Spraying for Mosquitos
06/16/2006 - The news wires have recently been carrying stories on how teens around the world have hijacked a signal in the near-ultrasonic range originally developed by a company in Wales to disperse groups of loitering youth. The signal has been turned instead into a ringtone for cellular phones. Apparently alert for a new market, the original company (Compound Security Systems, www.compoundsecurity.co.uk) is already selling "legitimate" ringtone versions of its "Mosquito" signal (U.S. customers are served by www.fork.com). What makes the 17-kHz CSS signal useful to teens is that while they can hear it, adults nearby including, and especially, their teachers, usually cannot! "Join the thousands of folks who can hear the ringtone their parents can't!" says the Fork site.
|The nasty-looking horn-loaded speaker that emits the official Mosquito signal to loitering youths|
That cellphone speakers are capable of reproducing near-ultrasonic signals should also be no great surprise. Their diaphragms are of necessity small, like those of dome tweeters in hi-fi speakers, which promotes, among other things, a wide angle of dispersion for the highest frequencies, including those in more conventional ringtones. Cellphones sound bad reproducing music not because they have bad highs but because they don't produce any lows, except maybe through their headphone outputs. Other sound-limiting factors include an uneven frequency response over their reproduced range as well as high distortion and limited volume capability.
Just how high is the 17-kHz Mosquito ringtone? Very. Here is an MP3 sequence I made of ringtones composed of pure sine waves. There are four "bursts" per tone, and they start at 500 Hz (very close to the B-natural above middle-C) and ascend by octaves (from one B-natural to the next higher one) until the Mosquito ringtone frequency range is reached at 16 kHz. If you are older than 25 or 30, don't be surprised if you can't hear the 16-kHz bursts - the age-related loss of the highest frequencies (presbycusis) may be have set in. If the 4-kHz burst sounds significantly softer than the others you might want to have your hearing checked, as this is the frequency region that often suffers the greatest damage from high-level sound.
I also was surprised - and pleased - to find in the news reports that teens could still hear frequencies upwards of 17 kHz, and this despite their visits to high-loudness dance clubs and rock concerts and despite worries that they are damaging their hearing by excessive iPod playback volume. In the "old days," the 15.7-kHz horizontal scan-line "whistle" emitted by a tube television served as a quick-and-dirty high-frequency hearing test. With the coming of computer monitors (which usually have much higher scan rates) and flat-panel displays (which don't really "scan" anything), this once common form of noise pollution is slowly disappearing. The Mosquito ringtone can substitute.
While it's good to find that teens are still capable of hearing 17 kHz, even youth has its limitations. Contrary to the wishes of the promoters of DVD-Audio and SACD, the ear's high-frequency performance usually falls off very rapidly above that point. The practical limits for a suitable audio bandwidth for music were established long before the Compact Disc system came out, and the conventional 20-kHz upper limit (easily reached by the CD format) is quite wide enough, even for teenagers. In fact, the actual highest possible frequency on a CD is 22.05 kHz, which is almost a half-octave higher than 17 kHz. Extending it further as done by DVD-Audio, SACD, and other so-called high-resolution audio technologies, has never made any engineering or musical sense.
A doubling the sampling rate from, say the CD's 44.1 kHz to 88.2 kHz, extends the audio range by only one octave. If you couldn't hear those 16 kHz bursts, then there's no way you'll ever hear anything one octave above them, much less the still higher frequencies that DVD-Audio and SACD can deliver. Doubling the sampling rate also doubles the amount of data that must be recorded while extending the audio bandwidth by only about one tenth on an octave basis - a very poor engineering tradeoff even assuming anything above 20 kHz is both audible and musically significant. Instead of widening the frequency range, any extension of audio data capacity is always better spent on increasing the number of channels, something that would be audible to the most unsophisticated listeners. Even your average teenager would sit up and pay attention. Extending the audio bandwith beyond 20 kHz has always been a musically ridiculous marketing ploy. End of 10th annual I-want-more-channels rant.