.

ALTHOUGH a number of comparative evaluations of open-reel audio tapes have appeared in the past, the number of such tapes on the market has increased substantially - as have, of course, the number of tape choices open to the serious recordist. Most of these newer entries (also um 1974 herum) can be classified as (for lack of a better term) "high-performance" tapes. As a group, they represent a significant advance in the state of the tape-recording art.
.

"Recording equalization" provides a part of the solution. The main purpose of equalization is to tailor the frequency response of the input signal - the "message," so to speak-to match the needs of the tape, or the "medium."

However, a secondary purpose is to adjust the treble response during the recording process so as to compensate for the various losses. The large amounts of treble boost (up to about 20 dB at 20,000 Hz) used during recording are possible only because the higher musical harmonics are normally much weaker than the middle and low frequencies.

However, there are circumstances in which this isn't so, and therefore the amount of permissible treble boost is limited. For example, it is possible to run into trouble when recording, say, a women's chorus, cymbals, or electronic music.

In such cases the high frequencies are not much lower in amplitude than the rest, so that giving them a large boost in the record-equalization stage can overload either the tape or the machine's electronics, thus producing distortion.

The obvious cure is to turn down the record level so that the VU meters don't hit 0 VU even on the loudest passages. Though this eliminates the distortion problem, the result is a recording that plays back at a lower average level and therefore runs the risk of an unacceptable signal-to-noise ratio.

Understanding the "trade-off" relationship between distortion, extended high-frequency response, and signal-to-noise ratio makes it clear what a "better" tape means. "Better" is improved performance in one area without loss of ground in the other two areas. For example, a tape with better high-frequency response (if used with increased bias, slightly less recording treble equalization, and slightly increased recording-signal level) can reduce noise (hiss).

To illustrate this, the curves in Figure 1 show four conditions: (a) the recorder has been adjusted for optimum performance using 3M's venerable "standard reference tape," Type 111; (b) the recorder has been readjusted for the requirements of one of that company's first (1962) "low-noise" tapes, Type 203; (c) the same "low-noise" tape with the machine set for "standard"; and (d) Type 111 used with "low-noise" recorder settings. From these curves several conclusions can be drawn.

First, within the limits of experimental error, it is clear that either type of tape can provide flat (±2 dB) frequency response from 20 to 20,000 Hz if the machine has the potential to reproduce all these frequencies at the outset and is adjusted specifically for the tape in question (curves "a" and "b"). Since the low-frequency response of both tapes was identical, only frequencies above 400 Hz are shown in Figure 1.

Second, when both tapes are recorded with "standard" settings, the "low-noise" tape produces slightly less overall output for a given recording input signal. This can be offset, however, by the fact that, for a given distortion level, one can correspondingly raise the recording level when using the "low-noise" formulation - which is to say that if your recorder is adjusted for a standard tape and you buy the low-noise variety, you will be able to run the VU meters a couple of decibels above 0 VU and obtain comparable output and distortion levels.

Third, the rising high-end response of the low-noise tape when used with the "standard" machine adjustments (curve "c" again), unless compensated for somehow, would cause the sound to be over-bright. Other than having adjusted the machine for low-noise tape in the first place, the best way to deal with this rising high end is to turn down the treble control on the amplifier. With that one simple step the correct musical balance would not only be restored, but the hiss level would be lowered also. All this should help make it clear that, in the table of comparative test results, all else being equal, the tapes with the most elevated frequency response at the high end of the spectrum are potentially the quietest.

Finally, curve "d" in Figure 1 shows what happens to a "standard" tape when it is used on a recorder adjusted for a "low-noise" type. It is obviously over-biased, with the results that the output level is lower (compare curve "a") and the high-frequency response is severely impaired. Clearly, then, if the owner's manual for a recorder indicates that it has been adjusted for a "low-noise" tape, one should not use a "standard" formulation if the highest-fidelity recording is desired.
.

Latest on the market today (1974 !!) are the "low-noise/ high-output" tapes. Starting with the same basic bias, equalization, and record-level requirements as conventional low-noise tapes, they provide, as Figure 2 graphically shows, a significantly higher output for the same level of input signal.

If you have a recorder that permits off-the-tape monitoring, you can verify this by setting your machine so that a conventional low-noise tape produces the same volume in "source" and "monitor" positions. Then put on one of the newer low-noise/high-output tapes and compare again. The playback volume will be significantly higher, which means, of course, that for a given listening level it will be possible to turn down the playback control, thus lowering the hiss level as well.

There is another advantage in the premium low-noise/high-output tapes. Many of today's home recordists use the slower 3 3/4 ips tape speed - obviously in the interest of tape economy. Audiophile recorders almost universally specify a 2 to 3 dB better signal-to-noise ratio at the 7 1/2-ips speed than at 3 3/4. This is because the higher the speed, the greater the signal output from the tape. But if a recorder is set up for a low-noise tape to begin with, Figure 2 shows that low-noise/high-output tapes, with their higher output, will give about the same S/N ratio at 3 3/4 ips as regular low-noise tapes do at 7 1/2.

Two of the tapes included in the present tests (3M 206 and Capitol 2) incorporated a new "back-treatment" process that is, I think, an indication of things to come. The back treatment involved in these cases consists of applying a roughening substance to the non-oxide surface of the tape, thus providing more grip for the drive mechanism and, in addition, making for a smoother wind at high speeds. I hope this addition becomes standard practice in the industry.
.

As for the tests themselves, the results appear in the accompanying table, together with annotations explaining the significance of the data. It should be mentioned also that the tapes were tested on a "Crown International SX-822" recorder and measured with a General Radio Company Model 1523 graphic level recorder and plug-in sweep oscillator; my thanks go to both these companies for their generosity in lending the equipment. Finally, some slight amplification of a few points is in order to help the tape user assess and make use of the data supplied in the table.

The experienced recordist knows that record-level indicators on tape machines are calibrated differently by different manufacturers, quite aside from the requirements of the manufacturer's "recommended" tape.

With high-quality machines, their usual "rule of thumb" is to set the VU meter to read 0 VU at a level 6 to 8 dB less than that which produces 3% THD (total harmonic distortion) of a mid-frequency test signal. (This allows for the fact that VU meter needles have a mechanical inertia that prevents their following peak signal levels accurately.) This was the procedure I followed, using 3M 203 tape as a typical example of today's low-noise types.

In the table, the column headed Signal-to-noise ratio contains figures computed mathematically from the measured data of the previous three columns. The procedure for each tape was to take the output level for a 0-VU recording level (column two) and add to it the additional recording level the tape will take before reaching 3 per cent distortion (column three). The resulting figure is the theoretical maximum output level of which the tape is capable at 400 Hz with the specified distortion.

Subtracting the weighted bias noise (column one) then gives a maximum signal-to-noise ratio for each tape, which is the figure listed in the fourth column. Though the results obtained from this calculation may not always agree completely with an actual measurement, they should be close enough to give a reliable indication of the tapes' relative performance capabilities.

Today's tapes are certainly better than their predecessors, and tomorrow's will be better still. Reel-to-reel tape is the backbone of all music recording, and though it has come a long way from the days of paper tape coated with so-called "barn paint," there is still room for improvements.

Judging by past performance, the tape industry can be relied upon to make the necessary improvements to keep pace with developments in the machines on which the tape is to be used.
□

The tapes were tested with the recorder's bias and equalization adjusted to obtain the flattest possible response with the 3M Type 203 "reference tape" (see Fig. 1, curve "b"). A recording level 8 dB below that which produced 3 per cent total harmonic distortion (THD) at 400 Hz with the 3M tape was chosen as the "0-VU" input signal. Each tape tested was carefully bulk-erased; all measurements were made near the center of the reel.

• Bias noise. This is the noise that results during playback after the tape has been run through the machine in the record mode but with no input signal being fed to the record head. The figures reflect each tape's response to the noise introduced by the bias signal and the recorder electronics during the record-playback process. The higher the negative figure, the quieter the tape The measurements were made with a filter conforming to the extended ASA "A" curve which places greatest emphasis on those noise frequencies likeliest to be audible.

• Output with 0-VU input. These figures indicate the playback levels obtained from each tape after recording with the same "0 VU," 400-Hz input signal, and are therefore a relative measure of each tape's sensitivity at that frequency, compared with that of the reference tape (0 dB).

• Input for 3% THD at Output. These figures are the recording levels above 0 VU (for a 400-Hz signal) each tape will tolerate before reaching 3 per cent total harmonic distortion (THD) on playback. The higher the figure, the higher the recording level the tape can take before overload distortion (magnetic saturation) occurs.

• Signal-to-noise ratio. As described in the text, these figures were computed from the data of the previous three columns. First the recording level for 3 per cent distortion for each tape was combined with its output for a 0-VU recording level to determine the tape's maximum output level at 3 per cent distortion. This output level was then compared with the weighted bias-noise figure to yield the signal-to-noise ratio at close to the maximum permissible recording level. A high ratio theoretically indicates a tape with superior performance.

• Frequency response. The figures given for six frequencies from 3,000 to 15,000 Hz show each tape's departure from flat frequency response when recorded and played back on the test machine adjusted for the reference tape. The 0-dB point for each tape is its output at 400 Hz. In general, a rising high-frequency response indicates a tape with a potentially better signal-to-noise ratio at the higher frequencies. Since the low-frequency responses of all the tapes were virtually identical, no figures for frequencies below 3.000 Hz are given. The recording level for the frequency-response measurements was the standard - 20 dB.
.