United States Patent (19)
`Akagiri et al.
`
`54 TWO-STAGE COMPRESSION AND
`EXPANSION OF COUPLNG PROCESSED
`MULTI-CHANNEL SOUND SIGNALS FOR
`TRANSMISSION AND RECORDING
`75 Inventors: Kenzo Akagiri, Tokyo, Japan; Mark
`Franklin Davis, Pacifica, Calif.; Craig
`Campbell Todd, Mill Valley, Calif.;
`Ray Milton Dolby, San Francisco,
`Calif.
`
`73 Assignee: Sony Corporation, Tokyo, Japan
`
`21 Appl. No.:
`446,689
`22 PCT Filed:
`Dec. 7, 1994
`86 PCT No.:
`PCT/US94/14267
`S371 Date:
`Nov. 20, 1995
`9
`S 102(e) Date: Nov. 20, 1995
`O
`O
`Foreign Application Priority Data
`30
`Dec. 7, 1993
`JP
`Japan .................................... 5-306898
`6
`51
`Int. Cl. ................................ HO4S 3/00; H04B 1/66;
`HO4H 5/00
`52 U.S. Cl. .............................. 7041500; 381/2; 704/227;
`704/229
`58 Field of Search .................................. 395/2.36, 2.38,
`395/2.39; 381/2; 704/227, 229, 230,500
`
`56)
`
`References Cited
`U.S. PATENT DOCUMENTS
`
`USOO5873065A
`Patent Number:
`11
`(45) Date of Patent:
`
`5,873,065
`Feb. 16, 1999
`
`
`
`5,490,170 2/1996 Akagiri et al. .......................... 375/240
`5,491,773 2/1996 Veldhuis et al.
`... 395/2.38
`5. 20. Eth AI - - -
`- - - - 3.
`2- - - 2
`OKllOTTC al. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`5,664,056 9/1997 Akagiri ................................... 704/229
`5,687,157 11/1997 Imai et al. .............................. 369/124
`Primary Examiner-David R. Hudspeth
`ASSistant Examiner T?livaldis Ivars Smits
`Attorney, Agent, or Firm-Limbach & Limbach LLP
`57
`ABSTRACT
`A multi-channel Signal compressor for compressing digital
`Sound Signals in the respective channels of a multi-channel
`Sound System. The apparatus comprises a first-stage com
`pression System and a Second-stage compression System. In
`the first-stage compression System, a coupling circuit per
`forms coupling between the digital Sound Signals of at least
`two of the channels to generate coupling-processed signals,
`one for each of the channels. A compressor circuit receives
`the coupling-processed signals from the coupling circuit and
`frequency divides each coupling-processed signal into fre
`quency range Signals in respective frequency ranges, and
`compresses the frequency range Signals obtained by dividing
`each coupling-processed signal to generate a first-Stage
`compressed signal. In the Second-stage compression System,
`a determining circuit receives the first-stage compressed
`Signal for each channel from the first-Stage compression
`System and determines an energy for each channel from the
`first-stage compressed signal of the respective channel. A
`channel bit apportionment decision circuit operates in
`response to the determining circuit, and apportions a prede
`t s".
`of bits ". t that to ap
`number oI DILS to eacn cnannel. Finally, an addiuona
`
`5,166,686 11/1992 Sugiyama - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 341/155
`
`compressor additionally compresses the first-Stage com
`
`5,185,800 2/1993 Mahieux - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 704/500
`
`pressed signal of each channel using, for each channel, the
`
`number of bits apportioned to the respective channel by the
`55. sE. E. et al. .......................... 7.
`5341457 S.E.4 R", al..."704s: channel bit apportionment decision circuit.
`5,471,558 11/1995 Tsutsui ..........
`... 704/219
`5,481,614
`1/1996 Johnston ..................................... 381/2
`16 Claims, 10 Drawing Sheets
`
`-20
`
`(i. 2
`9.
`t
`S
`
`20% |
`copressor
`215 :
`t
`
`20-
`------------------------------is-10 -20.
`o 220
`202 i. ETROPY to:
`CH
`COPRESSOR:
`AAPTIVE ---
`l
`:
`UANT2NG
`E,
`".
`*ist to it,
`|--
`|
`SUBSICAR
`EIH
`204
`-
`205,
`altropy coin, it. El
`ADAPTIVE
`27,
`2?.
`It
`it.
`SUBSIDIARY
`21,
`37.
`E. E.
`204,
`,
`SciRL
`205.
`i. 252 W208
`
`ENTROPY CODNG
`
`202
`COPRESSOR
`
`FIRST STAGE
`COMPRESSION
`SYSTEM
`
`N-214
`SUBSIDARY
`NFORMAION
`COMPRESSION
`
`SECON SAGE
`COMPRESSION
`SYSTEM
`
`Page 1
`
`NETFLIX, INC
`Exhibit 1015
`IPR2018-01630
`
`

`

`U.S. Patent
`
`Feb. 16, 1999
`
`Sheet 1 of 10
`
`5,873,065
`
`213
`
`- or a - - - - - - - - - - on - - as as a- - lf- 2051
`
`250
`
`200 - 20 Y20
`202
`ENTROPY cooING
`201, r2"
`COMPRESSOR
`ADAPTIVE
`C
`gym
`CH2
`M+.
`2012
`p-
`2013
`
`
`
`SUBSIDIARY
`INFORMATION
`COMPRESSION
`
`MULTI
`PLEXER
`
`NRPY CN --- - - - - - - - - - - - - - - -
`
`214
`2142
`
`216
`216
`
`ADAPTIVE
`QUANTIZING
`
`SUBSIDIARY
`INFORMATION
`COMPRESSION
`
`216
`209
`
`214
`215
`Ios
`SPECTRA
`ENVELOPE
`DETECTION
`
`CD
`
`201-1
`i
`f.
`c.
`- a
`d
`2018
`
`ZUZ
`
`COMPRESSOR
`
`
`
`
`
`COMPRESSOR
`
`
`
`215
`
`
`
`
`
`SUBSIDIARY
`INFORMATION
`COMPRESSION
`204
`
`FIRST STAGE
`COMPRESSION
`SYSTEM
`
`a- - - - - -
`
`- - - - - - - - - - - -
`
`- - - -
`
`- - - -
`
`SECOND STAGE
`COMPRESSION
`216
`SYSTEM
`7 -
`
`Page 2
`
`

`

`U.S. Patent
`
`Feb. 16, 1999
`
`Sheet 2 of 10
`
`5,873,065
`
`o
`
`108
`
`108
`
`108
`
`LB
`
`LB
`
`103
`
`Sw
`
`AUDIENCE
`
`110
`
`
`
`PROJECTOR
`
`100
`
`FIG. 2
`
`\-SCREEN
`109
`
`109
`
`109
`
`Page 3
`
`

`

`U.S. Patent
`
`Feb. 16, 1999
`
`Sheet 3 of 10
`
`5,873,065
`
`- - - - - - -
`
`- - -
`
`- - - - -
`
`-
`
`- -n a - - - - - m n - H -- - -
`
`202
`
`OHZ-11KHZ
`
`ADDITION
`
`
`
`OHZ-11KHZ
`
`2242
`
`2022
`
`COMPRESSOR
`
`224
`
`202
`
`COMPRESSOR
`
`OHZ-11KHZ
`
`11-22 KHZ
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Page 4
`
`

`

`US. Patent
`
`Feb. 16, 1999
`
`Sheet 4 0f 10
`
`5,873,065
`
`mm
`
`¢N
`
`mm
`
`mm
`
`mm
`
`.
`
`n
`
`
`
`
` NIéTEomxoo4mm.oszflzéo9"oz<282032E.L2o.u:mm>HE<o<LESLIE2:8on
`monmHomo
`
`
` m538%mn1523mm:x85mmozHNEZEOm”oz<20:53:Emu:mm>HE<o<NEFTNIJE_n..L
`
`rlllllIIIIIIIIIIIIIIIIIIIIIIIIIIIIIllllllllWWlllllllllllllllL__/0N:
`_1523
`
`__monmHomomV.3!n1523nmV5057mm_moZHNEzécmI_oz<28203:E“NONm:mm>HE<o<5.23125mn2_mmeu
`
`_
`
`Page 5
`
`Page 5
`
`
`
`

`

`U.S. Patent
`
`Feb. 16, 1999
`
`Sheet S of 10
`
`5,873,065
`
`22 kHz
`
`11KH
`
`X
`( )
`
`Z is
`
`C
`l
`1.
`
`Lll is O
`
`11.6ms
`
`TIME
`AIG. A
`
`X
`C
`Z
`L
`
`B
`
`L
`1.
`1
`
`22 kHz
`
`
`
`11kHz
`
`5.5 kHZ
`
`2.9ms
`O
`1.45ms
`
`5.8ms
`
`11.6ms
`
`TIME
`FIG. 5B
`
`Page 6
`
`

`

`US. Patent
`
`Feb. 16, 1999
`
`Sheet 6 0f 10
`
`5,873,065
`
`wow
`
`wow
`
`
`
`nowEsmkommm4<onm
`
`.°6Emozmmmo.5282535359%
`
`
`onp<0044<pHmmHZMHoHLLmoo
`
`
`
`oz<m
`
`
`
`EDMHommmmmHoz
`
`Hzmozmmmo
`
`onH<0044<HHm
`
`m.Em
`
`Page?
`
`mDDHHzo<2
`
`20HH<4304<0
`
`4<mkomamno
`
`mmmzzhoofim
`
`ZOHHDmHmeHoHHm
`
`moonmHomo
`magxwm_
`
`mAm<m24<HOH
`
`OHH<x.
`
`thm
`
`Page 7
`
`
`
`
`
`
`

`

`U.S. Patent
`US. Patent
`
`mmm
`
`onhommmoo
`
`onH<ZmngH
`
`mmm
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`eF
`
`9
`
`6.,0?,
`
`1Rmh.
`
`[gnuZOEMESzaséma203:3E5:205535
`
`mmmmmmmNmnme
`.mmszI2053928mezzo;
`
`
`mags:$8525an.qu92m
`
`
`
`
`
`wm
`
`wan
`
`5,873,065
`560,3
`5,
`
`mN«E
`
`EB
`
`Page8
`
`zsszHz
`
`m>mao>HH4HmHo2<ovm
`
`
`
`
`
`Page 8
`
`
`
`
`

`

`U.S. Patent
`
`Feb. 16, 1999
`
`Sheet 8 of 10
`
`5,873,065
`
`2
`
`2
`
`2
`
`O
`
`O
`
`i
`
`d
`
`d
`
`s
`
`dd
`dS
`o Ha,
`- LISNNI
`
`C
`
`Page 9
`
`

`

`U.S. Patent
`
`Feb. 16, 1999
`
`Sheet 9 of 10
`
`5,873,065
`
`
`
`ALISN.INI
`
`Page 10
`
`

`

`US. Patent
`
`FEb.16,1999
`
`Sheet10 0f10
`
`5,873,065
`
`_________________
`
`_________.u
`
`..
`
`>m<Honm2m
` mmoz<mxm
`
`ymompzm
`
`ozHoooma
`
`Ithzzmo
`
`mmeAm
`
`HHm
`
`onp<szng
`
`m£3.mlNzo
`
`£2
`
`
`
`.5252“.”95mm:m559.8%NE_IllllllllllllllllllllllllllllllllllllllllllllllllllllllllJ
`3FEw--.mm.o.z.<.n_m...............=1.......
`
`>momhzm
`
`ozHooomo
`
`>m<Honm3m
`
`onH<§xoqu
`
`mmoz<mxm
`
`Page11
`
`Page 11
`
`
`
`
`
`
`
`
`

`

`5,873,065
`
`1
`TWO-STAGE COMPRESSION AND
`EXPANSION OF COUPLING PROCESSED
`MULTI-CHANNEL SOUND SIGNALS FOR
`TRANSMISSION AND RECORDING
`
`2
`to each band. For example, the Spectral coefficients resulting
`from a modified discrete cosine transform (MDCT) are
`divided by frequency into bands, and the Spectral coeffi
`cients in each band are quantized using an adaptively
`determined number of bits.
`Two known adaptive bit allocation techniques will be now
`be described. First, in the technique described in ASSP-25,
`IEEE TRANSACTIONS OF ACOUSTICS, SPEECH, AND
`SIGNAL PROCESSING, No.4, August 1977, bit allocation
`is carried out on the basis of the magnitude of the Signals of
`the respective bands. Although this System provides a flat
`quantizing noise Spectrum, and minimizes noise energy,
`noise perceived by the listener is not minimized because this
`technique does not exploit the masking characteristics of the
`human Sense of hearing.
`On the other hand, the technique described in M. A.
`Kransner, The Critical Band Coder-Digital Encoding of the
`Perceptual Requirements of the Auditory System, ICASSP
`1980, uses the masking characteristics of the human Sense of
`hearing to determine the Signal-to-noise ratio necessary for
`each band to make a fixed quantizing bit allocation.
`However, this technique provides relatively poor results
`with a Single Sine-wave input because of its fixed bit
`allocation.
`As a high-efficiency System for compressing digital audio
`Signals, employing, for example, the above-mentioned Sub
`band coding System, a high-efficiency compression System
`called ATRAC is already used in practical applications. This
`System compresses digital audio Signals to about 20% of
`their original bit requirement by taking advantage of the
`characteristics of the human Sense of hearing using adaptive
`transform acoustic coding. ATRAC is a registered trademark
`of one of the present assignees (Sony Corporation).
`Multi-channel audio or Speech Signals in four to eight
`channels are not only encountered in, for example, com
`monplace audio equipment, but are also encountered in
`Stereo or multi-channel Sound Systems, Such as those found
`in motion picture theaters, high-quality television Systems,
`Video tape recorders, and Video disc players. In Such cases,
`the use of high-efficiency compression is desirable to reduce
`the bit rate required to represent the large number of audio
`Signals.
`In particular, in commercial applications, a tendency
`towards multi-channel digital Sound Signals and equipment
`handling eight-channel digital Sound Signals has developed.
`Typical of the equipment handling eight-channel digital
`Sound Signals are motion picture theater Sound Systems, and
`the apparatus that electronically reproduces the pictures and
`Sound of a motion picture film via various electronic media,
`in particular apparatus Such as high-quality television
`Systems, Video tape recorders, and Video disc players. In the
`Sound Systems of Such apparatus, the tendency is towards
`multi-channel Sound Systems of between four and eight
`channels.
`Motion picture theater Sound Systems have recently been
`proposed that record on a motion picture film the digital
`Sound Signals for the following eight channels: left, left
`center, center, right-center, right, left Surround, right
`Surround, and Sub-woofer. These Sound channels are respec
`tively reproduced by left loudspeaker, a left-center
`loudspeaker, a center loudspeaker, a right-center
`loudspeaker, and a right loudspeaker, all arranged behind the
`Screen; a Sub-woofer located behind or in front of the Screen;
`and a left-Surround loudspeaker and a right-Surround loud
`Speaker. For the left-Surround Speaker and the right-Surround
`Speaker, two groups of loudspeakers are respectively
`
`FIELD OF THE INVENTION
`This invention relates to a method and apparatus for
`compressing the Sound Signals of the multi-channel Sound
`System of, for example, a motion picture theater, a Video tape
`recorder, or a Video disc player, a method for transmitting
`the compressed Sound Signals of a multi-channel Sound
`System, a method and apparatus for expanding the com
`pressed Sound Signals of a multi-channel Sound System, and
`a recording medium on which the compressed Sound Signals
`of a multi-channel Sound System are recorded.
`BACKGROUND OF THE INVENTION
`Many techniques for compressing digital audio or speech
`Signals are known. For example, in Sub-band coding, a non
`block-forming frequency band dividing System, in which the
`input audio Signal is not divided in time into blocks, but is
`divided in frequency by a filter into plural frequency bands
`for quantizing. In a block-forming frequency band dividing
`System, Such as a transform coding System, the input audio
`Signal in the time domain is converted into spectral coeffi
`cients in the frequency domain by an orthogonal transform.
`The resulting Spectral coefficients are divided by frequency
`into plural frequency bands, and the Spectral coefficients in
`each band are quantized.
`A technique consisting of a combination of Sub-band
`coding and transform coding is also known. In this, fre
`quency range Signals produced by dividing the input audio
`Signal in frequency without dividing it into blocks are
`individually orthogonally transformed into Spectral coeffi
`cients. The spectral coefficients are then divided by fre
`quency into plural frequency bands, and the Spectral coef
`ficients in each band are then quantized.
`Among the filters useful for dividing a digital audio input
`Signal into frequency ranges without dividing it into blockS
`is the quadrature mirror (QMF) filter, which is described, for
`example, in R. E. Crochiere, Digital Coding of Speech in
`Sub-bands, 55 BELL SYST TECH. J. No.8, (1976). A
`technique of dividing the audio input Signal in frequency
`into frequency bands of an equal width is discussed in
`Joseph H. Rothweiler, Polyphase Quadrature Filers-a New
`Sub-band Coding Technique, ICASSP83, BOSTON (1983).
`Known techniques for Orthogonally transforming an input
`Signal include the technique of dividing the digital input
`audio Signal in time into blocks having a predetermined
`duration, and processing the resulting blocks using a fast
`Fourier transform (FFT), a discrete cosine transform (DCT),
`or a modified DCT (MDCT) to convert each block of the
`digital audio signal in the time domain into a set of Spectral
`coefficients in the frequency domain. A modified DCT is
`discussed in J. P. Princen and A. B. Bradley, Subband/
`Transform Coding Using Filter Bank Based on Time
`Domain Aliasing Cancellation, ICASSP 1987.
`As a technique for quantizing the Spectral coefficients
`obtained by frequency division, it is known to divide the
`Spectral coefficients by frequency into bands to take account
`of the frequency resolution characteristics of the human
`Sense of hearing. The audio frequency range of 0 Hz to 20
`or 22 kHz may be divided in frequency into bands, Such as
`25 critical bands, which have a bandwidth that increases
`with increasing frequency. The Spectral coefficients in each
`of the bands are quantized by adaptive bit allocation applied
`
`15
`
`25
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`Page 12
`
`

`

`25
`
`3
`arranged on the left side wall and the left part of the back
`wall of the auditorium, and on the right Side wall and the
`right part of the back wall of the auditorium. The two groups
`of loudspeakers on the Sides and back of the auditorium
`generate a Sound field rich in ambience to accompany
`Spectacular optical effects on the large-format Screen of the
`motion picture theater. For simplicity, these two groups of
`loudspeakers will from now on be referred to as the “left
`Surround loudspeaker' and the "right-Surround loud
`Speaker.”
`It is difficult to record on a motion picture film eight
`channels of 16-bit linear-quantized digital audio with the
`Sampling frequency of 44.1 kHz, Such as is employed in a
`compact disc (CD), because the film lacks an area capable
`of accommodating a Soundtrack wide enough for Such a
`signal. The width of the motion picture film and the width of
`the picture area on the film are standardized. The width of
`the film cannot be increased, or the width of the picture area
`cannot be decreased to accommodate a Soundtrack of the
`width required for digital audio Signals of this type. A
`Standard-width film, with a Standard picture area, a Standard
`analog Sound track, and Standard perforations, has only a
`narrow area in which digital audio Signals can be recorded.
`Accordingly, eight channels of digital Sound can only be
`recorded if the digital Sound Signals are compressed prior to
`recording on the film. The eight channels of digital Sound
`may be compressed using the above-mentioned ATRAC
`high-efficiency compression System.
`Motion picture films are Susceptible to Scratches, which
`can cause drop-outs if digital Sound Signals are recorded
`without any form of error detection and correction.
`Therefore, the use of error correction codes is essential, and
`this must be taken into account when the signal compression
`is performed.
`Optical discS have become popular as a medium for
`providing motion pictures in the home. It is desirable to be
`able to record multi-channel Sound with four to eight chan
`nels on an optical disc to provide more realistic Sound than
`conventional Stereo Sound. On an optical disc, the data
`Volume of the Video signal is as many as ten times that of the
`Sound Signal, and only a limited recording area is provided
`for the Sound Signal. Especially when the picture Signal is
`required to provide a high picture quality, as is required with
`the current trend towards larger-size Screens, as much of the
`recording area as possible is devoted to the picture Signal.
`Thus, the Sound Signal must be Subject to a high degree of
`compression if the desired number of channels is to be
`provided in the recording area available for the Sound Signal.
`When the above-mentioned ATRAC high-efficiency com
`50
`pression System proposed by one of the present assignees
`(Sony Corporation) is used in a stereo (two-channel) audio
`System, the audio signal in each channel is compressed
`independently of the other. This enables each channel to be
`used independently, and Simplifies the processing algorithm
`used to compress the audio signals. Operated this way, the
`ATRAC system provides sufficient compression for most
`applications, and the Sound quality obtained when an audio
`Signal is compressed and expanded using the ATRAC SyS
`tem is well regarded.
`However, because it compresses each audio signal
`independently, it cannot be said that the bit allocation
`process by which the present ATRAC system performs its
`compression operates at highest efficiency. For example, if
`the Signal level in one of the channels is very low, the Signal
`can be represented adequately using a Small number of bits.
`On the other hand, the Signal in another channel may require
`
`35
`
`40
`
`45
`
`55
`
`60
`
`65
`
`5,873,065
`
`15
`
`4
`a much larger number of bits to represent it adequately. Yet
`the present ATRAC system allocates the same number of
`bits to each channel, irrespective of the number of bits
`actually required to adequately represent the Signal in the
`channel. Thus, to provide its high quality of reproduction,
`there must be Some redundancy in the bit allocation per
`formed by the present System.
`
`SUMMARY OF THE INVENTION
`If a multi-channel Sound Signal is to be recorded on a
`motion picture film or on an optical disc, it is necessary to
`increase the compression ratio with a minimum deterioration
`in the Sound quality.
`It is therefore an object of the present invention to provide
`an encoding method and apparatus for further improving the
`compression ratio to enable multi-channel Sound Signals to
`be recorded on a motion picture film, an optical disc, and
`other media, or to be transmitted or distributed.
`It is another object of the present invention to provide an
`encoding method and apparatus for encoding Sound Signals
`of at least two channels with an extremely Small degree of
`deterioration in Sound quality, with excellent channel
`Separation, and with Stable Stereo imaging.
`The invention first provides a multi-channel Signal com
`preSSor for compressing digital Sound Signals in the respec
`tive channels of a multi-channel Sound System. The appa
`ratus comprises a first-stage compression System and a
`Second-Stage compression System. In the first-stage com
`pression System, a coupling circuit performs coupling
`between the digital Sound Signals of at least two of the
`channels to generate coupling-processed signals, one for
`each of the channels. A compressor circuit receives the
`coupling-processed signals from the coupling circuit and
`frequency divides each coupling-processed signal into fre
`quency range Signals in respective frequency ranges, and
`compresses the frequency range Signals obtained by dividing
`each coupling-processed signal to generate a first-Stage
`compressed signal. In the Second-stage compression System,
`a determining circuit receives the first-stage compressed
`Signal for each channel from the first-Stage compression
`System and determines an energy for each channel from the
`first-stage compressed signal of the respective channel. A
`channel bit apportionment decision circuit operates in
`response to the determining circuit, and apportions a prede
`termined number of bits among the channels. Finally, an
`additional compressor additionally compresses the first
`Stage compressed signal of each channel using, for each
`channel, the number of bits apportioned to the respective
`channel by the channel bit apportionment decision circuit.
`The invention next provides a method for compressing
`digital Sound Signals in the respective channels of a multi
`channel Sound System. In the method, coupling is performed
`between the digital Sound Signals in at least two of the
`channels to generate coupling-processed signals, one for
`each channel. The coupling-processed signals are frequency
`divided into frequency range Signals in respective frequency
`ranges. The frequency range Signals obtained by dividing
`each coupling-processed signal are compressed to generate
`a first-stage compressed Signal. An energy for each channel
`is determined from the first-stage compressed signal for the
`respective channel. A predetermined number of bits are
`apportioned among the channels in response to the deter
`mined energy for each channel to apportion a number of bits
`to each channel. Finally, the first-stage compressed signal
`for each channel is additionally compressed using, for the
`channel, the number of bits apportioned to the channel.
`
`Page 13
`
`

`

`S
`The invention also provides a recording medium on which
`there is recorded a compressed Signal generated by the
`above method for compressing digital Sound Signals in the
`respective channels of a multi-channel Sound System.
`The invention also provides a method of transmitting
`digital Sound Signals in respective channels of a multi
`channel Sound System via a transmission medium in which
`the first-stage compressed signal for each channel in the
`method described above is additionally compressed to gen
`erate a respective Second-stage compressed signal, the
`Second-Stage compressed signals of all the channels are
`multiplexed to provide a bit Stream, and the bit Stream is
`applied to a transmission medium.
`The invention next provides a multi-channel Signal
`expander for expanding a compressed signal representing
`the digital Sound Signals in the respective channels of a
`multi-channel Sound System in which the digital Sound
`Signals in at least two of the channels have been Subject to
`coupling prior to compression. The compressed signal
`includes main information and Subsidiary information rep
`resenting the digital Sound Signals of the at least two of the
`channels. The apparatus comprises a demultiplexer, a Second
`Stage expander and a first Stage expander. The demultiplexer
`demultiplexes the compressed signal to extract the main
`information and the Subsidiary information. The Second
`Stage expander expands the main information from the
`demultiplexer by converting variable-length codes to fixed
`length codes. The first-stage expander operates in response
`to the subsidiary information from the demultiplexer to
`additionally expand the fixed-length codes from the Second
`Stage expander to provide a reconstructed digital Sound
`Signal in each of the at least two channels. The fixed-length
`codes on which the first-stage expander operates in one of
`the channels represent a narrower bandwidth Signal than the
`fixed-length codes on which the first-stage expander oper
`ates in another of the channels, even though the digital Sound
`Signals in both of the channels have nominally the same
`bandwidth.
`The invention also provides a method for expanding a
`compressed signal representing the digital Sound Signals in
`the respective channels of a multi-channel Sound System in
`which the digital Sound Signals in at least two of the channels
`have been Subject to coupling prior to compression. The
`compressed Signal includes main information and Subsidiary
`information representing the digital Sound Signals of the at
`least two channels. In the method, the compressed signal is
`multiplexed to extract the main information and the Subsid
`iary information. The main information extracted from the
`compressed signal is expanded by converting the variable
`length codes of the main information to fixed-length codes.
`50
`Finally, the subsidiary information extracted from the com
`pressed signal is used to further expand the fixed-length
`codes obtained by converting the variable-length codes of
`the main information to generate a reconstructed digital
`Sound Signal in each of the at least two channels. The
`fixed-length codes that are further expanded in one of the at
`least two channels represent a signal having a narrower
`bandwidth than the fixed-length codes that are further
`expanded in another of the at least two channels, even
`though the digital Sound Signals in both of the channels have
`nominally the same bandwidth.
`Finally, the invention provides a multi-channel Signal
`compressor for compressing digital Sound Signals in the
`respective channels of a multi-channel Sound System. The
`apparatus comprises a first-stage compressor and a Second
`Stage compressor. The first-stage compressor receives the
`digital Sound Signals of the respective channels and com
`
`6
`presses the digital Sound Signals to provide respective first
`Stage compressed signals. The Second-Stage compressor
`receives the first-stage compressed signals from the first
`Stage compressor, determines an energy for each channel
`from the respective first-stage compressed signal, apportions
`bits among the channels in response to the energy deter
`mined for each channel to apportion a number of bits to each
`channel, and additionally compresses the first-stage com
`pressed signals to provide respective Second-stage com
`pressed signals. The Second-Stage compressed Signal in each
`channels uses the number of bits apportioned to the channel.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`FIG. 1 is a block diagram showing the construction of
`multi-channel compressor apparatus according to the inven
`tion for compressing multi-channel Sound Signals.
`FIG. 2 illustrates the loudspeaker arrangement of an
`eight-channel digital Sound System for use with motion
`pictures.
`FIG. 3 is a block diagram showing the construction of the
`coupling circuit and first Stage compressors constituting the
`apparatus shown in FIG. 1.
`FIG. 4 is a block diagram showing a practical example of
`one of the compressors used in the first-stage compression
`system of the apparatus shown in FIG. 1.
`FIGS. 5A and 5B show how each frame of the digital
`Sound Signal is divided into frequency ranges, and the Signal
`in each frequency range is divided in time into blocks in the
`compressor shown in FIG. 4.
`FIG. 6 is a block diagram of the adaptive bit allocation
`circuit for implementing a bit allocation technique employ
`ing both a signal spectrum-dependent bit allocation and a
`noise Spectrum-dependent bit allocation.
`FIG. 7 is a block diagram of the allowable noise level
`detection circuit for finding the allowable noise level in the
`spectrum-dependent bit allocation circuit shown in FIG. 6.
`FIG. 8 is shows an example of masking by the spectral
`coefficients in each frequency band.
`FIG. 9 is a chart showing the Signal spectrum, the masking
`threshold, and the minimum audible level curve.
`FIG. 10 is a block circuit diagram showing an example of
`a multi-channel expander according to the invention for
`expanding the compressed multi-channel Sound Signal gen
`erated by the multi-channel compressor shown in FIG. 1.
`
`DETAILED DESCRIPTION OF THE
`INVENTION
`Referring to the drawings, preferred embodiments of the
`present invention will now be described in detail.
`FIG. 1 shows the essential portions of the multi-channel
`Signal compressor apparatus according to the invention for
`compressing the digital Sound Signals of a multi-channel
`Sound System. The multi-channel compressor apparatus is
`designed to compress the digital Sound Signals in the chan
`nels CH to CH, to generate an output bit steam. In the
`apparatus, the first-stage compression System 1 includes the
`compressors 202 to 202, that respectively compress the
`digital input Sound Signals in the channels CH to CH. The
`Second-Stage compression System 2 includes the log spectral
`envelope detector 208, the channel bit allocation decision
`circuit 209, the entropy encoders 203 to 203, and the
`subsidiary information compressors 204 to 204.
`The log spectral envelope detector circuit 208 receives the
`Subsidiary information generated by the first-stage compres
`
`5,873,065
`
`15
`
`25
`
`35
`
`40
`
`45
`
`55
`
`60
`
`65
`
`Page 14
`
`

`

`7
`Sion System 1, and from the Subsidiary information detects
`the signal energy in each of the channels CH to CH, The
`channel bit allocation decision circuit 209 determines the
`allocation of the total number of bits corresponding to the
`output bit rate among the channels according to a ratio that
`Substantially corresponds to the output of the log spectrum
`envelope detector 208. In the entropy encoders 203 to 203,
`the adaptive quantizers 205, to 205, respectively adaptively
`quantize the compressed signals from the first-stage com
`pression System 1 using the number of bits indicated by the
`channel bit allocation information received from the channel
`bit allocation decision circuit 209. The entropy encoders
`convert the fixed-length words generated by the compressors
`in the first-Stage compression System 1 into variable-length
`words. The subsidiary information compressors 204 to
`204, adaptively compress the Subsidiary information, i.e.,
`the word-length information and Scale factor information
`generated by the compressorS 202 to 202, in the first-stage
`compression System 1 in the course of compressing the
`respective digital Sound Signals.
`The bit Stream resulting from compressing the digital
`Sound Signals in the respective channels by the multi
`channel Signal compressor according to the invention is
`recorded on a recording medium, or is transmitted via a
`transmission medium. Examples of the recording medium
`include a motion picture film, disc-shaped recording media,
`Such as an optical disc, a magneto-optical disc, a phase
`change type optical disc, and a magnetic disc; tape-shaped
`recording media, Such as a Video cassette; and Solid-State
`media, Such as a Semiconductor memory and an IC card.
`When the recording medium is a motion picture film, the
`Sound Signals in the channels CH to CHs are for feeding to
`the respective loudspeakers of the digital motion picture
`Sound System shown, for example, in FIG. 2. The Signals of
`the center channel C, the Sub-woofer channel SW, the left
`channel L., the left-center channel LC, the right channel R,
`the right-center channel RC, the left-surround channel LB,
`and the right-Surround channel RB are respectively fed to
`the center loudspeaker 102, the sub-woofer 103, the left
`loudspeaker 106, the left-center loudspeaker 104, the right
`loudspeaker 107, the right-center loudspeaker 105, the left
`surround loudspeaker 108, and the right-surround loud
`speaker 109.
`FIG. 2 shows the projector 100 that projects the picture in
`the picture area of the motion picture film onto the Screen
`101, and shows the placement of the various loudspeakers
`relative to the projector, the Screen, and the audience 110.
`The center loudspeaker 102 is located at the center of the
`Screen 101 on the side of the Screen remote from the
`audience 110 and generates the center channel Sound in
`response to the center channel Sound Signal. It generates
`centrally-located Sounds, and it is conventional that all
`dialog is reproduced by the center loudspeaker, irrespective
`of the position on the Screen of the actor or actress Speaking.
`The sub-woofer 103 generates low-frequency effect
`Sounds in response to the Sub-woofer channel Sound Signal.
`It generates Sounds which are often perceived as Vibration,
`rather than as low-frequency Sound, Such as the Sound of
`explosions and other large-scale Special effects.
`The left loudspeaker 106 and the right loudspeaker 107
`are respectively arranged at or beyond the left and rig