of phase subdominant signals, said variable mixing means being controlled with said control signal so as to mix said out of phase subdominant signals and thereby substantially cancel them.

2. A multisignal transmission apparatus as recited in claim 1, wherein said converting means comprises all-pass phase-shifting networks for giving said two composite signals a relative phase difference of 90*, a summing circuit for summing said phase-shifted two composite signals derived from said all-pass phase-shifting networks, and a subtracting circuit for differencing said phase-shifted two composite signals derived from said all-pass phase-shifting networks.

4. A multisignal transmission apparatus as recited in claim 1, wherein said control signal producing means comprises detector circuit means for receiving a first pair of said output signals which contain the same subdominant signals and a second pair of signals corresponding to the other pair of said output signals which contain the same subdominant signals and comparing circuit means for separately comparing said first pair of signals and for comparing said second pair of signals, respectively, and to produce separate control signals representative of whether said pairs of signals contain subdominant signals which are in phase coincidence or in phase opposition.

6. A multisignal transmission apparatus adapted to receive first and second composite signals LT and RT respectively containing dominant signals Lf and Rf in phase with each other and each including at least two subdominant signal components Lb and Rb, the apparatus comprising; a. circuit means including all-pass phase-shifting networks and combining networks for deriving first, second, third and fourth output signals containing in phase dominant signal components Lf, Rf, Lb and Rb, respectively, said first and fourth output signals also containing subdominant Lb and Rb signals with said Lb subdominant signal in each one of said first and fourth output signals being in phase opposition to said Rb subdominant signal in the other of said first and fourth output signals, said second and third output signals also containing subdominant Lf and Rf signals with said Lf subdominant signal in each one of said second and third output signals being in phase opposition to said Rf subdominant signal in the other of said second and third output signals, b. first, second, third and fourth channels for transmitting said respective first through fourth output signals, c. first and second variable mixing means connected between said first and fourth channels and between said second and third channels, respectively, for mixing said first and fourth and second and third output signals, respectively, d. control circuit means for comparing the sum and difference of said LT and RT composite signals to produce a first control signal when the sum of said LT and RT composite signals exceeds the difference of said LT and RT composite signals and a second control signal if the difference of said LT and RT composite signals exceeds the sum of said LT and RT composite signals, and e. means for supplying said first control signal to said first variable mixing means and said second control signal to said second variable mixing means so as to control the mixing conditions thereof in response to said first and second control signals, respectively.

7. A multisignal transmission apparatus comprising; a. first, second, third and fourth channels for transmitting, respectively, first, second, third and fourth composite signals each of which contains separate dominant signals a, b, c and d, respectively, said first composite signal also containing two subdominant signal components corresponding to portions of said c and d signals, said second composite signal also containing two subdominant signal components corresponding to portions of said c and d signals, said third composite signal also containing two subdominant signal components corresponding to portions of said a and b signals, and said fourth composite signal also containing two subdominant signal components corresponding to portions of said a and b signals, b. means for arranging the phase of said c and d subdominant signal components in said first composite signal so as to be in phase opposition with respect to said c and d subdominant signal components, respectively, contained in said second composite signal, c. means for arranging the phase of said a and b subdominant signal components in said third composite signal so as to be in phase opposition with respect to said a and b subdominant signal components contained in said fourth composite signal, d. first and second variable mixing networks connected between said first and second channels, e. third and fourth variable mixing networks connected between said third and fourth channels, f. comparing control means including means for comparing the magnitudes of said dominant a and b signals and for producing a first control signal when a - b >0 and a second control signal when a - b <0; means for comparing the magnitudes of said dominant c and d signals and for producing a third control signal when c - d >0 and a fourth control signal when c - d <0; means for applying said first and second control signals to said third and fourth variable mixing networks, respectively, to thereby cancel said out of phase c and d subdominant signal components, respectively, and means for applying said third and fourth control signals to said first and second variable mixing means, respectively, to thereby cancel said out of phase a and b subdominant components, respectively.

8. A multisignal transmission apparatus as recited in claim 6 wherein said control circuit means further comprises means for summing said LT and RT composite signals, means for subtracting said LT and RT composite signals, means for rectifying the outputs of said summing And subtracting means, and means for subtracting the rectified outputs to produce said first and second control signals.

9. A multisignal decoding apparatus comprising, first, second, third and fourth channels for transmitting first, second, third and fourth signals, respectively, as dominant signals, said first and fourth channels also containing oppositely phased subdominant signals corresponding to said second signal, said second and third channels also containing oppositely phased subdominant signals corresponding to said fourth signal, first comparing circuit means for producing at least a first control signal by comparison of the absolute magnitudes of at least said first and fourth dominant signals, first variable mixing means connected between said second and third channels for mixing said signals contained therein with each other so as to cancel said oppositely phased subdominant signals corresponding to at least one of said first and fourth dominant signals in response to said first control signal, second comparing circuit means for producing at least a second control signal by comparison of the absolute magnitudes of said second and third signals, and second variable mixing means connected between said first and fourth channels for mixing said signals contained therein with each other so as to cancel said oppositely phased subdominant signals corresponding to at least one of said second and third dominant signals in response to said second control signal.