Just a slight expansion on the NCT system(s) I mentioned earlier (11/27 above). The demo I saw was 12-15 years ago, and of course microprocessor chips (used by NCT then), as well as more advanced digital signal processing (DSP) chips, have advanced considerably. In the table-top demo I watched, a vibrating source of noise was sharply reduced; you could see the noise on a dual-trace scope, and the anti-noise, 180 degrees out of phase, which caused the sharp noise reduction. As I mentioned, NCT's patent(s) related to sampling the frequencies near the noise source, feeding them into their microprocessor, generating the anti-noise, then feeding the anti-noise signal back to a transducer near the source. I've always assumed industrial applications of noise cancellation, such as large fans, use techniques very similar to this. Whole-room noise cancellation, based on speakers in a room and aimed at a fairly narrow range of noise seems much more difficult to accomplish. (But with enough signal processing power, which seems to be the technique used by Cambridge Signals Technology also mentioned above, you can make the acoustics of an entire room disappear. A Nov. 21 Wall Street Journal article, "Psychoacoustic Chicanery," describes this system.) So, if I wanted to cancel the moderate noise coming from the two fans on my Philips 64-in. RPTV attached to a large, thin rear panel, I'd concentrate on sampling the frequencies, signal processing (a NCT-type computer), and careful placement of the anti-noise transducer(s) on the rear fan panel. Naturally I'd try specialized more-silent fans first.
[This message has been edited by John Mason (edited 11-29-2000).]