What is a sound masking system?

A sound masking system is a network of ceiling or under-floor emitters that broadcasts a precisely tuned background spectrum, typically 100Hz to 5kHz, calibrated to 42 to 48 dBA at ear level so private conversations stay private and ambient noise becomes less distracting in open-plan offices.

The technology works by lifting the noise floor of the room without being consciously perceived as noise. The masking spectrum is shaped to overlap the frequency range of human speech, which lowers the Speech Intelligibility Index (SII) between any two desks. The result is that conversations more than four to six metres away become unintelligible rather than merely quiet, so a colleague's phone call stops pulling your attention. Properly commissioned masking is barely noticed by occupants after the first few minutes, similar to the way you stop hearing a quiet HVAC fan.

A typical UK office deployment has three components. Emitters are small ceiling or plenum-mounted speakers spaced every 3 to 5 metres on a regular grid; modern systems use direct-field emitters facing down from the ceiling tile rather than older indirect designs that fired into the plenum. A controller generates the masking signal and applies zoning so a quiet boardroom can run at a different level than the open office outside it. Calibration is done with a sound level meter at ear height across the floor plate, tuning each zone until the spectrum is uniform within +/- 2 dBA.

Mainstream brands include Biamp Cambridge (formerly Cambridge Sound Management), K.R. Moeller LogiSon, Lencore and Atlas. Most systems also handle paging and background music through the same emitter network, which is why a sound masking system is often specified alongside other office audio infrastructure on a single project rather than as a standalone retrofit.

The biggest deployment swings come from ceiling type (open plenum vs sealed tile changes emitter spacing), zoning granularity (per-team vs whole-floor), and whether HR want to overlay music or paging onto the same network.

Quick reference: masking spectrum 100Hz-5kHz at 42-48 dBA, emitters every 3-5m, calibrated to +/- 2 dBA across each zone.