Author: Alfred W. Myers
Just a few years ago it was impossible to pick up a copy of any audio journal without some advertisement for a new piece of four-channel equipment or read of how quadraphonics was overtaking stereo and would soon replace all two-channel reproduction.
Obviously, this didn't happen, and today quadraphonic sound is largely dead in the hi-fi marketplace. No, it is not totally obsolete, but it certainly is not the thriving success that was predicted. The three main competing systems are still fighting it out down to the last tooth and nail. And work still is progressing on both the matrix and discrete formats, but, except for the recent introduction of the long-awaited Tate SQ ICs, system improvements and "breakthroughs" most usually remain in audio laboratories and are seen only at professional audio exhibitions.
The reasons for quadraphonic's near-demise are many and varied, and it is impossible to lay the blame at any one specific door. Suffice to say that three competing systems were introduced before final fruition and that the public was told an awful lot about a "wonderful new listening experience" to which performance capabilities did not fully measure up.
Indecision and inaction on the part of many hardware and software companies helped to stunt quadraphonic's growth. Finally, the total absence of standards for quadraphonic FM broadcasts only added to the public's confusion, helping to cement the "wait and see" attitude that most consumers adopted. Today the systems have improved and the FCC has completed a series of tests in an attempt to standardize on one four-channel format for FM broadcasts.
In 1972 the National Quadraphonic Radio Committee (NQRC) set out to determine which of the many competing systems would be best for quadraphonic FM. Although the original NQRC tests were meant to compare all formats, the two leading matrix systems (SQ and QS) were withdrawn from testing at the requests of CBS and Sansui. So, the NQRC evaluations were generally limited to discrete ( 4-4-4) and discrete/matrix ( 4-3-4) systems. One matrix format (the BMX from Nippon/Columbia-no relation to CBS) was used in the NQRC tests, but it was never commercially used in this country and is not a first rank contender. The primary difference between 4-4-4 and 4-3-4 (aside from equipment used) is the number and allocation of subcarrier signals used to unscramble the quadraphonic composite signal.
Between the time of the NQRC's tests and the present, major advances have been made in matrix technology.
Sansui developed their Variomatrix decoder, while SQ sources came up with a family of logic circuits culminating in the phase-cancellation technique used in the Tate, Scheiber, and CBS' own paramatrix decoders.
Seeing the improvements in matrixing, the FCC made the decision to conduct a new series of tests to determine if these matrix advances negated the need for "going discrete." In addition, the FCC received a petition from CBS to adopt SQ as the standard and a counter-proposal from Sansui in support of their own QS matrix. To add a bit more confusion to the whole quadraphonic FM scene, England's BBC came up with their own system (termed "Matrix H") and that too was tested.
In their laboratories in Laurel, Maryland, the FCC constructed a room measuring 15 feet wide and 22 feet long. One wall was covered with foam rubber, while one adjacent wall consisted of wood folding doors. The two remaining walls were of sheetrock. The FCC admits that such a room is far from acoustically ideal, but felt that such materials are representative of what might be found in an average living room-or at least would simulate the acoustic properties of average living rooms. Four Technics' SB-700A "Linear Phase" speakers were driven by a pair of Crown D-150A stereo amplifiers which, in turn, connected to a Bose 4401 four-channel preamp.
Crown and Ampex tape equipment was used as well as the various matrix encoders and decoders: SQ (phrase cancellation paramatrix), QS (Sansui QSD-1 Variomatrix), and experimental BBC-H gear. RCA provided the components for the 4-3-4 demonstrations.
To represent a discrete 4-4-4 broadcast, the FCC simply played a discrete open-reel tape, feeding the audio chain directly-a sore point with the matrix advocates, who say that no 4-4-4 broadcast will ever equal a master tape. But, right or wrong, that is the way the tests were performed.
Other than the complaint about the 4-4-4 simulation, there can be no doubt that the tests were conducted with the utmost fairness. The various system representatives selected and supplied the equipment used for their tests, ran their own tests to see that it was performing properly, and were given the option to check out their gear throughout the duration of the tests (not, of course, while actual listening tests were being made). At no time were the listeners (auditors) told which systems they were hearing, and in the case of system-to-system comparisons, they were told simply to choose between "A" and "B." Upon completion of the tests, the scorecard results were fed into the FCC's own Honeywell computer. After the test results were compiled and made public, the system proponents were given several months to make comments and defenses about their own systems and the overall test results. Naturally, the FCC commissioners will take all of this into consideration.
Quadraphonic Localization
Discrete quadraphonics has always been recognized as the "ideal" for channel separation, directionality, and localization. To determine how the three matrix systems stack up against a four-channel master tape, the FCC conducted two quadraphonic localization tests, simple and complex. Twenty-five numbered cards were placed around the listening room-16 at eye level around the room's perimeter and nine overhead. The auditor was seated in the center of the room and asked to determine which numbered card approximated the apparent source of the test signals. Both "real" signals (direct from a single speaker) and "phantom" signals (sounds seeming to emanate from a point between speakers) were applied. The simple tests consisted of pink noise from various points, while the complex tests involved electronic "chirp" sounds superimposed over pink noise coming from all four of the speakers. Although these tests are important, their degree of importance is open to question. Naturally, we listen to music-not pink noise and chirping sounds. In relating some of the auditors' comments, the FCC states that many of them prefer that the rear channels be reserved for ambient and reverberant sounds-in which case precise localization (for areas other than the front speaker array) would not be of prime importance. As expected, the 4-4-4 tape won the most "points" in quadraphonic localization.
Fig. 1 In the FCC's quadraphonic localization tests, the auditor was seated
in a room similar to the representation above. Facing the front of the room
(number 3 represents center front) he was asked to indicate which of the numbered
cards placed around the room and ceiling approximated the apparent source of
particular sounds (pink noise and electronic "chirping" sounds).
Auditors were instructed to remain facing the front during the bursts of sound,
but were permitted to swivel in their chairs to see which numbered card corresponded
with the sound's source.
Quadraphonic Musical Preference
This test is one that many people feel is the most important in dealing with quadraphonic FM reproduction as it is representative of the four-channel performance consumers can expect to get in their homes. In this test of system-to-system comparisons, one minute segments of five musical programs were played and repeated through various system combinations.
The five selections, commercially available on quadraphonic discs were: Swan Lake-Leonard Bernstein & New York Philharmonic
Young and Foolish
-Ray Conniff Singers
Moon River Hugo Montenegro Gates of Love-Mystic Moods Orchestra Stars and Stripes Forever Henry Mancini These are the same musical selections that were used earlier in the NQRC listener evaluations. Instead of a simple A or B choice, listeners were given five ratings to apply to each comparison so that their degree of preference could be known: A is greatly preferable to B A is preferable to B A is equal to B B is preferable to A B is greatly preferable to A The general results of this test showed 4-4-4 ahead of the others, but in the case of 4-4-4 against SQ, the lead was very slight, with 52 percent of the auditors preferring discrete and 48 percent opting for the SQ matrix. SQ came out ahead of all the other systems with (surprisingly) the Sansui QS matrix finishing last.
Stereophonic Compatibility
Because four-channel sound has not been such an overwhelming success, it is essential that any quadraphonic FM system be compatible with two-channel reproduction. Even FM broadcasters, who do make the switch to quadraphonics, don't want to lose their two-channel (and mono) listeners.
In determining how well the competing systems did in a stereo folddown, the FCC again ran localization and musical preference tests. This time only two speakers were used, placed 12 feet apart along the 22 foot wall of the listening room.
There are two distinct schools of thought on quadraphonic-to-stereo folddown-undoubtedly brought about by the performance and limitations of particular systems. In both the discrete 4-4-4 mode and in Sansui's QS matrix, the rear channels are simply pushed forward, that is, right-back to right-front and left-back to left-front.
The QS claim is that this type of fold produces the effect of widening the two-channel image. The CBS (SQ) claim is that such a fold causes an unnatural build-up at the two speakers and creates a hole-in-the-middle effect. In a two-channel fold of an SQ program, the rear signals are distributed to the phantom center front area, which CBS claims provides a more natural and pleasing listening experience. Because of these differing philosophies, the FCC did not have a ranking for stereo localization, although the stereo-fold imaging effects are shown in the accompanying illustrations.
Again, it is necessary to point out that we listen to music and most attention has been focused on quadraphonic-to-stereo musical preferences (how a quadraphonic musical program sounds when played through a two channel system). Unlike the four-channel musical preference tests in which 4-4-4 had a slight edge, in the stereo test SQ came out on top, followed by the 4-4-4 system, BBC's H, and again, in last place, QS. Because 4-4-4 and 4-3-4 systems have identical "folds," the 4-3-4 system was not tested.
Table I Quadraphonic Localization Results, Rank, Simple Localization.
Monaural Compatibility
Because there is no localization possible in mono (with all sounds coming from one speaker), the mono compatibility tests consisted only of musical preference. The rankings were identical to stereo: SQ, 4-4-4, H, and QS.
The FCC commissioners have the test results and the responses from the representatives from the competing systems and formats. Now that all the shouting is over (some, at least, temporarily), it is simply a case of waiting to see which way the FCC goes. But, the FCC is not bound by test results. It has several other options, such as:
-Permit discrete FM quadraphonics and prohibit matrix.
Permit discrete and one or more matrix systems.
-Permit one or more matrix systems and prohibit discrete.
-Allow only one matrix and no discrete.
-Tell the proponents that none of the systems are adequate and to come up with a better approach.
Do absolutely nothing and allow things to remain in the state of confusion that now exists, saying that there is not enough public interest in four channel sound.
We'll see what happens. One thing for sure-now is not the time to begin digging that hole in the backyard as a final resting place for quadraphonic equipment.
The accompanying charts and illustrations are taken from the FCC Project #2710-1 evaluation report on FM quadraphonic listening tests, published in August, 1977, by the Federal Communications Commission, Office of the Chief Engineer, Laurel, Maryland. Q
Fig. 2L and R are front speaker locations with C representing the center phantom
channel. The horizontal marks (=-) indicate where in the stereo speaker array
the sound seems to come from.
Table II Quadraphonic Localization Results, Rank, Complex Localization
(Source: Audio magazine, May 1978)
= = = =