DE2150168A1 - Multi-channel conversion system - Google Patents

Description

PATENT LAWYERS

DIPL.-ING. GÜNTHER KOCH DR. TINO HAIBACH

8 MUNICH 2, ⁷ * ° ^{kt #}

OUR MARK: J ^ QQ ^ _ Fk / Νβ

Industrial Patent Development Corp. Philadelphia, Pa. 19102 / USA

Multi-channel conversion system

The invention relates to a multi-channel conversion system and in particular to a system for encoding a plurality of audio frequency information sources into two audio frequency information sources or to decode the encoded information to create multiple sources of information subsequent playback as multi-channel audio frequency information.

The two-channel stereo method of recording and eventual Reproduction of audio frequency information, such as music, gives a good left-right characteristic for the Listeners. A two-channel stereo system provides a listener when properly recorded and reproduced using from correctly spaced speakers a complete sound panorama from left to right. Farther such a system gives a minimal depth perception (from front to back),

209817/0842

It has already been suggested, the sound depth concept by feeding the audio frequency information over more than two channels. For example, it is known that four Microphones arranged in suitable places, which are connected via four amplification systems and loudspeakers, the arranged around a listener (two in front of the listener and two behind him), the listener, compared to a common one Stereo system give the much heightened impression that he is sitting in a concert hall. Such a thing Commonly used direct system provides the listener with a conventional two-channel stereo reproduction through the two front speakers, while the rear speakers reproduce the concert hall environment (reverberation) as it passes through appropriately arranged microphones is detected.

The last-described system is impractical because it reproduces the sound instantaneously through a variety of amplification systems and speakers required. Furthermore, it cannot be used for radio broadcasting because existing frequency-modulated broadcast multiplex systems are only designed for two channels. Furthermore, there is such a thing The system is impractical because many people are able to listen to music using common recording media, such as tape or vinyl listen. The existing technology only results in the recording and reproduction of two information channels a record. Four information channels on a magnetic tape are technically feasible, but would be larger Require modifications to all existing magnetic tape players such as new tape head designs to be required. Furthermore, a four-channel band would not solve the problems encountered with four-channel broadcast systems given are.

From the foregoing, it can be seen that there is a need for a system that provides multiple audio frequency information. sources in two sources for transmission over two channels

209817/0842 ./.

transforms. There is also a need for a system that convert two sources of information into a plurality of sources for playback over multiple channels. The advantage of one Such a system consists in using existing two-channel equipment and methods for broadcasting, recording and can still be used to convert audio frequency information, although the end result is in the recording and playback consists of more than two sources of information.

The term "channel" has several meanings. In general it is used to define a given path, over which audio frequency information flows. For example, a stereo system has a left and a right channel. When viewed as a whole, each channel transmits information that is created by a specific source. This source is usually the information that is actually detected by a microphone and converted into electronic signals will.

The present invention takes the form of a method of converting four sources of information into two sources and then described back in four sources as well as in the form of an apparatus for carrying out the method. She relates therefore directly to the so-called four-channel systems, which are suitable for recording and playing back music. It has It has been found that four channels provide a sufficient number of audio frequency information sources to achieve the desired Depth presentation or the presentation of the concert hall environment to evoke. Furthermore, there is one related to four channels Description of an easy-to-understand left-right and front-back relationship. It will be understood, however, that this system is not limited to four channels. For the professional It is immediately apparent how this system is used to encode and decode three or more audio frequency information can be used.

209817/084?

According to the invention, four discrete audio frequency information sources can be encoded for two-channel recording, radio transmission or for similar purposes. The four information sources can be referred to as left front (L ^,), right front (R _f ), left rear (L _r ) and right rear (R _r ) information sources. Using the circuitry described below, the separate sources are encoded into two audio frequency information sources for two-channel recording. For the sake of convenience, the coded audio frequency information sources are described as "left coded source (L ¹ )" and "right coded source R ¹ )".

According to the invention it is further provided that the coded audio frequency information sources can be decoded into an equal number of audio frequency information sources. To simplify the designation, the decoded sources are referred to as "left front decoded source (LJ)", "right front decoded source (R ¹ _f ) "," left rear "decoded source (L ¹ )" and "right rear decoded source (R ¹ )." The foregoing represent four signals decoded ^{from the two-channel recording of L 1} and R ¹ Such decoded audio frequency information sources can be relayed over four channels of audio frequency equipment to produce four separate information sources which are played through four loudspeakers located at the left front, right front, left rear and right rear of the listener. Please note that the terms left in front, right in front, left behind and can be used throughout to simplify the description and only in exemplary form. The present invention is not limited to the coding and decoding of sources for these particular locations.

It is known that the measurable or discrete separation between two signals by an instrument is determined by their relative

209817/08 4?

Amplitudes is determined. However, the present invention is based on the principle that the apparent separation between the Signals or the acoustic effect produced in a human ear is magnified when the signals are out of phase are. As a result, a listener interprets signals with essentially the same amplitude, but with different amplitudes Phases as largely separate signals. This principle is used in the present invention to a predetermined degree of separation between a desired information signal and undesired information signals in to train a channel.

Briefly stated, the invention comprises the matrixing of signals derived from four spaced-apart audio frequency sources into two coded signals. The two encoded signals can be relayed using conventional two-channel broadcast, recording and conversion equipment. Decoding is achieved by matrixing the encoded signals to produce four decoded signals, each decoded signal containing the four original information signals. The desired degree of separation between an information signal in each decoded signal and the remaining information signals of this decoded signal is achieved by the parameters of the coding and decoding matrices. When these principles are applied to the left front channel (L ¹ _f ) it can be achieved that the original left front signal (L ^) predominates over the three other original signals In the channel in which the matrixing is carried out in such a way that the amplitudes of these other signals can be reduced with respect to the left front signal. Furthermore, if the phase of the original right rear signal (R) is reversed with respect to the predominant signal, a listener will interpret the separation as if it were greater than the actual or discrete separation. Therefore, using the above concepts, it is possible to improve the sound image of the left front and right front

209817/0842

Bring channels closer together and further spread the sound of the left rear and right rear channels, simulating the way sound is heard in a concert hall. Furthermore, the achieved contributes Separation from the quality of the perception of depth and / or the imagination a concert hall environment.

From the above it can be seen that an inventive built system encodes three or more audio frequency information sources in two audio frequency information sources, wherein the two audio frequency information sources are discrete combinations of multiple sources. The two audio frequency information sources are then routed over two channels and are ultimately considered to be three or more separate sources of information recovered that are at least approximate to the multiple original sources of information. Of course you can relaying the information through the channels by recording it on a vinyl record, tape or any other other recording medium may be delayed.

According to the invention there is provided a system for coding a number of (three or more) information sources into two information sources for use in recording (on tape or vinyl), broadcast or other means of transmission and transmission created. Furthermore, a system for decoding the encoded information sources into a conventional The number of sources corresponding to the number of encoded sources for the subsequent reproduction in acoustic Sources created.

The system according to the invention can record existing two-channel stereo recordings into a multitude of audio frequency information sources for subsequent reproduction as sound information convert to simulate an overall multi-channel effect.

209817/0842

The circuit used in the system of the invention is fully compatible with existing two-channel stereo equipment, without degrading the quality of such equipment.

Further details of the system according to the invention emerge from the patent claims and the following description an exemplary embodiment of the system according to the invention.

The invention is illustrated below with reference to an embodiment of the invention shown in the drawing Systems explained in more detail.

In the drawing show:

Fig. 1 is a circuit diagram of a in the invention System used to encode four discrete audio frequency information sources into two encoded sources Circuit*

FIG. 2 is a circuit diagram which basically corresponds to the circuit diagram of FIG. 1, but with the exception that it is used to decode the encoded audio frequency information sources into four audio frequency information sources is switched.

1 shows a circuit for coding four discrete audio frequency information sources into two discrete audio frequency information sources shown, which is designated as a whole by 10. Fig. 2 basically shows the same circuit as that generally designated 10 ', but with certain switch elements are converted to the circuit 10 in a circuit for decoding two audio frequency information sources to convert into four separate audio frequency information sources for subsequent playback. The coding circuit 10

209817 / nfU? ./.

will be explained first.

The circuit 10 has input connections 12, 14, 16 and 18, which can be, for example, sockets for coaxial cables. Using the designations given above, the original four discrete audio frequency information sources are applied to one port each. Thus the left front source (L ~) is connected to port 12, the right front source (R _f ) to port 16, the left rear source (L) to input port 14 and the right rear source (R _r ) applied to the input terminal 18.

The source L _f is fed to the amplifier 20 which is an amplifier whose output is in phase with the input. For example, an emitter-Polger amplifier fulfills this point. In the same way, the audio frequency source R ~ is fed to the amplifier 22, which is also an amplifier without phase inversion. The gain of amplifiers 20 and 22 can be set to 1 so that their outputs exactly match their inputs.

The signal L is fed to the amplifier 24 and the input R is fed to the amplifier 26. The amplifiers 24 and 26 correspond to amplifiers 20 and 22 in that their output is also in phase with their input. For this purpose, the Amplifiers 24 and 26 can be emitter-Polger amplifiers «They can also have a gain of 1.

As can be seen, the inputs L and R also become the amplifier 28 and 50 respectively. The amplifiers 28 and 50 cause a phase reversal of their respective output signals. Thus the output of each amplifier 28 and 50 is 180 degrees phase shifted with respect to the associated input. For this purpose, amplifiers 28 and 50 can be linear amplifiers. The gain of amplifiers 28 and 50 can be set to one.

209817/0842 ./.

The output of the amplifier 20 is connected to the output terminal 34 via a resistor 32. In the same way, the output of the amplifier 22 is connected to the output terminal 38 via the resistor 36. For example, the output ports 34 and 38 can be receptacles for coaxial cables. As described above, the two encoded signals have been arbitrarily referred to as L ¹ and R ¹ . For this reason, the output connection 34 ″ has been designated with L ¹ and the output connection 38 with R ¹ .

The output of the amplifier 20 Jsb is also connected to the output terminal 38 through a variable resistor 40. In the same way, the output of the amplifier is connected through a variable resistor 42 to the output terminal y * t . The variable resistor 40 is set so that the output signal of the amplifier 20 is reduced to a certain fraction of this signal, which is fed to the output terminal 34 which is connected to the voltage divider consisting of a resistor 32 and a resistor 44. Likewise, the variable resistor 42 is adjusted so that the voltage level of the output of the amplifier 22 is reduced to a certain fraction of the voltage level of the output which is applied to the terminal 38 which consists of the resistor 36 and the resistor 46. The output of the amplifier 24 is connected to the output terminal 34 ^ ^,

/ existing voltage divider across the resistors 48 and 50 / and across the switch 52 connected in the coding position E. The output of the amplifier 26 is connected to the output terminal 38 via the from the Resistors 5 ^ and 56 existing voltage divider and over the switch 58 in the coding position E connected.

The output of the phase reversing amplifier 28 is via a variable resistor 60 is connected to the voltage divider consisting of resistors 16 and 56. Starting from

./. 209817/0842

the voltage divider, the output via switch 58 is on the exit to the terminal 38. The output of the amplifier 30 is connected via variable resistor 60 to the voltage divider consisting of resistors 48 and 50, Based on these voltage dividers 48-50, the output becomes out via the switch 52 to the output connection 34.

From the previous description it can be seen that the output signal L ¹ appearing at connection y \ can be written as follows:

L ¹ = A _x L _f + A ₂ R ^ A ₅ L ₁₁ -A ₄ H ₁ , (1)

Where:

λ _ size of resistor 32

A ₂ =

^A 3 ⁼

total the Resistances 32 and 34 size of Resistance 42 total the Resistances 42 and 44 size of Resistance 48 total the Resistances 4Ö and 44 size of Resistance 62

^A 4 ⁼ sum of resistors 62 and 44

In the same way, that which appears at the output terminal R ' Signal can be written as follows:

R ¹ = B ₁ RfH-B ₂ L ^ ByI ₁₁ -B ₄ L ₁ , (2)

Where:

B, = size of resistor 36 Sum of resistors 36 and 46

τ. Resistance 40 size

B,

Sum of resistors 40 and 46 size of the resistance

3 Sum of resistors 54 and 46 size of resistor 60

Sum of resistors 60 and 46

209817/0842 ./.

- li -

The two coded signals L ¹ and R ¹ are thus obtained by matrixing the four original information _{signals L f} , L, FU and R. Each coded signal contains the four information signals with the specified phases and proportional to the values of the resistances specified above.

^{The signals L 1} and R ¹ generated at terminals 5 ^ and 38 comprise two sources of encoded audio frequency information ". These two sources can be recorded, transmitted or used in other transmission devices. With regard to recording, these signal sources can be recorded on magnetic tape or on a vinyl record, as desired.

From the above it can be seen that the circuit for coding four audio frequency information sources in two audio frequency information sources is provided. The resulting two audio frequency information sources can be via a common Two-channel system for recording, radio transmission or the like can be transmitted. The importance of the way on which the sources were encoded will be seen in the following when the decoding circuit is described.

An example of such a decoding circuit is shown in • Fig. 2, this circuit 10 * being ^{intended for decoding the coded signals L 1} and R ¹ '. ^{The circuit 10 1} shown in FIG. 2 corresponds to the circuit shown in FIG. 1, with the exception that all switches have been moved from the coding position labeled E to the decoding position labeled D and that certain settings have been made on the variable resistors . The change from position E to position D opens switches 52 and 58. In addition, switches 64 and 68, which have not yet been mentioned, are closed. In the circuit according to FIG. 2, the output connections 70 and 72 are now also used,

209817/0842

which were previously bridged by switches 52 and 58. the Output ports 70 and 72 can, if so desired, Be sockets for coaxial cables.

The circuit shown in Fig. 2 decodes the encoded signals L ¹ and R ¹ to provide L * _f - _s R ^f _f -> L ' _r and R' _r signals representing the "four decoded audio frequency information signals .

The signal L ¹ is connected to the amplifier 20 via the connection 12. The output of amplifier 20 is fed to the midpoint of the voltage divider formed by resistors J2 and 44. In the same way, the coded signal R ^{1 is} fed to the amplifier 22 via the input terminal 16. The output of amplifier 22 is connected to the midpoint of the voltage divider formed by resistors J> 6 and 46.

By closing switch 64, signal L ^{f is} also fed to the input of amplifiers 28 and 24. In the same way, signal R ^{1 is} fed from terminal 16 via switch 68 to the input of amplifiers J50 and 26. The output of amplifier 24 is connected to the voltage divider formed by resistors 50 and 48. The output of the amplifier 26 is connected via the voltage divider formed by the resistors 54 and 56.

The phase inversion output of amplifier 28 is via the variable Resistor 60 connected to the voltage divider from resistors 5 ^ and 56. The phase reversal output of the amplifier 30 is via the variable resistor 62 with the Voltage divider from resistors 48 and 50 connected.

Furthermore, the output of the amplifier 20 is variable Resistor 40 connected to the voltage divider from resistors 56 and 46. The output of amplifier 22 is

209817/0842

via the variable resistor 42 to the voltage divider from resistors 32 and 44 connected.

Each of the four voltage dividers mentioned above is connected to one of the output terminals labeled 34, 38, 70 and 72. The signals at the output terminals represent four audio frequency information signals which have been decoded ^{from the coded signals L 1} and R ^1. As mentioned above, the four recovered or decoded signals are represented by the symbols L ¹ ^, L ' _r , R ¹ _f and R ¹ . When these signals are converted into audible sound, they create the depth effect of the so-called four-channel stereo system.

From the above description it can be seen that the outputs at connections 34, 38, 70 and 72 with a suitable setting of the variable resistors 40, 42, 60 and 62 can be written as follows:

^R f ^{= D} 1 ^{L + D} 2 ^R W

LI T? T »TT TJ · fc \

Where:

_c _ Resistance 32 1 ~ size

1 Sum of resistors 32 and size of resistor 42

D.
D,
£

2 Sum of resistors 42 and _ size of resistor 40

1 sum of resistors 40 and

_ Resistance 36 size

2 ~ sum of resistances 3d and

_m size of resistor 48 1 "sum of resistors 4Ö and

E « size of resistance 62

2 3um of the resistors 62 and

"Resistance 60 size

IT.

Sum of resistors 60 and size of resistor 54

2 Sum of resistors 54 and 56 /

209817/0842

The equations given above apply to the resistances in the decoder circuit.

Equations (1) through (6) are made by choosing the various Resistances simplified:

A ₂ = B ₂ = X

A. = R. - W

C ₂ = D ₁ = Y
Ep = Pi ⁵ = Z

With this choice of resistances, the equations simplify to:

L '= L ₁₁₊ XR ^ L ₁₄ -WR ₁ , (1)'

R ¹ = R ₁₁₊ XL ^ R ₁₄ -WL ₁ , (2) »

R ₁ . ¹ = R '+ YL' (4) ' ¹

L ₁ Z-L'-ZR * (5) ¹

R ₁ , ¹ = R'-ZL ¹ (6) ^!

Here, the quantities W, X, Y and Z are dimensionless positive numbers less than or equal to 1, depending on these Resistance values assigned to variables.

By suitable substitution and algebraic conversion, the output ^{signals L 1} _£ , R ¹ , L ¹ and R * can be expressed in terms of the original four separate audio frequency information input signals as follows

209817/0842

L ' _f = (l + Xy) L _f + (X + Y) R _f + (l-YW) L _r + (YW) R _r (7)

R ' _f = (l + XY) R _f + (X + Y) L _f + (l-YW) R _r + (YW) L _r (8)

L ' _r (l-XZ) L _f + (XZ) R _f + (l + ZW) L _r - (W + Z) R _r (9)

R ' _r (l-XZ) R _f + (XZ) L _f + (l + ZW) R _i> - (W + Z) L _r (10)

The coefficients W, X, Y and Z can be selected to accommodate a wide variety of fades and separations from the original Information signals in each of the four playback channels is achieved by the respective decoded signals can be controlled. The choice of coefficients, however, depends on the taste of the listener, which creates a very subjective point of view results. However, in actual experiments, it was found that the following examples of coefficients pleasant results result:

example

No coefficient

W X Y Z

The preferred embodiment is Example No. III. Using the coefficients of this embodiment the decoded signals have the following form:

L _f ^f = 1.06L _f + 0.5R _f + 0.9L _r -0.5R _r

No. No. No. No. I. II III IV 1 1 0.5 0.5 0.3 0.5 0.3 0.5 0.2 0.2 0.2 0.2 1 1 0.8 0.8

R '= l.O6R "+ O.5L ,, + O.9R -O.3L (8)'

JL ι irr

^L r ^{! = 1} ^ V ¹ ' ⁵ V ⁰ -76L _f -0.5R _f (9)'

R _r '= l.4R _r -l.3L _r + O.76R _f -O.5L _f (10)'

209817 / 08A2

- 1β -

Taking into account the fact that the decoded signals are intended to drive loudspeakers which are arranged at positions indicated by the symbols used (for example, L ¹ ^, the signal that drives a loudspeaker which is arranged from the front left of a listener , etc.) * A consideration of the above equations reveals that the dominant information signal in each channel is the signal corresponding to that channel. For example, the information signal with the greatest amplitude in the left front playback channel is the original left front information signal. The discrete separations, measurable with the aid of instruments, between this signal and the right front, left rear and right rear information signals are 6 db, 1.5 db and 11 db, respectively, the separation being the ratio of the absolute values of the amplitudes of the signals. From a psychological point of view, the above separation between the dominant information signal in each of the two front channels and the information signals corresponding to the corner locations near the corner arrangement of the dominant signal results in a blend or dissolve that brings the virtual image of the two front sound sources somewhat closer and closer brings towards the listener than the actual sound sources. This amplification is particularly pleasant and eliminates the possibility of a "visual hole" at the position of the listener. While there is a momentarily measurable 11 + db separation between the dominant information signal in each of the two front channels and the information signal corresponding to the corner diagonally opposite the corner of the dominant signal, there is considerable actual separation. The existing phase inversion, however, gives a listener the impression of a much greater apparent separation and gives him the acoustic "feeling" of a concert hall.

Further consideration of the above equations appears to show that the discrete separation of approximately 1 db between

209817/0842

the dominant rear information signal in one channel on one side and the unwanted signal on the same Side in the two rear channels would result in an undesirable mix or dissolve. Although the actual breakup is small, the phase inversion present in this case causes a wide apparent separation, which for the Listener is comfortable.

The large number of matrixing operations available standing variables both during coding and during decoding enables the implementation of a comprehensive Coefficient selection. Such a selection enables preselected actual and apparent ones to be achieved Separations between the dominant information signal in a channel and the other information signals in this Channel.

A unique advantage of the present invention is that it can be used to simulate so-called four-channel stereo playback by generating four audio frequency information sources from a standard two-channel left-right stereo recording which has not been encoded in the manner described above can be used. Using the above The four derived symbols can be derived from a two-channel recording derived composite outputs can be written as follows:

L ¹ f - L + YR R ¹ f ⁼ R + YL L ' r L-ZR R ' r ~ R-ZL

where R and L represent the two recorded signals (usually left and right) and Y is a positive number, the is less than 1.

209817/0842

Examination of the four above equations shows that there is a reduced separation in the two ^{channels labeled L 1} ^ ₁ and R ', while an increased "separation results in the two ^{channels labeled L 1} and R ^1. Thus, four simulate With the help of the decoding circuit, signals actually generated the desired effects * i.e. the sound coming from the front seems to overlap and originate from sources at a short distance, which results in a good monophonic central signal, while the sound coming from the rear comes from two well spaced sources appears to originate, as would be the case with sound generated by a concert hall environment.

Claims

209817/0842

Claims (1)

Claims 1. A method of increasing the separation between a selected information signal of a plurality of Information signals and an undesired one of the other information signals in a composite signal at one Multi-channel conversion system, marked by the step of matrixing the information signals to produce a composite signal in which the undesired Signal has a phase inverted with respect to the phase of the selected information signal. 2. The method according to claim 1, characterized by the steps of matrixing the information signals to produce two encoded signals which are recorded broadcast or otherwise can be broadcast and then retrieved, and the matrixing of the two recovered coded signals for generating at least one decoded signal, wherein the matrixing steps are carried out so that the phase of the undesired information signal with respect to the phase of the selected information signal is inverted. 5. Method according to Claim 2, characterized in that four information signals are used and in that the two recovered coded signals are matrixed to produce the four decoded signals, wherein the matrixing steps take place in such a way that a different of the four information signals in each decoded signal has an amplitude greater than the amplitude of the rest Information signals in this decoded signal. 209817/0842 4. The method according to claim 3 »characterized in that the four signals are the signals L _f , L _p , R _r . and R- are derived from four spaced-apart sources, the signals being matrixed such that two encoded signals L ¹ and R ¹ result which are given by the equations L ¹ = K ₁ (L _f + XR _f + L _r -WR _r ) and R ¹ = K ₂ (R _f + XL _f + R _f ^ WXL _r ) are defined, where K. and K _{2 are} proportionality constants and 0.25 <W <1
0.1 f. X SO.6
0.1 <Y <0.4
0.7 <Z <1 Method according to Claim 4, characterized by the matrixing of the two coded signals in such a way that the decoded signals L ¹ _f , L ' _r , R ^f _f and R' ₂ result, these signals being given by the equations L ₁ , ¹ = L ¹ + YR ¹ R _f 'a R' + YL ' L _r 'β L ¹ -ZR ¹ R '= R ¹ -ZL ¹ are defined. 6. The method according to claim 5, characterized in that W = 0.5 X = OO Y = 0.2 Z = 0.8 209817 / 08Λ2. 7. Device for performing the method according to one of claims 4 to 6, characterized in that a first arrangement for matrixing the audio frequency signals to generate a first encoded signal L ¹ = K ₁ (L- + XR- + L -WR) and one second arrangement for matrixing the audio frequency signals to generate a second coded signal R ¹ = K ₂ (R ^ + XL ^ + R ^ WL) is provided, where K, and Kp are proportionality constants and o.i «x <o.6 0.255 W 51
is. 8. Apparatus according to claim 7 * marked thereby draws that X = OO and W = 0.5. 9. Apparatus according to claim 8, characterized in that a third arrangement for matrixing the first coded signal L ¹ with the second coded signal R ¹ for generating four coded signals L ^f _f , L ' _r , R ¹ ^ ₁ and R * is provided, where V - L ¹ + YR * ν = R ' + YL ' V - L ' -ZR ¹ R ' -XL ¹ 0.1 5 ? 0.4 o.7i Z 5 ι and where is. 10. The device according to claim 9, characterized geke η η - ze rejects that Y = 0.2 and Z = 0.8. 209817/0842