It now seems certain that discs and players for the Philips/Sony Compact Disc digital playback system will be generally available in the United States by the middle of next year. No event in recent audio history has been so eagerly awaited. In this field only four developments that have come since Thomas Edison made his first recording rank equally with the digital audiodisc in importance-electrical recording, the analog tape recorder, the long-playing 331/3 -rpm record, and stereo. Besides changing the way music was recorded (and, indeed, the types of music that were recorded), each of these developments has marked a new era in the quality of reproduced sound.
The Compact Disc system-digitally encoded, played by laser-has that potential too. Through the courtesy of Sony and Hitachi, I have been able to audition and experiment with pre -production samples of their first CD players (Sony's CDP-101 and Hitachi's DA - 1000) and a small number of CD recordings. I've also participated in an informal laboratory test of the Sony unit at Hirsch -Houck Laboratories. Without much fear of overstatement, I'm prepared to say that the Compact Disc system has the best potential sound quality yet to be offered to the home consumer.
It is expected that when the players are introduced here by the middle of 1983, they will cost $1,000 or less. The price and availability of the discs themselves are crucial issues, to be sure. As of October 1, CBS/Sony has released 112 titles in Japan. PolyGram plans to inaugurate its CD program in Europe with two hundred titles in March 1983 and to release thirty more per month thereafter. Of course, not all of the same CD's will be released in the United States. The price here is expected to be $18 to $20 per disc.
I had four CD's to listen to, three PolyGram sampler discs from Europe containing classical and popular selections and a disc of Mozart symphonies specially obtained in Japan. In every case I could hear no sonic problems that stemmed from the CD system. Compressed, equalized, raucous -sounding popular selections recorded on hissy multitrack analog tape machines sounded compressed, equalized, raucous, and hissy. The classical selections that were recorded directly onto a two - channel digital master tape were uncannily quiet, with stupendous dynamic range, wide and smooth frequency response, and rock -steady imaging. It's easier to cite the sonic problems the Compact Discs did not have. Unlike analog discs and tapes, the digitally mastered CD selections had:
- no ticks, pops, rumble, or "ocean roar,"
- no wow or flutter,
- no modulation noise,
- no end -of -side distortion,
- no variation in frequency response with time or signal level,
- no breakup, no mistracking distortion in loud passages,
- no sense of strain during heavily modulated music, and
- no pre- or post -echoes.
The analog -mastered CD selections were easy to distinguish from the digitally mastered ones, primarily by means of the higher analog -tape hiss level. Several of the selections said to be digitally mastered had more hiss than the others, probably because of noisy mixers or microphones used in the original recording sessions.
Some of the performances on the CD sampler were also readily available in this country on commercially released analog discs (this category includes the so-called "digital recordings"; see "Nomenclature" box on page 64). When I made direct A/B comparisons between a Compact Disc and its analog -disc equivalents, the CD playback was invariably quieter, a property that was especially noticeable as I increased playback volumes to live levels. I was surprised, however, at the close match in basic sound quality of equivalent CD and analog -disc recordings. This means two things. First, that analog -disc cutting and pressing can provide an extraordinary match to a digital master tape, provided good analog -disc playback equipment is used. Second, that a master tape that sounds bad is not made any better by having it available in a digital playback system. In theory, at least, CD players utilizing similar operating principles-like the Sony and Hitachi units-should sound almost exactly the same. And, in fact, these two did sound the same. 1 was not able to make a direct A/B comparison between them because I couldn't obtain two copies of any of the Compact Discs. I did, however, record the outputs of the two units playing the same CD on a two -track analog open - reel machine running at 15 ips. After much tape splicing to link together short excerpts of both players' sound, immediate "A/B" comparison was possible. Within the resolution and accuracy limits of this experiment, they sounded identical.
This is good news for the average music lover, since if most CD players sound the same, the consumer's choice is simplified to deciding which one has the features he wants at the right price. This result may also be bad news for CD -player manufacturers (and their advertising agencies), who must find some way to differentiate their products from those of other makers. Philips, co -developer of the CD system, has described a novel digital -to - analog conversion process utilizing specialized signal -processing techniques. This system, when and if it becomes available, stands the best chance of sounding "more different" from the Sony and Hitachi units.
The Record Itself
A Compact Disc is a small, shiny, entrancing object. Just 12 centimeters in diameter (a little less than 43/4 inches) and 1.2 millimeters thick (about 1/16 inch), it fits conveniently in the open palm of one hand. The CD shown on the cover of this issue is almost exactly full-sized. A CD weighs about 1 ounce and is stiffer than a typical 12 -inch analog disc. The CD center hole is 15 millimeters across (about 5/8 inch), a diameter that seems to have been chosen to let a CD fit easily on a finger. Although holding the disc surfaces with one's fingers causes no damage, the easiest way I found to hold a CD was either slipping it on a finger or gripping the center hole and disc edge with one hand. There is no need to touch the disc surface, even for someone with small hands.
The flat, plastic folding box the disc comes in contains a mounting hub that supports the disc so that it doesn't touch any other part of the package. The box measures 57/16 x 47/s x Vs inches. It is slightly more unwieldy than a typical tape -cassette box. Perhaps the most surprising aspect of the disc, aside from its size, is the playing surface. First of all, there is only one-the other side is covered with the label. The playable side will hold more than 60 minutes of nonstop music, and there are hopes of extending the playing time to around 75 minutes. As a Sony engineer told me, one of the informal requirements for the system was that one disc had to be able to hold Beethoven's Ninth Symphony, which runs slightly over an hour. This is a rare and laudable instance of old art influencing new technology.
The recorded information is carried on an aluminized "signal surface" that lies within the disc. It is protected on the unplayable side by a thin layer of plastic and the disc's label, on the other by a much thicker layer of plastic. This thick layer not only protects the signal surface from scratches, chemicals, and other damage but also reduces the deleterious effects of such damage by keeping it relatively distant from the signal surface inside. To the scanning laser focused on the imbedded signal, minor surface defects are out of focus and have little effect on the quality of the scanned signal
"The Compact Disc system has the best potential sound quality yet to be offered to the home consumer."
The laser beam in a Compact Disc player is aimed at a spiral track of "pits" impressed on the signal surface. (They are manufactured as "pits," but the laser scans them from the other side and sees them as "bumps.") These pits are incredibly tiny, about 0.5 micrometer wide, and they vary in length, depending on the signal, from 0.833 to 3.56 micrometers. (A micrometer, or micron, is abbreviated Am and is approximately equal to 1/25,000 of an inch.) The successive rotations of the spiral are only 1.6 Aim apart; sixty CD tracks would fit in the average analog - disc groove. The "information density" of a Compact Disc is from thirty to one hundred times higher than that of a conventional analog disc.
A CD track runs from the inside of the disc to the outside. As seen from the label side, the disc turns clockwise, but from the laser's "viewpoint" it spins counterclockwise. Its spin rate slows from about 500 rpm to about 200 rpm as the laser scans outward. This slowing keeps the track passing over the laser at a constant rate, since the digital -audio track contains a constant amount of information per unit of its length. With analog discs, which spin at a constant rate, the information content per unit of traveled groove increases toward the inside of the record, placing stringent demands on both cutting and playback styli near the center of an LP. Encoded in the length and spacing of the pits are the binary "ones" and "zeros" of sixteen -bit -quantized stereo sound But that's not all. The pit sequence also contains eight channels of "subcode" information. Only two of these channels have been used so far in the first CD pressings and players. Information carried in these "P" and "Q" channels includes a table of contents of the disc (with the total number of selections, their timings, and their starting locations); control codes governing player operation; a "music -start flag" that tells the player when a selection begins; track and indexing numbers (there can be up to ninety-nine "addressable" selections or indexing points on a CD); and a time code (for displaying elapsed times in minutes, seconds, and seventy-fifths of a second). As yet unutilized are six additional channels of digital subcodes. These might eventually be used to carry text, lyrics, librettos, and other graphic or verbal information for display on a TV or computer printout. (For you personal -computer hackers, the R -W subcode channels will each support a maximum data rate of 7.35 kilobits per second. If this rate is completely utilized, a large amount of text can be put on a 1 -hour CD: about four million words, assuming no data redundancy is used for error correction.)
Using a CD Player
Player operation, depending on the model, is very much a cross between using a normal analog turntable, a computer -controlled tape deck, and a videocassette recorder with advanced picture -search capabilities. The Sony CDP-101 and Hitachi DA -1000 have several features in common that lead me to make that statement. Both units have cueing facilities unmatched by any previously available audio medium. At the push of a few buttons a user may "access" any selection ("band" or "cut" in analog -disc terminology) on a Compact Disc. The whole operation takes no more than five to eight seconds regardless of the number of selections on the disc or what is playing when you change selections. The cueing is exact; the audio outputs are muted until the selection starts. Pausing while playing is no problem, and, unlike analog -disc players with their relatively inexact tone -arm cueing mechanisms, releasing a CD player from pause starts the music where it left off.
"electronically speaking, these are probably the most sophisticated and complex devices ever made available to the audiophile."
Skipping forward or 'backward to the next or a previous selection is also possible with the Sony and Hitachi front - panel controls. The players have slightly different auto -repeat functions. The Sony unit allows repetition of all or any part of a disc; the Hitachi seems to allow repetition only of whole selections or sequences of selections. Both players display elapsed time in minutes and seconds. The timing starts over with each new selection, but front - panel buttons allow momentary display of the elapsed time from the beginning of the disc. The selection number is also displayed, changing as the recorded sections pass by. The players both have headphone outputs, Sony's having an independent level control. Their rear -panel output phono jacks are driven at a standard audio -component line level to feed into a stereo system's auxiliary inputs. Hitachi's headphone output is controlled by a pushbutton volume control, which also affects the level of one pair of its four output jacks.
The description "compact" applies to both the Sony and Hitachi players. The former measures 123/8 x 137/8 x 4 inches and weighs 15 pounds. The Hitachi's dimensions are 9 x 121/2 x 53/4 inches, and it weighs 13 pounds. In bulk these players are smaller than most analog turntables or cassette decks, though they weigh about as much. To load a disc into the Sony CDP101, one presses an OPEN/CLOSE button on its front panel. A drawer then slides forward to receive the CD. The CD is placed face down, label up in the tray, the OPEN/CLOSE button is pushed again, and the tray slides back into the player. With the Hitachi DA -1000, pressing its OPEN/CLOSE button causes a door in the center of the unit's front panel to swing down. The disc is then inserted into a slot in the door; when the OPEN/CLOSE button is pressed again, the door closes and holds the disc ready for playing.
The Sony player has two features the Hitachi player does not: fast and slow backward and forward scanning. These controls are similar to the fast -forward and rewind functions on an analog tape deck, but using them to scan a Compact Disc does not produce a high-pitched squeal as on a tape deck. Instead, you hear brief, low-level snippets of the passing music at normal pitch but greatly increased tempo. These search controls, together with Sony's pause control, allow very rapid cueing of individual musical phrases if desired. The Sony unit comes with an infrared remote -control unit duplicating all the front -panel functions except the OPEN/ CLOSE button. The remote control also allows direct entry of a selection number on a numerical keyboard. As if to compensate, the Hitachi DA- 1000 has an interesting feature the Sony does not: programmable playing sequence. If you want to hear the selections on a disc in scrambled order, possibly repeating some too, you can enter the sequence on the Hitachi's front - panel controls. From the variety of different features on just these two units you can see the potential the CD system has for a myriad of convenience features. Each year's CD -player models will probably have a different array of features, as has happened with videocassette recorders. So if you can't find a feature you want on the initial offerings, hang around.
A Look Inside
Incredible as the discs themselves are, I am even more impressed by the technological achievement represented by the Compact Disc players. These machines must extract and transform all that digital information into usable form. A look at the inner works of two CD players has convinced me that, electronically speaking, these are probably the most sophisticated and complex devices ever made available to the audiophile.
That sophistication starts with the CD equivalent of a phonograph stylus: a solid-state laser. Similar in operating principle and construction to a light - emitting diode (LED), the laser used in a CD player emits a small, invisible beam of light in the near -infrared (wavelength = 780 nanometers). At the outer surface of the disc the beam is about 0.8 millimeter wide. By the time the beam reaches the signal surface, the combination of player optics and the light -bending properties of the disc's thick plastic base have focused the spot down to a 1.7-µm diameter.
"Surely any home system that is now capable of handling CX- and dbx-encoded discs will be able to take on Compact Discs."
This tiny spot of light must follow the spinning track of pits without benefit of a guiding groove. In both the Sony and Hitachi playing mechanisms, the coarse positioning of the laser beam is accomplished by a motor -driven assembly that carries the complete laser/optics portion of the player. Fine beam positioning is controlled by two small magnetic coils. Working like loudspeaker voice coils, they position the objective lens that aims the beam; one coil keeps the beam centered over the center of the pit track, and the other coil keeps the beam focused on the signal surface. The exact process by which the laser tracking -focusing system operates is a fascinating study in electromechanical servo systems. Much of the circuit - board space in the two CD players examined here is devoted to the laser tracking controls. The pit -tracking servo utilizes two auxiliary laser beams derived from the main beam and works to balance their reflections from the signal surface. The focusing servo monitors the shape of the reflected main beam and moves the objective lens until the reflected beam makes a circular spot on the player's "pickup," a light-sensitive device called a photodiode.
When the laser beam falls on a flat portion of the signal surface, it is reflected directly back into the optical system and onto the photodiode. If, however, the beam hits a passing pit (seen by the beam as a bump, remember), the beam is scattered and little light returns to the photodiode. These invisible "light" and "dark" flashes, corresponding to the presence and length of the pits scanned by the laser beam, generate in the photodiode circuits the "on" and "off," "one" and "zero" pulses of the recorded digital signal. This is a CD player's transduction stage. From the output of the photodiode until just before the final audio outputs of the player the signal path becomes very complicated. You can get a feeling for the complexity of the operation from a bald statement of what must be done: a digital bit stream with a data rate of 4.3218 megabits per second must be converted to two "analog" audio signals of high quality. First the photodiode signal is amplified. Then "clock regeneration" circuits lock onto the bit stream in order to synchronize the rotation of the disc with the crystal -controlled digital signal -processing circuitry. The next step of the process is "demodulation" of the bit stream; digital circuits convert the data from one digital code to another (from I4 -bit to 8 -bit "symbols"). In the demodulation process, the sub - channel signals discussed above are isolated and sent to the player's control and display circuitry. What's left is the encoded digital -audio signal together with its error -detection and correction data. Data errors caused by disc damage are then found and, if possible, corrected (see box on page 70). If the errors from a damaged disc are too numerous for exact correction, error -concealment circuits interpolate approximations of the missing data. These "cleaned -up" binary numbers are then demultiplexed to separate the left- and right -channel information. This is the final all -digital step. Digital -to -analog converters then receive the digital information for each channel and convert it into varying audio voltages. These signals are separately filtered to remove ultrasonic frequencies and fed to the player's audio- output jacks. Music at last! But no short summary can do justice to the circuits involved. The multitude of different signals coursing through the digital -decoding and servo -tracking circuits would be daunting to experienced computer designers. In fact, the Sony and Hitachi players contain computers in the form of specially programmed microprocessors. Directed by the front -panel controls, these microprocessors intercede between the human operator and the maze of signals within the player. Considering what goes on there, the astounding thing is not that CD players work so well, but that they work at all.
Given the v.ide dynamic range possible with the CD system, concern has been expressed about the compatibility of the system with typical home stereo Components. STEREO REVIEW has already published an article on this subject ("Superdiscs," December 1980) in which it was generally concluded that most home systems with reasonable output power and reasonably efficient speakers will be compatible with the output from a CD player. Surely any home system that is now capable of handling CX- and dbx-encoded discs will be able to take on Compact Discs. The wider dynamic range of CD playback did not encourage me to play music any louder than I play analog - disc recordings. Instead, I found that setting the same volume for digital- and analog -disc playback uses the low -noise properties of the CD medium to best advantage. If you are concerned about having enough amplifier power for the high peak levels a CD can generate, take a look at your amplifier's dynamic -headroom rating. This specification gives an idea of an amplifier's short-term peak -output ability. A high dynamic headroom (above 2.5 dB) will greatly reduce the chance of overload with CD playback.
Most conventional loudspeakers should have no trouble absorbing the high peak levels generated in CD playback, provided the system is played no louder than with analog discs. Of course, you can always use headphones, plugged directly into a CD player's headphone outputs (which all manufacturers should provide). Headphones will also protect your neighbors, if not your ears, if you do yield to the perfectly natural inclination to turn the volume up to louder -than -life levels.
Compatibility with present-day audio systems is actually the least significant obstacle to the success of the Compact Disc. The true problems with the Compact Disc system, all of which can greatly influence the record -buying habits of the audiophile or music lover, are subject to very little control by the consumer.
It is reported that the manufacturing costs of Compact Discs are very high, at least three times those of equivalent analog pressings. A CD mastering system costs about $1.8 million, not including the necessary digital playback equipment. A specialized CD manufacturing facility can cost many times that amount. These factors will certainly make it hard to keep CD costs down, but if the prices of CD's cannot be kept within the levels currently projected - 15 to 20 per cent higher than today's analog "audiophile" pressings-wide- spread adoption of the system will be slowed.
Software availability will probably be a significant problem for at least the first few months of the system's commercial life. At present there are only two CD pressing plants in operation, CBS/Sony's in Japan and PolyGram's in Germany. Two more are scheduled to go "on-line" soon, Denon's and Toshiba/EMI's, both in Japan. However, the United States consumes more than a third of the world's recordings, and at press time there were no announced plans by any U.S. company to start a U.S. pressing operation. Depending on the system's popularity in this country, the problem of software availability will range from severe to critical, at least at the start.
"The main musical problem with the Compact Disc is not its sound quality . . . [but] the large number of fine recordings of the recent past that will never make it to the new medium."
One of the reasons CD costs are higher than those of analog discs is that manufacturing problems have been keeping down the "yield" of usable discs. The CD's larger brother, the LaserVision videodisc, has been scandalously difficult to manufacture. CD's are smaller and have the advantage of a very comprehensive error correction and concealment system, but neither PolyGram nor Sony has released any disc -yield information. That's not very promising, and, considering the extraordinary audio quality of the discs that do work, it's quite unfortunate. The main musical problem with the Compact Disc is not its sound quality, as some opponents of digital audio seem to think. Instead, to my mind, it is the large number of fine recordings of the recent past that will never make it to the new medium. Stereo has already visited this fate on most recordings from the 78 -rpm and mono -LP eras. The same kind of musical extinction has occurred with each of the main technological revolutions in recording. I find that the Compact Disc's greatest sonic asset can also be its greatest liability: the system lets you hear exactly what's on the master tape. Many of the popular -music selections on the PolyGram samplers did not sound particularly lifelike, and they probably weren't intended to. But then neither did quite a few of the classical selections, and they should have. Because of the recording philosophy behind the latest Deutsche Grammophon digital recordings of Herbert von Karajan leading the Berlin Philharmonic, they have sounded unrealistic to me when I heard them in commercially released analog pressings. I had hoped that the excerpts from some of these sonic blockbusters (Shostakovich's Tenth Symphony, Holst's The Planets, and Strauss's Thunder and Lightning Polka) would sound less obviously multimiked, less artificial in perspective and ambience in digital playback from the PolyGram samplers. But they did not. Digital recording combined with digital playback creates a sound -reproduction chain that is unequaled in clarity. I hope recording producers and recording engineers will quickly learn to produce master tapes made with the intent of capturing a realistic sound quality in order to realize the full sonic potential of the CD system.
With these problems in mind, and with the schedule for U.S. introduction of CD players and software not established beyond being sometime in the first half of next year, it is too early to speculate on the system's ultimate success in the audio marketplace. By its sonic and convenience virtues alone it should do as well as product availability and pricing (neither yet specifically announced) will allow. What can safely be predicted from a first look at this new technology is where it can lead, how the potential of high -density digital -audio storage and playback can be fully realized.
Two areas the CD designers have had in mind since the system's inception are portable and automotive applications. The protective layers of a CD make it virtually immune to damage from normal handling in car or portable applications_ (I haven't yet tested its resistance to warpage from high pressure or the high temperatures commonly encountered in cars.) However, the inner works of the two players examined here will have to be much less bulky for in dash mounting. (The mechanisms seem to be stable enough; fairly violent shaking of the players while they were playing caused no audible trouble.) The IC's used in portable and car CD players will have to draw less current than those in the home -use models seem to do. The analog laser -tracking circuits will also have to be simplified or turned into easily integrated digital form before such uses for a CD player become practical. All these considerations pale in comparison with the question of whether 90 decibels of dynamic range is useful, or even desirable, in a car. This area, however, is one in which digital signal processing can make an important contribution by way of programmable compressors, and Sony is known to be working on such devices. Digital signal processing, operating directly on a CD player's digital signals before they are turned into audio outputs, may eventually substitute computer programs for a home stereo's volume, tone, balance, equalization, imaging, and ambience controls. Digital loudspeakers and microphones are also not impossible, so there might ultimately be an all -digital signal chain from recording -studio microphones to home listening -room speakers.
Less speculative are the control advances possible with the CD system. The precise cueing, programmable sequencing capability, and the predictable, repeatable output levels of CD players make them ideal for radio -station use, though few stations in the world can (or want to) broadcast a signal close to the quality possible from a CD player. The front -panel readouts of selection number and elapsed time should prove invaluable to radio DJ's.
That's all in the future. Within a few months the audiophile and music lover will be able to buy, at high but not unreasonable prices, a playback system and recordings that will together provide audio quality superior to anything available at present from any con- sumer -level system. The Sony/Philips Compact Disc system does indeed inaugurate a new era in the history of sound reproduction, the age of home digital playback. In the CD system the technological arts have given us a masterpiece of carefully considered engineering. I hope the Compact Disc's potential will be realized in a way that benefits the art form for which it was developed: music.