United States Patent c191
`Klank et al.
`
`(11] Patent Number:
`(45] Date of Patent:
`
`4,821,260
`Apr. 11, 1989
`
`(75]
`
`(54] TRANSMISSION SYSTEM
`Inventors: Otto Klank, Lehrte; Ernst Schroder,
`Hanover; Walter Voessing,
`Wennigsen, all of Fed. Rep. of
`Germany
`(73] Assignee: Deutsche Thomson-Brandt GmbH,
`Villingen-Schwenningen, Fed. Rep.
`of Germany
`(21] Appl. No.: 135,511
`(22] Filed:
`Dec. 16, 1987
`Foreign Application Priority Data
`(30]
`Dec. 17, 1986 [DE] Fed. Rep. of Germany ....... 3642982
`Int. CJ,4 ......................... H04J 3/00; H04L 27/20
`(51]
`(52] U.S. CJ .................................... 370/77; 370/110.4;
`375/67; 381/2
`(58] Field of Search .................... 370/77, 110.1, 110.2,
`370/110.3, 118, 110.4; 375/67; 381/2
`References Cited
`U.S. PATENT DOCUMENTS
`4,554,658 11/1985 Marte et al. ...................... 370/110.2
`4,750,167 6/1988 Meyer ................................... 370/77
`
`(56]
`
`FOREIGN PATENT DOCUMENTS
`0073979 8/1982 European Pat. Off ..
`0167849 6/1985 European Pat. Off ..
`
`0193143 2/1986 European Pat. Off ..
`3506912 8/1986 Fed. Rep. of Germany .
`3610398 10/1987 Fed. Rep. of Germany .
`
`OTHER PUBLICATIONS
`"An Introduction to Error-Correcting Codes'', Shu
`Lin, Prentice-Hall, Inc., Englewood Cliffs, N.J., pp.
`130-135.
`"Digital Sound Service for Direct Broadcasting Satel(cid:173)
`lites'', The Federal Minister of Research and Technol·
`ogy, pp. 76-79 and 106-108.
`
`Primary Examiner-Robert L. Griffin
`Assistant Examiner-Frank M. Scutch, III
`Attorney, Agent, or Firm-Spencer & Frank
`
`ABSTRACT
`(57]
`A system for transmitting and receiving digitalized
`audio signals, particularly via satellites, wherein data
`sequences are arranged in timely succession within
`frames, wherein before transmission, the digitalized
`audio signal is converted to a digital signal representing
`the momentary frequency spectrum and, during subse(cid:173)
`quent coding of the digital audio signal to be transmit·
`ted, portions of this transformed signal are given differ(cid:173)
`ent weights on the basis of psychoacoustic laws with
`respect to the accuracy of their representation.
`
`12 CJaims, 4 Drawing Sheets
`
`ADC
`
`TRANSFORMING
`CIRCUIT J
`
`aitl
`
`17
`
`f2
`20
`18
`dill t~~~~~5RMINGDECODER
`CIRCUIT
`
`bit)
`
`IPR2016-01710
`UNIFIED EX1025
`
`

`
`U.S. Patent Apr.11, 1989
`
`Sheet 1of4
`
`4,821,260
`
`c(t)
`
`S(tl
`
`ADC
`
`TRANSFORMING
`CIRCUIT 3
`
`4
`
`14
`
`15
`
`17
`
`19
`
`MUX
`
`e(t)
`
`11
`
`13
`
`t2
`
`21
`
`S1
`
`7
`
`6
`
`DE MUX
`
`6
`
`17 .
`
`17
`
`20
`18
`INVERSE
`DECODER
`d(t) TRANSFORMING
`CIRCUIT
`
`b(t}
`
`.J.
`
`S2
`22
`
`DAC
`
`Fig.1
`
`

`
`U.S. Patent
`
`Apr. 11, 1989
`
`Sheet 2 of 4
`
`4,821,260
`
`Kr-2-5~r------~-zr---------1
`M U X ----~--- MG---~
`INTERMEDIATE ·
`. FRAME MULTIPLIER
`
`e(t) 1
`e(tl 2
`eltl 3
`e(tl 4
`
`411
`!PT
`I
`PULSE SHAPING.
`LOW-PASS
`FILTER
`
`INTERMEDIATE
`FRAME MULTIPLEXER
`
`Fig.2
`
`43
`
`SYNCHR NIZER
`CLOCK PULSE
`DEMOD-
`DERIVATION
`ULATOR
`1.1 !
`!
`
`ERROR
`CORRECTION
`
`DEMUX
`
`PROGRAM SELECTOR
`SI. SCALE FACTOR
`
`ERROR
`CORRECTION
`
`Fig.3 CARRIER
`REGENERATOR
`
`CHANNEL
`SELECTOR
`
`43
`
`Fig.7
`
`

`
`U.S. Patent Apr.11, 1989
`
`Sheet 3 of 4
`
`4,821,260
`
`a(t)
`
`ADC
`
`WINDOWING STAGE
`
`PREPROCESSOR
`
`TIME
`
`FREQUENCY
`
`CODER
`
`TRANSMITTER
`
`TRANSMISSION LINK
`
`RECEIVER
`
`DECODER
`
`FREQUENCY
`
`TIME
`
`COMBINING STAGE
`
`DAC
`
`62
`
`63
`
`65
`
`66
`
`67
`
`68
`
`69
`
`70
`
`71
`
`72
`
`20
`
`b( t)
`
`Fig.4
`
`

`
`U.S. Patent Apr.11, 1989
`
`Sheet 4 of 4
`
`4,821,260
`
`ADC
`
`TRANSFORMING
`CIRCUIT
`
`k(t)
`
`a(t)
`
`c1(t)
`
`c2(t)
`
`Fig.5
`RECEIVER
`
`b(t)
`DAC
`
`l(t)
`DAC
`
`Fig.6
`
`PRECODER
`
`MUX
`
`80
`
`81
`
`TRANSMITTER
`
`ERROR CORREC;.
`TION CIRCUIT
`
`INVERSE TRANS(cid:173)
`FORMING CIRCUIT
`
`1---iDEMUX
`
`87
`
`'-------88
`
`CON
`
`CON
`
`INVERSE TRANS(cid:173)
`FORMING CIRCUIT
`
`

`
`1
`
`TRANSMISSION SYSTEM
`
`4,821,260
`
`2
`FIG. 5 is a partial block circuit diagram of a further
`embodiment of a transmitting station according to the
`invention for a single channel.
`FIG. 6 is a block circuit diagram of a tuner or receiv(cid:173)
`ing station for use with the station of FIG. 5.
`FIG. 7 is a further block circuit diagram showing a
`receiver for the station of FIG. 6.
`
`10
`
`5
`
`BACKGROUND OF THE INVENTION
`The present invention relates to a system for transmit-
`ting and receiving digitalized audio signals, particularly
`via satellites, in which data sequences are arranged in
`time succession within frames.
`The
`informational brochure, entitled "Digitaler
`Horfunk
`iiber Rundfunksatelliten" [Digital Audio
`Transmissions By Way Of Radio Satellites], 2nd revised
`edition, published by the German Federal Ministry of
`Research and Technology (BMFT), Heinemannstrasse
`2, D-5300 Bonn 2, Editor-Grad. Eng. P. Treytel, dis- 15
`closes a frame format in which 16 stereo programs
`available in parallel form are brought into a suitable
`serial frame format (page 106, paragraph 3.4, Frame
`Format). FIG. 47 at page 77 of that brochure shows the
`block circuit diagram for the entire multiplex and mod- 20
`ulation device for the Usingen radio transmission sta(cid:173)
`tion. At page 78, under Point 1.3, technical data are
`provided in which the number of audio channels is
`given as 16 stereo channels or 32 monaural channels,
`respectively.
`
`DETAILED DESCRIPTION OF THE
`PREFERRED EMBODIMENTS
`FIG. 1 is a general block circuit diagram of a commu(cid:173)
`nications channel for an audio signal a(t). The analog
`audio signal a(t) is converted into a digital signal in an
`analog/digital converter 1, hereinafter called an ADC
`(analog/digital converter). The digital signal travels
`through multi-conductor lines 2 to a code transforma-
`tion circuit or device 3, hereinafter called a transform(cid:173)
`ing member. Transforming member 3 transforms the
`digital audio signal from the time domain to a frequency
`domain, i.e. , transforming member 3 converts the digi(cid:173)
`talized audio signal to a signal representing the momen-
`tary frequency spectrum of the audio signal. Addition(cid:173)
`ally, in the transforming member 3, portions of the
`signal representing
`the momentary spectrum are
`25 weighted with respect to the accuracy of their represen(cid:173)
`tation on the basis of known psychoacoustic laws. Digi(cid:173)
`tal audio signals in the frequency domain are called
`spectral values. These spectral values are forwarded via
`a line 4 to a code converter 5. Code converter 5 pro(cid:173)
`vides the digital spectral values with additional bits, i.e.,
`parity bits. This is called later the external error code.
`Encoding as well as decoding proposals are disclosed in
`SHU LIN, "An Introduction to Error-Correcting
`Codes", Prentice Hall, Inc. (Englewood Cliffs, N.J.)
`page 131 to 135. This advance coding permits error
`correction in the receiving station.
`From code converter 5, the spectral values, which
`have been provided with redundance, are sent through
`a line 6 to one input of a switch 7, which can selectively
`switch its output between two input signal paths. In
`particular, either the digital audio signal directly pro-
`vided at the output of the digital/analog converter 1 on
`line 2, or the transformed digital audio signal from code
`converter 5 on line 7 is selectively provided at the out(cid:173)
`put of switch 7. The position of switch 7 is controlled
`via a control line 8 by means of a control signal c(t).
`Switch 7, depending on its position, forwards either the
`digitalized audio signals or the digitalized transformed
`audio signals via a line 9 to a multiplexer 10. In multi(cid:173)
`plexer 10 , the signal c(t) and possibly further audio
`signals are added to the audio signal at the output on
`line 9 to form a multiplex signal e(t). Signal c(t) informs
`a receiving station whether the digitalized audio signal
`or the digitalized transformed audio signal is being
`55 transmitted. The multiplex signal e(t) at the output of
`multiplexer 10 is fed via a line 11 to a transmitter 12.
`In transmitter 12, the multiplex signal e(t) is modu-
`lated to provide a radio signal. Transmitter 12 sends the
`radio signal, hereinafter called the RF signal, over a
`communications transmission channel or link 13. Trans(cid:173)
`mission channel 13 may be an HF radio link, a magnetic
`tape, a memory or the like. During transmission, the RF
`signal may be influenced by an interference signal S(t).
`From transmission channel 13, the RF signal travels to
`a receiver 14. Receiver 14 demodulates the RF signal
`and feeds a thus received modulation signal via a line 15
`to a demultiplexer 16. Demultiplexer 16 separates the
`modulation signal into a base band and a control signal
`
`SUMMARY OF THE INVENTION
`It is an object of the present invention to improve the
`transmission and reception of digital audio signal data in
`a system in which transmitted data sequences are ar- 30
`ranged in timely succession within frames. In particular,
`the digitalized data transmitted is to be reduced, prefer(cid:173)
`ably on a selective basis within a channel or channels of
`the system.
`The above object is generally achieved according to 35
`the present invention in that in a system for transmitting
`and receiving digitalized audio signals, particularly for
`satellites, of the type discussed above wherein data
`sequences are arranged in timely succession within
`frames for transmission, the digitalized audio signal is 40
`transformed or converted before transmission to a digi-
`tal signal which represents the momentary frequency
`spectrum of the audio signal and, during subsequent
`coding of this transformed digital audio signal to be
`transmitted, parts of this transferred signal are given 45
`different weightings on the basis of psychoacoustic laws
`with respect to the accuracy with which they are ulti(cid:173)
`mately reproduced at the output of a receiver.
`Various embodiments and features of the transmitting
`and receiving system according to the invention, 50
`whereby transformed and/or non-transformed digital
`audio signals are selectively transmitted in a communi(cid:173)
`cations channel and wherein error correction codes,
`internal and/or external to the coded transformed sig(cid:173)
`nals, are disclosed.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`FIG. 1 is, a general block circuit diagram of a com(cid:173)
`munications channel for a system according to the in-
`vention.
`FIG. 2 is the block circuit diagram of a multi-channel
`transmitting station for a system according to the inven(cid:173)
`tion.
`FIG. 3 is the block circuit diagram of a receiving
`station for use with the transmitting station of FIG. 2. 65
`FIG. 4 is a block diagram illustrating the basic
`method for reducing data in digitalized audio signals in
`a communication according to the invention.
`
`60
`
`

`
`4,821,260
`
`3
`d(t). If the signal in transmission channel 13 contains no
`interference, c(t) and d(t) are identical.
`From demultiplexer 16, the base band travels via a
`line 17 both to a second code converter 18 and to one
`input of a switch 22. Code converter 18 checks the 5
`incoming data and separates the additional bits pro(cid:173)
`vided by the code converter 5 from the information
`signal. If the transmitted audio signal contains interfer(cid:173)
`ence S(t), code converter 18 attempts to reconstruct the
`audio signal which contains the interference S(t). Sim- 10
`ple interference errors are detected in this code con(cid:173)
`verter 18 and the transformed digital audio signal is
`reconstructed. Difficult interference errors cannot be
`reconstructed and are covered up by a concealment
`circuit. Code converter 18 corrects errors up to a cer- 15
`tain number per block, e.g. up to two; beyond this and
`reliably up to a limit of five errors, error detection with
`subsequent concealment (averaging over several sam(cid:173)
`pling values) performed. A code converter which can
`be advantageously used for code converters of blocks 5 20
`and 18 is disclosed in European Pat. No. 73 979 B 1 of
`Dec. 12th, 1985.
`The spectral values of the digital audio signal present
`at the output of code converter 18 are transferred via a
`line 19 to an inverse transforming member 20. Inverse 25
`transforming member 20 transforms the weighted spec(cid:173)
`tral values from the frequency domain back to the time
`domain. Inverse transforming member 20 feeds the
`inversely transformed digital audio signals via a line 21
`to a second input of switch 22. The control signal d(t) 30
`from demultiplexer 16 is fed to the switch 22 to control
`whether a transmitted and received audio signal is to be
`picked up directly from the multiplexer 16 or whether
`the transmitted and received audio signal is to be picked
`up after having been sent through transforming member 35
`20. At the output of switch 22, the audio signal is pres(cid:173)
`ent in its basic digital form. From the output of switch
`22, the audio signal travels via a line 23 to a digital(cid:173)
`/analog converter 24, hereinafter called DAC (digital(cid:173)
`/analog converter). The analog audio signal b(t) is 40
`available at the output of DAC 24.
`The analog audio signal b(t) at the output ofDAC 24
`is identical with the original analog audio signal a(t) if
`the transformation in transforming member 3 and the
`inverse transformation in inverse transforming member 45
`20 have been omitted from the transmission path and if
`the signal in transmission channel 13 contains no inter(cid:173)
`ference S(t). If, however, the audio signal a(t) does pass
`through the transformation and inverse transformation
`circuits 3 and 20, respectively, along its transmission 50
`path, then transformation circuit 3 suppresses portions
`of the digital audio signal which are irrelevant and
`redundant for the listener, and thus a(t) and b(t) are
`different.
`FIG. 2 is a block circuit diagram of a complete multi- 55
`plex and modulation device (ZT) for a radio station, i.e.,
`corresponding to a transmitter 12. One channel unit 25
`is available for four monaural signals e(t) (see FIG. 1).
`Monaural signals e(t) are digitalized audio signals or
`transformed digitalized audio signals. A channel unit 25 60
`processes either two digitalized audio signals e(t), one
`digital audio signal and two transformed digital audio
`signals e(t), or four transformed digitalized audio signals
`e(t). After processing of signals e(t), two 14-bit code
`words for four monaural channels or two stereo chan- 65
`nels are available at the outputs of each channel unit 25.
`In the illustrated embodiment, the channel portion
`(KT) of the system includes sixteen channel units 25,
`
`4
`with each group of eight channel units 25 forming one
`channel portion. The data from each group of eight
`channel units 25 are transferred via a respective data bus
`26 to a respective block coder 27. In each of the block
`coders 27, the audio signals are coded in BCH (binary
`coded hexadecimal), and the audio signals coming from
`the respective group of eight channel units 25 are multi(cid:173)
`plexed. Via respective lines 28, a scale factor is transmit(cid:173)
`ted to respective ZI (intermediate) frame multiplexers
`29 for the respective block coders 27. The scale factor
`has a function similar to that for commanding and as(cid:173)
`signs sample data words to certain level regions. (See in
`this connection page 19 of the above-cited informa(cid:173)
`tional brochure.) Via a bus 30, V (verification) bits are
`passed from channel units 25 to an SD frame multi(cid:173)
`plexer 31 which processes special services (SD). These
`special services indicate whether the signal e(t) is part of
`an opera broadcast, sports news, general news, etc.
`Additionally, in the system according to the invention,
`the control signal c(t) is handled as a special service bit.
`In a scrambler/main frame multiplexer 32, to which the
`output signals of the block coders 27 and the SD frame
`multiplexer 31 are fed, successive double blocks are
`scrambled, provided with a sync word, and combined
`to form a main frame A' or B', respectively. Block cod(cid:173)
`ers 27, ZI frame multiplexers 29, SD frame multiplexer
`31 and main frame multiplexer 32 form a multiplexing
`device (MUX).
`The three output lines 33 to 35 emanating from main
`frame multiplexer 32 for a frame A', a frame B' and a
`clock pulse signal Tare fed to a difference coder 36. Via
`lines 37 to 39, signals for frames A", B" and clock pulse
`signal T' are fed to two pulse shaping lowpass filter
`units 40. In the pulse shaping portions of units 40 , uni(cid:173)
`polar NRZ signals are converted into bipolar signals
`and the lowpass filter portions of units 40 serve to limit
`these signals in bandwidth. These bandwidth limited
`band signals serve as modulation signals for a 4-PSK
`modulator 41. PSK stands for phase shift keying (En(cid:173)
`glish) or "Phasenumtastung" (German). A quartz oscil(cid:173)
`lator (not shown) produces the required 70 MHz carrier
`for the modulator 41. An IF output filter (not shown) of
`the 4-PSK modulator 41 suppresses the mixed products
`generated during modulation. Thus, a PSK signal mod(cid:173)
`ulated with 70 MHz is present at output 42 of the PSK
`modulator. The modulation device (MG) thus includes
`the difference coder 36, the pulse shaping lowpass filter
`units 40 and the PSK modulator 41.
`FIG. 3 shows a receiver 14 for 4-PSK modulated
`signals. Such 4-PSK modulated signals are present at
`input 43. A channel selector 44 selects a receiving fre(cid:173)
`quency band, for example the band with the 70 MHz
`carrier frequency, and forwards the signal to an ampli(cid:173)
`fier and bandwidth limiter 45. In a demodulator 47, the
`carrier signal and the useful signal are separated from
`one another. For coherent demodulation, a carrier re(cid:173)
`generator 46 furnishes a carrier signal which has been
`regenerated from the RF signal. The channel selector
`44, the amplifier and bandwidth limiter 45, the carrier
`regenerator 46 and the demodulator 47 make up the
`analog portion (AT) of the receiver 14.
`Two bit streams corresponding to the two bit streams
`that were combined in the PSK modulator are again
`available at the output of the analog portion AT , i.e., at
`the output of the demodulator 47. In a clock pulse deri(cid:173)
`vation circuit 48, a clock pulse is generated which is
`synchronous with the given bit sequence in the bit
`streams. After the clock pulse derivation, the bit
`
`

`
`5
`streams, which now also contain a difference code, are
`present at the output but, depending on the latching
`state of the carrier oscillator, they may be inverted or
`exchanged or appear inverted at an output. These faulty
`positions as well as differential coding are removed in 5
`the subsequent difference decoder 49 so that the origi(cid:173)
`nal bit streams are available at its output. To demulti(cid:173)
`plex the original bit streams, synchronization must be
`performed accurately in a synchronizer 50, the bit
`streams must be descrambled in a descrambler 51 and it 10
`must be possible to detect the beginning of a frame in
`the endless data streams. Then, with the aid of counter
`circuits, a demultiplexer equipped with a control unit 52
`can be used to couple out groups of bits belonging to a
`signal e(t), they can be selected by a program selector 15
`53 and can be processed further. At the same time, the
`scale factor is considered by a scale factor detection
`circuit 54. In an error correction circuit 55, the addi(cid:173)
`tional bits , hereinafter called parity bits, of the external
`error code are evaluated and the signals are corrected as 20
`much as possible. If errors occur which can no longer
`be corrected, these errors are covered up in a conceal(cid:173)
`ment circuit 56. With the aid of the output signal of
`scale factor evaluation unit 54, the signal representing
`the sample values is shifted back into the original value 25
`domain according to the transmitted scale factor. From
`concealment circuit 56 the signals are fed to a demulti(cid:173)
`plexer 16 (See FIG. 1).
`FIG. 4 shows the process sequence of the procedure
`employed for data reduction in the audio signals ac- 30
`cording to the invention generally shown in FIG. 1.
`The analog signal a(t) which represents an audio signal,
`such as, for example, voice or music, is converted in
`ADC stage 61, i.e. corresponding to ADC 1 of FIG. 1,
`into a corresponding digital audio signal. In stage 62, 35
`so-called windowing of the signal occurs by means of
`timely successive and overlapping time windows. The
`signal is thus subdivided into time sequential blocks
`containing 1024 sample values per block,.each having a
`duration of 20 ms per block in such a manner that the 40
`signal of each block can be processed further separately.
`In stage 63 the signal is preprocessed under consider(cid:173)
`ation of suddenly occurring sound events. (See in this
`connection DE-OS No. 3,506,912 published Aug. 28th,
`1986.) In stage 64, the digital signal of a time window or 45
`of a block is transformed into a frequency spectrum.
`Thus, during the timely successive blocks a signal ap(cid:173)
`pears at the output of stage 64 which, for th duration of
`a time window or block, represents the spectral compo(cid:173)
`nents of the signal over the entire frequency spectrum. 50
`Stage 64 thus converts the signal in the time domain to
`the signal representing the spectrum in the frequency
`domain.
`The signal from stage 64 is fed to a coder 65, wherein
`it is coded according to psychoacoustic aspects. That 55
`means that spectral components which are not dis(cid:173)
`cerned in any case during playback, particularly be(cid:173)
`cause of masking effects, are given lower weights or are
`omitted during coding. Such processing of the momen(cid:173)
`tary spectrum is possible, for example, with the aid of a 60
`computer. Stages 62 to 65 correspond to transforming
`member 3 of FIG. 1. A transforming member which can
`be advantageously used is disclosed in European patent
`application No. EP 193 143 A2, published Sept. 3rd,
`1986.
`The thus coded signal travels via a transmitter 66
`which corresponds to the transmitter 12 of FIG. 1 to a
`communication transmission channel or link 67. The
`
`65
`
`4,821,260
`
`6
`resulting reduction in the average bit rate permits this
`communication channel to be dimensioned with a cor(cid:173)
`respondingly narrow band. Communication transmis(cid:173)
`sion channel 67 is followed by receiver 68 which essen(cid:173)
`tially performs functions inverse to those performed by
`the transmitter 66.
`The output signal of the receiver 68 initially reaches
`a decoder 69 which, corresponding to coder 65, per(cid:173)
`forms decoding. In stage 70, the thus obtained signal
`representing the spectrum in the frequency domain is
`reconverted to a digital signal in the time domain. In
`stage 71 the signal is combined again into a uniform,
`continuous digital signal and the preprocessing per(cid:173)
`formed in stage 63 is considered. Then the signal is fed
`to digital/analog converter 72. Converter 72 again fur(cid:173)
`nishes analog signal b(t). This signal b(t) is not identical
`to signal a(t) because during coding in coder 65 spectral
`components were given different weights or were sup(cid:173)
`pressed. However, the difference between analog sig(cid:173)
`nals b(t) and a(t) is such that it will not be noted by the
`listener during playback. Thus, only irrelevant informa(cid:173)
`tion, inaudible for the listener, is removed from the
`signal to reduce the required bit rate during transmis(cid:173)
`sion over communications channel 67. In FIG. 4, the
`stages 62 to 65 correspond to the transforming member
`3 of Fig. 1, stage 66 corresponds to the transmitter of
`FIG. 2, stage 68 corresponds to the receiver of FIG. 3,
`and stages 69 to 71 correspond to the inverse transform(cid:173)
`ing member 20 of FIG. 1.
`Today's satellite radio transmissions employ a 14-bit
`sampling value for transmission. Of these 14 bits, 11 bits
`are error protected and 3 bits remain unprotected. Ac(cid:173)
`cording to the present invention, it is proposed to utilize
`only the 11 protected bits for transmission by means of
`the described data reduction process. The already exist(cid:173)
`ing error protection then permits the reduced data to be
`less protected.
`Thus, with a sampling frequency of 32 kHz, 11 bit/(cid:173)
`sampling values can be transmitted in a monaural chan(cid:173)
`nel. If a coding method as disclosed in DE-OS No.
`3,506,912 (published Aug. 28th, 1986) is employed, 4
`bits per sample value and per monaural channel are
`sufficient. If a coded stereo signal is transmitted (2 X 4
`bits) in an originally monaural channel (11 bits), 3 more
`bits remain which can be utilized for error protection,
`hereinafter called internal error code.
`FIG. 5 is a partial block circuit diagram of a further
`embodiment of a transmitting station. In this embodi(cid:173)
`ment, two audio signals a(t) and k(t) are processed in
`two signal paths each including a respective analog/(cid:173)
`digital converter 1 and a transforming member 3. In a
`multiplexer 80, the two transformed signals from the
`transforming members 3 are multiplexed and sent via an
`output line to a precoder 81, which adds an internal
`error code to the multiplexed transformed signals. This
`code may be a BCH code. Such a code is described, for
`example, in F. J. Furrer, "Fehlerkorrigierende Block(cid:173)
`codierung fiir die Dateni.ibertragung" [Error Correct(cid:173)
`ing Block Coding For Data Transmission], published by
`Birkhiiuser Verlag, 1981. From the precoder 81, a signal
`path leads to a switch 7 which sends either the precoded
`multiplexed signal from precoder 81, or the digitalized
`received audio signal a(t) from the associated ADC 1 to
`a transmitter 12'. The position of switch 7 is controlled
`by a control signal cl(t). Signal cl(t) is also fed to a
`transmitter 12' to indicate whether two multiplexed
`transformed signals or a digital signal is being fed to
`transmitter 12' from switch 7. In the transmitter 12',
`
`

`
`4,821,260
`
`7
`control signal cl (t) is added to the data stream as a
`special service (SD) bit and is recognized by a receiver.
`See in this connection Federal Republic of Germany
`published patent application No. 36 10 398 Al, pub(cid:173)
`lished Oct. 1st, 1987.
`Transmitter 12' may be configured as a channel unit
`25 and transmit either a total of four transformed sig(cid:173)
`nals, two digital signals, or one digital signal and two
`transformed signals. For this purpose, a further a con(cid:173)
`trol signal c2(t) informs transmitter 12' whether the 10
`signal from a further switch 7 where output is con(cid:173)
`nected to another input of transmitter 12' also includes
`two transformed digital signals or one nontransformed
`digital signal. An external error code is added in trans(cid:173)
`mitter 12'. As indicated, transmitter 12' has four inputs, 15
`82-84, i.e., one input 84 for the control signal cl(t), one
`input 85 for the control signal c2(t) and two inputs 82
`and 83 for digitalized audio signals from the respective
`switches 7. Although only partially shown, it is under(cid:173)
`stood that the circuit connected to inputs 83 and 85 is a 20
`duplicate of the circuit connected to inputs 82 and 84, so
`that a total of four digital/analog converters 1, four
`transforming members 3, two multiplexers 80, two pre(cid:173)
`coders 81 and two switches 7 are provided.
`FIG. 6 is the block circuit diagram for a tuner or 25
`receiving station for the transmitting arrangement of
`FIG. 5. At input 86 of receiver 14' there then is present
`a signal which is demodulated, descrambled and demul(cid:173)
`tiplexed in "receiver 14'. Signals x(t), v(t) and r(t) are
`then present at the outputs of receiver 14'. Signal x(t) is 30
`a signal transformed in the transmitting station which is
`forwarded in the tuner to an error correction circuit 87.
`Error correction circuit 87 employs the internal code
`provided by precoder 81 to perform an error correc(cid:173)
`tion. After the error correction, the error corrected 35
`transformed signal reaches a demultiplexer 88. Demulti(cid:173)
`plexer SS-demultiplexes the error corrected signal and
`forwards two signals to two inverse transforming mem(cid:173)
`bers 20 which transform the respective received spec-
`tral values into time domain signals.
`From the two inverse transforming members 20, the
`inversely transformed signals reach respective conceal(cid:173)
`ment circuits 84 whose outputs are fed via respective
`lines to respective switches 22. The switches 22 are
`connected in synchronism and in one position both 45
`switches release the path between the respective con(cid:173)
`cealment circuits 84 and respective DAC's 24. A mon(cid:173)
`aural signal b(t) or a monaural signal l(t) can be picked
`up at the outputs of the DAC's 24. Both DAC's 24
`together furnish a stereo signal of signals b(t) or l(t).
`The signal path from error correction switch 87 to
`switches 22 processes two transformed signals which
`are inversely transformed in this path. Non-transformed
`signals are forwarded together with signal v(t) via a line
`to the other input of both switches 22. If the received 55
`signals are not transformed, the switches 22 are brought
`into their other position so that signal v(t) is sent from
`receiver 14' via the switches 22 to the respective DAC's
`24. The separation of signal v(t) into the two output
`signals b(t) and l(t) is effected by means of control 60
`signals r(t). Receiver 14' puts out three control signals
`rl(t), r2(t) and r3(t). Two of these control signals, rt(t)
`and r2(t), one for each one of the two DAC's 24, are
`used to clock the respective DAC's 24. The digital
`audio signal v(t) is multiplexed, i.e., both signals b(t) and 65
`l(t) are present in signal v(t) in multiplexed form, and
`this multiplexed signal v(t) is separated in the two
`DAC's 24. Depending on when the DAC's 24 are
`
`8
`clocked, only signal b(t) or only signal l(t) reaches the
`outputs of the respective DAC's 24. Control signal r3(t)
`controls the switching of the two switches 22.
`FIG. 7 is the block circuit diagram of a receiver 14' of
`5 FIG. 6 and shows the outputs for signals x(t), v(t) and
`r(t). As can be seen, this receiver has the same basic
`configuration as that of FIG. 3. However, in this em(cid:173)
`bodiment, the transformed signal x(t) is picked up be-
`tween the error correction circuit 55 and the conceal(cid:173)
`ment circuit 56 due to the presence of the correspond(cid:173)
`ing subsequently connected concealment circuits 84
`(see FIG. 6). Error correction circuit 55 utilizes the
`external code to perform an error correction. After
`utilization of the external code, the signal x(t) is picked
`up at the error correction circuit 55. Following conceal(cid:173)
`ment circuit 56, the multiplexed signal v(t) is put out.
`The demultiplexer and control unit 52 puts out the three
`control signals r(t) which serve to control switches 22
`and DAC's 24. A demultiplexer and control unit which
`can be advantageously used is disclosed in European
`Patent Application No. 167 849 A2, published Jan. 15th
`1986.
`The present disclosure relates to the subject matter
`disclosed in Federal Republic of Germany patent appli(cid:173)
`cation No. P 36 42 982.l, filed Dec. 17th, 1986, the
`entire specification of which is incorporated herein by
`reference.
`It will be understood that the above description of the
`present invention is susceptible to various modifica(cid:173)
`tions, changes and adaptations, and the same are in(cid:173)
`tended to be comprehended within the meaning and
`range of equivalents of the appended claims.
`What is claimed is:
`1. In a system for transmitting and receiving digita(cid:173)
`lized audio signals including a transmitting station hav(cid:173)
`ing respective analog to digital converter means for
`converting respective analog audio signals to respective
`digital audio signals, means for multiplexing the respec(cid:173)
`tive digital audio signals to arrange the resulting data
`sequences in timely succession within frames, and trans-
`mitting means for transmitting the multiplexed signal,
`and a receiving station having receiving means for de(cid:173)
`modulating and for demultiplexing a received transmit(cid:173)
`ted signal, and means for converting the received sig(cid:173)
`nals into corresponding analog audio signals; the im-
`provement comprising: respective transforming and
`coding means, disposed in said transmitting station and
`each connected between the output of a respective one
`of said analog to digital converter means and said multi(cid:173)
`plexing means, for transforming a respective said digital
`audio signal at the output of a respective one of said
`analog to digital converter means to a respective trans(cid:173