The present invention relates to audio systems and more particu-larly to a motion picture format and sound system in which more than two playback channels are provided while employing a compatible two track release print and overcoming the problems of prior art multi-channel systems.
Prior art multi-channel motion picture sound systems for enhancement of realism and dramatic or special effects date back at least to the 1930's film "Fantastica" where discrete optical tracks were used for each channel.
subsequently, with the introduction of wide screen processes, such as Cinemascope and Cinerama, multiple track magnetic sound tracks were used to provide as many as six or more discrete channels. More recently, the films "Tommy" and "Pink Floyd" employed multiple channel magnetic release prints to provide multi-directional sound with two speakers in the rear theatre corners and two or three in front (left, center, right).
Heretofore, multi-channel systems have generally employed discrete sound tracks for each channel. While compatible stereophonic optical sound tracks have recently increased in popularity, most multi-channel systems employ more than two channels and have generally employed magnetic tracks which relatively few theatres are capable of reproducing due to the extra cost in projection and sound equipment. Consequently, standard optical prints were also released for performance in the majority of theatres. In addition to the problems of the limited number of theatres equipped for mag-netic release prints, the cost of producing magnetic prints is substantially higher than for optical prints and the number of useful plays of magnetic prints is more limitedO m eatres equipped to handle multi-channel release prints have generally employed low quality secondary or rear ("surround") channel speakers, chosen on a minimum cost basis, rather than for fidelity.
--1-- ' ;
ll~)v881 In the case of recent magnetic multi-channel efforts, two high quality rear channel speakers were employed, however, this approach is also disadvantageous for several reasons cost; the rear corner theatre seats receive a predominance of left back or right back sound; theatres are designed acoustically for sound to travel from the screen to the rear (absorption by cushioned seat fronts and/or viewers) - audio point sources at the rear can cause excessive slap echoes from reflections off smooth hard seat backs; and the aesthetics of large speaker enclosures at the theatre rear.
To avoid the cost and compatibility problems of magnetic release prints, multi-track optical prints can be considered. However, there are disadvantages in having more than two optical sound tracks on motion picture film. Significant among these are: more complicated and expensive replay optics and pick-up cells; deteriorating signal-to-noise ratio caused by narrower tracks and increased total guard band area; printing tolerance problems from narrower tracks and/or guard bands; increased relative boundary noise; questionable compatibility when all tracks are scanned simultaneously to achieve monophonic reproduction; increased playback electronics expense; and increased optical recorder expense.
Two syste~s for providing three front channel reprod~tion(left, center, right) while using a two track release print are disclosed in commonly assinged United States application, 603,670 by Robert A. Berkovitz and Kenneth J. Gundry, filed August 11, 1975 , now United States Patent NoO
~ X g ~ issued Fe~ ~Y, ~q7 ~ ~ and in commonly assinged United States Patent No. 4,024,344 to Ray M. Dolby and Philip C. Plunket, issued May 17, 1977.
, ` 11()~}~381 In the Berkovitz et al patent L, C and R signals are encoded onto dual sound tracks: a portion of the C signal is summed with the L and R
signals, respectively. In playback, each track feeds an L or R speaker and a summing network which drives the C speaker at reduced volume relative to the R and L speakers. Delay circuits in the R and L speaker feeds cause the center channel information to appear to issue from the C speaker while pre-serving the stereophonic presentation from the R and L speakers.
In the Dolby et al patent a further L, C and R playback system is disclosed for use with two channel sources. In one embodiment means are provided for delaying the left and right channel playback signals relative to the center channel. However, these patents do not address the unique problems in providing rear channel sound.
In order to overcome these and other problems in the prior art, the present invention contemplates a multi-channel motion picture sound format in which stereophonic front sound channels are encoded along Wit]l at least one secondary or rear "surround" sound channel on a compatible dual bilateral or stereo variable area optical sound track format. The latter format is increasingly in use in the industry and will provide full compatibility, allowing a release print in the format of the present invention to be used not only in theatres having traditional monophonic optical projection equip-ment and sound systems but also in those theatres employing stereo playback equipment and the equipment according to the present invention.
According to one aspect of the present invention, there is provided a theatre sound system including main channel speakers forwardly located and a plurality of rearwardly located surround channel speakers for reproducing a two track motion picture sound track onto which at least two main sound channel signals and at least one surround sound channel signal are encoded, . --3--. ~ .
, ~
- : ~ - :
: - : . .
.
3Bl the combination comprising sound track playback means for generating first and second signals representative of the information on said two tracks, matrix decoder means receiving said first and second signals for decoding said first and second signals to generate said at least two main sound channel signals and said at least one surround sound channel signal, said at least one surround sound channel signal containing cross-talk components from said main sound channel signals as a result of imperfect separation characteristics of said matrix decoder means, means receiving said at least one surround sound channel signal for processing said at least one surround sound channel signal, said processing means including time delay means having a delay time of about 75% of the sound path delay from the front to rear of said theatre, whereby perceived cross-talk is supressed from said main sound channel signals, and means for applying said decoded main sound channel signals to said main channel speakers and for applying said processed at least one surround sound channel signal to said plurality of rearwardly located surround channel speakers.
According to another aspect of the present invention, there is provided a motion picture sound recording and theatre sound playback system using a two track sound track onto which at least two main channel signals and at least one surround channel signal are encoded, the theatre playback system including main channel speakers forwardly located and a plurality of rearwardly located surround channel speakers, the system comprising encoder means receiving said at least two main sound channel signals and said at least one surround sound channel signal for generating first and second encoded signals carrying said main sound channel signals and said at least one surround sound channel signal, sound track recording means for applying said first and second encoded signals to said two motion picture sound tracks, respectively, sound track playback means for generating first and second signals representative of the information on said two tracks, matrix decoder means receiving said first and second signals for decoding said first and second signals to generate said at least two main sound channel signals and said at least one surround sound channel signal, said at least one surround sound channel signal containing cross-talk components from said main sound channel signals as a result of imperfect separation characteristics of said matrix decoder means, means receiving said at least one surround sound channel signal for processing said at least one surround sound channel signal, said processing means including time delay means having a delay time of about 75% of the sound path delay from the front to rear of said theatre, whereby perceived cross-talk is suppressed from said main sound channel signals, and means for applying said decoded main sound channel signals to said main channel speakers and for applying said processed at least one surround sound channel signal to said plurality of rearwardly located surround channel speakers.
In the preferred embodiment a single "surround" ~secondary) sound channel is provided in addition to left and right front ~main) sound channels.
Preferably, the surround channel drives substantially more than two multiple surround speakers located along the rear wall and the rear portion of the side walls of the theatre. This arrangement provides integration of sound sources over a wider area to reduce slap echoes and the volume level of each speaker can be relatively low so that the sound from a given surround speaker does not dominate the other sound fields heard by a person sitting near that surround speaker. Sound localization can also be improved by locating relatively directional surround speakers high up on the theatre side walls so that the direct field on the opposite side of the theatre auditorium (having -`
- ~ .
llV(~-8~31 more high frequencies~ will compensate for proximity effects.
The present invention overcomes a serious problem encountered in matrix encoding the surround channel into the main two audio channels: all matrix encoded sound systems suffer from cross talk in the decoded -5a-.
signals due to the encoding process. While the introduction of some sur-round channel information into the main chalmels is not of concern, spurious front channel cross talk in the surround channels is highly undesirable due to the Haas or precedence effect: when two physically separated sound sources emit substantially the same sound whether cne or both is heard and the apparent direction of the source as perceived by the listener depends on the intensity and relative time of their arrival at the listener's ear. If the time is the same and the intensities equal, the sound source appears to be between the two physical sources. If one source is delayed so that its arrival at the listener is from 5-35 ms after the other, then the sound appears to come only from the undelayed source. For delays of 35-50 ms, the second sound is perceived, but the sound still appears to come from the undelayed source. For longer delays in a low reverberation room, the listener hears a distinct echo. Thus, front channel cross talk in the surround speakers will undeSirably create the impression that the front channel sound is coming from the rear of the theatre audit~rium. The reverse situation of surround cross talk in the front channels is less likely to be noticeable since the audience sits closer to the surround speakers and thus the surround speaker generated surround sound will reach the audience before the cross talk from the main channel speakers. Also, some surround channel information is normally purposely mixed into the front channels when the film is produced to accommodate the case where surround speakers are not in use at a theatre.
9 In order to overcome the aforementioned problem of main channel cross talk in the surround channel speakers, the invention provides for a time delay in the surround channel reproduce apparatus: the delay is _~ .
, .
chosen in a preferred embodiment to be about 75 % of the front to rear path length of the particular theatre auditorium. Such delays will typically range from about 25 ms to 120 ms, the velocity of sound being roughly 1 ft./ms.
Because the information carried in the surround channel tends commonly by its nature to be steady-state background sounds and/or reverbera-.~ f~
tion signals f~ simulating spaciousness, the quality of the surround channel reproduction chain need not be of as high a quality as the main channel.
Consequently, any surround channel signal degradation by delay means and lower quality speakers does not materially affect the sonic effect in practical use.
On the other hand, the main channel signal information tends to include sub-stantial transient detail, particularly human voices, which require good fidelity forrrealistic reproduction. The present invention actually enhances the fidelity of the main channels by eliminating audience perceived cross talk from the rear channels, thereby stabilizing the main channel signals, particularly high transient content signals which desirably are perceived as reproduced by the high quality main channel reproduction chain. Bass frequencies are not well separated by typical single band matrix endoding/
decoding systems. Hence substantial bass cross talk occurs among the main and surround channels. However, that result is acceptable since bass fre-quencies are essentially non-directional and the resulting effect is to make the overall listener bass perception essentially uniform throughout the theatre.
The invention will now be described in detail with reference to the accompanying drawings, in which:
Figure 1 is a block diagram of the recording portion of a system embodying the invention; and 8~1 Figure 2 is a block diagram of the playback portion of a system embodying the invention.
Referring now to Figure 1, the record portion of the present in-vention is shown in the context of a motion pictule sound system. A means for encoding more than two audio signals onto two channels such as a conven-tional matrix encoder 2 receives two main sound channel signals "LF" (left front) and "RF" (right front) at one pair of inputs, respectively, and a processed secondary sound channel input at a further pair of inputs "LBn (left back) and "RB" (right back). The matrix encoder 2 oan be any one of several quadraphonic type encoders known in the art, such as "SQ" or "QS"
for example. The designations "LF", etc. are commonly used in quadraphonic sound systems and in connection with such encoders.
Matrix encoder 2 can receive its LF and RF inputs from a 3:2 logic circuit 3 which has LF, C (center) and RF discrete signals applied thereto.
Logic circuit 3 can be of the type disclosed in the above mentioned Berkovitz et al United States patent.
The main sound channel stereophonic signals "LF" and "RF" derived from a discrete multi-track master (not shown) carry the voice dialogue and other transient rich information intended for reproduction by speakers loca-ted behind or adjacent the theatre screen. The secondary channel input signal is a so-called "surround" signal carrying material that is ordinarily slowly changing and lower level background and reverberation information intended for reproduction by surround speakers located at the rear and rear sides of the theatre auditoriumO
While discrete secondary channel signals can be applied to the matrix encoder LB and RB inputs, it has been found satisfactory to use a ..
. , ' ,- ':
single signal to the surround speakers and hence, the same signal is applied to both B inputs.
The secondary sound channel or surround input signal is also de-rived from a discrete multi-track master and is processed prior to applica-tion to encoder 2. Since encoder 2 is less effective at high frequencies, the signal is applied to a low pass filter 4 having an upper frequency limit at about 7kHz, for example, to reduce cross talk of high frequency components into the surround channels as a result of azimuth error. Following the filter 4, a noise reduction encoder 6, preferably of the type known in the art as a Dolby B-type encoder (matched by a decoder in the playback circuitry), provides further processing before the signal is applied to the LB and RB
encoder inputs. Noise reduction in the secondary or surround channel is particularly desirable because of random phase noise on the film that tends to be more predominant in reproduction of the surround channel.
The encoded outputs of encoder 2 on two lines "LT" (left track) and "RT" (right track) are applied to conventional sound track recording apparatus. While the present invention contemplates principally that the recording medium is the stereo bilateral varlable area optical sound track, the invention is also usable with other types of recording medi~m having two tracks.
` In the playback system of Figure 2, conventional sound track play-back apparatus 10 reads the encoded tracks and provides LT and RT signals to the matrix decoder 12. Decoder 12 matches the encoder 2 and provides LF, RF~ LB a~dl RB outputs. m e LF and RF main channel outputs may be applied directly to respective channels of conventional playback electronics and _g_ '' ' ' . ~ :
. ' ~ ' . . ':
llUl~8~1 speakers (not shown) or, alternatively, as shown, to a 2:3 logic circuit 14 of the type in the above mentioned Berkovitz et al United States Patent, which derives a center channel signal CF for appl:ication to playback electronics and speaker(s) (not shown). The 2:3 logic circuit 14 alternatively can be placed prior to decoder 12. If desired, the teachings of the aforementioned United States Patents of Dolby et al or Berkovitz et al can be employed in the record and playback apparatus with respect to the LF, CF and RF signals.
A combiner 16 receives the LB and RB decoder outputs and sums them for application to a low pass filter 18 which preferably has an upper fre-quency limit at substantially the same frequency as the filter 4 of FigUre`1.
The signal can then optionally be applied to high pass filter 20 having a lower frequency limit on the order of 80Hz to prevent any high energy low frequency energy from clipping the surround speakers. Next, a delay means 22, delays the signal by a time on the order/of 75% of the front to back path length of the theatre auditorium in which the system is used. In prac-tice, delays in the order of 25 to 120 ms are likely to be required, depend-ing on auditorium size. Such large delays are well within the current state of the art by a number of alternative means. High p~ss filter 20 can be placed after decoder 24 so that the input to decoder 24 more closely comple-ments the signal applied to encoder 6 (Figure 1). Also, such a placement of filter 20 will also tend to remove noise introduced by delay line 22. Low pass filter 18 may also be located after decoder 24.
A noise reduction decoder matching ~ncoder 6 of Figure 1 follows the delay 22. The output then drives playback electronics and surround speakers.
.
- ' '' ', llV~381 In practice, a normal surround sound 100% level is set at -3dB with respect to 100% on the film. The Dolby B-Type encoder level is set at 50%
on the film.
In non-cinema applications of the invention, the delay 22 will be adjusted to fit the room size. In the case of quadrophonic sound systems for home use, the invention could employ a fixed delay on the order of 12 to 30 ms to acco~lmodate a typical room.
.
.