Tape Guide (Jun 1980)

Pitch and Frequency

Q. In "All About Tape Recorder Equalization" in the October, 1972 issue of Audio, you stated in connection with noise frequencies, "Clearly there are more frequencies between, say, 3,000 and 6,000 Hz than in the preceding octave of 1,500 to 3,000 Hz." This is incorrect. A mathematician will tell you that both intervals contain the same number of frequencies (although the number is unaccountably infinite). In fact, there are as many frequencies between 0 to 1 Hz as between 0 and infinity.

-William Anderson, Gaithersburg, Md.

A. Your criticism of my statement about the number of frequencies per octave is justified. A better explanation would be as follows: White noise consists of random frequencies distributed throughout a stated bandwidth during a stated time.

The bandwidth in our case is the audio range of about 20 to 20,000 Hz. At a given instant, each frequency within the bandwidth has equal chance to occur. Correspondingly, at a given instant a noise voltage may equally well occur at any frequency, i.e. anywhere in the bandwidth. If we subdivide the bandwidth, the probability of a noise voltage occurring at a given instant within the subdivision varies with the width of the subdivision. If we subdivide the total bandwidth into two equal parts, say. 20 to 10,000 Hz and (for exactness) 10,000 to 19,980 Hz, there is equal chance of a noise voltage occurring in either subdivision. If one subdivision is twice as great as the other, say 20 to 6,660 Hz and 6,660 to 19,980 Hz, at a given instant there is twice as much chance of a noise voltage occurring in the larger subdivision as in the smaller one. If we divide the audio range into octaves, where each successive bandwidth is twice that of the preceding bandwidth, within a given period of time twice as many noise voltages will occur within a given octave as in the one just below. In technical terms, the spectral voltage density doubles with each octave.

We perceive pitch essentially on the basis of ratios between frequencies.

For example, the pitch difference between 80 and 160 Hz seems to human ears to be of about the same order as the pitch difference between 40 and 80 Hz, because both involve the same ratio (2:1). Thus, successive octave intervals represent about equal pitch differences. But if each ascending interval contains double the random energy as the preceding interval, the total effect is that of high-pitched sound.

In sum: An infinite number of random frequencies can occur within each octave. But within a finite time span, the number of random frequencies (noise voltages) that do occur tends to double with each ascending octave.

Setting Standard Level

Q. I have three tape recorders and have found that it is not easy to know if I am putting the correct level of signal on the tape. The meters in one of the machines are too far out of balance to be of much value, but if I have a tape that I know is recorded with the right levels, I can adjust the meters for proper readings. The damping of the meters is different from one machine to the other. One of my machines does have good meters, properly adjusted, and I can use it for checking the tapes from the other two machines. I have an Ampex full-track test tape. What VU reading should the meters produce when playing the "standard level" tone on the test tape?

-William Benjamin, San Jose, Calif.

A. When referred to a test tape bearing a "standard level tone," the VU meter should ordinarily read about 0 VU in recording. Specifically, one does the following. (1) Play the test tape's standard level tone. Note the meter reading in playback or note the reading of an external meter connected to the machine's output. (2)

Record a test tone of the same frequency as the standard level tone, which is usually in the range of 400 to 700 Hz; this test tone is supplied by a signal generator or (less desirably) by a phono disc. Record this signal at a level such that when the tape is played back, the playback meter reads the same as for the reference tape. (3)

When recording at such a signal level, adjust the VU meter to read 0 VU.

Poor Treble Response

Q. I am interested in improving the high frequency response of my tape recorder. Playback of tapes I have recorded and of prerecorded tapes all lack adequate high frequency response. When recording I set record level to just below the distortion point; in terms of output this compares favorably with the output levels of prerecorded tapes. Can I increase bias current to improve response? Will a change in bias affect playback of prerecorded tapes? Playback of tapes I have recorded lack good high frequency characteristics when played back on a friend's tape machine. Help!

- Christopher Battista, New York, N.Y.

A. The problem may be due to some extent to heads that haven't been adequately cleaned and demagnetized.

Somewhat more likely is the possibility that high frequencies are attenuated in recording due to excessive bias (not too little bias). That is, you might try reducing bias. You also appear to have a playback problem in view of the fact that prerecorded tapes sound dull. The fault here may be a playback head with gaps that are too wide owing to wear or other reasons. Playback equalization may be incorrect. Bias does not affect playback.

Open-Reel vs. Cassette Decks

Q. Isn't it true that any better than average open-reel deck will out-perform any cassette deck in terms of sound quality and perhaps reliability?

-Gerald Woods, Santa Rosa, Calif.

A. While true or nearly true some years ago, this is certainly not true today, except at speeds of 7 1/2 ips and higher for open-reel decks. Thanks to continuing improvements in cassette transports, heads, electronics, tapes, tape shells, etc., a truly high-quality cassette deck today out-performs any open-reel deck operating at 1 7/8 ips, rivals or out-performs many open-reel decks operating at 3 3/4 ips, and comes close to the performance of some open-reel decks operating at 7 1/2 ips so far as the human ear is concerned.

This is especially true when metal tape or one of the most advanced ferric-oxide or cobalt-modified tapes is used.

At the time this was written, at least one manufacturer had already achieved response beyond 20,000 Hz in the cassette format at 1 7/8 ips, along with low distortion, high signal-to-noise ratio, low wow and flutter, etc. In fact, the same manufacturer had achieved response out to 15,000 Hz at 15/16 ips, along with good performance in the other aspects.

When you take into account other important developments in the cassette format, such as automatic adjustment of bias, record equalization and record head azimuth, the superiority of open-reel decks becomes less and less apparent to more and more listeners.

Head Protection

Q. The typical quality recorder today has three or more expensive heads. Since many of us spend much time playing rather than recording, all heads but one are being worn down needlessly. Would it be feasible to install a small head cover or pin to divert the tape from unused heads during playback? That way, the recording head would be preserved longer. The cover could trigger a microswitch recording interlock so one could not attempt recording with the cover in place. In future models, perhaps the tape path could actually be changed automatically in accordance with the record interlock.

-John Stith, Seabrook, Md.

A. I suppose you could work out a mechanical arrangement to divert the tape from the record and erase heads during playback, but I doubt that it is worth the effort. In changing the tape path you risk a change in response at the very low frequencies, where the entire head, not merely the gap, responds to the tape; the angle of approach of the tape to the head affects this response. You also risk an increase in wow and flutter. Finally, you risk poorer contact between the playback head and the tape, resulting in treble loss. Further, keep in mind that the head whose performance is usually the most adversely affected by wear is the playback head. With use, the gap may widen, resulting in treble loss.

Therefore, measures to protect the erase and record heads are not as important as would be the case if they were equally susceptible to the effects of use.

One reader has suggested that, instead of a head cover, one might wrap an appropriate thin, plastic (non sticky) material about the heads not in use. However, this would have to be very smooth to avoid an increase in wow and flutter.

The Reel Truth

Q. What are the relative advantages of the following three types of 7-inch reels? (1) Regular 2 1/4-in. diameter hub, plastic reel; (2) low-torque, 4-in. diameter hub, plastic reel, and (3) 2 1/8-in. diameter hub, metallic reel.

-Paul Gilchrist, Chicago, Ill.

A. Metallic reels have the advantage of being least subject to warp. As the tape unwinds from the outside diameter to the inside, this tends to affect the torque (pull) exerted by the takeup reel as well as the resistance exerted by the supply reel. The greater the difference between the outside and inside diameters, the greater the change in torque and resistance, which in turn can affect such things as tape speed, wow and flutter, and tape to-head contact. Hence a reel with a wide-diameter hub results in less change in torque and therefore helps achieve better performance.

If you have a problem or question on tape recording, write to Mr. Herman Burstein at AUDIO, 1515 Broadway, New York, N.Y. 10036. All letters are answered. Please enclose a stamped, self-addressed envelope.

(Audio magazine, Jun 1980; Herman Burstein )

= = = =