How good is the prerecorded cassette today? Are manufacturers doing everything possible to insure quality?
by Ralph Hodges
It was in 1983-1984 when sales of prerecorded music cassettes overtook those of LP's and passed them pulling away. Already there is bickering over why the revolution chose just that moment to occur, with debates encompassing car stereo, the Walkman phenomenon, and a growing public acceptance of miniaturization in technology. But largely unaddressed so far is the question of whether the prerecorded cassette has earned its new eminence by any demonstration of sonic merit. Is it audibly worth the serious listener's attention? Can it now sound com parable to-or even better than-a well-made and well-reproduced LP, after many years of sounding consistently worse? And if it can, how? The Hard Path Upward You could argue, with plenty of historical evidence to back you up, that the Philips cassette was never meant to sound as good as an LP. It was meant to reproduce music, even from the very beginning, but with heavy emphasis on portability and convenience, and with not much more than toy-like pretensions to serious fidelity.
On the other hand, for more than a decade it's been possible-with genuinely good tape and a precision cassette shell to house and guide it, a sophisticated cassette deck, and painstaking attention to recording levels-to handcraft a tape copy of an LP that could stand up well under the sternest sort of A/B comparison. In the early Seventies, when Advent briefly offered prerecorded cassettes made with comparable devotion to detail, it was claimed that the only thing that couldn't be captured accurately from a studio master was a truly stentorian cymbal crash-something that badly shook up many phono cartridges as well. So cassettes, prerecorded and otherwise, have not been held back by lack of potential but by failure of execution. Prerecorded cassettes have, up to now, just not been made very well.
Economic factors take most of the blame. An LP consists of three parts, the disc itself and two labels, and it is produced in two steps, molding and trimming. Even the simplest, cheapest cassette of the sort that has been widely used for prerecorded product has tape, head and tall leader tape, splicing tape or adhesive, tape hubs, hub clamps, pressure pad, pressure-pad spring, shells (two halves), two windows, labels, a hum shield, and probably slip-sheet liners (two). A premium cassette will add even more parts, up to beyond two dozen.
Assembly entails numerous steps and much expensive automation, and even when you're virtually done, you still have to face the matter of getting a recording into that cassette and then of making sure that it's a satisfactory recording (proper visual inspection can actually work for LP's, but not for tape formats). In other words, there is no logical reason why a record-store prerecorded cassette should sell as cheaply as an LP, and if it does, there's every reason to suspect that something is wrong. Unfortunately, something often is.
First of all, there's the tape, which historically has not been the best, either magnetically or physically. If factors affecting its magnetic recording properties are not all in order, the final recording will be defective. It might be hissy because the tape will not accept a high enough signal level before overload to over come its inherent noise level. It might be distorted because the duplication operation that made the cassette tried to force the cassette to retain that high level anyway, and overdid it. It might be deficient in high frequencies because at tempts to record high-level short wavelengths have tended to erase the tape rather than record it. It might be variable in level and frequency response because the tape is itself variable in its magnetic properties. Or all of the above.
------------
A premium tape cassette can have as many as two dozen parts--or more.
---------------
If the tape is physically impaired, which today usually means that it is not uniform in thickness or it is slit crookedly rather than in a perfectly straight line, it will tend to ride high or low through the capstan pinch-roller and over the heads, or it might even weave and squirm as it travels. You might hear a loss of high frequencies because the edges of the recorded tracks are not properly perpendicular to the playback-head gap.
Or you might get loss of level and an increase in noise, indicating that tape-and-head alignment is so bad that tracks and gaps are beginning to miss each other completely. Or wow and flutter, because ragged tape winding is beginning to cause the cassette's tape pack to bind. Or all of the above. Or even silence, which means that the tape pack has bound and jammed, and that there is probably an unholy tangle of tape wrapped around the capstan.
While these difficulties may be indicative of poor tape (and prerecorded cassettes have traditionally been the consumer's readiest source of poor tape), they can just as easily indicate a poor cassette shell, the shell being the guidance system that ultimately controls the tape's mechanical behavior in the machine. Premium cassette shells like those provided with premium blank cassettes have precision roller guides, stationary guides, and pressure-pad and spring assemblies, and their all around physical tolerances are very low. Prerecorded cassettes cannot normally afford such precision, and they risk sounding like it.
A Matter of Speed
Heaving dealt with the raw materials, let's look at the way the usual prerecorded cassette acquires its recording. Most prerecorded cassette duplication takes place at high speed, at 32, 64, and even 128 times the 1 7/8-ips speed at which the product will be played. For such rap id duplication the tape cannot be in shells. Instead, it is worked with in bulk rolls called pancakes, which are loaded onto high-speed "slave" recorders and afterward chopped up into pro gram-length segments of tape to be wound into tapeless "C-0" cassette shells. The slave recorders (they are typically used in multiples for large production runs) receive a signal from a master playback machine employing a recorded continuous loop of 1/2- or 1-inch tape, which it rips along at such a speed that conventional tape reels have been abandoned in favor of bins. Tape coming off the transport is dropped in loose curls into a deep bin and retrieved at the bottom of the bin for another break neck pass over the play back head. Hence the interchangeable terms "bin master" and "loop master" for both the tape and the machine.
Speed can kill recording quality quite handily. Consider, in the case of 128 X duplication, that the cassette stock is zipping along at 240 ips. That's already a very high speed, but worse is that a bin master, if recorded at 7 1/2 ips, is hitting 960 ips, or not much under 60 mph.
What can happen at such velocities is a loss of tape-to-head contact because of an air film that tends to build up at the contact point, or progressive outright destruction of the tape if head contact is maintained.
Accordingly, bin masters are recorded at more conservative speeds, such as 7 1/2 ips only for 32X application, 3.75 ips for 64X, and 1 7/8 ips (!) for 128X. These speeds are workable from a day-to-day production standpoint, but in terms of fidelity, a copy of anything made from a 1 7/8-ips tape of the sort usually used for bin masters is probably not worth owning.
Nevertheless, most authorities in the field of high speed duplication agree that, in principle, a cassette tape doesn't know and doesn't care whether it's been recorded slowly or quickly. John Baxter of Accurate Sound, a manufacturer of duplication equipment, feels strongly that a prerecorded cassette made with good tape stock can equal the master no matter what, within reason, the duplication speed. Terry O'Kelly of BASF suggests that high-speed motion of the tape probably promotes mechanical stability and alignment, which may be why Pat Weber of MCA has occasionally found that his high-speed product is preferred to dubs made on high-quality home equipment at real time. The consensus is that, aside from low fidelity of bin masters, the only weakness of high-speed duplication is that its objective is a high production rate, and careful quality checks take time and may defeat that objective.
Fast Forward
Questions involving quality assurance alone are probably enough to justify the activities, increasingly vigorous, of "audiophile" prerecorded sup pliers dedicated to quality above all else except price (their offerings occasion ally hover in and above the CD price range). They do tend to come and go with bewildering rapidity, but just now companies such as Bose, Mobile Fidelity Sound Labs, and Nakamichi are forging ahead with demonstrated staying power and sufficient clout to interest big time record labels in catering to their small but passionately interested markets.
The assembly of a cassette entails numerous steps and expensive automation.
These companies are in a position to solicit 15-ips Dolby A master tapes of reasonably prestigious material and artists. They normally make a transfer to digital tape and employ that as a signal source for the slave recorders, which are not high-speed devices but rather high-quality consumer-type machines, frequently modified. Tape is duplicated, in the shell, at 1 7/8 ips, or in "real time." Bose and MFSL use top-quality chromium-dioxide tape, and Nakamichi uses a TDK metal-powder formulation. TDK also supplies Nakamichi's shells, while Bose favors a Shape shell, and MFSL has devised a proprietary shell of its own. Because the tape is recorded in the shell, tape-guidance anomalies for which the shell may be responsible are presumably factored into the final recording, something that may be significantly advantageous, but which is difficult to assess except on a cassette-to-cassette-and even a machine-to-machine-basis.
The three companies watch signal-processing and developments closely, but guardedly. Dolby B noise reduction is used by all three, but while MFSL and Nakamichi are favorably inclined toward Dolby C, reported compatibility problems between first- and second-generation Dolby C processors have dictated some caution so far. Bose, strongly oriented toward the automotive market, is sticking with Dolby B, which is the prevalent system in cars. (Bose has also opted for some slight compression to make soft passages more audible in automobiles.) Dolby HX Pro headroom extension, a recording pro cess that uniquely increases a tape's ability to capture and hold high-frequency energy without requiring consumer purchase of special equipment to enjoy its benefits. is regarded with even more caution. Nakamichi takes the position that HX Pro is not needed with metal tape, while in MFSL's view the system is not yet quite satisfactory. Although some of Bose's slave recorders are fitted with HX Pro, Bose has not set forth any defined policy on its use.
Should there be any doubts that Mobile Fidelity, Nakamichi, and Bose are capable of making a superb product, far beyond the norm of prerecorded cassettes and, it is claimed, nearly indistinguishable from the master, the expenditure of between (gulp) $13 and $20 will dispel them promptly.
But what are the implications for the bigger sup pliers who provide most of the recorded music consumed in the world? Do they care that a cassette can be made to sound better, particularly if it can't be demonstrated that it will thereby sell better?
On this front there is a little bad news, much more good news than usual, and flocks of rumors that tend in both directions.
On the good side, a brisk transition to better tape stock is taking place. A&M's adoption of BASF chromium dioxide has been well publicized, but it is an open secret that others are experimenting with and even actually running it and similar formulations. Rumors are strong that Columbia is again deeply involved in a chrome project of its own. Capitol has developed a chrome-equivalent (cobalt-doped ferric oxide) tape for its prestige XDR line of prerecorded tapes, and although names cannot be named, TDK advises that its chrome-equivalent products-even the premium SA X-are in the hands of more duplicators than we suspect. The PolyGram group, including the Deutsche Grammophon, London, Philips, and Mercury labels, has been working with chrome and chrome-equivalent tape for some time now, and virtually all its classical music cassettes are so released, with the pop material not lagging far behind.
Unfortunately, however, there is yet little evidence that similar pains are being taken with the preparation of bin masters for high-speed duplication-a real pity because, according to a spokesman, the main reason MFSL is not running faster than real time is the impossibility of controlling the quality of bin masters from outside record companies or their subcontractors. For the XDR line, Capitol's stated policy is to accept only bin masters no further than one generation away from the studio master. And MCA and PolyGram say that HX Pro is used not only for the cassette copies, but for the recording of the master (a number of experts believe that a bin master recorded at slow speed may need HX or the equivalent even more than the cassette itself).
Otherwise, record companies tend to keep the origins of master materials to themselves, which makes it hard for an outsider to guess what a given prerecorded cassette will sound like. Needless to say, it cannot sound better than the master it was derived from.
On cassette shells, the news is encouraging. Major duplicator manufacturers Cetec Gauss and Electro Sound have both hosted well-attended conferences recently, a principal topic being the vital contribution of the shell to prerecorded-cassette performance. Columbia is investing serious research to discover how to sort shell guidance errors from machine guidance errors, while several cassette-shell suppliers now think they have a firm handle on the magical combination of structural simplicity and performance reliability. The general activity level has been so high that 1985 was referred to as the Year of Azimuth (azimuth, or the angle of tape-to-head contact, is the aspect of shell-dependent tape guidance that maintains high-frequency response during tape playback). But on the troubling side, although record companies are now determined to provide better shells, they are equally determined not to pay anything extra for them. If they are forced to pay extra, we will be too. In the meantime, we may as well enjoy the headiness of rapid progress.
The Overall Situation The best news is that the prerecorded-cassette industry seems sincerely committed to making significantly better products, and that the efforts to do so began some time ago and can be heard on some new releases even today. Where new technology has not been available to plug in immediately, old-fashioned quality assurance of the check-it-every-day and even check-it-every-hour variety has been filling in. All manufacturers willing to speak about their duplication projects have reported gratifying new successes in preventing inferior cassettes from escaping the plants, and many, such as Capitol with its controversial test-tone bursts at the head of every cassette, have described new procedures and resource allocations to help them understand what the duplication lines are producing at any given moment.
Meanwhile, the prerecorded cassette of today has so far surpassed the prerecorded cassette of the early Seventies that no one, by listening, could guess that the two pieces of tape had anything fundamental in common. It has been progress not by lightning bolt but by aggregation of numerous little things-the "thousand steps," as Mobile Fidelity Sound Labs' Greg Schnitzer calls them-that collectively add up to something remark able. Progress so hard won is rarely lost.
--------------------
Picking the Winners
After reading this article, the editors of STEREO REVIEW posed a question: Of the tapes sold in record stores today, how many reflect the innovations de scribed here? An excellent question. Unfortunately, it's a little like asking: Of the breakfast cereals in supermarkets today, how many contain less than 10 percent sugar? The answer is the same: It's hard to tell, because the labels don't necessarily say.
For most major record companies (there are some noteworthy exceptions), the use of premium tape and cassette shells is still in the experimental stage. They are beginning to test the results they can get with costlier procedures and materials, but they are still hoping they can find a way to equal those results with the older, cheaper methods. At this moment, a number of manufacturers are certainly shipping cassettes that contain any thing from chromium dioxide (or a chrome-equivalent) tape to garden-variety Fe2O3 to something sent in by a dear old friend who has just ventured into the tape business. A mix of all ) three is probably the rule (as it has been in the past), but the labeling won't reflect this, for obvious reasons, and the audible differences between the three tape types may not prove dramatic anyway if all three have been recorded with competence. The worth of a premium tape lies more in consistency than in suddenly bringing out ten times more cymbal than you hear with a lesser formulation.
Ironically, for those of us accustomed to switching cassette decks to the "high bias" (or Type II) position when using chrome or chrome-equivalent tapes, even that clue to tape identity is denied us by most prerecorded cassette labeling. While a chrome or chrome-equivalent tape must be recorded with a higher bias than a normal ferric one, bias is not a factor in playing a cassette. And as for the 70-microsecond equalization also introduced by the Type II position, none of the prerecorded tapes except the audiophile variety use it, no matter the tape type. All the major la bels employ good old 120-micro second Type I EQ. and since no special switch position is called for, their packaging doesn't refer to any.
(The use of 120 microseconds sacrifices a bit in signal-to-noise ratio, but gains in high-frequency headroom. The trade-off is reasonable for much of the recorded music in demand today.) Well, the editors testily asked, how likely is it that today's average prerecorded cassette will have good sound quality? To this I could reply, with pleasure: Quite likely, and growing more likely every minute. It seems we're really on a roll here.
R.H.
====================
Source: Stereo Review (USA magazine)