How Recordings are Made [part 2]

<<prev.

TAPE

The controversy over reel-to-reel versus cassette is all but dead now, with cassettes the clear victors in the consumer marketplace. In fact reel-to-reel has been so thoroughly vanquished that the last source of prerecorded reel music tapes, Barclay Crocker, closed down in mid-1986.

However, shortly thereafter, Teac brought out an “economy” model of a reel-to-reel deck, comparable in price to a good cassette deck. Open reel’s superior sound may still keep it alive.

When the cassette deck was first introduced by Philips in 1964, it was never intended to be a high-fidelity medium. Then tape formulations were improved and, most importantly, noise reduction was introduced. These factors significantly improved cassette’s sound, but many still disputed its candidacy as a hi-fi medium. Many still do. Nonetheless, 1984 was the first year that prerecorded cassettes outsold LPs— largely thanks to portable cassette players. Both tape and tape deck quality have continued improving to the point that cassette is becoming a worth while medium for quality listening.

While cassette recordings are still sonically inferior to open reel or LP, their convenience and economy ensured rapid success. Cassette players can be taken just about anywhere for recording or listening, and they are about half the cost of open reel to operate.

All tape, whether for reel-to-reel or cassette, is a ribbon of thin, flexible plastic coated with a film of magnetic particles suspended in a binder or glue. Signals are recorded by varying the patterns in which those particles are magnetized. In playback, the process is reversed and the varying magnetic field on the tape is translated back into an electric signal.

Fidelity in tape recording is determined by the speed at which the tape is recorded and played back, the width and thickness of the film backing, and the thickness and quality of the coating, which is the actual recording material. On all of these counts, cassette tape is substantially inferior to open reel. In addition, commercially recorded cassettes are generally less good than a recording you carefully make yourself on a good machine. In one area only does cassette tape currently exceed the quality of open reel and that is in the quality of the coating material itself. Exotic oxide formulas that would be exceedingly expensive on open reel are practical when applied thinly on the narrower, shorter cassette tape.

Reel-to-reel is the oldest magnetic tape format and of course is still used commercially for recording. It is simply a spool or reel wound with tape that, during playback, simultaneously unwinds from the one reel and winds up onto an empty reel—whence its name. It is not only sonically the best but also provides the best reliability, durability, and, for editing, flexibility. Because higher speeds are used than with cassette tape, more tape passes under the recording head in a given time, and a given “quantity” of music signal has more tape over which to spread itself. The signal is in effect stretched, with more room for de tails to register, which is another way of saying that because there is more room on the tape, the signal is less compressed and so the fidelity is better. This is similar in principle to LPs, where the less music per side, the more room the grooves have to spread out, and so the less compressed the music signal.

The cassette format, in contrast, while cheaper, more compact, and more convenient, does sacrifice some fidelity. Because there is less tape in a cassette, it must be run more slowly to cover a reasonable amount of time—generally 1 7/8 i .p. s. compared to open reel’s 7½ to 15 i.p. s. on domestic equipment, or 15 to 30 on commercial equipment— and so all the problems of wow, flutter, and splice audibility that are alleviated by higher speeds are now clearly exposed.

A small tape imperfection will at slow speeds loom comparatively larger than the tiny width of a moment’s worth of signal. Also, as a signal varies from low to high, the corresponding change magnetically impressed on the tape from millimeter to millimeter is more drastic or rises more steeply than when stretched out over 5 or 10 millimeters. Thus, with any slight jerk or lag in the movement of the tape (more commonly known as wow or flutter), the tape head will “see” a much sharper change in the signal than if it had been laid down with a more gradual increase per millimeter on fast-moving reel-to-reel tape.

Since the area of tape that the recording head influences is not infinitely thin, but has a certain width, the more slowly it moves by the varying signal source, the more the pattern of each instant will overlap with its neighbors and be less distinctly “itself.” Sharp distinctions of each moment’s detail are the essence of high fidelity. But as the moments are piled one on top of each other by slow-moving tape, their individuality is forever lost. This is known as compression.

A cassette is similar to open reel except that the two reel hubs and tape are miniaturized and enclosed in a polystyrene shell. The tape is about half the width of open reel tape, and also thinner in order to fit enough into the small housing. Because of the thinner plastic film, the tape can stretch or break more easily. There is also a higher incidence of print-through than with open reel’s more durable, thicker coating on a thicker tape base. Print-through occurs when the signal on one section of tape bleeds-through onto the adjacent piece of tape, resulting in “pre echo.” You hear the music that will be coming up before it actually does come up, sort of like a preview of coming attractions.

A disadvantage of commercially prerecorded cassettes is their manner of production. Cassette tapes are gang-produced at a superfast speed, as much as 64 times normal listening speed, on special mass-production high-speed duplicators that add noise and distortion. In general, the higher the reproduction speed (as opposed to recording speed), the greater the loss of high frequencies and the greater the tape hiss. Cassette producers seeking better fidelity will reproduce either in real time (the most expensive method since a tape with 36 minutes of music will take 36 minutes to duplicate), in half time, or at a still relatively low ratio of 8 or 10 times playback speed.

This same attitude represented by gang production can be found in other areas of quality control. As many as half of the prerecorded cassettes you buy are likely to have one defect or another. Because the cassette is such a convenient medium, and most people listen on low- grade decks, they expect neither good sound nor good QC. Poor quality is accepted almost without notice or comment. Manufacturers, receiving only this feedback of indifference, continue happily to produce low- grade cassettes.

A serious drawback—and there is some suspicion that this happens more with commercial recordings than with home recordings—is that the recorded signal tends to deteriorate quickly. After some months, the high frequencies become duller and rolled off, the overall sound more veiled, and pre-echo more pronounced. Print-through is more likely to occur when a tape is recorded at a high speed. Most of it occurs during the first few days of storage and is more severe when stored under conditions of high temperatures and if subjected to a slight magnetic field.

When buying a prerecorded tape, you don’t have any choice as to which grade of tape you buy—if you want the recording, you have to accept whatever tape it comes on. But if you make your own recordings, blank tapes are offered in a potentially confusing array of choices. The least expensive tapes often end up being actually very expensive because the housings jam and malfunction easily. The most expensive, on the other hand, may offer more than you need. For example, taping off AM radio seldom requires the best tapes as the fidelity of the AM signal leaves a lot to be desired anyway. On the other hand, if you want a good recording of a live performance off FM, a high-fidelity tape is worthwhile. Match the fidelity to the end use. Don’t buy cheap cassettes—if you’ve put in the effort to record something, presumably you want to preserve it, at least for the short term. You don’t want your tape to turn out defective.

By the way, you can protect against accidental erasure by breaking out the plastic tabs in the cutaway holes in the top (thin edge) of the cassette. These engage a safety interlock in the recorder that governs the recording and erasing circuits. With the tabs removed, the deck will neither record nor erase. If you later change your mind and decide to erase or record over the existing recording, just tape over the tab holes. If you want to protect (or unprotect) just one side of the tape, here’s how to tell which tab goes to which side: When you hold the cassette so that side A faces you, the side A tab is the one on the top of the cassette near the left end. The other one is for side B.

Unlike LPs, cassettes have no single, standardized tape format. There are four different tape types, requiring four different bias and equalization settings. In addition, precise bias-setting requirements for recording vary slightly from brand to brand, even within the same tape type. The instructions for your tape deck will generally specify which tape type will produce the best-quality recordings on that particular ma chine. But the manufacturer may not have experimented with everything available, so it’s probably worth your while to experiment with a variety until you’re satisfied. You can use up all the less successful tapes for recording radio shows or some other lower-fidelity use.

Tape Formulations

The three basic formulations of cassette tape are ferric, chrome, and metal. Tape formulations are also broken down into four “types” assigned by the International Electrotechnical Commission. Each basic tape formulation, whether ferric, chrome, or metal, has a different ability to receive and retain an audio signal and each requires a different magnetic strength, or bias setting, to record the signal properly. These characteristics in turn determine each tape type’s frequency response, dynamic range, signal-to-noise ratio (how much hiss will be present), and saturation levels (how much audio information the tape will hold before distortion or signal loss occurs).

Bias setting adjusts the magnetic field strength applied to the cassette tape during recording. Normally, magnetic particles don’t respond linearly in a magnetic field. Adding the correct amount of an inaudible, very high frequency signal ensures that the particles all respond linearly, and thus prevents distortion.

Because bias needs can vary from brand to brand even within the same tape type, it’s a good idea either to stick with the deck manufacturer’s recommended brand for which the deck has been specifically calibrated, or else to do some serious experimentation and come up with a brand that sounds the best. Or if you already have a brand you prefer, the deck can be recalibrated to match.

EQ or equalization corrects for different tapes’ varying responses to different areas in the frequency range. Equalization “equalizes” the frequency response across the audio band. Most tapes are marked to show which EQ to use; those that aren’t almost always are Type I and require a 120-microsecond setting.

Noise reduction capabilities, such as the various Dolby methods, will reduce hiss and improve the signal-to-noise ratio. Signal-to-noise characteristics vary from tape to tape. All cassette tapes have a certain amount of noise or high-frequency hiss, which can be heard during quiet passages, blank spots, or when the volume is low.

Dolby HX Pro, developed jointly by Dolby and Bang & Olufsen of Denmark, is a recording process that requires no special decoding for playback. This is not a noise reduction system but a headroom extension circuit that increases the high-frequency storage capacity of a tape.

The four IEC tape types are as follows:

Type I: commonly referred to as “standard-bias” or “normal-bias” ferric or ferrichrome tape. Requires the normal 100 percent bias current and 120-microsecond EQ. This is the oldest and most economical formulation and can be used successfully on any cassette deck, from the cheapest boom box to the best professional machine. Top of frequency response falls within the range of 10,000 to 15,000 cycles.

Type III: seldom-used ferrichrome. Low fidelity. Requires 70-microsecond EQ and a low to medium bias—use low-bias setting.

Types II and IV are considered to give better fidelity and are also more expensive. They are “high-bias” tapes, requiring bias starting at 150 percent and 70-microsecond EQ. They have better tape saturation, lower hiss, greater ability to reproduce high-frequency sound, and greater ability to reproduce music of wide dynamic range. In general, the higher the bias required, the better these characteristics. The main drawback to these chrome and metal tapes, aside from their greater expense, is that the backings are highly abrasive and tend to wear out the heads faster. They can also be successfully played back only on machines that offer the capability of playing them.

Type II: chrome and pseudochrome or chrome-equivalent formulations, requiring high bias (roughly 150 percent) and 70-microsecond EQ. Originally formulated with chromium dioxide, this was developed and patented by Du Pont, but competing companies quickly came out with chrome substitutes. Type II represents a compromise between cost and fidelity.

Type IV: In these tapes, the magnetic materials are made of particles of pure metal alloy rather than of metal oxides and so can be magnetized more intensely than ferric or chromium tapes. Metal tape is also more expensive. It requires 70-microsecond EQ and high bias settings. Can be played on any deck but, in order to record with these, the deck must have the metal tape position.

COMPACT DISCS

Digital technology is in its infancy, even though it has been under development for about 15 years. Commercial release of the CD was pre mature—perhaps out of need to recoup the enormous sums put into development, or from eagerness to revive a flagging recording industry—and the cost was to the CD’s credibility. The mid-fl press em braced the medium without reservation, and Herbert von Karajan became so enthused as to proclaim that “all else is gaslight.” But high-end listeners and press, used to enjoying far better fidelity from their analog systems, gagged on the sound and said so vociferously. We have these people to thank for the subsequent advances in CD sound.

Those digital proponents who acknowledge that digital recordings are flawed—now a steadily growing group—generally try to place the blame anywhere but on the digital method itself. First the miking techniques were blamed as being unsuited to digital (interestingly enough, the simple miking techniques favored for “purist” analog recordings are now increasingly employed for digital, a trend started by Jack Renner of Telarc). Then the playback electronics were faulted—most players contain an analog playback section in their circuit that would put even the cheapest boom box to shame. Players in which this section has been upgraded do indeed sound better. In the haste to ship vast quantities of a product whose technology was barely understood even by those whose job it was to use it, many mistakes were made. Unfortunately, nearly all digital recording equipment still has an awful analog section.

The CD is forcing a reassessment of sound quality, even though it is not setting the pace. It has renewed the public’s interest in audio. Compact discs’ virtues are undeniable. And for the many people who have been listening on shoddy turntables and/or poorly set-up tables, CDs can represent a great sonic step forward.

Curiously, people who come into a store to buy a CD player often find, on comparing CD and LP, that they prefer the analog recording, and so are instead walking out with a top-quality turntable. Many of the early buyers of CD players who had moderately good tables (like an AR or Thorens) were disappointed with the CDs and realized that a good table can still provide the maximum return for money invested. It is also more expensive to gain entry-level high-end sound from digital than from analog, though prices even out somewhat above entry level. An important factor is that the cost of compact discs is double that of LPs and is unlikely to come down soon (though there are indications that the price of LPs is being increased to reduce the gap).

While the high-end players are improving in their ability to play a CD, where improvement is desperately needed is in the digital recorders. All the understanding the high-end designers have put into the players has yet to be applied to the recording equipment—it is the weakest link in the digital chain. This is a worse offense than poor players, because the “software” is the heart and soul of the music system. With LPs, the software was better than the hardware for a long time. But then as the hardware slowly improved, it was able to reveal more and more of what was captured on the recordings. If CDs are poor, the better the players become, the more they will reveal the failures of the recordings. If the recording itself is bad, no amount of improvement in the playback can do a whit of good. Unrepeatable works of art—whether performances, compositions, or artists recorded only once or only on digital—are being stored by a medium that does not do them justice.

The Advantages of Compact Discs

The three strongest features used to sell CDs are their convenience (unquestioned), virtual indestructibility (alleged), and fidelity (occasional).

CDs are certainly convenient—as easy to use as cassette tapes and requiring none of the fuss and bother associated with LPs. Pop a small disc into its receptable, push a few buttons, and sit back and listen for 40 minutes or more.

The claim of CD indestructibility appeals greatly both to those who take a certain decadent joy in not bothering to care for their recordings but also, more positively, to those who do in fact care deeply about preserving music for future generations. It would be wonderful if music could be stored on a medium less subject than most things to the frailties of life and the second law of thermodynamics (which, roughly translated, states that all things left alone will gradually drift into chaos). Even if CD is a more enduring storage medium, the question still has to be considered whether it stores the MUSIC as well as does analog, or whether it simply stores the bad sound well.

One of the most exciting claims for digital is that there is no loss of information or sound quality between generations because the music is recorded in numbers, which cannot alter. This theory may be impeccable, but recording engineers and listeners have regrettably reported audible degradation between generations, just as with analog. In fact, digital, in both tape and compact disc form, is a more fragile medium than generally recognized. Compact discs are impervious to “stylus wear” since the laser beam makes no physical contact with the disc. But this single fact has been exploited to imply that CDs are impervious to any wear or damage.

Just because the laser makes no contact with the disc, and there fore neither causes wear not “reads” scratches as pops and ticks, doesn’t mean you can use your CDs as bar coasters. Despite their plastic coating, they are as subject as anything else to damage. On the label side, which is the side all the information is encoded on, there is no hard plastic covering—only a very thin lacquer barrier protects an aluminum reflective layer less than seventy millionths of a millimeter thick. Scratch this and you’re wiping out information that no “error correction” system can successfully compensate for.

Listening to a damaged CD with missing information is like trying to read a damaged book from which some of the pages have been torn. You can still get the gist of the story and can even fill in the gaps for yourself (as does error correction) but if there are a lot of pages missing, you’ll lose the author’s structure, even misapprehend the meaning through incorrect interpretation of missing information. A worn LP, on the other hand, is more like a used, rather than a damaged, book. LPs can be come scratched, the groove walls worn, and the surface noisy. But the music remains in the record. A different stylus, riding below the worn area of the groove walls, can capture much of the music. Properly cleaned and played, the LP preserves the music in its entirety and so can still provide a deep emotional experience. Finial Technology of California has developed a laser turntable that plays records without any stylus contact. This would eliminate one of the main disadvantages of LPs.

This basic difference—that on CD information is lost totally, while on LP it fades or is covered over by a fine layer of noise—applies also to the digital master tapes, which are much less durable than their analog counterparts, tape, like everything else, is an imperfect medium. Analog master tapes gradually fade over time, like a print exposed to sunlight that slowly bleaches away. While the music, if the tape is very faded, may become a little ghostly, and the highs will be gone, nonetheless the music is preserved. Digital master tapes, on the other hand, are encoded in a sequence of numerical bits. Because this is not the infinite sampling rate of analog, but only a discrete sampling rate, there are gaps between the bits, which during playback are essentially “re constructed.” But, when a piece of the metal oxide flakes off with aging, a bunch of numerical bits are taken along with it and there’s now in effect a gaping hole in the music.

Apparently earlier tape formulations, which had larger particles, are storing well. The modern ones, especially American tapes, with very small particles scrunched together for low noise, are fading more. And the digital master tapes are not storing well at all. The tape designer for 3M says 6 dB of erasure and it’s gone. Some smart engineers are copying their digital tapes over each year—but how many people know about this, and how many major companies are likely to take such trouble? So music being recorded digitally may turn out to be far more transitory than analog. Digital may turn out to be actually offering “imperfect sound for only a little while.”

CD “silence’ ‘— “The music wells forth from a black velvet back ground’ ‘—is a major selling feature. If digitally recorded—and many CD releases are analog reissues, not digital—then CDs are indeed quiet, with no tape hiss or surface noise. But who hears a black velvet back ground at a live performance? The real importance of this in terms of the music is questionable. Out of that silent background can come some impressive sonic effects, but analog places real players with real instruments in a real room, something digital has yet to be able to do.

On more recent digital recordings, a form of audible background noise called dithering, which sounds very much like tape hiss, is intentionally added to improve digital fidelity. In our view, surface noise and tape hiss on LPs is a very minor issue—this noise is clearly separate from the music and easy to ignore, at least on any good system. Mediocre turntable systems exaggerate noise, but a good table system minimizes it so much that one wonders what people can hear in CD’s great claim of silence.

The very first novelty of digital has worn off and its mere differences from LP and tape are no longer absorbing all attention. Perspective and balance restored, listeners are now focusing attention on the sound of the music, on what digital can do and what it has lost that analog does so well.

The cost of the discs themselves is a major drawback. Though this may eventually come down, the manufacturing process is an exacting and expensive one, unforgiving of shortcuts. To put together a library of CDs costs at least twice as much as LPs, and you are starting from scratch. With the price this high, CDs may discourage people from having several versions of a classical work and so exposure to differing interpretations will be limited. Also, what music you can listen to is likely to be restricted to selections chosen for being “safe” for larger sales.

Compact Disc Manufacture

A compact disc is only 4¾ inches (12 centimeters) across, made of thin plastic substrate coated with a reflective layer of aluminum (or, rarely, silver) and then wrapped in a thin layer of acrylic sealant. The information is encoded as binary numbers—strings of 1’s and 0’s—represented by microscopic pits and smooth places in a spiral about 3 miles long.

As the disc spins, the laser beam reads the digital code and translates it into electric signals that are fed into an amp. CDs have only one playing side, which can hold up to 75 minutes of music. Legend claims that the CD 75-minute format was chosen in order to accommodate Beethoven’s Ninth Symphony, a Sony top executive’s favorite music. If true, this represents one of the rare occasions when technology tips its hat to art.

In manufacture, accurately reproducing the billions of bits of in formation (each disc has a capacity of more than 8 billion bits) by which the music is encoded on a CD requires extremely low levels of contamination in both the raw material and in the processing plant. The tolerance is as small as for the manufacture of microchips. In theory, it is possible to make discs with very few defects, at least under laboratory conditions. But in practice, the majority have quite high levels of errors, caused by such problems as pinholes and dust. The disc players’ tolerance for faults is limited. Error correction is regularly required to compensate for poor production techniques, resulting in musical degradation even if one is not consciously aware of it.

CD mastering, occurring under clean room conditions, begins with an optically polished glass disc. Coated with a layer of photo-resist, it is then “etched” by a laser with the required pattern of pits. The photo-resist is then developed and silver-coated. From this glass master is taken a metal master, from the metal master the mother, and from the mother the stampers. CDs to this extent follow the same pattern as LPs.

Pressing involves three principal stages. First, using a combination of compression and injection molding, the pattern of the pits is impressed by a stamper on the transparent base material of polycarbonate plastic, similar to Lexan or Plexiglas. This process is similar to LP pressing except that it must occur in a “clean air” environment to pre vent dust contamination. The pressed surface is then coated with an aluminum reflective layer less than seventy millionths of a millimeter thick. Finally a very thin acrylic resin barrier is applied to protect the reflective surface.

This is the basic procedure. There are two variations, each one having its effect on sound quality—injection molding and injection / compression molding. Injection molding, usually identifiable by the see- through center of clear plastic, results in sharper definition of the pits. This is important because the detection of digital data on a CD depends not on the presence or absence of a bump but on the transition between the two. The drawback of this method is that the center hole is pressed into the disc from the stamper at the beginning, so correct centering requires very exacting accuracy from the pressing molds. While the pits may be easier for the laser to read accurately, eccentricity in the disc will cause problems.

A combination of injection and compression molding results in a disc with slightly less sharp definition of pit edges but more accurate center-hole punching, as this can be done at the end of the production with great accuracy. These discs can usually be identified by being metallized all over and having a less squared-off edge.

A new pressing technique is in the works from Teldec, as a variation of its dmm (direct metal mastering) technique. This would allow a CD master to be embossed by a stylus. Clean room conditions would be unnecessary. Done this way, CD masters could be cheaper, allowing the production of small quantities of “noncommercial” CDs, and the artistry and skills developed over years of cutting LPs could be applied to producing CDs.

The transparent sealant is intended to shield the disc against dust, dirt, scratches, and other damage. Though in certain ways less fragile than LPs, CDs must still be handled with great care. They should be held by their edges only and returned to their jewel boxes or jackets immediately after playing. All ad copy to the contrary, scratches and dust and fingerprints definitely interfere with the laser beam’s ability to track the pits correctly. Because scratches, for example, aren’t heard as pops or ticks as on LP, there is a tendency to believe mistakenly that no damage has occurred. But any alteration of the surface will cause loss of information, which can at best be disguised by the error correction circuitry but which nonetheless results in sonic degradation of the music.

There are currently no industry forces promoting better CDs. The prevailing attitude, at least publicly, is that CDs are exquisite just as they are and represent the millennium. The few CD-processing plants, overwhelmed with business, have reneged on promised production for the small recording companies that might have been able to show what could be done with CDs. As more plants are being built in the United States, this may turn around again. The major recording companies have no more interest in good-quality CD recordings than they have had in good-quality analog recordings over the past ten years and more. CD reviewers on the whole are doing nothing to help—all CDs, without discrimination, sound excellent to them. “Perfect sound forever” lives on in the minds of the many. If only we could all enjoy it.

Few record stores have the space and the capital to stock LPs, CDs, and cassettes in any depth. As the range of CD titles has broadened, so has the space allotted to them in the stores. This means that one of the other formats must yield room and that has primarily been the LP.

Meanwhile, CD production remains limited, with few plants worldwide able to perform the highly sophisticated processing procedures. Though plants are being built, production is expected to remain limited for some time. Limited production for a product in strong demand does not bode well for quality manufacture.

Large recording companies can afford the sizable initial investment of setting up plants or at least of booking up large production runs. By selecting “big-name” recording artists and “proven-successful” music, they recoup their investment quickly through turning out lots of CDs. This is not a promising approach for experimentation and creativity.

Small companies cannot set up their own CD-processing plants as might be possible with vinyl discs. Vinyl LPs are very low tech and production can be fairly easily set up—at one time, LPs were even hand pressed in small workshops. With CD, control of the recorded art form is being taken out of the hands of the small companies, which often have a stronger interest in quality and in recording obscure artists who would otherwise remain undocumented. Small companies must go to one of the few CD plants, and wait on the sidelines until all the major labels’ needs have been fulfilled.