`Dolby et al.
`
`[11] Patent Number:
`[45] Date of Patent:
`
`4,815,068
`Mar. 21, 1989
`
`[54] AUDIO ENCODER FOR USE WITH MORE
`THAN ONE DECODER EACH HAVING
`DIFFERENT CHARACTERISTICS
`
`[76]
`
`Inventors: Ray M. Dolby, 3340 Jackson St., San
`Francisco, Calif. 94118; loan R.
`Allen, 18 Hemway Ter., San
`Francisco, Calif. 94117; Randolph G.
`Pauker, 1015 Euclid Ave., #2,
`Berkeley, Calif. 94708
`[21] Appl. No.: 82,651
`[22] Filed:
`Aug. 7, 1987
`Int. Cl.4 .......................... GllB 5/09; G11B 5/02;
`[51]
`H04N 5/94
`[52] U.S. Cl ....................................... 369/175; 369/54;
`358/336; 360/27
`[58] Field of Search ....................... 369/48, 49, 54, 59;
`358/336, 340; 360/27
`References Cited
`U.S. PATENT DOCUMENTS
`4,101,849 7/1978 Blackmer .............................. 333/14
`4, 136,314 1/1979 Blackmer et al. ................... 328/167
`4,433,347 2/1984 Sugiyama ............................ 358/342
`4,490,691 12/1984 Dolby ................................... 333/14
`4,641,204 2/1987 Sugiyama ............................ 358/341
`
`[56]
`
`FOREIGN PATENT DOCUMENTS
`0206746 12/1986 European Pat. Off ..
`
`OTHER PUBLICATIONS
`"A Noise Reduction System for Consumer Applica(cid:173)
`tions" by Dolby, Audio Engr. Society, New York, Oct.
`1970.
`"Dolby B-Type Noise Reduction System" by Berko(cid:173)
`vitz et al., Audio, Sep.-Oct. 1973.
`"A 20 db Audio Noise Reduction System for Consumer
`
`Applications" by Dolby, J. Audio Eng. Soc., vol. 31,
`No. 3, Mar. 1983, pp. 98-113.
`"Optimum Use of Noise Reduction in FM Broadcast(cid:173)
`ing" by Dolby, J. Audio Eng. Soc., vol. 21, No. 5, Jun.
`1973, pp. 357-362.
`"Dolby B-Type Noise Reduction for FM Broadcasts";
`by D. P. Robinson, J. Audio Eng. Soc., vol. 21, No. 5,
`Jun. 1973, pp. 351-356.
`
`Primary Examiner-William L. Sikes
`Assistant Examiner-Akm E. Ullah
`,;lttorney, Agent, or Firm-Thomas A. Gallagher
`
`[57]
`ABSTRACT
`A "hybrid" encoder has a characteristic action suitable
`for substantially full complementary playback in an
`audio signal transmission or recording system using
`decoders designed for one type complementary en(cid:173)
`code/decode system while permitting compatible de(cid:173)
`coding with decoders designed for use with another
`type complementary encode/decode system, or play(cid:173)
`back without any special decoders at all. The hybrid
`encoder is a series or parallel arrangement of two en(cid:173)
`coders that are modifications of standard encoders used
`in the two systems. In the preferred embodiments the
`two systems are A-type noise reduction and spectral
`recording. A variable degree of hybrid encoding is
`provided by varying the characteristic controlling cir(cid:173)
`cuit parameters in the two encoders, or in a preferred
`embodiment, by providing a variable combining ar(cid:173)
`rangement that selects selects from the outputs of a
`hybrid encoder and an unmodified encoder (spectral
`recording).
`
`27 Claims, 7 Drawing Sheets
`
`40
`
`42'
`
`I
`
`AUDIO
`OUTPUT
`SIGNAL
`
`I
`I
`I
`TRANSMISSION
`OR
`RECORDING
`CHANNEL
`10'
`~--------------L--~
`14' l
`FIRST PATH 18' i
`12'
`s~Jf Wf L :
`f ~P~~
`~~~~~Ef
`I
`RECORDING
`ENCODER
`ENCODER
`I
`I
`L ________________ .J
`,...-
`
`1
`
`I
`:
`I
`:
`
`-;i
`
`SIGNAL
`
`22'
`~-SP-EC-T-RA_L,_f ~
`RECORDING
`ENCODER
`
`.[
`
`I
`
`SECOND PATH
`20'
`
`'.1!U
`260 ~- FJ"-24'
`
`I
`
`.
`,
`30
`
`CONTROL
`
`CONTROL
`
`-
`
`-
`
`-
`
`-
`
`-
`
`-
`
`-
`
`-
`
`-
`
`-
`14'
`
`110'
`r- -
`- - ' - - - ..,
`:
`12·:
`MODIFIED
`MODIFIED
`1
`1
`A-TYPE ---~
`SPECTRAL
`I
`:
`RECORDING
`ENCODER
`:
`ENCODER
`I
`I
`L-- - - - - - - - -------J
`22'
`
`SPECTRAL
`RECORDING
`ENCODER
`
`IPR2016-01710
`UNIFIED EX1023
`
`
`
`Ul
`~
`"' 00
`~
`
`"' 0 °' 00
`
`-.l
`~
`0
`~
`~
`
`rJ) =-~
`
`~
`
`~
`"'
`~
`N
`~
`a::
`
`~ ""*-a
`
`•
`rJ'J
`~
`
`SIGNAL
`OUTPUT
`AUDIO
`
`DECODER
`RECORDING
`SPECTRAL
`
`--1
`
`TRANSMISSION OR
`
`CHANNEL
`RECORDING
`
`OUTPUT
`
`I
`
`A-TYPE OR 1--
`
`16'
`
`FIG.-2.
`
`L __ --------------___ J
`I
`I
`
`-
`
`I
`I
`,~
`____ J ___ ,
`
`,~
`
`ENCODER
`RECORDING
`SPECTRAL
`MODIFIED
`
`I
`I
`I
`I
`r-------~--
`
`.'
`
`ENCODER
`A-TYPE
`MODIFIED
`
`FIG._I.
`
`1
`
`SIGNAL
`OUTPUT
`AUDIO
`
`OR 2ND CHAR. ACTION
`(
`COMPLEMENT OF 1ST
`)
`CHARACTERISTIC
`
`DECODER
`
`16
`
`TRANSMISSION OR -1 ACTION SIMILAR TO
`
`CHANNEL
`RECORDING
`
`OUTPUT
`
`I
`I
`/ 14 I
`---'--,
`/'IO
`
`ACTION
`CHARACTERISTIC
`~ SECOND
`)
`ENCODER
`SECOND
`
`L ___________________ J
`I
`I
`I
`I
`I
`I
`I
`I
`I
`
`r---------~12 -----
`
`CHARACTERISTIC
`)
`('.
`
`ACTION
`FIRST
`ENCODER
`FIRST
`
`SIGNAL
`INPUT
`AUDIO
`
`SIGNAL
`INPUT
`AUDIO
`
`
`
`00
`0\
`Q
`....
`Vt
`~
`00
`....
`.&;;..
`
`-....J
`~
`N
`(t> s:a.
`::r'
`r.ri
`
`~
`
`~
`\0
`~
`~ ...
`N
`
`~ ~
`~ ,.......
`T/1
`~
`
`SIGNAL
`OUTPUT
`AUDIO
`
`16'
`
`DECODER
`RECORDING
`SPECTRAL
`
`A-TYPE
`
`OR
`
`SIGNAL
`OUTPUT
`AUDIO
`
`FIRST OR SECOND
`COMPLEMENT OF
`ACTION SIMILAR TO
`CHARACTERISTIC~
`
`CHAR. ACTION
`
`(
`
`DECODER
`
`16
`
`~30
`--_ rJ
`1 ~ RECORDING SSION
`L
`26 1
`~-
`
`CONTROL
`
`CHANNEL
`
`'OUTPUT TRANSMI
`
`----_ _J
`I
`I
`;
`I
`i--IO
`I
`--14'
`
`(
`CHARACTERISTIC
`)
`
`ACTION
`SECOND
`ENCODER
`SECOND
`
`r--------------
`
`12
`
`CHARACTERISTIC
`)
`
`ACTION
`FIRST
`ENCODER
`FIRST
`
`(i
`
`L __
`I
`1
`:
`
`SIGNAL
`INPUT
`AUDIO
`FIRST PATH 18 i
`
`1
`
`1
`
`II
`r---=---,22 -----
`
`---
`
`~ ENCODER
`3~
`
`28': OUTPUT TRANSMISSION
`.L-.
`,24'
`
`OR RECORDING
`
`CHANNEL
`
`1
`
`FIG.-4.
`
`26'_L..,
`
`-
`
`I
`
`-
`
`CONTROL
`1---------~ r3o·
`L_ -f J
`:
`----_J ,... -
`
`L ---------· -----
`I
`
`,~--------•
`r22'
`
`II'
`
`ENCODER
`RECORDING
`SPECTRAL
`
`SECOND PATH 20· I
`
`I
`~-10'
`I
`~----__._,14'1
`-..,
`FIG._3.
`
`ENCODER
`RECORDING
`SPECTRAL
`MODIFIED
`
`-
`
`-
`
`-
`
`-
`
`,-------------
`
`'
`
`12
`
`SIGNAL
`ENCODER
`INPUT
`A-TYPE
`AUDIO ~: MODIFIED
`FIRST PATH 18' :
`
`!
`
`-;.;.;..;;_;:.:..:!!!!:....!:· A~CT~IO~N
`Oii ~~J~H~JO 1sr'
`SECOND PATH 20-' I 3RD. CHAR. ACTION J
`
`
`
`00
`0--.
`0
`-...
`th
`li-il
`00
`~ -...
`
`-.....!
`~
`w
`i:a.
`~
`
`~
`
`\0
`00
`\0
`~
`'"
`~
`N
`
`"""'"'"
`"""'"'" ~
`~
`'flJ
`~
`
`110"'
`
`L ________________ J
`I
`I
`I
`I
`I
`I
`I
`I
`I
`...--
`louTPUT
`I
`I
`I
`I
`I
`I
`I
`I
`------------_l __ ..,
`
`14'
`
`I •
`12'
`
`ENCODER
`RECORDING
`SPECTRAL
`MODIFIED
`
`ENCODER
`A-TYPE
`MODFIEO
`
`I
`
`,--
`
`FIG._6.
`
`FIG._5.
`
`SIGNAL
`INPUT
`AUDIO
`
`+
`
`14
`
`~
`
`CHARACTERISTIC
`
`ACTION
`FIRST
`ENCODER
`FIRST
`
`~
`
`'
`,--------------~--
`
`12
`
`1IO"
`
`L----------______ .J
`I
`I
`1
`1
`1
`:
`1
`:OUTPUT
`32 1
`l
`I
`I
`I
`:
`
`...------_._-..
`
`~ SECOND ~
`
`CHARACTERISTIC
`
`ACTION
`
`ENCODER
`SECOND
`
`•
`
`SIGNAL
`INPUT
`AUDIO
`
`
`
`00
`0\
`0
`-...
`""""" Ul
`00
`~ -...
`
`-I
`~
`,&;o..
`~
`~
`g2
`
`~
`~ ft a
`
`~
`
`~
`"'
`~
`N
`
`~
`~
`
`SIGNAL
`OUTPUT
`AUDIO
`
`SIGNAL
`OUTPUT
`AUDIO
`
`1
`I
`'
`I
`r--------------L~
`
`~
`I
`1
`38 I
`r34
`
`TO FIRST
`
`TO SECOND
`
`36
`
`CHARACTERISTIC
`(COMPLEMENTARY)
`SECOND DECODER
`
`ACTION
`
`L __________________ J
`I
`I
`1
`1
`
`CHARACTERISTIC
`(COMPLEMENTARY)
`FIRST DECODER
`
`ACTION
`
`FIG._8.
`
`TRANSMISSION
`
`CHANNEL
`RECORDING
`
`OR
`
`TRANSMISSION
`
`CHANNEL
`RECORDING
`
`OR
`
`I
`I
`~
`----------'--...,
`
`~
`I
`I
`1
`
`I
`L _______________ _
`I
`I
`1
`1
`:
`I
`
`DECODER
`RECORDING
`SPECTRAL
`MODIFIED
`
`DECODER
`A-TYPE
`MODIFIED
`
`~
`
`r34'
`
`r-----
`
`FIG._7.
`
`L------------------~
`I
`I
`I •·~~~~~~-A
`I
`I
`1
`
`ENCODER
`A-TYPE
`MODIFIED
`
`-
`
`I
`I
`I
`
`I
`I
`I
`14'1
`--'---:i
`
`,10·
`
`ENCODER
`RECORDING
`SPECTRAL
`MODIFIED
`
`-
`
`-
`
`-
`
`,--------12.-
`
`L-----------------~
`1
`:
`1
`
`CHARACTERISTIC
`
`ACTION
`
`ACTION
`
`FIRST v ~ SECOND v
`
`1
`I
`I
`:
`:
`I
`1
`---------------L--
`
`ENCODER
`SECOND
`
`14
`rlO
`
`12
`
`ENCODER
`FIRST
`
`CHARACTERISTIC
`~
`
`SIGNAL
`INPUT
`AUDIO
`
`SIGNAL
`INPUT
`AUDIO
`
`
`
`00
`O'\
`0
`-..
`tit
`~
`00
`~ -..
`
`-....J
`~
`th
`~ ::a.
`rJ1 ::r
`
`[
`
`_J
`
`~:--24
`i
`
`-L, ....r-26
`
`FIG._9.
`
`CONTROL
`
`CONTROL
`
`30
`
`----
`30
`-'
`I
`
`II
`
`-
`
`2ND CHAR. ACT
`SIMILAR TO tST OR
`3RD CHAR ACTION
`3RD ENCODER
`
`~
`
`lo-I "° ~
`
`"'
`lo-I
`N
`
`[
`~
`r;JJ
`~
`
`1
`
`-
`
`SIGNAL
`OUTPUT
`AUDIO
`
`4
`
`,10
`
`142
`
`CHARACTERISTIC --CHARACTERISTIC
`(r
`)
`FIRST ENCODER
`
`-..,
`ACTION
`!__ ~ SECOND
`~
`12 i
`l SECOND ENCODER -l
`r--------------'--...,
`
`L_ -----------_____ _J
`1
`:
`1
`1
`
`ACTION
`FIRST
`
`
`11
`
`F
`'
`
`I
`I
`I
`I
`I
`I
`I
`I
`
`TRANSMISSION
`
`I
`I
`I
`
`OR
`
`CHANNEL
`RECORDING
`
`
`
`00
`O'\
`0
`--..
`'JI
`~
`00
`~ --..
`
`-...J
`~
`O'I
`~
`~ rt)
`
`~
`""'*"
`""'*" E§
`~
`•
`rJ1
`~
`
`\0
`~
`1--l
`'"
`1--l
`N
`
`ENCODER
`RECORDING
`SPECTRAL
`
`22'
`
`1
`
`12~
`
`ENCODER
`A-TYPE
`MODIFIED
`
`14'
`
`_______ _) __ _
`
`110'
`
`SIGNAL -
`
`OUTPUT
`AUDIO
`
`142'
`
`[
`
`ENCODER
`RECORDING
`SPECTRAL
`MODIFIED
`
`L
`1
`:
`1
`:
`
`-
`
`-
`
`-
`
`-
`
`r--
`I
`
`TRANSMISSION
`
`t
`CHANNEL
`RECORDING
`
`OR
`
`I
`I
`I
`
`ENCODER
`RECORDING
`SPECTRAL
`MOD
`
`12'
`
`FIG._10.
`
`II'
`
`-
`
`I
`I
`~---'-( ~
`22'
`
`ENCODER
`RECORDING
`SPECTRAL
`
`SECOND PATH i
`
`20'
`
`,---------------L--
`,IQ'
`
`ENCODER
`A-TYPE
`MODIFIED
`
`:
`
`SIGNAL
`INPUT
`
`AUDIO t :
`
`FIRST PATH 18':
`
`
`
`00
`O'\
`Q
`Ul
`~
`~ 00
`~
`
`~
`
`-....I
`~
`-....I
`~
`~
`g2
`
`\0
`00
`\0
`~
`~ -..
`N
`~
`~
`
`~ ......
`~ fl"'+.
`~
`~
`
`AUDIO OUTPUT
`
`SIGNAL
`
`-_J
`I
`
`I
`
`(14
`
`ACTION
`SECOND
`
`I
`CHARACTERISTIC
`)
`,_, S~EC-ON-D E---NCO_D ___ ER_..
`
`~
`
`-
`
`,, 44 I
`
`-
`
`-
`
`-
`
`-
`
`-
`
`-
`
`AUDIO OUTPUT
`
`SIGNAL
`
`1
`I
`:
`I
`'
`I
`I
`1
`I
`I
`r----------1--,
`
`12
`
`,10"
`
`CHARACTERISTIC
`~
`)
`FIRST ENCODER
`
`ACTION
`FIRST
`
`/ 44
`
`40
`
`-
`
`-
`
`-
`
`FIG.-11.
`L -
`1
`
`~----'4'• I
`
`ENCODER
`RECORDING
`SPECTRAL ,
`MODIFIED
`
`L _____________ _J
`I
`
`ENCODER
`A-TYPE
`MODIFIED
`
`I
`
`--12·-·
`
`L
`1'10"'
`
`----
`
`:
`I
`.------
`
`'
`1 CHANNEL
`: RECORDING
`1 ~
`•TRANSMISSION
`
`I
`I
`
`42 I
`
`L -
`,10"'
`
`14'
`
`-
`
`r
`
`----__ J
`
`-
`
`ENCODER
`RECORDING
`SPECTRAL
`MODIFIED
`
`L ______ -
`I
`I
`I
`I
`I
`I
`
`ENCODER
`A-TYPE
`MODIFIED
`
`------
`
`-
`
`-
`
`r--
`
`I
`1
`:
`I
`I
`
`iTRANSMISSION
`
`:
`I
`
`,10·
`
`(i
`CHARACTERISTIC
`)
`FIRST ENCODER
`-----12-_ L
`
`ACTION
`FIRST
`
`-
`
`r--
`
`CHANNEL
`RECORDING
`
`I
`I
`
`OR
`
`L-------------
`I
`I
`I
`I
`I
`· •
`
`CHARACTERISTIC
`~
`SECOND ENCODER
`r--------L.,.14
`
`~ SECOND
`
`ACTION
`
`FIG._12.
`
`SIGNAL
`INPUT
`AUDIO
`
`SIGNAL
`INPUT
`AUDIO
`
`
`
`AUDIO ENCODER FOR USE WITH MORE THAN
`ONE DECODER EACH HAVING DIFFERENT
`CHARACTERISTICS
`
`BACKGROUND OF THE INVENTION
`The present invention relates generally to audio noise
`reduction and signal processing systems. In particular,
`the invention relates to the problem of encoding audio
`signals in such a way that substantially full complemen- lO
`tarity is obtained when using decoders designed for one
`type complementary encode/decode system while per(cid:173)
`mitting compatible decoding with decoders designed
`for use with another type complementary encode/(cid:173)
`decode system, or without any special decoders at all, 15
`while minimizing audibly objectionable side-effects.
`In its preferred embodiments, the invention is di(cid:173)
`rected to the processing of audio signals recorded in
`motion picture sound tracks. However, the principles of
`the invention are not limited to the motion picture 20
`sound environment and also may be applied to other
`audio recording and reproduction environments.
`Within the past ten years or so an increasing number
`of motion pictures have been made with sound tracks
`encoded with A-type noise reduction, a complementary 25
`system developed by Dolby Laboratories which re(cid:173)
`quires a decoder to obtain its full benefits. The use of
`A-type noise reduction for motion picture sound tracks
`is described in "The Production of Wide-Range, Low(cid:173)
`Distortion Optical Soundtracks Utilizing the Dolby 30
`Noise Reduction System," by loan Allen, J. SMPTE,
`September 1975, Vol. 84, No. 9, pp. 720-729. Over ten
`thousand motion picture theaters around the world are
`equipped with A-type noise reduction decoders. Cur(cid:173)
`rently, about forty percent of all motion pictures pro- 35
`duced in the United States have A-type encoded sound
`tracks.
`The basic elements of A-type noise reduction are
`described in "An Audio Noise Reduction System," by
`Ray M. Dolby, J. Audio Eng. Soc., October 1967, Vol. 40
`15, No. 4, pp. 383-388. Various A-type noise reduction
`products (encoders, decoders, encoder/decoders) are
`manufactured and sold by Dolby Laboratories. A-type
`noise reduction employs four frequency bands: band 1,
`80 Hz lowpass; band 2, 80 Hz to 3 kHz bandpass; band 45
`3, 3 kHz high-pass; and band 4, 9 kHz high-pass.
`Recently, the originators of A-type noise reduction
`introduced and began marketing an improved audio
`signal processing system, spectral recording. This new
`system is described in "The Spectral Recording Pro- 50
`cess," by Ray Dolby, J. Audio Eng. Soc., Vol. 35, No. 3,
`March 1987, pp. 99-118. Various spectral recording
`products (encoders, decoders, and encoder/decoders)
`are manufactured and sold by Dolby Laboratories.
`Spectral recording employs two frequency ranges with 55
`a broadly defined crossover frequency of 800 Hz, such
`that there is a substantial overlap.
`Spectral recording bears some similarities to A-type
`noise reduction. For example, both are complementary
`systems in which a main signal path is primarily respon- 60
`sible for conveying high level signals and a side chain or
`side path signal with the system characteristic (A-type
`. or spectral recording, respectively) is additively com(cid:173)
`bined with the main signal in the encoding mode and
`subtractively in the decoding mode, whereby an overall 65
`complementary action is obtained.
`In spectral recording, a multi-stage series arrange(cid:173)
`ment is used with staggered regions of dynamic action.
`
`1
`
`4,815,068
`
`2
`The high-level and mid-level stages have both high
`frequency and low frequency sub-stages with a cross(cid:173)
`over frequency of 800 Hz. The low-level stage has only
`a high frequency sub-stage, with an 800 Hz high pass
`5 characteristic. In the spectral recording encoded, each
`stage has a low-level gain of about 8 dB, such that when
`the outputs of the stages are combined with the main
`signal path a total dynamic effect of about 16 dB is
`obtained at low frequencies and 24 dB at high frequen(cid:173)
`cies. The reciprocal characteristic is provided in the
`spectral recording decoder.
`In the A-type system, a single stage is used in which
`the outputs of the four bands are combined with the
`main signal path in such a way as to produce a low-level
`output from the encoder which is uniformly 10 dB
`higher than the input signal up to about 5 kHz, above
`which the level increases smoothly to 15 dB higher at
`15 kHz. The reciprocal characteristic is provided in the
`A-type decoder.
`A further difference between A-type noise reduction
`and spectral recording is the manner in which dynamic
`action is provided. In the A-type system, the dynamic
`action in each of the four frequency bands is provided
`by a fixed band circuit in which the signal gain varies
`essentially uniformly across each particular band in
`response to signals within the frequency band. In other
`words, in an A-type expander, the dynamic action
`within each band is a variable, but flat, low level cut
`across the entire band.
`In the spectral recording system, the dynamic action
`is provided by an action substitution technique that
`combines, in a synergistic manner, the characteristic
`actions of fixed band and sliding band (variable filter)
`circuits operating in each of the sub-stages. The action
`substitution technique and the use of single-pole filters
`to allow a broad overlapping of action above and below
`the 800 Hz crossover frequency, provides an overall
`dynamic action that is highly conformable to signals
`virtually anywhere in the frequency band. In other
`words, the spectral recording encoding action is highly
`frequency selective and adaptive by virtue of its action
`substitution of fixed band and sliding band elements
`operating in broadly overlapping frequency bands; the
`overall effect is essentially that of variable width and
`variably positioned frequency bands, an almost infi-
`nitely variable characteristic that adapts itself to both
`the level and frequency content of the signal. In con(cid:173)
`trast, the A-type system, which employs non-varying
`frequency bands, each having fixed band dynamic ac(cid:173)
`tion, has a characteristic that adapts itself only in a
`limited way to signal level and frequency content.
`Another difference in the characteristics of spectral
`recording and the A-type system is that spectral record(cid:173)
`ing employs level dependent low- and high-frequency
`anti-saturation, providing in the encoded signal a gentle
`roll-off in the low and high frequency regions that in(cid:173)
`creases as the signal level rises in order to reduce the
`possibility of overloading the medium on which a spec(cid:173)
`tral recording encoded signal is recorded or transmitted
`at frequency extremes where the ear is less sensitive to
`noise. Also, spectral recording employs low- and high(cid:173)
`frequency spectral skewing, an abrupt and deep reduc(cid:173)
`tion in the low-and high-frequency extremes of the
`encoded signal, primarily for the purpose of reducing
`the susceptibility of the spectral recording decoder to
`any uncertainties in the low- and high-frequency ex(cid:173)
`treme regions of the recording or transmission medium.
`
`
`
`4,815,068
`
`40
`
`3
`Both anti-saturation and spectral skewing are comple(cid:173)
`mentary in the spectral recording system; complemen(cid:173)
`tary de-anti-saturation and spectral de-skewing are pro(cid:173)
`vided in the decoder.
`In order to benefit from the improved performance 5
`and characteristics of spectral recording, such as its
`improved dynamic range, lower noise modulation, im(cid:173)
`proved transient response, and greatly reduced low(cid:173)
`and high-frequency saturation, it would be desirable to
`employ that new system rather than A-type noise re- 10
`duction in the production and playback of motion pic(cid:173)
`ture sound tracks. Of particular benefit for use on mo(cid:173)
`tion picture optical (photographic) sound tracks is the
`substantially improved low- and high-frequency over(cid:173)
`load margin provided by spectral recording.
`Adoption of spectral recording for motion picture
`sound tracks would present no problems if spectral
`recording encoded sound track motion picture films
`were supplied only to motion picture theaters having
`spectral recording decoders. However, two factors 20
`severely restrict that approach: (1) motion picture pro(cid:173)
`ducers prefer, whenever possible, to release a film in
`"single inventory" (for example, all prints of a specific
`film are A-type encoded, even those prints supplied to
`theaters not having A-type decoding equipment), and 25
`(2) in view of the very large number of theaters having
`A-type decoders, single inventory films must be com(cid:173)
`patible with those A-type decoders.
`In view of the various differences between the A(cid:173)
`type and spectral recording systems, it would appear, 30
`upon first analysis, that the systems are not compatible,
`in the sense that reproduction of spectral recording
`encoded audio signals by A-type decoders would likely
`result in subjectively annoying audible effects. It would
`also appear that the reproduction of spectral recording 35
`encoded audio signals without any special decoding
`(non-decoded playback) would likely also result in sub(cid:173)
`jectively annoying audible effects. However, in accor(cid:173)
`dance with the teachings of the invention such apparent
`incompatibility can be overcome.
`Listening tests indicate that audible effects relating to
`system compatibility problems may be characterized as:
`(1) apparently steady-state effects, namely, changes in
`frequency content of the reproduced signal as, for ex(cid:173)
`ample, low frequency or high frequency emphasis, and 45
`(2) dynamic effects, usually referred to as "pumping,"
`whereby signals and/or noise in one part of the fre(cid:173)
`quency spectrum vary in level in accordance with the
`level of a signal in another part of the spectrum. The
`extent to which the ear tolerates such effects is, of 50
`course, level dependent: if the effect is at a sufficiently
`high level it is not acceptable.
`Preferably, dynamic effects should be eliminated or
`minimized because such instability in the reproduced
`signal is more readily perceived by the listener than are 55
`steady-state effects. Steady-state effects are less likely to
`be noticed by most listeners because there is no chang(cid:173)
`ing sound to attract the ear's attention. Even to critical
`listeners steady-state effects may seem attributable to
`differences in the sound mix. Of course, a direct A/B 60
`comparison between
`fully
`complementary
`en(cid:173)
`coding/ decoding and a partially complementary "com(cid:173)
`patible" arrangement would reveal some differences in
`the reproduced signal. However, in a practical situa(cid:173)
`tion, such a comparison is not available and the only 65
`audible cues are those in the reproduced signal itself.
`For example, it has been found that balanced steady(cid:173)
`state low- and high-frequency response effects tend to
`
`4
`be overlooked by the ear. Thus, to the extent that the
`spectral recording roll-off of the low- and high-fre(cid:173)
`quency regions of the encoded signal is not restored by
`A-type or non-decoded playback, the balanced or sym(cid:173)
`metrical effect on the frequency spectrum makes the
`resulting playback acceptable to most listeners.
`In cases where dynamic stability is not achieved and
`there are low level dynamic effects in the presence of
`primary signals, it has been found that the most subjec(cid:173)
`tively uncomfortable audible effects are those resulting
`from signal deficiencies rather than excesses in the por-
`tion of the reproduced signal suffering such low level
`dynamic effects. Such signal deficiencies are often re(cid:173)
`ferred to as a "suck-out" effect. Under such conditions
`15 level variations or pumping that causes low level signals
`to drop further in level, as from an audible to an inaudi(cid:173)
`ble level, are particularly disturbing to the ear. Thus, a
`generalized statement of an underlying principle of the
`present invention is that if the encoder always provides
`at least as much or a surplus of signal at each frequency
`when the signal is decoded as in the original signal
`before encoding, the ear tends to be satisfied. This prin(cid:173)
`ciple may be referred to as the principle of signal suffi(cid:173)
`ciency. In other words, if there is any decode error, it
`should be positive so as to provide an excess of signal;
`the ear is more tolerant of an excess rather than a defi-
`ciency of signal.
`The encoding characteristics of the spectral record(cid:173)
`ing system provide an excellent starting point for gener(cid:173)
`ating an encoded signal that meets the requirements of
`the principle of signal sufficiency. This is because spec-
`tral recording provides highly frequency selective com(cid:173)
`pression during encoding: the compressor tends to keep
`all signal components fully boosted at all times; when
`the boosting must be cut back at a particular frequency,
`reduction in boost essentially is not extended to low-
`level signal components at other frequencies. The audi(cid:173)
`ble effect of this type of compression is that the signal
`appears to be enhanced and brighter but without any
`apparent dynamic compression effects (the ear detects
`dynamic action primarily by the effect of a gain change
`due to a signal component at one frequency on a signal
`component at some other frequency, somewhat re(cid:173)
`moved). As a consequence, spectral recording encoded
`signals reproduced with no special decoding whatso(cid:173)
`ever are free of pumping effects for nearly all signal
`conditions because of the compressor's frequency adap(cid:173)
`tiveness (dynamic action occurs substantially only at
`frequencies requiring such action and nowhere else) and
`thus are discernible to a critical listener as compressed
`signals only because there are changes in the low fre-
`quency and high frequency emphasis (a maximum of 16
`dB compression at low frequencies and 24 dB at high
`frequencies).
`In accordance with the underlying principle of this
`invention, it has been found that signal deficiencies in
`the order of a few decibels, say 2, 3, or 4 dB oflow level
`signals, such as low level ambiences, in the presence of
`primary or dominant signals, are audibly acceptable to
`most listeners, but that larger deficiencies, in the order
`of 6 or 12 dB are not acceptable to most listeners. In
`contrast, it has been found that signal surpluses of 10,
`12, 15 dB or even more of such low level signals in the
`presence of primary signals are generally acceptable to
`most listeners. Thus, in accordance with the principle of
`signal sufficiency, the playback arrangement may be
`considered to be compatible with an encoder if at any
`frequency or time the low level signals or ambiences in
`
`
`
`4,815,068
`
`5
`the presence of primary signals in the reproduced signal
`are no less than a few dB below the original signal and
`are no more than some 10 to 15 dB above the original
`signal. This question of compatibility relates to low
`level signals in the presence of primary or dominant 5
`signals because it is such low level signals that are pri(cid:173)
`marily manipulated by the A-type and spectral record(cid:173)
`ing systems. In those systems high level primary or
`dominant signals are substantially unaffected.
`If a spectral recording encoded signal is played back 10
`using an A-type decoder the results conform generally
`to the above stated principle except for certain signal
`conditions that cause audible pumping and/or signal
`suck out in the 80 Hz to 3 kHz A-type band 2.
`Consider, for example, the application to a spectral 15
`recording encoder of input signals including signals in
`the 80 Hz to 3 kHz range. For low level signals below
`its threshold, the spectral recording encoder ( compres(cid:173)
`sor) provides 16 dB of boost at low frequencies and 24
`dB of boost at high frequencies. Boosting of low level 20
`signals narrows the dynamic range, resulting in signal
`compression. The amount of signal compression is re(cid:173)
`duced in accordance with the spectral recording encod(cid:173)
`er's compression law in a relatively narrow frequency
`range at ad near the frequency of signals exceeding the 25
`encoder threshold. The highly frequency selective or
`adaptive nature of the spectral recording encoder as(cid:173)
`sures that the encoder's dynamic action is restricted to
`a relatively narrow frequency range in which dynamic
`action is required due to the presence of signals above 30
`threshold in that frequency range.
`With respect to A-type playback, for low level sig(cid:173)
`nals below its threshold, the A-type decoder (expander)
`in band 2 provides about 10 dB of signal reduction
`uniformly across the 80 Hz to 3 kHz band. Reducing the 35
`level of low level signals widens the dynamic range,
`resulting in signal expansion. The amount of signal ex(cid:173)
`pansion throughout band 2 is reduced in accordance
`with the A-type decoder's expansion law in response to
`the presence of signals above threshold anywhere 40
`within band 2. Since band 2 is relatively wide, it is likely
`that signals above threshold located in one part of band
`2 that control the dynamic expansion action will cause
`audible pumping of other low level signals and noise
`contained with the original signal in the frequency 45
`range of band 2 prior to encoding, as the gain through(cid:173)
`out band 2 varies uniformly across the band. Such
`pumping, in the direction of signal suck out, is likely to
`be audible for certain signal conditions because.band 2 is
`so wide in frequency that the signal controlling the 50
`dynamic action cannot effectively mask the modulation
`of other signals and noise in band 2 as the gain of the
`entire band varies.
`As mentioned above, many A-type sound track en(cid:173)
`coded motion pictures have been released in single 55
`inventory, despite the fact that not all theaters have
`A-type decoding equipment. Such non-equipped thea(cid:173)
`ters typically employ sound systems designed to play
`films conforming to the so-called "Academy" mono(cid:173)
`phonic (mono) format, developed in the 1930's. As is 60
`well known, a considerable amount of treble cut is ap(cid:173)
`plied when optical sound tracks are played back in such
`theaters. This high frequency roll-off, referred to as the
`Academy characteristic, produces an attenuation of at
`least 20 dB at 9 kHz. Subjectively, the roll-off provided 65
`by the Academy characteristic brings the tonal charac(cid:173)
`ter of a A-type encoded track essentially back to nor(cid:173)
`mal. The low level boosting of high frequency compo-
`
`6
`nents in the encoding process adequately compensates
`for the high frequency Academy roll-off. Low fre(cid:173)
`quency, low level signals will be left in the boosted
`condition, but this effect is noticeable only when a track
`is switched directly from A-type decoding to Academy
`replay. Consequently, A-type encoded films sound ac(cid:173)
`ceptable to most listeners when played in "Academy
`mono" theaters and film-makers often make the judg(cid:173)
`ment that a single inventory release is appropriate.
`Spectral recording encoded films, although having a
`greater amount of compression than A-type encoded
`films, also sound acceptable to most listeners when
`played in "Academy mono" theaters. In fact, the
`greater compression may be beneficial in the environ(cid:173)
`ment of most "Academy mono" theaters in that such
`theaters tend to have high ambient noise levels from
`noisy air conditioning systems and/or are part of a mul(cid:173)
`ti-screen theater complex in which sound from adjacent
`auditoria is audible. In addition, the highly frequency
`selective encoding in the spectral recording compressor
`results in less likelihood of audible pumping than with
`an A-type encoded sound track.
`
`SUMMARY OF THE INVENTION
`In accordance with the teachings of the present in(cid:173)
`vention, encoding arrangements are disclosed that sub(cid:173)
`stantially retain the improved characteristics of the
`spectral recording process when spectral recording
`decoders are used, while achieving reproduction audi(cid:173)
`bly acceptable to most listeners when A-type decoders
`are used. Simply stated, the invention provides, at least
`for certain input signal conditions, for the addition of
`some degree of the A-type encoding characteristic to a
`modified spectral recording characteristic so that when
`the signal is reproduced by an A-type decoder the prin(cid:173)
`ciple of signal sufficiency is not violated. The problem
`overcome relates mainly to the deficiencies occurring in
`the reproduced signal due to the action of band 2 of the
`A-type decoder, as described above. By adding a suffi(cid:173)
`cient amount of A-type encoding, pumping and suck
`out effects in band 2 are restricted to an extent essen(cid:173)
`tially unnoticeable to the ear.
`In addition, the encoding arrangements of the present
`invention continue to permit reproduction audibly ac(cid:173)
`ceptable to most listeners for sound tracks played back
`by Academy mono systems. Since A-type encoding per
`se and spectral recording encoding per se provide gen(cid:173)
`erally acceptable playback by Academy mono systems,
`a degree of A-type encoding along with modified spec(cid:173)
`tral recording encoding also provides the same sort of
`acceptable results.
`As with current practices, the decision regarding the
`release of a particular film in a single inventory format,
`employing an encoding arrangement according to the
`invention, would be based on the artistic and business
`judgment of the film-maker.
`According to one aspect of the present invention, a
`modified A-type encoder and a modified spectral re(cid:173)
`cording encoder are placed in series to provide the
`encoded audio signal for application to an audio trans(cid:173)
`mission or recording channel, such as a motion picture
`sound track. The encoders are modified principally
`with respect to their dynamic action, such that each
`provides roughly two-thirds of the maximum compres(cid:173)
`sion provided in their standard, unmodified form. Alter(cid:173)
`natively, a modified A-type encoder and a modified
`spectral recording encoder are placed in parallel to
`provide an encoded audio signal. According to this
`
`
`
`7
`parallel alternative, the encoders are also modified prin(cid:173)
`cipally with respect to their dynamic action and, in
`addition, the A-type encoder may also be modified so as
`to operate only with a single band, band 2. For conve(cid:173)
`nience, these arrangements in general will be referred to 5
`as "hybrid" encoding.
`In practice, continuous encodin