Binaural Four-Channel Stereophony

The present invention relates generally to stereophony, and in particular to four-channel stereophony using binaural signals generated from microphones mounted in dummy heads simulating the human head.

The known binaural sound recording system, is a closed circuit type of sound reproducing system in which two microphones, used to pick up the original sound, are each connected to two independent corresponding transducing channels which, in turn, are coupled to two independent corresponding earphones worn by the listener. The microphones are mounted in a dummy simulating the human head in shape and dimensions and at the locations corresponding to the ears of the human head. The listener is transferred to the location of the dummy head by means of a two-channel sound reproducing system. Because of the direct transfer of signals, the listener has spatial impressions as if sitting at the location of the dummy. However, because of the inconvenience that the listener has to wear the earphones (headphones), two-channel loudspeaker reproduction using a dummy head has been proposed to take advantage of the spatial impressions of the binaural system. However, a stereophonic pair of signals is usually reproduced by two loudspeakers standing on the right and left in front of the listener. The listener then receives at his left ear not only the wanted left signal, but an unwanted right signal as a result of sound diffraction at the head. Because of the symmetry of the system, this process is called acoustical crosstalk. This acoustical crosstalk can be eliminated by an electronic circuit as proposed by P. Damaske (Head-Related Two-Channel Stereophony with Loudspeaker Reproduction, The Journal of the Acoustical Society of America, Vol. 50 No. 4 Part II pages 1109-1115). Reproducing the binaural signals of a dummy head with two loudspeakers is still problematic in that virtual sound sources can only be produced between the two loudspeakers and in that a slight movement of the listener's head gives him an impression that the sound sources have been dislocated.

Therefore, the primary object of the invention is to provide four-channel stereophony with loudspeaker reproduction in which at least two dummy heads are used to generate front and rear binaural sound signals to provide localization of virtual sound sources between four loudspeakers located around the listener.

A four-channel stereophony in accordance with a first aspect of the invention comprises first, second and third dummy heads each simulating the human head in shape and dimensions and provided with microphones mounted in positions corresponding to the ears of the human head to pick up front, lateral and rear sound sources, respectively. The signals representing the front and rear sound sources are applied to a first and a second acoustic crosstalk cancellation circuit, respectively, while the signals representing the lateral sound source are coupled through a pair of variable attenuators to the first and second cancellation circuits at an attenuation ratio determined by the direction of the lateral sound source from the location of the second dummy head. Sound sources are precisely localized between the four individual speakers.

A four-channel streophony in accordance with a second aspect of the invention comprises first and second dummy heads with microphones mounted therein. Both dummy heads have their faces oriented in the same direction with the second dummy located exactly behind the first dummy with a predetermined spacing therebetween and located in a sound field of the same sound sources. Preferably, the first and second dummy heads are separated by a partition to produce a sound separation effect so that the first dummy picks up primarily front sound sources and the second dummy picks up primarily rear sound sources. The signals derived from both dummy heads are coupled to a first and a second crosstalk cancellation circuit, respectively.

The invention will be further described in detail by way of examples in conjunction with the accompanying drawings, in which:

FIG. 1 is a schematic circuit diagram useful for describing the principle of the invention;

FIG. 2 is a detailed circuit diagram of an acoustic crosstalk cancellation circuit;

FIG. 3 is a graphic illustration of acoustic transmission characteristics of filter and delay circuits of FIG. 2;

FIG. 4 is a graphic illustration useful for describing the adjustment of variable attenuators employed in the circuit of FIG. 1;

FIG. 5 is a variation of the circuit of FIG. 1;

FIG. 6 is a detailed circuit diagram of a modified from of the acoustic crosstalk cancellation circuit;

FIG. 7 is a graphic illustration of acoustic transmission characteristic of the filter and delay circuits of FIG. 6;

FIG. 8 is a first embodiment of the invention;

FIG. 9 is a second embodiment of the invention; and

FIG. 10 is a modification of the second embodiment;

Before describing the embodiments of the present invention, reference is first had to FIG. 1 in which the principle of the invention is illustrated. A sound pickup arrangement 1 picks up binaural sound signals from an original sound source (not shown). Microphones 3R and 3L are mounted in a dummy 2 simulating the human head in shape and dimensions and at the locations corresponding to the ears of the human head. The sound signals are amplified through amplifiers 4R and 4L and applied to a front converter or compensator 10f and thence to a four-channel transducer 5. The amplified signals are also applied to a rear converter or compensator 10r and thence to the transducer 5. The transducer 5 may be an amplifier, a radio transmitter and receiver, a phonograph recorder or reproducer. Front loudspeakers 11f and 12f and rear speakers 11r and 12r are located in front and rear respectively of a listener M and connected to the outputs from the transducer 5. A set of ganged variable attenuators 15 and 16 is interposed in the circuit between the outputs from amplifiers 4R, 4L and the inputs to the converter 10f. Similarly, a set of ganged variable attenuators 17 and 18 is interposed between the amplifier outputs and the inputs to the rear converter 10r. In accordance with the invention, the front speakers 11f and 12f are used in the reproduction of the original sound sources located in front of the dummy head 2, while the rear speakers 11r and 12r are used in the reproduction of the original sound sources located rearward of the dummy head 2, and both front and rear sound channels are used simultaneously in the reproduction of the original sound located laterally of the dummy head. Localization of the original sound sources is achieved with reference to FIG. 4 which graphically illustrates the input gains of converters 10f and 10r as a function of direction of localization from the listener M in which arrows a, b, c, and d designate the directions toward speakers 11f, 12f, 11r and 12r, respectively.

Forward sound localization is achieved by adjusting the attenuators 15 and 16 to reduce their losses to provide a maximum input gain for the converter 10f as indicated by the arc of a solid-line curve which extends between lines a and b, while adjusting the attenuators 17 and 18 to maximize their losses to provide a minimum input gain for the converter 10r as indicated by the dotted-line curve. Conversely, rearward sound localization is achieved by adjusting the attenuators 17 and 18 to reduce their losses to provide a maximum input gain for the converter 10r as indicated by the arc of the dotted-line curve which extends between lines c and d, while adjusting the attenuators 15 and 16 to maximize their losses to provide a minimum input gain for the converter 10f as indicated by the solid-line curve. Localization of a sound source located laterally of the dummy head in a direction as indicated by arrow e, attenuators 15 to 18 are adjusted by the ratio of line segments on line e divided by the solid- and broken-line curves: in this case the attenuators 15 and 16 and attenuators 17 and 18 are adjusted such that the input gains for converters 10f and 10r are 60% and 40%, respectively, of the maximum level.

Under forward localization, the converter 10f provides cancellation of acoustic crosstalk resulting from sound diffraction at the head of the listener M. The converter 10r also provides crosstalk cancellation resulting from sound diffraction at the rear of the head of the listener M under rearward localization.

Right and left sound signals received at the ears of the listener M are expressed by the following Equation: ##EQU1## where, Le and Re are the sound signal levels at the left and right ears of the listener M, respectively, Lsp and Rsp are signal levels at the left and right loudspeakers 11f and 12f, respectively, and A is given by ##EQU2## where, a.sub.11 is a transmission characteristic of sound from loudspeaker 11f to the left ear of listener, a.sub.12, a transmission characteristic of sound from loudspeaker 12f of the left ear of the listener, a.sub.21, a transmission characteristic of sound from loudspeaker 11f to the right ear of the listener, and a.sub.22, a transmission characteristic of sound from loudspeaker 12f to the right ear of the listener.

Signals Lsp and Rsp have the following relation with the signals Ls and Rs delivered from amplifiers 3L and 3R, respectively: ##EQU3## where, T is a delay time and A.sup.-1 is an inverse matrix of A. By substituting Equation (3) into Equation (1), signals Le and Re are given by ##EQU4## Equation (4) implies that the signals received by the ears of the listener M are the signals received at the ears of the dummy head 2 delayed by time T if the latter is converted in accordance with Equation (3). The conversion coefficient T.A.sup.-1 is expressed by ##EQU5##

In FIG. 2, each of the front and rear converters 10f and 10r is shown and comprises a right channel 19R connected between the output of the amplifier 4R through input terminal 8 and an input of the transducer 5 through output terminal 13 and a left channel 19L connected between the output of amplifier 4L through input terminal 9 and an input of the transducer 5 through output terminal 14. A filter 20R having a transmission characteristic ##EQU6## is placed in the right channel 19R and a filter 20L having a transmission characteristic ##EQU7## is interposed in the left channel 19L. Numerals 21R and 22R are a delay element and a filter, respectively, to provide attenuation of signal on the right channel 19R by the factor of a.sub.12 /a.sub.11. The delay element 21R and filter 22R are connected in series through an inverter 23R between the output of filter 20R and one input of an adder 24L connected in the left channel 19L. Similarly, the output from the filter 20L is attenuated by the factor of a.sub.21 /a.sub.22 by series-connected delay element 21L and filter 22L connected between the output of fileter 20L and one input of an adder 24R connected in the right channel 19R through an inverter 23L. The output from the filter 20R is directly applied to the other input of the adder 24R to provide summation of the two input voltages to generate an output Rsp which is given by ##EQU8## Likewise, the output from the filter 20L is directly applied to the other input of the adder 24L so that adder 24L delivers an output Lsp which is given by ##EQU9## The right and left channel signals are expressed by Equations (6) and (7) are fed into the transducer 5 and will be eventually fed to the right and left loudspeaker 12f and 11f, respectively. FIGS. 3A and 3B illustrate graphically the transmission characteristic of the filter 20R and the combined characteristic of delay element 21R and filter 22R, respectively, given that the propagation axes of sound waves transmitted from the front left and right loudspeakers 11f and 12f are inclined at 45.degree. to each other and the listener M is seated at the vertex of a triangle formed by the lines interconnecting the loudspeakers 11f, 12f and the listener M. Since the listener is seated at equal distances from the loudspeakers 11f, 12f, the transmission coefficients a.sub.11 and a.sub.22, are of roughly equal values, and also a.sub.12 and a.sub.21 are substantially of equal values. Therefore, similar transmission characteristic curves may be obtained for the filter 20L and the delay element 21L and filter 22L (the curves of FIGS. 3A and 3B are obtained for a delay time of 0.4 milliseconds over the entire spectrum of frequencies).

Localization of front sound sources is therefore effected by passing the signals Rs and Ls through the front converter 10f while suppressing the signals to be converted by the rear converter 10r, and the sound waves resulting from the converted signals Rsp and Lsp as emitted from the loudspeakers 12f and 11f produce a binaural effect as if the listener M is seated in the position of the dummy head 2. Localization of rear sound sources which come from the rear of the dummy head 2 is also achieved by allowing the rear converter 10r to pass the signals Rs and Ls while suppressing the signals to be converted by the front converter 10f, with the result that the rear sound waves emitted from rear loudspeakers 11f and 12r produce a similar binaural effect to the ears of the listener M as if he is seated in the position of the dummy head 2. In FIG. 4, solid-line curve illustrates a pattern of sound level measured when the phantom sound sources are localized in front of the listener M, while broken-line curve illustrates a pattern of sound level measured when the phantom sound sources are localized at the rear of the listener, and "M" denoting the location of the listener and "a" and "b" designating the lines interconnecting the front left and rear right loudspeakers and the front right and rear left loudspeakers, respectively.

In the embodiment of FIG. 1, it is important to note that the original front and rear sound sources are localized independently from each other on a one-at-a-time basis. Therefore, the original sound sources are not present simultaneously.

The attenuators 15 to 18 may be connected between the output of converter 10 and the input to the transducer 5 as illustrated in FIG. 5. In this case, a single converter 10 is commonly used for conversion of the binaural signals Rs and Ls, rather than two converters for conversion of front and rear binaural signals.

FIG. 6 shows a modified form of the converter 10 which is derived as follows: From Equation (1), Lsp and Rsp can be given by ##EQU10## The conditions which satisfy that Le = Ls and Re = Rs can be obtained by substituting Ls and Rs for Le and Re, respectively, of Equation (8), thereby resulting in Equation (9) ##EQU11## In FIG. 6, the converter 10 comprises a pair of paths 19R and 19L for right and left channels, respectively, as previously described. A filter 30R having a transmission characteristic 1/a.sub.22 is connected in the right path 19R and a filter 30L having a transmission characteristic 1/a.sub.11 connected in the left path 19L. Adders 34R and 34L are interposed in the right and left paths. A first feedback compensation path 35 is provided between the output of adder 34R and one input terminal of the added 34L, and a second feedback compensation path 36 is provided between the output of adder 34L and one input terminal of the adder 34R. The first compensation path 35 includes a filter 31R, a delay element 32R and an inverter 33R all of which are connected in series. The filter 31R and delay element 32R have a combined transmission characteristic of a.sub.12 /a.sub.11 so that the adder 34L provides an output Rsp which satisfies Equation (9). Similarly, the second compensation path 36 includes a filter 31L, a delay element 32L and an inverter 33L all of which are connected in series to generate a signal Lsp at the output of added 34L which also satisfies Equation (9). In practice, the filters 30R and 30L are realized by taking into account delay time T so that their characteristics become T/a.sub.11 and T/a.sub.22, respectively. FIG. 7 illustrates the transmission characteristics of the filters 30R and 30L on the assumption that the listener M is seated at equal distances from the loudspeakers of interest (front or rear loudspeakers) with the direction of sound wave propagation oriented to the listener, as previously described.

FIG. 8 illustrates an embodiment of the present invention which comprises a front sound pickup arrangement 1a, a lateral sound pickup arrangement 1b and a rear sound pickup arrangement 1c, all of which are similar in configuration to that shown in FIG. 1 with the exception that buffer amplifiers 51 to 54 are connected between the attenuators 15 to 18 and the inputs of the converters 10f and 10r and that the outputs of the buffer amplifiers of the same reference numeral are connected in common to the respective converters in order to provide simultaneous recording of various sound sources. Front sound sources S.sub.1a, S.sub.2a and S.sub.3a are located in front of the dummy head 2 of front sound pickup arrangement 1a, a lateral sound source S.sub.1b to the left of the dummy head of arrangement 1b, and rear sound sources S.sub.1c, S.sub.2c and S.sub.3c at the rear of the arrangement 1c. These sound pickup arrangements are acoustically isolated by means of partitions 40 and 41 schematically illustrated. In the illustrative embodiment of FIG. 8 attenuators 15a, 16a are adjusted to provide a maximum input gain for converter 10f, while attenuators 17a and 18a are adjusted to provide a minimum input gain for converter 10r. Likewise, attenuators 17c, 18c are adjusted to provide a maximum input gain for converter 10r, while attenuators 15c, 16c are adjusted to provide a minimum input gain for converter 10r. It will be understood that if the front and rear sound sources are permanently associated with the pickup arrangements 1a and 1c, respectively, attenuators 15a to 18a and 15c to 18c as well as buffer amplifiers 53a, 54a, 51c and 52c can be dispensed with. Atenuators 15b to 18b are adjusted in a manner as previously described in connection with FIG. 4. Therefore, the converter 10f receives binaural signals representing the frontal components of the overall signal through buffer amplifiers 51a, l 52a and 51b and 52b, whereas the converter 10r receives binaural signals representing the rearward components of the overall signal through buffer amplifiers 53b, 54b and 53c and 54c. All the binaural signals passing through the converters are recorded by the four-channel transducer 5 in a known manner or transmitted through a communication medium to a receiving end where loudspeakers 11f to 12r are provided.

In FIG. 9, there is shown a second embodiment of the present invention which comprises a front sound pickup arrangement 1f and rear sound pickup arrangement 1r located at a predetermined spacing behind the front pickup arrangement 1f in the same sound field. Each of the arrangements comprises a dummy head 2 and right and left microphones 3R and 3L mounted within the dummy as in the previous embodiment. The dummy heads 2 of both pickup arrangements have equal dimensions and shapes, are supported at the same height from the floor level and have their faces oriented in the same direction with their noses exactly aligned along a line "l" which bisects both dummy heads. The front pickup arrangement 1f is used to primarily collect sound signals from front sound sources and feeds the front speakers 11f and 12f through front converter 10f and transducer 5, while the rear pickup arrangement 1r is used to primarily collect signals from rear sound sources located at the rear end of the rear pickup arrangement 1r to feed the rear speakers 11r and 12r through rear converter 10r and transducer 5. Amplifiers 4fR and 4fL are connected between the microphones 3R, 3L of front pickup arrangement 1f and the input terminals of converter 10f to provide amplification of the front binaural signals. Similarly, amplifiers 4rR and 4rL are provided between the microphones 3R, 3L of rear pickup arrangement 1r and the input terminals of rear converter 10r to provide amplification of the rear binaural signals. Front and rear binaural effects are thus reproduced through front and rear channels. A lateral sound source, if present, is picked up by front and rear pickup arrangements 1f and 1r and reproduced through respective channels.

A partition 60 may be provided between the front and rear sound pickup arrangements as illustrated in FIG. 10. The effect of the partition 60 is to provide a certain degree of isolation between the front and rear sound sources so that the listener M is able to clearly distinguish therebetween. The dummy heads 2 are preferably provided with earlaps since they enhance the degree of distinctness between the front and rear sound sources.

An experiment showed that each sound source or group of sound sources can be localized between adjacent ones of the four speakers and thus the binaural effect as produced by a closed-circuit type binaural sound reproducing system are simulated without the inconvenience and disadvantages of the prior art system.