ABOUT THE COVER PHOTO (above): This star-lit stack of front-loading cassette decks represents the new heights to which tape recorder designers have pushed their art. Find out how well cassette decks really stack up in our lead article this month (below).
By George W. Tillet
ONLY TWO OR THREE years ago, the pundits were prophesying the early demise of the open-reel tape recorder, but (as with Mark Twain's famous remark) the reports of its death were highly exaggerated. New models are seen at every hi-fi show, ranging from inexpensive stereo models to semi-professional and quadraphonic machines with such features as logic tape-motion controls, multi-track sync facilities, line and mike mixing, plug-in head assemblies, and much more.
In the same period, however, cassette recorders have been improved almost beyond belief, and the first-line models-those in the $300 to $500 bracket-cheerfully invite comparison with most any open-reel machine. Recent innovations include monitor heads, mixing circuits, variable speed controls, greater choice of bias and equalization, and easier head adjustments. And there are even some luxury machines in the $500 to $1200 range! So, let's see how the cassette deck really stacks up, whether it will kill off the open-reel recorder in home applications.
Before we try to sum up the pros and cons, it would be well to set down the various requirements of a tape recorder, and here are some--not necessarily in order of importance:
1. Wide frequency response;
2. Low distortion;
3. High signal-to-noise ratio;
4. Low wow and flutter;
5. Ease of operation;
6. Monitoring facility;
7. Editing facility, and
8. Additional features, such as peak limiter, mixing, sound-on-sound, track sync, etc.
There are, of course, other factors to consider when actually buying a recorder such as styling, guarantees, servicing, and price, but for the moment we'll ignore them.
Basic Parameters
Frequency response, distortion, and noise are all closely related so they will be considered together. As is well known, the performance of any tape recorder is greatly determined by the tape area and speed. Thus, a cassette recorder, with narrower tapes and working at the slow speed of 1 7/8 ips, is basically handicapped relative to the open-reel recorder. Examining the track configuration of cassette tape compared with the larger tracks used by open reel tape and taking into consideration the speeds, the tape area (or the number of magnetized particles, if you will) for the cassette is only 15 percent of that used by the open-reel machine working at 7 1/2 ips. Thus, it follows that the inherent problems of tape hiss and random particle noise are more difficult for the designer of cassette machines.
Fig. 1-Frequency response of a moderately priced, three-speed, open-reel
tape recorder at 0 (dashed line) and-20 VU (solid line).
Fig. 2-Frequency response of a typical top-quality cassette deck with three
different tapes at -20 VU.
Fig. 3-Frequency response of a Nakamichi 1000 at -20 VU with CrO2 tape.
Assuming that the signal output is in direct proportion to the track area, it is easy to see that if the cassette output is amplified to the open-reel level, the noise level is, almost certainly, going to be a lot higher. Figure 1 shows the frequency response of a typical quarter-track, open-reel machine costing around $350.00 and using a low-noise tape such as Maxell UD, Scotch HO/LN, or BASF LH Super. Figure 2 shows the frequency response of a typical top-quality cassette deck, and it will be seen that its high-frequency response is somewhat reduced. Note also the difference between the Normal, Low Noise, and CrO2 tapes. High-frequency response might be extended to some extent with the new ferric-chrome, dual-layer hybrid tapes, but not all machines can use them to advantage at present. Thus, it would seem that 18 kHz is about the top limit for cassette machines at present.
But this is not the whole story. One of the limiting factors presently is the head itself, as most cassette decks use a single head for recording and playback. But the requirements of the different functions are conflicting, as the gap on the recording head needs to be large for efficient signal transfer, and the playback head gap must be small to give a good output at high frequencies (or put another way, the gap must be small compared with the recorded wavelength). A limit is reached when the recorded wavelength is equal to the gap length because the two oppositely-magnetized half wavelengths will cancel and the induced voltage is zero. The wavelength of a 20-kHz signal at 7 1/2 ips is about 8.5 microns, but reducing the tape speed to 1 1/4 ips brings down the wavelength of the same frequency signal to about 2.2 microns! Although the effective magnetic gap is a little larger than the actual physical gap, if we want a cassette deck to have a linear response to 20 kHz (without too much high-frequency boosting), then the head gap should be less than half wavelength or about 1 micron. But in this case, recording efficiency would be low. Therefore, machines using combined record-playback heads usually settle for a compromise gap between 2 and 4 microns.
Recently, several three-head recorders have become available, and they do have a rather better high frequency performance. Figure 3 shows the frequency response of a Nakamichi 1000, which uses a 5-micron recording head with a playback head having a remarkably small 0.07 micron gap.
The response extends to over 20 kHz, and results are apparently only limited by the tape medium itself.
Frequency response is usually measured at low levels, either at -20 VU or sometimes at -30. The dashed line in Fig. 1 shows the frequency response of our open-reel recorder at 0 VU, and it will be seen that the response at high frequencies falls relative to the response at-20 VU (shown as a solid line). This phenomenon gets progressively worse as the tape speed decreases, and it is ca led tape saturation because all the magnetic particles have been affected and further signal increases can do nothing except erase what is there already! At 15 ips, the effect is quite small, but it becomes significant at 3 3/4 ips, and, as you might expect, it is even more important with cassette decks working at 1 1/4 ips. What does this mean in practice? Simply that care must be taken to keep that VU meter pointer well below the 0 VU mark when making recordings of music having large transient peaks or the sound will lack definition and brilliance. But, there is a snag; if the overall level is reduced by 10 VU as shown on the meter, then the overall signal-to-noise ratio will also suffer. Some cassette machines have peak limiters which do help, but special noise reduction systems, such as the Philips DNL, JVC's ANRS, and Dolby, provide the real answers.
How do we score the two sorts of tape recorders on these basic parameters? Well, top marks have to go to the very best open-reel machines, but we also must note that many of the first-line cassette recorders offer a quality of performance superior to the less-sophisticated open reelers. It's a case of paying your money and taking your choice.
Fig. 4-Action of the DNL system. Note that it is inactive above levels of
-40 dB.
Fig. 5-Distortion versus recording level at 1 kHz for a typical open-reel
machine costing about $350.
Fig. 6-Distortion versus frequency at 0 VU for the same typical open-reel
deck.
Noise Reduction
Most readers are at least moderately familiar with noise reduction systems, so that a long and involved technical explanation would be inappropriate here. In brief, however, the Dolby and ANRS systems work by increasing the signal level of frequencies above 500 Hz during recording, if they are low, and then reversing the procedure for playback, thus reducing the high frequency signal level to the amplitude of the original and reducing the noise by a like amount. There are two Dolby systems, A and B; the latter is the one used in domestic machines. The amount of high frequency lift is determined by both amplitude and frequency, and the circuitry is quite complex. Although the ANRS also operates above 500 Hz, it is not a dynamic, continuously-controlled system, and so recordings made with the two are not directly compatible. The Philips DNL system is not a two-way type, as it functions on playback only to attenuate signals as shown in Fig. 4. Note that the high level signals are not affected and the maximum effect is in the region--of 10 to 12 kHz.
A cassette deck using a Dolby system can show an increase of up to 10 dB in signal-to-noise ratio over a non-Dolby recorder, which means that recordings can be made with a lower level and you would still have a low background noise with less danger of tape saturation. A typical high quality cassette recorder thus would have a signal-to-noise ratio of 50 dB without Dolby and 60 dB with (weighted). On the other hand, an open-reel recorder would probably have another 6 to 8 dB at least, more if it also had a built-in Dolby system. But, of course, the cassette's 60 dB does represent a very good signal-to-noise ratio indeed, quite adequate for most purposes. The user, then, must decide for himself whether he has to have that few extra dB greater signal-to-noise ratio. Incidentally, a S/N has to be related to a reference point, and unfortunately there is no fully-accepted standard. Some manufacturers use 0 VU, some prefer the 1 percent distortion point, while still others opt for 3 percent which makes The figures look better. Therefore, care must be taken when making comparisons as there might well be as much as 4 dB difference between the cassette machine figures and even more for open-reel machines. It might be thought that 0 VU seems pretty conclusive as a standard, but it isn't. What is 0 VU on one deck might be +2 VU on another, and further confusion is caused by some manufacturers who try to avoid tape saturation by tailoring the VU meter response so it reads higher above 5 kHz or so! All this is be cause volume units (VUs) only measure relative levels and not absolute levels.
Distortion
Figure 5 shows the distortion (THD) with a typical open reel recorder, and it will be seen that the 3 percent level is not reached until the signal gets to a level of +7 VU, while at 0 VU the distortion is about 0.5 percent. In terms of distortion vs. frequency, there would be a slight rise at each end of the frequency range (see Fig. 6). The unfortunate cassette deck designer does not have the same leeway to work with, and many cassette decks have a distortion characteristic like that shown in Figs. 7 and 8. In other words, the distortion is higher and the headroom lower, once again emphasizing the importance of a noise reduction system in effectively giving a wider dynamic range. So, in terms of frequency response, distortion, and signal-to-noise ratio, an open-reel recorder working at 3 3/4 ips can have a 3 to 4 dB better signal-to-noise ratio, an extra 2 or 3 kHz in frequency range, and a bit lower distortion than an average cassette deck fitted with a Dolby system. At 7 1/2 ips, the disparity is greater, and if the open-reel machine also has a built-in Dolby system, then there is scarcely any contest. But we are here talking about a machine that would probably cost at least half again as much as the cassette deck! In actual practice, though, a cassette machine can make tapes from discs or FM that sound identical in an A-B comparison to tapes made by an open-reel recorder using 33/4 ips or even 71/2 ips, but greater care is necessary when using the cassette deck. When it comes to making direct recordings with top-quality microphones, the open-reel recorder will win, especially if the 15 ips speed is used and an ultra wide dynamic range required.
Fig. 7--Distortion versus recording level at 1 kHz for a typical cassette
deck.
Fig. 8--Distortion versus frequency at two recording levels for a typical
cassette deck.
Wow and Flutter
Wow can be defined as a slow variation in speed below 5 or 10 Hz, while flutter is a faster variation. Measurements are now usually made using the DIN or IEEE standard, which give a combined and weighted figure. Wow is frequently caused by capstan shaft eccentricity or by an unstable drive tension system, and flutter can be produced by erratic tape motion as it passes over heads and guides or between various constricting portions of the cassette. Poorly designed or produced motors can also cause speed variations, but these days both open-reel and cassette decks use hysteresis-synchronous motors which are frequency controlled from the power supply. Both use servo-controlled motors in the higher-priced models, with the speed electronically controlled by feedback circuits which automatically correct for any fluctuations. The big difference between the recorders, cassette and open reel, is the way the tape is held steady so it passes smoothly over the heads. Some open-reel recorders use two motors, others have three, though at least three European recorders use a single heavy-duty motor with a tension clutch arrangement. All use a kind of back torque system to keep the tape under tension. Few cassette decks use anything but the simple pressure pads of the cassette to maintain smooth head-to-tape contact. However, if properly designed, this relatively simple method can work very nicely at the low speed involved. The cassettes themselves often present problems, since unlike tape reels where both tape and reel turn together, the cassette tape moves but the plastic case does not. While the various sorts of liner sheets, steel pin rollers, etc. do help with the friction, it and other variables are still not fully under the control of the deck designer.
Not only will an unsteady tape movement cause annoying flutter, but worn pressure pads or dirt on the heads can produce a poor contact with the tape resulting in a loss of high frequencies. It can also cause modulation noise-a kind of IM distortion. Further, it is not generally realized just how serious this sort of spacing loss can be at the slow speeds. A speck of dust only one-eighth of a mil will produce an attenuation of 54.5 dB or about 99 percent. Furthermore, the loss is compounded when the same head is used for recording.
In practice, it is unlikely that the spacing will be that large, but it doesn't take much oxide build-up to produce a loss of 6 dB. The inference is obvious; cassette recorders need a lot more care, TLC, to get the best results, and the heads must be kept clean, really clean. And for truly top-grade recordings, only the best cassettes should be fed to the recorder.
The cassettes should also be stored in their protective plastic boxes, just as with records, to prevent the accumulation of dust either on the tape itself or in the sprockets and hubs.
A look at the measured wow and flutter figures for open reel recorders shows results from 0.04 to 0.09 percent, while cassettes run (no pun intended) between 0.04 and 0.17 percent, depending on price. It should be noted that these figures for cassettes are a good more variable than they are for open-reel machines. Even figures for the very same test cassette and recorder combination will vary from day to day.
The main point to remember, however, is that the performance by cassette machines in this area make them worthy of serious consideration by the enthusiast.
Ease of Operation
Now here is where the cassette deck really scores! Anyone--even my Aunt Agatha--can load a cassette into a machine without getting into a tangled mess! And the decks are smaller and more portable. Some, such as the Yamaha TC-800GL, the Nakamichi 550 or Sony/Superscope 152 SD, will work on batteries too, and most have easy-to-use push-button controls. The more recent front-loading models, such as the Pioneer CT-7171, Sony/Superscope TC-177 SD, and Technics RS-676 US, are not only simple to operate but they provide facilities previously found only on open-reel decks. For example, the Technics recorder has a memory circuit that can get the tape rewound back to a predetermined point before switching the machine back into the play mode. It also has a Dolby 25 and standard 75 µS de emphasis network switch for recording from tuners with either time constant, a peak or normal VU reading switch, and provision for Cr0 2 low noise, and ferric-chrome tapes. The bar switches have built-in indicator lights, a useful refinement for the non technical user.
Monitoring
In my opinion, this is an essential feature for the truly serious recording enthusiast, and most open-reel machines over the $300 mark do have the three heads necessary.
Among the cassette decks with this facility are the Nakamichi 1000 and 700, AKAI GXC-352D, Sony/Superscope TC 177 SD, and the Technics RS-279 US. Curiously, the Technics model uses a separate head for monitoring, while retaining a combined record-replay head. A word of warning; because a deck has three heads, it doesn't always mean that one is available for monitoring. For example, the third head in the Toshiba PT-490 is a second erase head for the reverse tape direction.
Editing, S-O-S, etc.
I stated earlier that it is difficult to edit cassette tapes, and so it is. While it can be done, the process does require a great deal of patience. The 3M Company makes a repair kit, and Editall, among others, makes the right size blocks and tabs for splicing. None of the special effects like sound-on sound, multitrack sync, and echo are possible with cassette decks, although certain effects are possible with the aid of external units.
The Future
We will undoubtedly see more cassette decks with monitoring facilities. In fact, several were introduced at the recent Consumer Electronics Show in Chicago. Many of these had provision for ferri-chrome tapes and servo motor control. Automatic bias and equalization adjustment for Cr02 tapes, actuated through sensing an indentation on the cassette, is now standard practice. More and more sophisticated cassette decks will appear with such features as variable speed control, but the biggest advance will be in the tape itself.
The originator of the cassette medium, Philips, recently introduced a super fine-grain iron oxide formulation which they say gives an improvement of 10 to 12 dB in the signal-to-noise ratio. Both TDK and Maxell have even more recently introduced cassette tapes making use of cobalt for improved performance. TDK's is called Super Avilyn (SA) and uses a cobalt ion added to an extremely fine ferric oxide particle by absorption. Maxell's UDXL uses cobalt ferrite epitaxially grown on extremely fine ferric oxide particles.
Bias and equalization characteristics of these two tapes are similar to Cr0 2, so we will be spared a proliferation of bias/eq switches. However, I don't see cassette machines challenging open-reel units very much more seriously than they do now because both benefit from technological advances.
So, then, back to our opening question; how do the cassette recorders of today stack up? And the answer, obviously at this point, is very, very well!
(Audio magazine, Sept. 1975)
OPEN-REEL vs. CASSETTE (April 1977)
Fighting Distortion in Tape Recording (April 1977)
= = = =