Active Noise Cancellation in the 21st century

No flying cars (yet). No robot maids (yet). No Mars colonies (yet). At least we have active noise cancellation (ANC), right?

Experiments with cancellation of sound by sound started as early as 1878. So why in the 21st century aren’t we cancelling all unwanted sounds with sound? Well, physics, that’s why.

The principle seems simple enough:

Place an “error” microphone at the location where you want cancellation to occur.
Create a precise inverted copy of the sound wave using smart electronics and play back through a nearby loudspeaker.
The atmosphere pushes, the loudspeaker pulls. Net result: no sound pressure. By definition, silence.

If only it was that easy in practice. By way of simple explanation:

1) The arriving wave must be an exact opposite copy of the incident wave. If it’s too strong or too weak, too early or too late, cancellation is partial or non-existent. Too far out of whack and the ANC system can amplify sound instead, or even feed back!

2) The cancelling wave is correct at the error mic ONLY, so the amplitude and/or timing are guaranteed wrong at other locations. As a result, the “sweet spot” has a radius of about ⅛-wavelength around the error mic. Maybe the size of a grapefruit.

3) A single mic/speaker combination cancels decently within the “grapefruit” around a single point. Therefore additional mics, speakers, and computers are required to expand the cancellation zone. In a real room, the mics would be positioned on-grapefruit-sized centers throughout the space!

4) The cancellation wave takes time to travel from the loudspeaker to the error mic, so the control system has to predict the future. A consistent waveform, like a steady tone, is easier to predict than one with fluctuations, as in many if not most sounds.

So in what applications does broadband ANC work well?

In small, grapefruit-sized volumes (like closed cup headphones).
When the “grapefruit” completely fills a duct or tube (as in automobile exhaust).
Where either the listener or the source are well “inside” the “grapefruit”.

Therefore even in the 21st century, it’s still best to prioritize noise control at the source: