Two-channel, Four-component Stereophonic System

Abstract

A two-channel stereo signal including left and right channel responses (L and R, respectively), deriving from only two amplifiers, is fed to four speakers located approximately at the corners of a quadriliateral area, and are positioned to face toward the interior of the area. The network drives the speakers with four different signals so that sound from the speakers appears to a listener positioned interiorly of the area to come from the four sides of the area, rather than from the four corners. In one embodiment, speakers in the left and right front corners respond to (L + R/2 and (R + (L/2), respectively, while in a second embodiment these speakers respectively respond to (L) and (R), respectively. Speakers in the left and right rear corners (directed toward the listener's back) respond to signals proportional to L - (R/2) and R - (L/2), respectively.

Parent Case Text

This application is a continuation-in-part of my application for U.S. Pat. filed May 6, 1970, Ser. No. 35,018, entitled Two-Channel, Four-Component Stereophonic System.

Description

BACKGROUND OF THE INVENTION

In my copending application for United States patent, filed May 6, 1970, supra, signals received by left and right microphones are assumed to contain front information, i.e., information common to the two microphones, and back information which may be taken to represent hall ambience, or reflections from the back wall directly or via side walls. The latter are of random phase, and therefore include cophasal and contraphasal components. I use a single rear speaker which responds differentially to the right and left signals, so that front information cancels completely, right-left information cancels insofar as it is cophasal, but much of the ambience signal combines additively.

SUMMARY OF THE INVENTION

Left and right front speakers respond without any change of signals of a two amplifier stereo amplifier system. To this extent the system is conventional. Two rear speakers are added to the conventional system. These are located left and right of a listener, and are driven on the left by the left channel signal minus half the right channel signal, and on the right by the right channel signal minus half the left channel signal. Thereby the common cophasal components of the two channels subtract in both speakers, by 6db, but the ambience representing components of the signal add, insofar as they are non-cophasal in both speakers.

BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 is a circuit diagram of one embodiment of a reproduction system in accordance with the present invention, wherein negative blending of signals coupled to rear speakers is provided;

FIG. 2 is a circuit diagram of a reproduction system in accordance with the present invention wherein positive blending of signals fed to front speakers and negative blending of signals fed to rear speakers is provided;

FIG. 3 is a circuit diagram of a modification of embodiment of FIG. 1, employing a fifth speaker;

FIG. 4 is a circuit diagram illustrating a fifth speaker located remotely of the four speakers illustrated in FIGS. 1 and 2;

FIG. 5 is a circuit diagram of a modification of the system of FIG. 1, employing autotransformers for intercomplying speakers;

FIG. 6 is a circuit diagram of embodiment of the invention utilizing autotransformers for blending signals supplied only to rear speakers;

FIG. 7 is a plan view of a hall arranged for recording signals capable of reproduction according to the systems of FIGS. 1-6, incl.;

FIG. 8 is a circuit diagram of microphone circuitry employed in the hall illustrated in FIG. 7.

DETAILED DESCRIPTION OF THE DRAWING

Reference is now made to FIG. 1 of the drawing wherein there is illustrated a two-channel stereophonic source for deriving output signals respectively denominated as L and R, for the left and right channels, respectively, of the stereophonic signal. Stereophonic source 11 can be a standard stereophonic signal, of the type derived from conventionally recorded stereophonic sources, such as disc records or magnetic tapes, or from a stereophonic f.m. multiplex receiver.

In the embodiment illustrated in FIG. 1, the right and left channels derived from source 11 are applied to input terminals of conventional stereo amplifiers 12 and 13, each of which includes a signal output terminal referenced to ground.

The right and left channel output signals of amplifiers 12 and 13 are fed to speakers 21, 22, 23 and 24, positioned approximately in the corners of a quadrilateral listening area 25, which generally is defined by the walls of a room. Speakers 21 and 22 are at the front of the listening area so that a listener in the interior of area 25 is generally positioned so that he is facing them, while speakers 23 and 24 are located in the rear of listening area 25 so that they are facing toward the back of the listener. Speakers 21-24 preferably have matched impedance characteristics, although this is not a necessary requirement, as seen infra, as long as front speakers 21 and 22 and rear speakers 23 and 24 are matched in pairs.

In accordance with the embodiment of FIG. 1, speakers 21 and 22 are responsive only to the output signals of amplifiers 12 and 13, respectively. The left and right channel signals are differentially combined at speakers 23, 24, with -6 db of crossover, by feeding the output signals of amplifiers 12 and 13 to one input terminal of each of speakers 23 and 24, respectively, and by connecting the other terminal of the speakers to ground through T network 28 and impedance 29. To achieve -6 db crossover, impedance 29 has a value corresponding with the impedance of either of matched rear speakers 23 and 24. To this end, impedance 29 can be a resistor if speakers 23 and 24 have a substantially flat impedance versus frequency response over the entire audio spectrum, or the impedance can be a circuit simulating and matching the impedance characteristics of speakers 23 and 24. In the alternative, impedance 29 is a further speaker having impedance characteristics matched with those of speakers 23 and 24.

Crossover can differ from - 6 db if the relative crossover between the two channels can theoretically vary so that the responses for speakers 23 and 24 are respectively represented as (L - KR) and (R - KL), where K is a constant between zero and one. In practical situations the value of K will be between 0.25 and 0.75; for -6 db of crossover K = 0.5. The negative crossovers from the right channel into the left rear speaker 23 and from the left channel into right speaker 24 provide a differential effect in speakers 23 and 24 relative to the left and right channels to simulate the reflective or reverberative characteristics of walls of a concert hall wherein the stereophonic signal was originally recorded. In addition, the predominant signals derived from speakers 23 and 24 are from the left and right sides of the concert hall. The amplitude of signals derived from the rear speakers 23 and 24 is less than the amplitude of the signals derived from speakers 21 and 22. This is desirable because the listener is usually positioned towards the rear of listening area 25, in greater proximity to rear speakers 23 and 24 than to front speakers 21 and 22. The ratio between the responses of the rear and front speakers can be relatively varied, if desired, by controlling the relative value of resistors in network 28.

T network 28 includes three variable resistors 31, 32 and 33 which are provided to control the amplitudes of sounds from front speakers 21 and 22 relative to rear speakers 23 and 24 for a listener positioned towards the rear of the listening area 25. In addition, network 28 enables the same amount of negative cross coupling between the left and right channels and speakers 23 and 24. Resistors 31-33 are ganged together so that the values thereof are varied together and are always the same regardless of the direction in which a controller for the values thereof is moved. Resistors 31-33 are increased to attenuate the rear speakers in order for the listener to be able to sit closer to the rear. Without attenuation, and with four loud-speakers of equal efficiency, the listening position is close to the center of the room. If impedance 29 has a value different from the impedance of speakers 23 and 24, the value of resistor 33 is always maintained equal to the value of resistor 31 or 32 times the impedance ratio of impedance 29 to speaker 23 or 24.

To balance the responses of speakers 21-24 so that the responses of left speakers 21 and 23 are the same as the responses of right speakers 22 and 24, the same signal, a monophonic signal, is fed to the input terminals of amplifiers 12 and 13. To test if the speakers are balanced, normally closed, spring biased switch 27 is open circuited and a listener determines if he is able to hear sound from any of the speakers. If speakers 21-24 are balanced, no sound is derived from them because the same signal is applied to each of them and they are connected in a differential pair.

If, in certain instances, it is desired to disable rear speakers 23 and 24 and provide a stereophonic signal only from front speakers 21 and 22, this can be accomplished by opening switches 35 and 26.

To enable sounds from the front speakers to be derived with greater directional characteristics so that sounds appear to be derived to a greater degree from the front, as well as left and right sides, the system can be modified by positively blending the left and right channels with +6 db of crossover, as illustrated in FIG. 2. In FIG. 2, the right and left stereophonic channels derived from source 11, prior to being fed to amplifiers 12 and 13 are fed to a blending network 37 including resistors 38, 39 and 40. Resistors 38 and 39 are respectively connected to the left and right output channels of stereophonic source 11 and the input terminals of amplifiers 12 and 13 while resistor 40 bridges the input terminals of the amplifiers. Each of resistors 38-40 has the same value to provide 6 db of positive crossover between the left and right channels, whereby the output signals of amplifiers 12 and 13 can be respectively represented as (L + (R/2)) and (R + (L/2)). If the amount of positive crossblending amongst front speakers 21 and 22 is desired to be other than 6 db, the values of resistors 38-39 are changed accordingly. The output signals of amplifiers 12 and 13 are applied directly to the terminals of speakers 21-24 in the same manner as illustrated and described with reference to FIG. 1. In the circuit of FIG. 2, however, to obtain 6 db of crossover between thr rear speakers, whereby the responses of the left and right rear speakers are respectively represented as, (L - (R/2)) and (R - (L/2)), the value of impedance 29 is selected to be four times the impedance of each of speakers 23 and 24, whereby the value of resistor 33 is always four times as great as the value of resistor 31 or 32. With the value of impedance 29 four times greater than the impedance of each of speakers 23 and 24, the average response of rear speakers 23 and 24 is one-third that of front speakers 21 and 22.

In the system of FIG. 2, the same degree of directivity is derived from rear speakers 23 and 24 as in the embodiment of FIG. 1 because the right and left signals are fed to the speakers with the same relative amplitudes as in the FIG. 1 system. The responses from the front speakers 21 and 22 in the FIG. 2 system, however, provide enhanced directivity. This is because the left and right channel responses are additively combined in speakers 21 and 22.

Reference is now made to FIG. 3 of the drawings wherein five speakers are interconnected, the fifth speaker 41 performing the function of impedance 29, FIGS. 1 and 2. Corner speakers 21-24 are responsive to the output signals of amplifiers 12 and 13, in the same manner as described in conjunction with either FIG. 1 or FIG. 2. The corner speakers are also positioned as described supra, except that the front corner speakers 21 and 22 may be positioned at a slight acute angle with respect to the front of the listening area. The fifth speaker 41 is connected in circuit in exactly the same manner as impedance 29, FIGS. 1 and 2, and is positioned in close proximity with the front of listening area 25 and approximately midway between the left and right sides of the listening area. Speaker 41 responds to the sum of the currents supplied to rear speakers 23 and 24 and thereby derives a response proportional to the sum of the left and right channels, i.e., proportional to L + R. Since the response of speaker 41 is proportional to the sum of the left and right channels, the acoustic output thereof can be considered as similar to the output derived from the front speaker of the diamond array descrived in my aforementioned copending application to simulate the response of a microphone located in the center front of the concert hall.

If a speaker is substituted for impedance 29, it can also be located in a listening area remote from listening area 25, as illustrated in FIG. 4. In such an instance, the four corner speakers can be located in one room, while the fifth speaker 42, FIG. 4, can be located in another room of a dwelling where stereophonic reproduction is not desired. Since speaker 42 is responsive to the sum of the two channels, all of the acoustical information originally in the stereophonic source is reproduced by speaker 42.

In accordance with another embodiment of the invention, +6 db of blending for the front speakers and -6 db of blending for the rear speakers is provided by connecting autotransformers 51 and 52 to the output terminals of left and right channel amplifiers 12 and 13, as illustrated in FIG. 5. Amplifiers 12 and 13 are directly responsive to the right and left channels of the stereophonic signal as derived from source 11 and illustrated in FIG. 1 so that no blending is provided in the input circuits thereof. Autotransformers 51 and 52 are designed to have suitable amplitude versus frequency responses over the audio spectrum of interest and include terminals 48 and 49, respectively, connected to the output terminals of amplifiers 12 and 13. Each of autotransformers 51 and 52 includes three equally spaced taps and an unconnected terminal at the other end of a winding from the terminal connected to amplifiers 12 and 13. Center taps 53 and 54 of transformers 51 and 52 are both grounded so that at taps 55 and 56, midway between terminals 48 and 49 and the center taps, there are respectively derived voltages indicative of (+L/2) and (+R/2). At taps 57 and 58, midway between the center taps and ungrounded terminals 59 and 60 of transformers 51 and 52, there are derived voltages equal to (-L/2) and (-R/2), respectively. To provide 6 db of positive blending, front left speaker 21 is connected between the output terminal of amplifier 12 and tap 58 of transformer 52, while the terminals of speaker 22 are connected between the output terminal of amplifier 13 and tap 57 of transformer 51. The voltages thereby applied across speakers 21 and 22 can therefore be respectively written as [L - (-R/2)] and [R - (-L/2)] , which can be rewritten as (L + (R/2)) and (R + (L/2)). To provide -6 db of crosscoupling for rear speakers 23 and 24 the terminals of speaker 23 are connected between the output terminal of amplifier 12 and tap 56 of transformer 52, while right rear speaker 24 is connected between the output terminal of amplifier 13 and tap 55 of transformer 51. Speakers 23 and 24 respond to the voltages applied across their terminals to derive responses respectively proportional to (L - (R/2)) and (R - (L/2)) to establish the desired amount of negative crosscoupling.

A transformer arrangement can be employed to provide -6 db of respectively, or crosscoupling for the rear speakers and no crosscoupling for the front speakers by employing the circuit configuration illustrated in FIG. 6. In FIG. 6, the output terminals of amplifiers 12 and 13, having input terminals respectively driven only by the left and right channels of source 11, feed front speakers 21 and 22 in parallel with tapped autotransformers 61 and 62. Tapped autotransformers 61 and 62 have the desired amplitude versus frequency characteristics over the audio spectrum to achieve the desired response from the four speakers 21-24. Autotransformers 61 and 62 include center taps 63 and 64, respectively at which are derived voltages L/2 and R/2, respectively. To enable speaker 23 to derive an L - (R/2) output response, its terminals are connected between the output terminal of amplifier 12 and tap 64 of autotransformer 62. The R - L/2 response for speaker 24 is obtained by connecting its terminals between the output terminal of amplifier 13 and tap 63 of autotransformer 61.

It may be preferred to drive the systems of FIGS. 1-6 from specially recorded stereophonic signals. The special recording arrangement considered most desirable involves placing microphones effectively at approximately the four corners of a concert hall, as illustrated in FIG. 7. In FIG. 7, source 71 of acoustic radiation is positioned at the front of the hall and corner microphones 72, 73, 74 and 75 are positioned between source 71 and the rear wall of the hall. Front microphones 72 and 73 are positioned so that their apertures are directed towards the front of the hall in proximity to the left and right edges of source 71. Microphones 74 and 75, which may be located near the rear of the hall, are positioned so their apertures are directed toward the rear of the hall so that they are responsive to ambience and reverberative sounds reflected from the rear wall of the hall.

With microphones 72-75 positioned approximately in four corners of the hall to transduce acoustic radiation in the manner illustrated by FIG. 7, it is possible to define effective responses for the four walls of the hall in terms of linear combinations of the responses of the four microphones. To this end, assume that the responses of microphones 72, 73, 74 and 75 are respectively denominated as A, B, C and D. The front response can be defined as the sum of the acoustic radiation transduced by microphones 72 and 73, (A + B); the left side acoustic energy can be defined as the sum of the acoustic radiation transduced by microphones 72 and 74, (A + C); the right side acoustic radiation can be defined as the sum of the acoustic radiation transduced by microphones 73 and 75, (B + D); and the rear acoustic energy can be defined as the difference between the acoustic energy transduced by microphones 74 and 75, (C - D). The acoustic energy for the rear wall is defined as the difference (C - D) because the rear acoustic energy includes ambience information due, e.g., to reflection, reverberation and audience noise, such as applause, which is of random phase and derived with greater accuracy by a differential combination.

By defining the front, rear, left and right sides in the manner stated, left and right channels can be formed as if the microphones were located in a diamond array as described in the aforementioned copending application. Thereby, the left channel of the stereophonic signal can be defined as the acoustic signals transduced from the sum of the left, front and rear of the hall, while the right channel of the stereophonic signal can be defined as the sum of the front and right sounds minus the rear sounds. By appropriate substitution, the left and right channels can therefore be expressed in terms of responses from microphones 72-75 as Equations (1) and (2), supra.

One system that can be utilized for forming the left and right channel signals is illustrated in FIG. 8. Acoustic signals transduced by microphones 72-75 are fed to amplifiers 76-79, each having an output terminal referenced to ground. The output signals of amplifiers 76-79, which all have the same relative phase as the signals derived by microphones 72-75, are fed to a combining matrix 81. Combining matrix 81 includes eight resistors 82-89, each having the same value. Resistors 82-87 are connected in pairs to form three summing networks that feed input signals to amplifiers 91, 92 and 93. Amplifier 91 responds to the A and B output signals of amplifiers 76 and 77, as coupled through resistors 82 and 83 to derive an (A + B) signal indicative of the acoustic radiation at the front of the hall. Amplifier 92 responds to the output signals of amplifiers 76 and 78, as coupled through resistors 84 and 85, to derive an (A + C) signal indicative of the left side acoustic radiation. Right side acoustic radiation is derived by amplifier 93 by virtue of the connection of its input terminal to the output terminals of amplifiers 77 and 79 through resistors 86 and 87, respectively. To form the back acoustic radiation signal (C- D), the output terminals of amplifiers 78 and 79 are connected through resistors 88 and 89 to ground and feed plus and minus input terminals of differential amplifier 94.

To derive the left channel of the stereophonic signal, the front, left and back output signals of amplifiers 91, 92 and 94 are additively combined in summing network 95 which derives an output signal in accordance with (1) L = 2A + B + 2C-D. The right channel of the stereophonic signal is derived by additively combining in summing network 97 the output signals of amplifiers 91 and 93 and an inverted replica of the output of amplifier 94, as derived by inverter 96. Thereby, summing network 97 derives an output signal commensurate with the sum of the responses derived from amplifiers 91 and 93 subtracted from the response of the signal derived at the output terminal of amplifier 94 to produce an output signal in accordance with (2) R = A + 2B - C + 2D, where

A = sound picked up by a microphone in front left corner;

B = sound picked up by a microphone in right front corner;

C = sound picked up by a microphone in left rear corner; and

D = sound picked up by a microphone in right rear corner.

The right and left channels of the stereophonic signal derived at the output terminals of networks 95 and 97 are applied to a suitable two-channel output device 98. Output device 98 may be a standard two-channel stereophonic device, such as a tape recorder or a stereophonic disc recorder, as well as a stereo f.m. broadcast station.

If a two-channel stereophonic signal having characteristics indicated by Equation (1) is supplied to a reproduction network including both 6 db positive blending for the front speakers and 6 db negative blending for the rear speakers, as illustrated by the reproduction systems of FIGS. 2 and 5, for example, excellent results are provided with regard to separation between the channels. In particular, each corner speaker has a response that is predominant for itself and includes components common to the responses of the two speakers with which it is adjacent, but no components in common with the predominant component of the speaker in the opposite corner. For example, speaker 21 in the system of FIG. 2 derives a predominant response associated with the left front corner microphone of the hall illustrated in FIG. 7, as well as lower amplitude responses indicative of the responses from the right front and left rear microphones of the hall, to the exclusion of components derived from the microphone in the right rear corner of the hall. This is particularly desirable to achieve a directional effect while preserving the feeling of deriving sounds from all sides of the listening area.

Mathematically, it can be shown that if the left and right channels of stereophonic source 11 are as indicated by Equations (1) and (2) the system of FIG. 2 produces output responses at its four speakers in accordance with:

L + (R/2) = (5/2) A + 2B + (3/2) C; speaker 21 (3); R + (L/2) = 2A + (5/2) B + (3/2) D; (4); ker 22

L - (R/2) = (3/2) A + (5/2) C - 2D; speaker 23 (5); and R - (L/2) = (3/2) B - 2D + (5/2) D; (6). ker 24

Analyzing Equations (3) - (6) it is noted that the response for each corner is devoid of the predominant response of the diagonally opposite corner and includes lower amplitude components for the adjacent corners. In addition, it is noted that the responses for the front speakers are greater than those for the rear speakers, the desired results since the listener is generally in closer proximity to the rear speakers than the front speakers.

While there have been described and illustrated several specific embodiments of the invention, variations in the details of the embodiments specifically illustrated and described may be made without departing from the true spirit and scope of the invention as defined in the appended claims. For example, a balancing network similar to the network illustrated for FIGS. 1-4 can be employed in conjunction with the systems of FIGS. 6 and 7. In addition, the six db of positive blending between the right and left channels, as illustrated in the reproduction embodiments of FIGS. 2 and 5, can be obtained by combining the responses of networks 95 and 97 prior to feeding them to putput device 98, FIG. 8. The latter feature, in many instances, is not desirable, however, because it is not compatible with standard stereophonic reproduction systems having only two speakers. If the signals derived from summing networks 97 and 95 are directly recorded there is complete compatibility with standard two-speaker stereophonic reproduction systems.

The recording system of the present invention has been stated, see Equations (1) and (2), to provide L and R signals, which with change of constants are:

L = A + (B/2) + C - (D/2)

R = (A/2) + B - (C/2) +D

The 50 percent cross mixing represented by the factor 1/2, is cancelled by corresponding negative cross feeding in the reproduction system of FIG. 2, popularly known as a Quadaptor (T.M.) This implies that if a front left signal is recorded, no right rear signal will be reproduced, and similarly for a front right signal no left rear signal will be reproduced.

There is a complete separation between A and B with 6 db of separation negative between C and D, for a pure left signal, so that a signal on left input only has much more separation in front than in back, in the Quadaptor. This is acceptable since the listener hears the rear speakers from a wide angle, and because phase reversal in the rear widens the apparent separation.

The 50 percent cross feed system of Equations (1) and (2) can be modified in respect to cross feed factor, i.e., 50 percent is not critical or essential. It is advantageous to equalize front and back separations, for example, and this can be optimumly accomplished with cross feed 2 - .sqroot. 3, or 26.8 percent. Separation in both front and rear, apart from outphasing in the rear, is then about 11 db.

The matrixing system, with parameters rounded off, then becomes

L = A + (1/4) B + C - (1/2) D (7) and R =(1/4) A + B - (1/2) C (8).

Employing the suggested matrix of Equations (7) and (8), differentiation of direction of better than 10 db both front and back, and about 10 db left to right, and the matrix is optimum for Quadaptor reproduction, but also quite suitable for normal stereo use, or for monophonic use.

While the specific value of cross feed, 26.8 percent, is optimum, in general a range of cross feeds should be made available in the recording studio, to provide desired musical effects, and to accomodate various microphone locations and physical recording environments.

Claims

What I claim is:

1. An audio reproduction system driving four speakers in response to only two left and right stereophonic audio power amplifiers, said amplifiers being left and right amplifiers and the audio signals provided by said amplifiers respectively having audio signal levels denominated as L and R, a different one of said speakers being positioned approximately at each of the four corners of a quadrilateral area, which may be occupied by a listener, said speakers directing sound towards the interior of the area, wherein first and second ones of the speakers are respectively positioned in the left and right rear corners of the area behind the listener and third and fourth ones of the speakers are positioned in the left and right front corners of the area facing the listener, each of said speakers including a signal and a ground terminal, comprising means responsive to said left amplifier for applying a first audio signal to said signal terminals of the front left and rear left speakers, means responsive to the other one of said amplifiers for applying a second audio signal to said signal terminals of the front right and rear right speakers, means grounding the second terminals of the two front speakers, and passive impedance network commonly in series with said rear speakers connecting the ground terminals of the two rear speakers to ground, the impedance of said impedance network having a substantial value capable of audibly modifying the relative amplitudes of the responses of the rear speakers relative each other.

2. The audio reproduction system of claim 1, wherein the value of said impedance network is selected so that the responses of the rear left and rear right speakers are respectively approximately proportional to (L - (R/2)) and (R - (L/2)).

3. The audio reproduction system of claim 2, wherein the impedance of said network has a value of the order of the impedance of each rear speaker.

4. The audio reproduction system of claim 2, wherein the first and second audio signals are respectively commensurate with (L + (R/2)) and (R + (L/2)), and the impedance network has an impedance approximately equal to four times the impedance of each rear speaker.

5. The audio reproduction system of claim 1, wherein said impedance network comprises a fifth speaker.

6. The audio reproduction system of claim 1, wherein the impedance network comprises a fifth speaker located at the front of the area approximately midway between the two front speakers.

7. The audio reproduction system of claim 1, wherein the impedance network comprises a fifth speaker located midway between said front speakers.

8. The audio reproduction system of claim 1, wherein said impedance network includes first, second and third variable resistors respectively connected, in the order recited, from the rear left speaker, and from the rear right speaker and as a common impedance extending from the first and second variable resistors to ground, and means for gang controlling the values of said resistors so that the first and second resistors always have approximately the same value and the third resistor has a value equal to that of one of the other resistors multiplied by a predetermined constant.

9. An audio reproduction system driving four speakers in response to only two stereophonic power amplifiers, said amplifiers being respectively denominated as left and right, a different one of said speakers being positioned approximately at each of the four corners of a quadrilateral area, said speakers directing sound towards the interior of the area, wherein first and second ones of the speakers are respectively positioned in the left and right rear corners of the area behind the listener and third and fourth ones of the speakers are positioned in the left and right front corners of the area facing the listener, said speakers each having a signal terminal and a ground terminal, comprising means responsive to one of the amplifiers for applying a first signal to the signal terminals of the front left and rear left corner speakers, means responsive to the other one of the amplifiers for applying an audio signal to the signal terminals of the front right and rear right corner speakers, means grounding the second terminals of the two front speakers, means connected to the second terminals of the two rear speakers for providing a common substantial resistance from the ground terminals of both rear speakers to ground, and means individually grounding the ground terminals of the front speakers.

10. An audio reproduction system driving four speakers in response to only left and right power amplifiers, a different one of said speakers being positioned approximately at each of the four corners of a quadrilateral area, said speakers directing sound towards the interior of the area, wherein first and second ones of the speakers are respectively positioned in the left and right rear corners of the area behind the listener and third and fourth ones of the speakers are positioned in the left and right front corners of the area facing the listener, said speakers each having a drive terminal and a ground terminal, comprising means responsive to the left and right amplifiers for deriving responses from the left and right front speakers respectively proportional to the outputs of said amplifiers, means responsive to said left and right amplifiers for deriving audio responses from the left and right rear speakers respectively proportional to (L - KR) and (R - KL), where K is a constant greater than zero and less than one, L is the output of the left amplifier and R is the output of the right amplifier.

11. The audio reproduction system of claim 10, where K is approximately 1/2.

12. The audio reproduction system of claim 11, wherein the means for deriving responses from the left and right rear speakers comprises a common impedance connecting the ground terminals of only the rear speakers to ground, said common impedance having a value approximately equal to the impedance of one of said rear speakers.

13. An audio reproduction system driving four speakers in response to only two stereophonic amplifiers, said amplifiers being respectively denominated as left and right, a different one of said speakers being positioned approximately at each of the four corners of a quadrilateral area, said speakers directing sound towards the interior of the area, wherein first and second ones of the speakers are respectively positioned in the left and right rear corners of the area behind the lister and third and fourth ones of the speakers are positioned in the left and right front corners of the area facing the listener, comprising means responsive to said amplifiers for deriving responses from the left and right front speakers respectively proportional to (L + K.sub.1 R) and (R + K.sub.1 L), means responsive to said amplifiers for deriving responses from the left and right rear speakers respectively proportional to (L - K.sub.2 R) and (R - K.sub.2 L), where K.sub.1 and K.sub.2 are constants greater than 0.25 and less than 0.75, where L is the audio output amplitude of the left amplifier and R is the audio output amplitude of the right amplifier.

14. The audio reproduction system of claim 13 wherein K.sub.1 = K.sub.2 = 1/2.

15. In a four speaker two amplifier stereophonic reproduction system, wherein left and right front speakers are connected to left and right amplifiers of said reproduction system, respectively, the combination wherein is provided left and right rear speakers connected respectively to be driven directly by said left and right amplifiers, and a passive circuit connected directly between said rear speakers and ground for introducing at least 25 percent of negative cross coupling of said rear speakers.

16. In a four speaker audio system having only left and right power amplifiers, four loudspeakers having each a signal input terminal and a grounding terminal, said loudspeakers being left front, right front, left rear, and right rear loudspeakers, respectively, means connecting said left power amplifier in parallel with the signal output terminals of said left front and left rear loudspeakers, means connecting said right power amplifier in parallel with the signal input terminals of said right front and rear loudspeakers, a switch normally connecting the grounding terminals of said rear loudspeakers to ground, a passive impedance network normally commonly connecting the grounding terminals of said rear speakers to ground via said switch, said impedance network comprising a common resistance connected with said rear speakers in parallel to audibly cross couple said rear loudspeakers.

17. The system according to claim 16 wherein is further provided control means for at will disconnecting only said signal input terminals of said rear speakers from said power amplifiers.

18. The system according to claim 16, wherein said impedance network is a resistive T-network including said common resistance and further separate resistances respectively connecting said grounding terminals of said rear speakers to said common resistance and thence via said switch to ground, and a gang control means for commonly adjusting all the resistances of said T-network.

19. The system according to claim 16, wherein said common impedance includes a fifth loudspeaker.

20. A four speaker stereophonic system driven solely from a left and a right power amplifier, said four speakers including left and right front speakers and left and right rear speakers, means driving said left speakers in multiple directly from said left amplifier, means driving said right speakers in multiple directly from said right amplifier, an adjustable resistive passive cross coupling network, said resistive cross coupling network being connected commonly between said rear loudspeakers and ground and providing at least 25 percent cross coupling between said rear loudspeakers.