SURROUND SOUND - a new era in
home music reproduction
We now seem to be at the beginning of a new era in home music reproduction. SACD, DVD-A and surround sound ‚ is this really the revolution weíve been waiting for? For a little over 20 years the compact disk (CD) has ruled the recorded music market. Now it is joined by the Super Audio Compact Disk (SACD) and the audio Digital Versatile Disk (DVD-A), two apparently competing technologies that claim not only to sound better than the CD, but give us the capability for true surround sound. The CD, and digital audio, entered our lives with the kind of hoopla that suggested we now had the holy grail of perfect sound ‚ all that could possibly change after this was how much more efficiently the digitised information could be stored and disseminated. But all change has its detractors, some simply against it, and others who say the change hasnít gone far enough. Those who claimed the CD audio standard was inadequate have never fully established their case, but there is no argument that the CD was intended primarily for stereo, and not multi-dimensional (surround) sound. This was no oversight, however. The industry at the time was still recovering from a botched attempt to introduce surround sound via the long playing gramophone record (LP), when the two major competing technologies failed in the marketplace. This failure was attributable more to consumer reluctance to make a choice, and implicitly pick the ëwinner,í than to any serious shortcomings in either technology.
SACD and DVD-A could also obliterate each other if past lessons arenít learnt. Coexistence and compatibility should be the watchwords rather than competition. Both these technologies have the advantage of the growing home theatre market, so already there are many surround sound systems for them to strut their stuff on. But there is also the still-dominant CD market. Music lovers who have spent many years building their CD collections might well feel aggrieved and resistant to the prospect of replacing their sound systems and CD collections if either or both of the new technologies take hold. But once again, coexistence and compatibility should smooth things out. Itís important also for music lovers not to be swayed by the sweeping claims for the new technologies and to use their own hearing and good judgement to make their choices. Particularly with reissues of stereo analog recordings, thereís ample evidence that the generally much cheaper CD provides exemplary results. If in fact SACD or DVD-A reissues of these recordings sound ëbetterí, one should question whether the same masters or remastering processes were used. That is, are we comparing apples with apples?
These and other issues are here discussed in more depth, to try and illuminate what is needed to move forward, and to ensure that what is truly worthwhile will survive and bring even more enjoyment to our lives. I offer my reflections and advice as one who has spent many years in, observing, and consuming the products of the music recording industry. For the sake of continuity in this discussion, I have shamelessly merged fact, opinion and fuzzy recollection.
Whatís wrong with the CD? CD sound quality has been a hotly debated issue for the 20 or so years this medium has been with us. Traditionalists repelled by the very idea of digital sound have been all too willing to find fault. Many early CDs were justifiably criticised for cold, hard, grainy sound ‚ often termed 'digital glare' ‚ and reissues of analog recordings sounded congested and lifeless. But looking back now, can we simply say that itís been a case of garbage-in, garbage-out? I believe we can.
As a freelance sound engineer during the 1970s and early 1980s, I found that many of the condenser microphones commonly used for recording and broadcasting, although analog, had 'digital glare' sound. Put them over a violin or body of strings, and the result could be truly unpleasant. Frequently too many were used too close by recording engineers, driven by time and cost pressures. There was also a tendency to over-use signal processing devices such as equalizers and compressors during mastering, which ultimately degraded the sound. Itís worth noting too that the advent of CDs coincided with the early days of the original instrument movement, when it could be very difficult to decide whether the uningratiating sound was the fault of the recording or the performers!
Consider also that the analog 'masters' used for many early reissues were not the original session tapes, but n-th generation edited copies. Take for example the first CD reissues of the Klemperer Beethoven symphony cycle on EMI, which had tired, constricted sound, and truncated reverberation after each movement. Comparison with the LP set suggested that the same tape copies had been used for both. Since then, with the re-masterings for the EMI Klemperer Legacy series, the quality of the original tapes has been far better revealed. And what a difference! By contrast, the Bruno Walter cycle issued on Sony/CBS at about the same time as the Klemperer cycle sounded well from the outset, and has hardly been bettered in subsequent remasterings.
How did these aberrations occur? For analog recordings, with the prospect of reissuing them over and over again with incremental ëimprovementsí, the cynical might say that the record companies had little need to get it right first time. Let's be charitable however and concede that the companies were dealing with a new medium, and were probably only just ahead of the listeners in learning its pros and cons. For analog reissues, it soon became apparent that copies of the master tape prepared for LP mastering (such as in the Klemperer/Beethoven example above) were simply not good enough for CD mastering - the deficiencies were ruthlessly exposed. Only by going back to the original sources could the full potential of the CD be tapped.
Having mentioned LPs, itís worth noting that while the CD is technically superior in almost every respect, the LP has the dubious advantage of masking low-level recording and mastering faults with its own imperfections, such as surface noise. LP loyalists who cite among other things greater dynamism and warmth are only in fact describing the electromechanical effects of overshoot and even harmonic (ëpleasantí) distortion. Having had LPs cut from my own master tapes, I well recognise these effects. Listening to the current crop of CDs, itís obvious that lessons have been learnt. As a source of both original and reissued recordings, one wonders now if todayís state-of-the-art existed in 1980 whether we would be so readily contemplating change (the surround sound issue is separate, but does not necessarily exclude the CD). Donít be beguiled however into thinking the "20-bit", "24/96" or other hi-tech claim on your latest CD means anything really special has happened ‚ weíre still dealing with the same basic replay technology. What these labels do indicate though is the additional effort that is now going into mastering process to ensure that the very best quality is available for transfer to CD.
For those who remain dissatisfied with CD sound, it is worth remembering that when the CD was first released numerous listening trials were held, including by Decca, where typically an invited audience was asked to choose between an analog master tape and a simulated CD recreation of it. Under controlled conditions where neither source was identified and individual preferences were secretly recorded, it was common for the audience to be statistically unable to tell the difference. While such trials have their own validity, they are often quite unsuccessful in changing beliefs and perceptions, even for those who take part. Many will resolutely believe that they can still hear a difference. Which is not say that no audible degradation is introduced by the CD. But it is apparently beyond the hearing acuity of most people, and beyond the resolving power of mainstream audio systems. The connection however between the quality of a reproducing system and the number of people who will hear an effect is perhaps the key. If the new media ‚ SACD and DVD-A ‚ with their higher resolution require a new generation of audio systems to reveal all, then so be it. As a caveat, any such advances must become quickly available and affordable to the mass market.
Of course the CD player itself has figured in the sound quality debate. With the alleged deficiencies of the CD, and that mystical view held by audiophiles, abetted by the audio press, that somehow nirvana is attainable from any source material so long as you throw enough money and esoteric electronics at reproducing it, the manufacturers have had a field day churning out players that offer silk-purse solutions to every perceived CD ill. The list of ills is endless, and it would be pointless to detail any here. Again, let it rigorously be shown that listeners can actually tell the difference between competently designed CD players if this is to be a valid factor.
Whatís new about SACD and DVD-A? Both before and after CD technology was launched on the market, there was considerable technical discussion and criticism of the basic parameters chosen for it ‚ digital sampling of analog signals at 44.1 kHz and storing each sample as a 16-bit digital word, giving a frequency range of 22 kHz, a dynamic range of 96 dB, and a maximum resolution of 1 in 32,767 (or minimum distortion of 0.003 %). At first glance these figures look impressive, but there are pitfalls both in the recording and replaying of CDs. For example, the above dynamic range is actually a sound window that the signal must fit into, without exceeding the top of it, or falling below the bottom. Recording at too low a level means that quiet sounds such as ambience can be lost altogether.
Why not go for a higher sampling frequency and a greater word length, the critics carped. Not enough space on the CD for a reasonable playing time, would be the stock response. Personally Iíve never bought the latter argument. If the consortium that put the CD on the market truly had a need to fit more data on it without changing anything else, they would have gone back into the laboratory and done it. The major recording studios almost immediately built their own systems to record at higher sampling rates and word lengths from 18 to 24 bits. This did not necessarily mean dissatisfaction with 16/44 CD sound quality, but it gave the studios greater flexibility and future proofing for their recordings. The greater word length in particular gave them a much wider sound window to work within. (A 24-bit word means 144 dB of dynamic range ‚ a deafening prospect!) For CD mastering, the digitised sound was down-converted to 16/44. So effectively from the outset of digital audio, there has been use of sampling values beyond those of the CD standard. The 24-bit, 96 kHz standard for DVD-A has in fact been around for some time, and could easily have been considered for the original CD standard ‚ thereís nothing technologically innovative about it, but would have more than trebled the storage requirement of CDs for the same playing times. As noted above, however, I believe the extra capacity would still have been achievable.
SACD technology uses 2.8 MHz sampling of an audio signal, generating a digital bit-stream of ë1ís and ë0ís based on the relative amplitude of successive samples. For replay, this bit-stream switches the polarity of a fast ramp generator to re-create the original audio waveform. Once again, thereís nothing particularly novel about this, and the technology has been around for quite a while. As with DVD-A, however, it requires a much higher storage capacity than CD. So if on paper there is nothing at all new about SACD and DVD-A, we might ask therefore why we should be contemplating replacement of our CD collections simply because the consortium that launched the CD in the early 1980s opted for a lesser specification? We might also ask, with even greater cynicism, why largely the same consortium is now pushing SACD?
But there are always dangers of judgement in hindsight, so letís now look forward. There are dangers here also, not least of which is the human trait to see the "emperorís new clothes", or in other words, that a priori SACD and DVD-A must be audibly better than CD. Implicitly so far Iíve been talking about stereo recordings, and make one further point before moving on to multi-dimensional (surround) sound. Iíve described above the evolution of good remastering practices and the statistical evidence that CD transfers seem good enough for human hearing. In the stereo domain, therefore, it is questionable whether there should be any audible differences between CD and SACD or DVD-A transfers of, say, analog tapes where identical sources and remastering procedures have been used up to the final point of digitisation. I certainly havenít heard any, and would like to statistically test those who claim to hear differences. (Of course one has to know that the sources were identical, and therein lies the difficulty, not least if the industry has decided that it wants these media to sound different to reinforce perceptions.) Most consumers however are not concerned about these kinds of nuances, but will almost certainly say ëwowí when they hear good surround sound, so letís look at that feature of SACD and DVD-A.
Surround sound is one of those things that nobody has ever denied would be a great and desirable improvement, to bring that ëmissing dimensioní into sound reproduction. What we have of course is still not the complete picture ‚ the surround sound models that are commonly used only introduce a second dimension (front-side-rear), rather than a full 3-dimensional sound field that would include the vertical dimension. While 3-D models have been proposed, it is questionable whether enough information would be added to justify the extra equipment, complexity, space consumption and, needless to say, expense. While SACD and DVD-A now offer us everything from straight stereo to six channels of surround sound, itís worthwhile to reflect on what has happened before.
The last serious push for marketable surround sound systems was in the late 1960s and early 1970s when technologies were devised to produce a 4-channel LP that was also compatible with stereo and mono replay. What eventuated was a tussle between two quite different competitors: the Sony/CBS SQ system, with phase-encoded rear channel information, and the JVC/RCA CD-4 system, which created four discrete channels on the LP. Each system required its own replay decoder if you wanted surround sound, and the CD-4 system additionally required a special stylus and cartridge. Without doubling up, therefore, you had to make a choice. As history records, consumers chose neither, and both systems disappeared very quickly. Writing in the 1974 Hi Fi Year Book, shortly before the collapse of LP quadrophony, Donald Aldous comments:
"There is little doubt that at the present time the general public is completely in the dark about four-channel audio and most people still regard anything with four speakers as ëquadraphonicí, whether it is discrete, matrix, ambiophonic or just extra speakers connected in parallel. It is also true that many audio enthusiasts and people in the hi-fi business are equally confused about the different systems. This is probably the main reason for the marked lack of enthusiasm for quadraphonic equipment Ö, in spite of its enormous commercial possibilities. So far the big manufacturers in this field have concentrated too much on the differences between the systems ‚ which are often quite small ‚ and on the supposed advantages of their own methods compared with those of their competitors .Ö the basic need at present is for some first-class, simple explanations and really good demonstrations to let consumers appreciate the increase in realism and the additional musical pleasure that good quadrophony provides."
Sound familiar? My feeling at the time was that neither SQ nor CD-4 produced sufficiently impressive results to justify the cost and domestic disruption of adding the rear channels. Much more impressive were the quadraphonic tapes that appeared to tap the general interest in surround sound, but requiring an expensive and bulky 4-channel tape machine to replay them, they disappeared as well. The good news however is that during this era the major companies made many hundreds of quadraphonic recordings of great performances that we should now be able to enjoy in their full glory through the new surround sound technologies. Many of those recordings have of course already been issued, and reviewed on this site.
The salutary lesson for SACD and DVD-A is that once again we have two different technologies competing in the same market niche. Should it be competition or coexistence, though? Itís not only a matter of backward compatibility with the CD (no inferiority implied!), but ensuring transparency and universality in the replay and decoding systems.
To this extent both SACD and DVD-A have the advantage of an established and growing home theatre market. But while surround sound comes with the deal, the movieís the thing, and I wonder how well the average system owner both appreciates and uses the surround sound component. My own experience, including as a hi-fi salesman, is that for the relatively few that have stereo systems of any quality, even fewer have their speakers placed to produce a good stereo image. I suspect therefore that many of those with surround sound systems are in even greater spatial disarray. As an even smaller market segment, how many will fully appreciate the benefits of multi-dimensional sound, communicate it effectively, and create real demand?
The other ëproblemí with home theatre systems is the perception left by movie sound itself, with its distorted tonal balances, and the synthesised bangs, thumps and rumblings that pass for bass information, grotesquely reproduced by run-of-the-mill systems. Ambience consists of intermittent sound effects from satellite speakers, often lo-fi in quality. Surround sound for music demands much better than this, and needs the kind of public demonstrations that will show consumers it's not just ëmovies without pictures.í
This is perhaps getting too pessimistic, but the market for surround sound by itself may be the smallest and most vulnerable of all. Any perception of direct competition between SACD and DVD-A could be fatal to it. However it is not for this reason that I donít intend to critique here the relative merits of SACD and DVD-A. That I have implied so far by saying that their results will be as good as how they are made and played. Each of course will have its fervent advocates and critics, but that discussion is for another time, and if the advice above from 1974 is heeded, perhaps best avoided. Of more immediate concern is what is needed for them to survive in the market, and bring back surround sound to stay.
Before leaving this part of the discussion, though, it should be said that from the outset CD could itself have been a surround sound medium, and of course it has subsequently offered this facility in Dolby-encoded versions, but often with less than ideal results. That it was not originally developed for discrete-channel surround sound, and marketed as such, was almost certainly because the industry was still recovering from the failure of LP quadrophony.
What effect will the mass market have? We wouldn't have CDs or DVDs if the mass market didnít like them. Or at least they wouldn't be as affordable or as plentiful. The mass market will also determine the success or failure of SACD and DVD-A. But what motivates the mass market isn't always what excites the aficionados. While belief and perception largely drives all preference, the mass market arguably makes the more rational choices, or at least those more concerned with the obvious than the obscure. In the case of the CD, it was not the promise of better sound that moved the mass market away from the LP ‚ it was the convenience of the CD and the means of playing it. And of course its durability. We music lovers and audio enthusiasts should naturally be grateful that it happened. But if we also believe that SACD and DVD-A promise a better sound experience, what else do they have that will succeed in the marketplace? Convenience is no longer a differentiator because SACD and DVD-A are physically identical to the CD, and can be stored and handled in the same way. The market has certainly embraced DVD video, but will it take up DVD audio, or any other substitute for the CD simply because of an arguably better sound quality, and the potential addition of surround sound? What proportion of the home theatre market is also the surround sound market, and what additional surround sound market is there that doesnít have a primary interest in home theatre?
CD quality is clearly good enough for a mass market that takes its music through walkmans, micro systems and the like. It is also still a predominantly stereo market, with lifestyles that largely preclude surround sound for practical listening. And as already asserted, the subjective quality of CD sound for stereo listening is effectively limited only by the quality of the recording and mastering. While the golden-eared fraternity may passionately dispute this, they are not the mass market, and marginal as opinion leaders.
This is a numbers game. The surround sound market, which SACD and DVD-A appear certain to dominate if they survive, is likely to remain small and vulnerable for some time. The CD market, with something like 2 million titles available worldwide, is simply too large and too entrenched to expect that SACD or DVD-A, for any conceivable reason, will cause any rapid obsolescence.
We must ensure that surround sound survives this time around. To do that, we must ensure the survival of the new surround sound media. Itís probably detrimental that SACD and DVD-A seem to be in direct competition, because the main game is to make inroads into a music market dominated by the CD, and relatively insensitive to the nuances of sound quality and reproduction. Coexistence is clearly possible and desirable, supported by the availability of affordable and universal CD/SACD/DVD-A players and multi-format disks. Those manufacturers who currently choose not to make them are doing neither the industry nor themselves any favors. The industry must show more leadership, not just by providing quality and consumer-friendly products that transparently accept all compatible media, but by doing more than just dipping their toe into the software side of the new market. We need them to release many more recordings, both new productions and the gems from those huge back catalogs, that fully demonstrate the glories and realism of surround sound. But we donít need the ëpremium priceí mentality, particularly for reissues, that tries to make us believe that the extremely well-oiled disk manufacturing industry is somehow cost-burdened by the new formats.
Also, disparaging the CD in comparison with SACD and DVD-A is neither helpful nor justified. Itís a distraction most consumers will neither understand nor greatly care about, but theyíll certainly smell snake oil if the argument persists. Any differences must be demonstrated beyond doubt. Itís probably better for the present to promote SACD and DVD-A as complementary to the CD, rather than as competitors. The mass market will ultimately make the choice, but perhaps the success of SACD and DVD-A depends on there not being a choice. That is, if through technological and price compatibility the general consumer does not necessarily have to care that he or she is buying a CD or SACD or DVD-A, because it will play anywhere, and in multi-dimensional sound if the disk and reproducing system are programmed for it.
Those who ëdiscoverí surround sound this way and create more demand for it should inexorably swing the market toward a new reality. Just think of a future when we have the choice of surround sound titles by the millions!
Des Hutchinson is a "music junkie" and audio enthusiast who for some years supplemented his day job as a physicist by working freelance as a recording engineer. While attached to a public FM broadcaster in his home town of Perth, Western Australia, Des took on such projects as the complete Bach organ works with Gillian Weir, and concert series by the Lindsay String Quartet. Through a chance meeting with legendary Decca producer John Culshaw, Des was able to attend sessions at recording industry temples such as Kingsway Hall and Abbey Road, and observe the inner workings of the BBC. Des now has his spare time fully occupied by family duties, with five children under 14. While his sound engineering days are over (for the moment at least), Des reports that family tastes are further broadening his musical horizons.