JPS582519B2 - How do you know what to do? - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to the reproduction of multidirectional audio performance information that is more directional and immersive than conventional stereo reproduction, and to the recording and/or transmission of such performance information.

In particular, the present invention provides methods for coding or mixing directional acoustic information into a number of recording or transmission channels that is (at least normally) smaller than the number of sound sources being reproduced and for reproduction psychoacoustically similar to the original performance. It concerns the decoding of the contents of these channels or the processing and transmission of signals to reproducers that differ in number or location from the sound source location.

Although the invention in its broad scope is not limited to any particular number of sound sources (typified by microphones) or reproducers (typically speakers), this application refers to a "four-channel audio system" or simply " This relates to a playback format commonly known as ``4-channel'', which extends the stereo technique using two speakers to multidirectional playback of performance information using four speakers to increase the sense of presence. .

Four-track tape recordings are often used for playback through four speakers arranged around the listener.

Recently, schemes have been devised for obtaining such reproduction on transmission media conventionally used for stereo reproduction and having only two transmission channels, and four (or more) in such devices.
The original directional sound source channels are mixed or combined by encoding into two signal channels and then decoded to generate four playback signals to feed the loudspeakers, such a scheme is 4-2-4. Described as a mixing device.

A requirement for widespread adoption of any four-channel device is that the current mono or stereo playback devices used continue to be dual-purpose.

In other words, recordings made for 4-channel playback should preferably be able to be played satisfactorily by a mono record player and a regular stereo record player, and 4-channel FM transmissions (either live or from recordings) should be available there. must be reproduced equally as mono or stereo by listeners located at the same location.

A number of encoding and decoding devices for both mono and stereo applications have been proposed, some of which have been the subject of experimentation and are in preliminary form for commercialization.

A relatively important weakness or imperfection of each of these known devices is that their proponents differ in their evaluations, but none of the previously proposed 4-2-4 devices
The performance of channel playback is quite different, and therefore it has not been adopted as a standard encoding and decoding device for use in stereo disc recording and stereo FM broadcasting equipment.

Other recently proposed formats that approach the problem include the use of audio-modulated ultrasound carrier techniques, whereby two additional transmission channels are added to the conventional stereo signal, i.e., mono and stereo reproduction. It is a 4-4-4 device with varying degrees of ambidexterity.

However, such 4-4-4 devices known in the art are based on questionable and even erroneous assumptions as shown below, which
The most satisfactory use of the transmission channels is the assumption that these channels are necessarily used for direct unmatrixed feeding to the loudspeakers as well as for the playback of conventional four-track recordings.

The present invention not only provides substantial improvements to 4-2-4 matrix devices, but also allows for even greater numbers of 4-2-4
It provides a substantial improvement in the use of transmission channels in playback.

Before describing how these improvements are made, the general nature of the deficiencies found in conventional devices will be briefly discussed.

In any 4-2-4 device, including this arrangement, sound originating from a single point (particularly a point in a position that activates only a single microphone) ends up coming from different speakers rather than from a single speaker. The same original sound source is reproduced from a plurality of arranged speakers but at relatively different amplitudes, so that from a psychoacoustic point of view the reproduction from a single direction is satisfactorily similar.

However, in the best of the 4-2-4 devices heretofore devised, the satisfaction of this illusion of directionality is not uniform in all directions.

The nature of anomalies or directional ambiguities in signals that are heard by a listener as coming from a particular direction in conventional devices is not exactly the same in each case, and conventional matrix devices do not allow for the effects of such anomalies. The direction of the sound that occurs is also not the same.

A typical example, however, is out-of-phase reproduction from rear speakers and similar anomalies, in which sound directed toward the front is reproduced more faithfully than sound directed toward the rear.

In some cases, this anomaly is psychoacoustically negligible for most performance information, but becomes significant for information with "sound effects" specifically designed for 4-channel playback, in which case the rear Sound is not just an aid.

Additionally, prior art devices (including 4-4-4 devices) assume one specific location of the speaker relative to the listener.

The required loudspeaker orientation is specified and known with respect to the orientation represented by the four separate transmission channels or the coating used for two-channel transmission or recording, while at the same time maintaining satisfactory reproduction.
There is no provision for using different directions of the speakers or different numbers of speakers.

Unlike the encoding and decoding of the prior art, the encoding and decoding of the present invention can create the same playback characteristics in all directions,
It can be said to be "directionally symmetric."

The meaning of the terminology used here is most easily understood by considering the simple example of sound reproduction, in which a sound source is moved sequentially to successively activate each of four orthogonally arranged microphones. Ru.

In devices with directional symmetry, the corresponding sequential 9
A listener rotating through an angle of 0° hears the sound source in exactly the same way as it is moved correspondingly through the four positions.

This property is inherent in four-channel audio systems having separate signal channels for the transmission of each microphone output, but is not available in prior art 4-2-4 matrix systems.

As indicated above, this is particularly important in order to be able to use new forms of performance information, where independent sound sources (e.g. voices or choruses) can be authenticated in a localized position behind the listener. It conveys a close impression.

Moreover, to eliminate this practical limitation, the directional symmetry of the matrix, which is a further feature of the present invention, allows a simple signal conversion, whereby the placement of the loudspeaker arrangement is relative to the assumed loudspeaker placement in signal generation. Completely satisfactory reproduction of recordings or FM transmissions designed for a ``2+2'' loudspeaker orientation (for example) by means of a reproduction scheme with the loudspeakers rotated and arranged in a diamond-shaped ``1-2-1'' shape or vice versa. I can do it.

The present invention provides a method and apparatus whereby the number and arrangement of loudspeakers used for playback can be decoded into a playback signal that is "universal" and for feeding into any desired number and orientation of loudspeakers. It is not essentially the same thing as in signal transmission.

The two encoded transmission channels are decoded to produce six loudspeaker signals for, for example, hexagonally arranged loudspeakers, so that the listener receives approximately six separate signals.
It gives you the feeling you get from transmitting individual channels.

Such advantages can be obtained simply by replacing the two-channel matrix of the present invention in various previously known "4-2-4" devices, but by extending the number of transmission channels to a larger number, such as four. If so, the present invention has further unique properties.

The two-channel matrix device described below generates a reproduced signal to a loudspeaker, which signal is shown to contain the maximum possible azimuthal localization that can be carried on two directionally symmetrical transmission channels.

Such devices cannot be made "separately".

That is, even if the psychoacoustic sound localization is satisfied with reduced amplitude and phase relationships, a sound that is represented as coming from an azimuthal direction that coincides with that of the loudspeakers will always be delivered by two adjacent loudspeakers. It will be done.

This form of regeneration can be sharpened, i.e. by simply adding one or more further transmission channels to the directional sharpness, the center of regeneration in the undesired direction or the fractional ratio in the desired direction can be can be reduced.

With a 4-channel speaker arrangement, if you add two more transmission channels to the basic device, you will get completely separate playback, i.e. the same amount of sound that can be played on a single speaker using a regular 4-track tape recording. Also, with this device, it is possible to reproduce music from one speaker without having to listen to other speakers.

For any performance information (rarely encountered) that consists primarily of sound played directly from the four main speaker positions, the performance is therefore a four-channel system in which the speaker signals are completely separated. It is psychoacoustically indistinguishable from that which occurs in .

However, unlike the latter type of four-channel separation device, the separation matrix of the present invention is easily made to be fully compatible with any form of conventional mono or stereo reproduction.

Moreover, the four-channel device of the present invention produces substantially better psychoacoustic localization of directional sound images that do not correspond to "virtual" azimuth positions, i.e., actual speaker positions, than conventional four-track reproduction. It is known to do.

Separation or independent operation of loudspeakers placed directly in the direction of the "sound source" is obtained in the four-channel matrix device of the present invention at zero, which is the total reproduction of the signal detected as being at any given loudspeaker position. 90° of the figure, 18
Appears at 0° and 270°.

In addition to the already mentioned advantages of conventional separate transmission of similar channels, the device allows highly restricted frequency ranges to be transmitted in two complementary channels without compromising the improved directional localization they provide. It offers the following advantages:

Bandwidth requirements are therefore substantially reduced for transmission in the medium, where each transmission channel would be required to transmit the entire audio range if other devices were used.

Both the basic two-channel matrix arrangement of the present invention and the addition of additional channels as described above for the sharpening of directional reproduction are best understood with reference to the following description and the embodiments in the accompanying drawings. It will be.

1 to 3 illustrate a four-channel audio system that can be advantageously utilized in the present invention.

1 and 2 show some of the same prior art, except for the matrices for encoding and decoding the transmitted signals.

These exemplary devices are illustrated and described to facilitate an understanding of the advantages of the present invention and the broad utility of the coding and decoding matrices (also referred to as rematrixes) described below.

The device of FIGS. 1 and 2 is an alternative type of four channel device conventionally used with various matrix devices.

In each case, an arrangement or type of microphone 20 or 20a orthogonal to the performance position is shown in the listening space surrounding the listener 23, together with a corresponding orthogonal arrangement or type of loudspeaker 22 or 22a.

In the device shown in FIG. 2, a microphone 20a and a speaker 22
a is the front left LF, front right RF, and rear right R of the performance and listening space, respectively.
are arranged to receive and reproduce sound from positions 20 and 2 in FIG.
2 is on the front F, right R, rear B and left L.

Encoders or matrices 24 and 24a generate two transmission signals at 26 or 26a, which signals in turn
8 or 28a to generate a reproduction signal that drives a speaker at a corresponding position.

LR, LF, LB and RB in Figure 2 and L, F in Figure 1,
The display of speaker positions in R and B is so-called "2+2"
and "1-2-1" are more or less representative of the positions actually used in what is called a 4-channel orientation, but for those skilled in the art, the microphones shown correspond to 4-channel playback.
It will be appreciated that this is a highly simplified representation of the actual arrangement of microphones normally used, especially for recording.

Although a simple device such as that shown, a four-way (principal type) microphone, can and is sometimes used in the usual positions of listeners in concert halls, for example, more complex microphones may be used for the effect judged to be most pleasing. It is common to use arrays and mix the outputs of different microphones, and indeed, as is the case in common stereo recording and broadcast studio techniques, multidirectional signals can be Can be synthesized or assembled from soundtracks.

It will therefore be understood, both from the drawings and from further discussion, that a representative audio signal of sound from a particular direction is completely synthesized as directional information.

As will be seen, the present invention additionally provides simple means for such synthesis.

It will also be understood by those skilled in the art that the illustrated examples of FIGS. 1 and 2 represent signal forming and processing operations that are performed to create a synchronous reproduction of live performance information, but are not typically ie, recording the signal at one or more points in succession.

Typically, the two transmission signals are ordinary Sdereo disks.
either on the "left" and "right" groove walls of the recording or on the corresponding audio channels of the stereo broadcast.

This, of course, is due to the current burden of the two
This is a channel limit.

The matrixing or coding of the present invention is also advantageously used in simple fixed position devices such as that of FIGS. 1 or 2 because of the directional symmetry provided by the invention.

However, a further advantage of the matrixing method and apparatus of the present invention lies in its wide utility.

The present matrixing not only can be easily applied to both the devices of Figures 1 and 2, but also provides a coding for very satisfactory use of speaker placement or orientation that does not in any way "match" the source signal placement or orientation. can.

One embodiment of such an overall device is shown in FIG. 3, where a microphone device (or composite directional signal representing a sound source) 20a
and the output of the encoder 24a in FIG.
and is reproduced by the speaker 22 in FIG.

As can be seen below, the present matrixing or coding and decoding device not only provides excellent reproduction from all angles with such a "rotated" arrangement, but also allows the listener to use any number of speakers desired. It is even possible to use more spread out source signals and playback arrangements, such as:

For confirmation, FIG. 4 shows certain angular relationships used in the matrixing and rematrixing or coding and decoding of the present invention.

In the present invention, the amplitude and phase of each sound source signal appearing in each reproduced signal are completely and independently determined by the angular relationship between the direction or position represented by the sound source signal and the direction or position of the speaker to which the reproduced signal is supplied. It can be decided.

If the entire reproduction matrix (the result of the encoding and decoding matrices) is as follows, that is, the amplitude and phase of each sound source signal in each reproduction signal (relative to the original amplitude and phase) is based on only this angular relationship. If it is a function, perfect directional symmetry can be achieved with any device such as that of FIGS. 1-3.

The angle between any given sound source (actual or synthetic microphone arrangement) and the speaker position is designated as α as shown in FIG.

The overall value of α is the same in Figures 1 and 2,
It will therefore be seen that the same overall matrix for both of these arrangements is generated by the present invention as will be seen later.

However, although the encoding matrices 24 and 24a in each figure are not numerically the same, they are preferably chosen to preserve stereo compatibility, ie to allow stereo playback in devices without a decoder.

This is done by determining the encoding matrix coefficients at 24 or 24a according to the angle θ that defines the directional angle of each sound source with respect to the lateral intermediate (front or rear) direction and forming the reproduced signal, as will be detailed later. If the reproduction angle .phi. of each speaker is determined with respect to an intermediate position in the lateral direction, the matrix coefficients of the decoder 28 or 28a are determined in accordance with the azimuth angle .phi. of each speaker with respect to the intermediate position in the lateral direction.

As shown in FIG. 4, and as currently employed in the description of the present invention, the intermediate reference position in the lateral direction is considered to be the front position, and the angle is measured clockwise, where "left", "right"
It will be understood that references to and similar terminology are used for convenience of expression rather than as a specific limitation.

As is clear, matrixing and rematrixing cannot produce a complete overall matrix in the same sense as would be possible if there were no need to compress the number of signal channels for transmission.

However, the requirement of a complete global matrix is met more closely than previously known by the use of the matrix of the present invention.

A "universal" encoding matrix forming two transmitted signals TL and TR from any number n of sound sources is where Sk is the kth sound source signal and θk is the position and lateral direction of the sound represented by it. The azimuth angle from the central reference position, j, is the square root of -1.

It will of course be understood that the symbols in the equations hereafter indicate proportionality and that uniform changes in absolute values in signal processing are irrelevant.

The phase coefficients of the respective transmitted signals TL and TR are shown in FIG. 5 for the particular source positions discussed earlier.

The signal from the left, i.e., L (i.e., appears as "from" upon reproduction), is reproduced with full amplitude and the original phase of the TL signal, but is zero for the TR signal, and vice versa. .

Signals from other azimuths appear in both transmitted signals, but always in a quadrature phase relationship, one leading and one lagging relative to the reference phase, and this is maintained in the L and R signals.

The magnitude of each component decreases with increasing relative phase angle (positive or negative) for each 90° phase angle (18
becomes zero at 0° difference).

The following mixing equation, calculated from the ``universal'' matrix described above, allows for a fixed four-microphone configuration to accommodate other signals such as ``on-mic'' modification signals, which are often added for soloists and other special effects. With or without additions, it may be used to utilize the present invention.

The numerical values of the mixing formula for the sound source direction of 1-2-1 are as follows.

The numerical values of the mixing formula for 2+2 sound source directions are as follows.

Fixed circuitry that generates the desired mixing for one or both of these fixed microphone locations is configured with or without the microphone direction pattern if so desired.

Additional insertion of signal information is then performed using additional mixers as shown schematically in FIG. 6 to simulate performance at arbitrary positions.

As shown in FIG.
The phase divider 30 generates positive and negative reference phase signals and positive and negative 90° phase shifted signals.

The reference and phase-shifted signals are attenuated (and reversed in polarity when used) with azimuth-set sine and cosine potentiometers 32 and 34, and signal S is similarly inserted.

The positive reference signal and the potentiometer output are mixed in combiners 36 and 38, the outputs of which are then inserted as components of signals TL and TR, respectively, according to the basic encoding equations previously described.

In principle, a mixer as shown in FIG. 6 could be used for each sound source direction.

However, if a significant number of microphones or soundtracks are desired for recording or broadcasting at easily selectable angular positions, the number of phase shifters required may be increased using a configuration such as that shown in FIG. It can be reduced to width.

As shown, each signal S1, S2, etc. is supplied to a polarity divider (phase inverter) 40.

The positive or in-phase and negative or anti-phase signals are fed to a sine-cosine potentiometer that generates positive and negative signals of amplitude and polarity determined by the setting angle of the potentiometer.

The unattenuated positive signals from all sound sources and the negative sinusoidal signals (which are of course in positive phase for angles with negative sinus values) are mixed in the combiner 44.

The positive cosine signals are mixed in a synthesizer 46.

The output of the latter is 90 at 48 with respect to the output of the synthesizer 44.
The phase is advanced by .degree. and the two are mixed or summed by 50 to form the signal TL.

(As will be recognized by those skilled in the art, the output of summer 44 must be fed to summer 50 via the reference phase portion 52 of phase shifter 48, which is a shift of any currently available frequency independent phase shifter.) Phase is the phase difference between the phase shifted output and the output of the reference phase channel as shown at 52, rather than the phase difference between the output and the input.

) In a similar manner, the positive input signals and positive sine function signals from all sound sources are mixed in a combiner 54 and the negative cosine function signals are combined in a combiner 56.

The latter integrated output is the reference phase 6 of the output of the synthesizer 54.
The phase is advanced by 90° at 58 relative to 0 and these are likewise summed at 62 to form the transmission signal TR.

It will be appreciated by those skilled in the art that the functions performed by some of the elements shown in FIG. will admit it immediately.

For example, microphone sensitivity patterns of known types may readily be used in place of directed attenuation potentiometer circuitry for some or all of the signals S1, S2, etc. in the signal mixing device of FIG.

An orthogonally placed dipole microphone directly captures the signal attenuated according to the sine and cosine of the azimuthal angle of the incident source, with one omnidirectional microphone in close proximity to it used to generate a single or unattenuated component. It can be used because it occurs.

As indicated earlier, the positioning signals TL and TR can be recorded on any conventional medium, in particular a stereo disc or tape recording, or a four-channel FM using two audio channels provided in an ordinary stereo. Either it is used for simultaneous playback such as in broadcasting.

Next, consider a method of decoding the signal shown in FIG.

Decoding is very similar to coding, except that the coefficients added to the transmitted signal in the formation of each reproduced signal are a function of the angle φ, the listening spatial azimuth of the loudspeaker from which each reproduced signal is formed.

Each reproduced signal Pi is formed from the transmitted signals by mixing in amplitude and phase relationship.

Here, φi is the azimuth angle between the reproduction position and the horizontally central reference position, and j is the square root of −1.

Each regenerated signal is therefore formed by multiplying each transmitted signal by the complex conjugate of the multibriar or the coefficient used (or used) for inserting the signal from that azimuth in forming the transmitted signal, and then combining The respective products obtained are added.

Therefore, each synthesized reproduced signal Pi is as follows.

Therefore, the reproduced signal PL for the speaker in the left position is unchanged by the signal TL as shown in FIG. 5, and TR is similarly unchanged in the formation of the reproduced signal for the speaker in the R position (if there is one). Given.

The reproduced signals for the other positions are exactly the same in appearance as the phase diagram, except that the represented position is midway between the 180° shown in FIG. 5 for the L and R signals.

Except for the source signal diametrically opposite to the playback point, the total source signal appears in all playback signals, and its magnitude varies from the maximum as a function of the magnitude of the angle between the source direction and the playback direction. Continuously changes up to zero.

For the 1-2-1 speaker orientation, the numerical value of the decoding formula (2) is equivalent to the numerical value of the decoding formula (2) for the 2+2 speaker orientation.
), the overall transmission or reproduction is unaffected by selecting the transmitted signals TL and TR to correspond to the left and right directions, respectively, to the two channels with their original phase and amplitude. It can be seen that any choice of diametrically opposite directions for transmission gives the same result.

However, the combination of the transmitted signals in the left and right directions provides ambidexterity to conventional stereo equipment.

A reproducing device which allows the individual loudspeaker positions to be at any desired angle is devised along the same lines as the signal preparation device previously described.

However, such equipment is generally wasteful because the actual speaker placement is not as diverse as the microphone placement, where the balance between front and back is optional rather than required.

The fixed reproduction signal output for the eight positions shown is sufficient to cover the user's needs and preferences, while the 15° spacing is perfectly suitable for virtually any practical use.

One configuration of the decoder is shown in FIG. 8, which can be used for a very wide variety of speaker arrangements.

The respective transmission signals TL and TR are transmitted through 90° transporters 70 and 7.
2, each having a positive and negative reference phase and a phase-shifted output.

These outputs are fed to a fixed mixing network 74 which consists of a voltage divider that attenuates the input signal and distributes the attenuated signal to a synthesizer that produces an output according to equation (2).

The output from the fixed output terminal 76 is
It is played through speakers at arbitrarily selected multiples of 5° intervals (or any other interval to provide output).

The number and arrangement of speakers is therefore selected according to user preference (including economic constraints).

In general, the loudspeakers are usually spaced equidistantly in azimuth and preferably equidistant from the listening position, ie in a square or regular polygonal configuration.

However, in many cases room geometry and acoustics and personal preference may lead to other arrangements.

Another type of decoder with selectable fixed position output terminals is shown in FIG.

The transmitted signals TL and TR are provided to a summation and difference circuit 80 to generate a summation signal TΣ and a difference signal TΔ.

The difference signal is processed in the same manner as at 70 or 72 in FIG. 80, and separate phase shifter channels for reference phase 82 and phase shift 84 are again shown in FIG.

The respective polarities of the reference and phase-shifted difference signals TΔ are applied to fixed voltage dividers 86 and 88, and the attenuated outputs are used as references to generate reproduced signal outputs for preselected angles φ1, φ2, etc. A phase channel 91 provides a total signal TΣ to a synthesizer 90, which is designed with taps in attenuators or voltage dividers 86 and 88 for the preselected angle.

If the present invention is used in a standard stereo FM broadcast, the total and difference circuit shown at 80 in FIG.
Operating in mixing, circuit 80 can be omitted in decoding.

It will be appreciated that such total and difference signals TΣ and TΔ can be formed directly from the source signal and used as transmission signals without forming signals TL and TR, such signals being substantially as follows. It is formed.

The pair of transmission signals TL and TR is the pair of transmission signals TΣ and TΔ
and these signal sets can be easily converted from one format to another in either direction without any change in the available information content.

Although these two formats of the same signal information are usually the most useful and simplest for implementation of the device, other transmission signal pairs can be devised that have the same overall information content and that can be converted at any time to and from these specific formats. It will be understood that this is included in the above expression.

If so desired, after the individual reproduction signals have been formed, they can be "touched" according to the listener's preferences.

For example, a particular listener may further phase shift one or more of the reproduced signals after forming them to obtain a more pleasing overall characteristic (not shown).

As another example, directional effects are conventionally used in conjunction with other coding and decoding devices to increase the apparent contrast or localization of a sound source for certain types of sounds, such as the amplification of amplifiers that feed a particular loudspeaker. The same type of auxiliary signal processing, such as varying degrees, is emphasized.

It will be appreciated that the encoded transmission signal can be easily used in current playback devices that do not provide for decoding multidirectional signals.

The sum of the two transmitted signals is the pure sum of the entire source signals in their original phases.

Therefore, when the transmission signal is used for mono-compatible matrixing of the total and difference stereo FM broadcasts or for the reproduction of encoded stereo disc recordings on a mono phonograph, a completely mono reproduction results.

When using two encoded channels as left and right channels for conventional stereo playback, conventional stereo coding (four channels with two-channel transmission)
Slightly less apparent left-right separation results than in conventional arrangements for channel encoding and decoding.

If desired, the decoder may be provisioned to artificially encode an ordinary stereo signal containing no directional information, so that such performance information is generally similar to the reproduction of the signal in a multidirectional loudspeaker system. The direction information is further encoded.

A normal stereo signal corresponds to a sound source signal at the front left LF and the front right RF.

FIG. 10 shows an adapter that may be substituted for the totalizing and difference circuit 80 of FIG. 9, for example, by switching a decoder to convert a direction-encoded signal having source signal components only from these directions. A similar listener effect or sensation occurs.

The conventional stereo signal is applied to respective 45° phase dividers 92 and 94 to produce a reference phase total signal TΣ and a quadrature phase difference signal TΔ.

It was mentioned earlier that the selection of the two transmission signals for direct reproduction of the L and R positions respectively is important only for the compatibility of conventional stereo equipment without provision for decoding.

The invention has a broader application in applications where stereo compatibility is not important.

For example, the present invention may be used in the common type of recording currently performed on four or more separate tape channels for the sole purpose of conserving tape space and thus extending playing time.

Greater utilization of tape space is made by compressing the information into two recording channels and then decompressing it during playback.

In such uses of the invention, the azimuth reference direction used for encoding is chosen more or less arbitrarily, and the directions represented by the two transmitted signals are similarly arbitrary, so long as they are chosen diametrically opposite. be.

The specialized embodiments of the invention described above are therefore limited to those specified for use in two-channel transmission equipment that is currently standard for stereo broadcasting and disc recording.

However, the broad scope of the invention has wide applicability.

The two-channel transmission described above is therefore a special application of the principles advantageously used for the reproduction of directional audio information in more than one channel.

By use of the present invention, the results obtained with transmission limited to two channels as used in conventional stereo are maximized.

However, a large number of channels can also be advantageously used according to the invention.

The first use of multiple channels is to sharpen the directional pattern for any given speaker arrangement, i.e. to reduce the crosstalk that is an inevitable consequence of using a larger number of speakers than the number of transmission channels. .

However, these features further enable the present invention to produce a ``2+2'' signal with loudspeakers arranged in a ``1-2-1'' orientation, even when the number of transmission channels is equal to or greater than the number of required reproduction signals. It has great advantages not only for purposes of allowing "rotation" of the playback signal, such as in the playback of four-track tape recordings for speaker playback, but also for purposes discussed below.

An understanding of the application of the invention to channel counts greater than two will be facilitated by first considering certain features of operation and the underlying principles of the two-channel device described above.

The essence of advantageous novelty can be seen by considering that it arises from the following facts.

That is, the total product of all the functions of θ added to the source signal at an azimuth angle θ in the form of each transmitted signal and the function of φ added to the transmitted signal in the formation of each reproduced signal at an azimuth angle φ is the standard. angles having a maximum absolute value at the corners of the difference, a relatively small absolute value at the diametrically opposite corners of the difference, and an absolute value at the intermediate corners symmetrical about the axis thus defined. The fact is that it is a function of a single variable, the difference between θ and φ.

This is reflected in the decoded speaker pattern of the transmitted signal.
It is a characteristic that provides "rotability" or "universality."

All features that satisfy these criteria are relevant to the simulation of psychoacoustic effects when directly reproducing directional audio information using separate transmission channels.
Their strengths are not exactly the same.

Regarding the directional pattern, all other factors being equal, it is desirable to satisfy the following conditions:

(b) The magnitude or absolute value of the entire reproduction function becomes zero in an azimuth of 180° from the maximum position.

(b) In the amplitude pattern, θ and φ are equal, that is, the angle of difference α
decrease as quickly as possible from the maximum value that occurs at zero.

(c) The phase difference between crosstalk components due to arbitrary predetermined sound sources appearing in reproduction speakers in various directions should be minimized.

That is, the entire reproduction should have a minimum relative phase difference.

For example, it is a daily experience that when crosstalk components that are close to the opposite phase are reproduced from other speakers, the sound image localization becomes unclear and the user feels uncomfortable.

In actual operation, if these three elements are not satisfied, an illusion may be given to the position of the sound image, which is an important matter from a psychoacoustic perspective, but cannot be quantitatively evaluated at present.

The two-channel matrix regeneration using a complex function consisting of a 0° phase shift and a 90° phase shift in the embodiment of the present invention is useful for other matrix regenerations using in-phase and anti-phase functions that have the same purpose but do not use a complex function. It has been empirically established that it is more satisfactory in comparison.

The present invention satisfies the above conditions.

As is clear from Figure 13, the pattern is completely zero at 180° azimuth, and 90° azimuth (or 270° azimuth, these are convenient points to indicate the "sharpness" of the measured pattern).
shows an amplitude reduction of -3 dB.

Also, as is clear in Figure 5, 1 from any other
There are no components with a phase difference as large as 80°, and any components that are reproduced with a phase difference close to 90° from their original relative phase have essentially negligible amplitude.

It can be shown from information theory and sampling theory that the reproduced information provided by the mixing and remixing coefficients determined as above is as accurate as possible in the case of only two transmission channels; It is therefore within the broad scope of the invention to use the entire functionality of sacrificing the other.

The properties of the operating elements just described can be substantially further improved by applying the same general principles in the configuration of the transmission signals for three or more transmission channels.

Such applications of the principles of the invention can be broadly divided into two categories with respect to the two-channel embodiments described.

(1) Apparatus for using one or more auxiliary channels or transmission signals in addition to the two transmission signals already described;
2) A device that uses three or more transmission signals that exhibit the same type of mutual symmetry as the TL and TR signals of a two-channel device.

Three channel (and even higher channel) devices of the first type described above may be said to be "compatible" with two channel devices.

One of the common advantages of such embodiments of the invention is the production of three-track or four-track tape recordings, which recordings can be played back by any desired multiple arrangement of speakers with appropriate decoding, or Instead, it is played back as a regular stereo recording by devices that do not have access to auxiliary recorded channels.

However, greater utility is a substantial advantage of the present invention when it is implemented in conjunction with effectively fully separated four-channel signals for playback on media such as FM broadcasts and disc recordings.

The encoding and decoding of each auxiliary channel is of course such as to retain the essential characteristics of the overall transmission or reproduction function described above.

If we create the desired overall reproduction function in which the angle of difference α is a single variable, we can devise a composite decoding for all channels apart from the two-channel decoding to It is in principle possible to devise additional channel transmissions that can be decoded together with the transmission signal.

However, it is more desirable to retain the same method of decoding the two fundamental channels and simply add to each reproduced signal the auxiliary information contained in the encoded and decoded auxiliary transmission signal.

To do this, each encoding and decoding of the auxiliary transmission signal is required to generate an added component to the reproduced signal that is itself simply a function of the difference angle with a maximum value at a zero difference angle. It is.

The simplest and most desirable way to utilize the additional channels is to use an encoding function of θ for the generation of each auxiliary transmission signal, which itself becomes the unity of the difference angle when multiplied by the conjugate decoding function of φ. It yields a product that is a function of the variables, and this function increases the sharpness of the maximum amplitude in the pattern when added to the reproduced signal function resulting from the two-channel transmission.

These requirements for the auxiliary channel are met by using an appropriate exponential function of θ for encoding and its conjugate of φ for decoding.

A mixing coefficient proportional to each signal Sθ and each reproduced signal P
The addition of a third transmission signal using a remixing coefficient proportional to φ to form an additional component results in a product function of the angular differences, which when added to the basic two-channel regeneration function Substantially sharpens the sexual effect.

The overall playback signal is as follows.

With this overall reproduction function, all sound sources appear in all loudspeakers at their original relative phase and the amplitude for an angular difference of 90° (or of course 270°) is approximately 10 dB below the maximum at 0°. few.

The auxiliary signal TT formed in this way can be used with either the transmission signals TL and TR of equation (1) or the transmission signals TΣ and TΔ of equation (4) above.

The overall regeneration function obtained in this way improves the pattern in the respect just mentioned, but with a signal of the same magnitude at 180° as the signal component at 90°, i.e. -10 dB from the maximum at 0°. produce components.

This "backlobe" can be removed with a simple modification.

When considering the transmission format of TΣ and T△, if T△ and T
If the T signal is attenuated by the square root of 1/2, but TΣ is not changed, and is decoded by the conjugate function of φ, the combined overall transmission function is as follows.

This overall regeneration function has no phase shift and is zero at 180°,
The amplitude at 90° results in a reproduced signal that is 6 dB down from its maximum value at 0°.

Of course, the same result can be obtained with a suitable partial mixing of the aforementioned signals TL and TR and a similar attenuation of TT.

If the basic pair of transmitted signals is transformed in this way, the third
The operation of devices without channel availability is clearly impaired.

Therefore, recorded or broadcast sets of transmitted signals will normally not include this modification.

Preferably, a modified set of transmitted signals is generated at a decoder from a signal that is unmodified when recorded or broadcast.

The overall transmission or regeneration equation described above resulting from the modified transmission signal has the coefficients of the equation shown in square brackets, which may also be written as:

It can be seen that this is in the same form as the coefficients in equation (5) for the three untransformed transmission signals, and each overall reproduced signal can be expressed as follows.

Here, m is the square of the damping coefficient used to form the deformed TΔ and TT.

Some variation in the details of the reproduction characteristics can be obtained by selecting m.

When m is changed in the range of 0.5 to 1.0, the above
The ``backlobe'' is reintroduced, but the ``90° separation'' is simultaneously improved as previously shown numerically.

When m is the intermediate value of 0.707, "90° separation" is 7.66
dB and the level of the "back lobe" is 28.3 dB below the maximum value at 0°.

The selection of the constant m therefore involves a transformation of the desired pattern characteristics, which cannot be evaluated with respect to the psychoacoustic effect on the listener and which is within the range of 0.5 to 1.0 for a three-channel decoder. Preferably, there is a means for the user to adjust the factor m defined above.

If a factor √m is introduced into the transmitted signal at the decoder and a conjugate function is immediately applied to it for decoding, then of course the latter also contains a factor 5, but the two sequential attenuations by only a factor √m are not transformed. It may be replaced by a single attenuation by only a factor m, such as by linked attenuator potentiometers on the input side for the signals TΔ and TT, allowing the user to set the value of m to 0.5 and 1.0. It is recognized that one can choose between.

In order to further increase the "contrast" between the reproduction amplitude of the sound source signal from a loudspeaker located at a position corresponding to the original position of the sound source and its reproduction amplitude from other loudspeakers, i.e. (5)
The same general principle can be used when adding an additional fourth channel to sharpen the overall reproduced signal function of the equation.

Adding a fourth channel to the above-mentioned third channel satisfying the stated standards consists in forming the transmission signal TQ (with positive or negative exponent) as follows.

Such a function converts the outputs of two 4-microphones (each with one dipole pattern out of phase with the other) rotated by 45° relative to the output of one of the 4 microphones. This can be generated by shifting the phase by 90° and integrating them.

Alternatively (or as a supplement), mixing circuits obtained by suitable variations of these previously mentioned may be used.

An attenuation factor equal to the square root of the constant m can be applied to the transmission signals TT and TQ in the formation of these auxiliary transmission signals and in the conjugate function decoding.

The format of the entire reproduced signal in this case is as follows.

The positional relationship of the sound source signal components in each reproduced signal is the same as in the two-channel case, except for the relatively small point described later.

However, as we now know, 90° separation has been greatly improved.

The effects of changing m in the case of adding a fourth channel are generally similar in broad terms to those of the corresponding changes in the three-channel case described above.

For a value of 0.33 for m, the pattern is heart-shaped, but at 90° away from the maximum at 0°, 9d
It is B.

For a value of 0.5 for m, the zero at 180° is retained, but a very small backlobe (-23.9 dB) appears at adjacent locations, and 12.9 dB at 90° away.
It is 6dB.

For a value of m of 1.0, zero is a point moved by 90° in both directions and 18
Appears at 0°, but approximately 1 in each direction from the maximum value of 0°
At 30 degrees, the magnitude is noticeable (-11.3 dB).

Unlike the three-channel case, the insertion of the coefficient m does not affect the two fundamental transmission channels, so it can be inserted either in the encoding device, in the decoding device, or in a combination of both.

The coefficients involved in the selection of m by the listener are generally similar to those in the three-channel case, and the listener can adjust this coefficient in the range 0.33 to 1.0 if he so desires. obtain.

For this purpose, TT and TQ are transmitted without attenuation of the m coefficients which are attenuated immediately before decoding and are then decoded by use of the conjugate function rematrixing described above.

However, a further advantageous use of the effect resulting from the variation of the coefficient m in the previous equation (7) is to be made with respect to reducing the frequency range requirements of the two auxiliary channels without impairing the reproduction.

FIG. 11 shows all the encoding, transmission, and decoding equipment used in the 4-channel multiplexing of the present invention, where the input signal sound sources are Sθ1, Sθ2, etc., and the output or reproduced signals are Pφ1, Pφ2, etc. It is indicated by
Each input signal and each output signal are treated as having the same azimuthal direction.

(The azimuth angle is in this case measured counterclockwise from a reference direction to the listener's right, in order to facilitate the display of the actual bilateral symmetry inherent in the matrix, thus appearing here in the display format conventionally used.) plus an angle of π/2.

) To form the transmission channel signal TΣ, all of the source signals are simply mixed additively without changing the phase.

To form the transmission channel signal T△, each source signal undergoes a phase lag for each of its frequency components with respect to the corresponding TΣ equal to the azimuth (actual or total) of the source, without changing the amplitude. The phase is shifted as follows.

The auxiliary channel signal TT is formed similarly to the TΔ signal, except that the direction of phase shift is reversed.

The fourth channel signal TQ is formed in the same way as TΔ or TT, except that each phase shift angle is doubled.

Prior to transmission, the TQ and TT signals are filtered with a bandpass filter to limit their content to only intermediate range frequencies, such as from 130 Hz to 3 KHz.

Since bandpass filtering of the auxiliary channel also inherently introduces a phase shift in the passband, phase equalizers are used on the TΔ and TΣ channels to maintain a favorable phase relationship for the signals as they are transmitted.

(The term "transmission" as used throughout refers, of course, to the variety of recordings or signal storage used for later playback, such as phonograph records or tapes, as well as the immediate transmission used in media such as FM broadcasting. ).

It will be appreciated that the fundamental signal pair TR and TL or TΣ and TΔ incorporates each source signal at the same azimuthal phase that differs from the reference phase by a phase angle equal to the phase of the reference and the azimuth of the source.

This is most obvious in the case of TΣ and TΔ, where one channel carries only the reference phase component and the other only the same-phase component.

However, although linear operations are required to separate these distinct signal components, other forms of transmission pairs carry exactly the same signal information.

The third channel TT incorporates each signal with a relative phase equal to and opposite to the same phase and the fourth signal TQ except that the phase differs from the reference phase by twice the phase angle for the corresponding signal source. It is formed in the same way as TΔ and TT.

The generalized illustration of FIG. 11 is of course applicable to broadcasting, recording or any other audio reproduction medium.

This is particularly advantageous when the required broadcast bandwidth is an issue, since the limited frequency range of the auxiliary channel allows for reduced bandwidth requirements compared to a 4-channel device; In the case of a channel device, each of the four transmission channels is used for the reproduction signal at the position assigned to that channel.

The utility of the four encoded channels in the various recording and FM broadcasting schemes previously proposed for the latter form of transmission is clear and the figures show:
It will be appreciated that the direct feeding of the transmission channel to the decoding part is shown very schematically, and the direct connection usually refers to recording or FM broadcast production and playback using a multiplexing device for the auxiliary channel.

(Channels TΣ and TΔ in FIG. 11 would normally replace TL and TR in a recording device, and it will be clear that this format is more suitable for this device.

) In the preferred use of the apparatus of FIG. 11 for FM broadcasting, the two fundamental channels are transmitted as the sum and difference signals of a conventional mono-compatible stereo broadcast.

The auxiliary signals are alternated at a sampling frequency of 9.5 KHz and multiplexed together as a composite modulation of 38 KHz carriers in quadrature as known in the art for quadruple FM broadcast on the same loudspeaker, but with limited bandwidth. This was objected to because of the requirements.

In another preferred embodiment, the four channels (using TL and TR as the basic stereo channels) are
It can be replaced with the same speaker channel in a 4-channel disc playback device described in "Audio Engineering Society Magazine" Volume 19, page 576.

The processing of each of the four transmission channels in a decoding device (e.g. associated with a PM receiver) before addition to form each reproduced signal ensures that the phase angle corresponds to the position of the loudspeaker rather than the sound source. Except that the transmission channel is paired with the processing for which it was created.

The TΣ signal again remains unchanged in phase and continues to serve as a reference phase, its suitability for this purpose being limited by the shift of the reference required by the operation of the frequency-independent phase shifters used for other channels. maintained again by the phase.

For a sound source in the same orientation as the reproduction loudspeaker, the altered outputs of the four transmission channels are all applied in phase and directly.

In the final step of creating the reproduced signal, each reproduced signal passes through an amplitude equalizer or band-attenuated p-wave filter, which is generally similar to the band-pass filter used to form the auxiliary channel signal, except for the amplitude attenuation. They are complementary.

In the frequency range that is completely unaffected by the bandpass filter, each equalizer attenuates the signal by 3 dB, and the compensation of the auxiliary channel power share is applied to the adjacent upper and lower passbands of the filter used for transmission. The same is done in the damping section.

Sharpness of the bandpass filter characteristics and the corresponding equalization filter characteristics are not required for reasons explained below.

The encoding matrix, decoding matrix and overall playback matrix (for fully transmitted mid-range frequencies) values are shown in Figure 12, corresponding to those previously used for conventional 4-track playback. The four-track playback is the same for each transmission channel with a specific speaker in a "2+2" speaker arrangement.

Encoding and decoding matrices are described at this point for convenience in understanding the scope of compatibility with current mono and stereo playback devices, as well as a description of the relationship between the two basic channels of the present invention and between the auxiliary channels. are shown separately.

Similarly, for clarity of understanding, the entire regeneration matrix is shown in an unsimplified form, and as can be seen on examination of it, the majority of terms disappear in the expansion.

As shown by the disappearance of all other coefficients in the entire reproduction matrix, the unattenuated signal frequency in TT and TQ is limited to only one loudspeaker since the other three loudspeakers are in each case at the aforementioned zero. appear.

Since all of the reproduced signals of this type are of reference phase, the reproduction of these frequencies is exactly the same as if each loudspeaker signal were transmitted directly from the corresponding sound source.

The matrix coefficients in FIG. 12 are of course the signals of frequencies attenuated by TT and TQ, or the four equally spaced signals shown in the diagram.
It cannot be applied directly to any signal encoded for any azimuthal direction other than the individual speaker positions.

Regarding the former, frequencies completely outside the filter passband (and corresponding equalization attenuation band) are only transmitted and regenerated in basic two-channel devices.

FIG. 13 shows in polar diagram form the amplitude and phase angle of reproduction for four-channel and two-channel devices as a function of the angle between each speaker position and the source position.

Furthermore, the figure shows in dotted lines the reproduction amplitude of a typical "transitional" frequency component, i.e. a partial attenuation (and subsequent relative rise) or attenuation region corresponding to one intermediate or fractional value of m. .

Therefore sharpness of the frequency bands transmitted in TΣ and TΔ (or TL and TR where this format is used) is not required and simple types of filters can be used without the need for sharp cutoff limits.

From Figure 13, we can see that the small “
It will be seen that the backlobe is in antiphase from the corresponding portion of the two-channel pattern, even though both patterns are the same within a 90° difference between the source orientation and the speaker orientation.

However, it can be seen that the aforementioned limit to a maximum of 90° phase difference in diametrically opposed loudspeakers holds and no unpleasant psychoacoustic effects occur.

Listening tests have shown that the reproduction of performance information utilizing band-limited channels, which is the scope of the present invention, is at least psychoacoustically acceptable to that of conventional full-bandwidth, separated four-channel equipment.

Since psychoacoustic detection of azimuth direction is unreliable at least at very few frequencies or frequencies with small wavelengths compared to the size of the human head, the substantial loss of directional information may be due to frequency limitations in the auxiliary channel. Does not occur.

It has been found that a listener's ability to judge sound sources emanating (or pretending to emanate) from between the speaker positions as being exactly the same in direction is enhanced compared to traditional four-channel playback. 4-channel playback typically applies such signals to two loudspeakers in phase and with emphasis on amplitude to approximate such sound.

Compared to conventional 4-channel stereo, in which sound image localization is synthesized based on the amplitude ratio of signals of in-phase or anti-phase components leaking from adjacent speakers, in the reproduction according to the present invention, the sound image localization is synthesized based on the amplitude ratio of signals of in-phase or anti-phase components leaking from adjacent speakers. Phase and amplitude seem psychoacoustically useful, at least to many listeners, because they allow for more precise sound localization.

It is therefore the intent of the present invention that a device such as that of FIG. 11 may be used to generate 4-track tape recordings, etc. (of course the bandwidth limitations and equivalency compensation features are omitted for that use). .

The principles of the present invention can be extended to even larger numbers of channels, while for playback in devices that cannot use auxiliary channels, two
It will be clear that the basic transmission signals TL and TR or TΣ and TΔ are retained.

In addition to such addition of an auxiliary channel for the initially mentioned two-channel device, which can be considered as "ambidextrous" in addition to the basic two channels, the principles of the present invention can be used for multiple channels, such that all such multiple channels The signals TL and TR are equated with directional "left" and "right", i.e. the signals are directly reproduced or decoded for a particular azimuth angle in the same general way as they are decoded for other speaker positions. Related.

An understanding of this aspect of the invention may be advanced by first considering the theory of the basic two-channel embodiment described above.

In the discussion up to this point, it is assumed that the azimuths θ and φ for all channels are measured from the same reference azimuth.

As already mentioned, this results in slightly different encoding functions for the "left" and "right" transmitted signals TL and TR, which remain unchanged in the generation of the reproduced signals for these oppositely located loudspeakers. be.

Angle θ for each sound source measured from the "left" direction or azimuth to TL and from the "right" direction or azimuth to TR
′, the encoding function or mixing coefficient function for both of these transmission signals can be expressed as follows.

For simplicity, it is convenient to rewrite it in the following form.

(Conventionally, this representation of a fraction is simple in cases where its existence may be strictly required; for example, (1) and (
It has already been recognized that when formula (3) is introduced from formula 2), the multiple ``2'' is simply ignored.

Using the direction in which the transmitted signal corresponds as a reference to the azimuth of the sound source, the above expression for the formation of each transmitted signal from the sound source signal can be denoted by f2(θ'), where the letters written at the bottom represent the two-channel The encoding function for each is shown.

In the same way, the decoding function f2(φ') for each transmission channel in the formation of each reproduced signal can be stated as follows.

Here, φ' is the azimuth angle of the reproduced signal with respect to the corresponding direction of the transmitted signal.

The overall or total reproduction signal function is of course unaffected by this change in the reference points used for encoding and decoding and is the same as the encoding function except for the argument α of the difference angle.

Similar encoding functions (and conjugate decoding functions) may be used for multiple transmitted signals corresponding to equally spaced azimuthal angles.

The encoding function for the three channels spaced 120° is f3(θ')=1/3+2/3 cos θ', and the decoding function for φ1 is the same (there are no complex terms in the functions).

The overall reproduction signal function is again the same as the encoding function except for the argument of the difference angle.

This entire function is the untransformed “
You'll find that compatibility is exactly what you get with a 3-channel device.

It will be appreciated that the "ambidextrous" and "equally spaced" transmission signals can be derived from each other by linear combinations of each other, i.e. by reversible methods of linear combinations that do not result in any change in the information.

For four channels or transmission signals (the first transmission signal at each 90°), each encoding function is as follows, and again the entire reproduction is performed using the previously discussed "ambidextrous" four-channel device ( (untransformed) and related to linear transformations.

It will therefore be understood that any particular matrix formula as applied herein or in the claims includes linear transformation variations thereof.

Of course, such encoding may be used to create separate signal four-track recordings that can be played back on any desired speaker arrangement, or the playback signals can be reproduced using the improved sound localization of the conventional separate channel playback apparatus discussed above with respect to FIG. or for other purposes that do not require mono-stereo compatibility of the type of the invention described above.

A further extension of this particular method of encoding and decoding according to the invention is made for any n channels representing equally spaced azimuthal angles according to the formula, and for any odd channel By the following formula, and for any even number of channels, by the formula below.

The greater the number of available channels, the greater the variety of special encoding and conjugate decoding functions that have been devised to implement the basic scope of the invention and the embodiments described above are particularly suitable for transmission channels. When the number is 2 or more, it is only shown in the simplest way.

More complex functions for use in coefficient factors of various components, somewhat similar to the single coefficient m discussed earlier, can be devised as well as in the form of linear combinations.

Accordingly, many modifications of the specific embodiments of the invention chosen for illustration and description herein will readily occur to those skilled in the art.

For example, the interchange of j and -j in arbitrary coding and decoding is an exemplary obvious equivalent, and others like it may be devised.

The protection afforded to this invention should extend to all uses as limited in the claims and their equivalents.

[Brief explanation of the drawing]

Fig. 1 is a schematic illustration of one type of 4-channel sound reproduction device, Fig. 2 is a similar illustration of a different type of 4-channel sound reproduction device, and Fig. 3 is a similar illustration but with a sound source. Unlike the previous diagram, the signal and playback position do not correspond.
FIG. 4 is a schematic diagram showing a certain angular relationship used in the present invention, FIG. 5 is a phase diagram of a transmission signal formed by the present invention,
FIG. 6 is a block diagram of an exemplary single signal encoder embodying the present invention; FIG. 7 is a block diagram of a "universal" encoder for multiple audio signals embodying the present invention; FIG.
The figure is a block diagram of one type of decoder combined with the present invention, FIG. 9 is a block diagram of another type of decoder of the present invention,
FIG. 10 is a block diagram of an adapter circuit for using a conventional stereo signal in the decoder of FIG. 9, FIG. 11 is a block diagram of the entire 4-4-4 mixing and transmitting device according to the present invention, and FIG. Fig. 11 shows a matrix representation of encoding, decoding, and playback of the method, and Fig. 13 shows the basic 2-channel transmission device and 4-channel transmission device, with each direction of playback and the direction from the listener as the origin. FIG. 3 is a polar diagram showing the amplitude and phase of the reproduction of a directional acoustic signal as a function of the azimuth angle between the two, and the dotted line showing the characteristics described below. In the figure, 20, 203 are microphones, 22, 22a
is the speaker, 23 is the listener, 24.24a is the encoder, 26.26a is the two transmitted signals, 28.28a is the decoder, α is the angle between the microphone and the speaker, θ and φ are respectively relative to the horizontal intermediate position Azimuth angle of microphone and speaker, L, R, F, B, LF, LB. RF and RB are respectively left, right, front, rear, left front, left rear, right front, right rear, 30 is 9
0° phase divider, 32.34 are sine and cosine potentiometers, 36.38 are combiners, TL and TR are transmission signals, 40.42 are polarity dividers and sine and cosine potentiometers, respectively, 44, 46, 50 ,54.56.62
48 and 58 are phase shifters, 52 and 60 are reference phase shifters, 70 and 72 are phase shifters, and 74 is a fixed mixing network.
76 is a fixed output terminal, 80 is a total and difference circuit, 82,
84 is a reference phase and a reference phase shift, respectively, 86 and 88 are fixed voltage dividers, 90 is a synthesizer, 91 is a reference phase channel,
92 and 94 are phase dividers, Sθ,...Sθ4 are input signals, Pφ1...Pφ4 are output or reproduction signals, respectively.
TΣ and TΔ represent the total and difference signals, respectively, and TT and TQ represent the third and fourth transmission signals, respectively.

Claims (1)

[Claims] 1. A direction characterized in that a transmission signal is generated by matrixing a plurality of sound source signals from different azimuths θk or (here, Sk is the k-th sound source signal) by a pair of equations. How to generate sexual audio information. 2. Generate a transmission signal by matrixing a plurality of sound source signals from different azimuth angles θk or (where Sk is the k-th sound source signal) by a pair of equations, or (where φi is the reference position and the reproduction position). A method for transmitting directional audio information, characterized in that the audio information is reproduced by rematrixing according to the formula of (azimuth angle). 3. An audio recording method comprising a transmission signal formed by the method according to claim 1.