White Noise in Open Office Environments
The Radius That Describes a Room
Across twenty-one open-plan offices, the radius within which one person’s speech stays intelligible enough to distract ran from 2.5 metres to 18 metres. In the 2.5-metre office, 6.4 per cent of occupants reported high disturbance; in the 18-metre one, 45.2 per cent did. A room can sit anywhere in that range, and where it sits is the first thing worth knowing about it.
The quantity has a standard definition. The international standard for open-plan offices defines distraction distance as the radius around a talker within which the Speech Transmission Index stays at or above 0.50 — that is, the area inside which that person’s speech is intelligible enough to be distracting. The figures above come from a meta-analysis combining room measurements with surveys of 883 people across those twenty-one offices, which found that the quantity predicts perceived disturbance: every additional metre of distraction distance raised the odds of someone reporting high disturbance by ten to fifteen per cent.
That same analysis gives target values. Fewer than thirty per cent of occupants reported high disturbance where distraction distance was under seven metres, speech level at four metres was under 46 decibels, and background level was above 40 decibels. Note the direction of that last one: distraction distance correlated strongly and negatively with background level, at minus 0.83. A higher noise floor is the cheapest way to shrink the radius.
A Layout That Did Not Deliver
Open-plan offices were sold on collaboration, and the best measurement of that claim found the opposite. Two field studies tracked employees at corporate headquarters before and after a move to open space, using wearable sensors and email and instant-message server logs rather than surveys. Face-to-face interaction fell by about seventy per cent in both cases, with electronic messaging rising to replace it. The authors read this as people withdrawing socially from officemates when the walls came down.
That is the backdrop for the acoustic problem. The layout produced less of the talking it was meant to produce, and the talking that remains is the single thing most responsible for people finding these rooms unworkable.
Speech, Not Loudness, Is the Complaint
Open offices are not especially loud. The reason they are hard to work in is that the dominant sound carries language, and the experimental literature is specific about which property of speech does the damage: the effect on performance is driven by how intelligible the speech is, not by how loud it is. Reducing the level of a conversation you can still follow does not help much; making the same conversation unintelligible does.
This is why the fix is additive. Masking works by raising the background so that speech stops being decodable, which means the remedy for a room full of unwanted sound is more sound. That is counterintuitive enough to be worth saying plainly, and it is also why the remedy has a ceiling, discussed below.
A meta-analysis of sixty-five reading experiments puts a number on the general cost: background noise, speech and music all carry a small but reliable penalty, with intelligible speech and lyrical music the most disruptive conditions tested. A laboratory study simulating an open office compared a high-noise and a low-noise condition and found that people in the noisier one remembered fewer words, rated themselves more tired, and were less motivated.
Masking Has a Low Ceiling
Raising the floor indefinitely does not keep working. The authors of the twenty-one-office analysis recommend against masking above 45 decibels, because beyond it people start raising their voices to be heard, which puts more speech energy into the room and undoes the benefit.
Independent work on speech privacy in open-plan offices gives a consistent figure from the other direction: a background level of 45 A-weighted decibels, which drives the Speech Intelligibility Index down to about 0.20, is the criterion for acoustic comfort. So the usable masking range is narrow — roughly 40 to 45 decibels — and it is a property of the room, set by whoever installs the system.
What Headphones Actually Change
Here is the part that most advice gets wrong. Distraction distance is a room quantity. It is set by absorption, screens and the installed background level, and it describes the acoustic conditions for everyone in the space. Putting on headphones changes none of it. Your colleague two desks away is still inside the radius; what changes is what arrives at your own ears.
That is a real and worthwhile change, but it is a different intervention with different limits. Closed-back headphones attenuate before you add anything, so the sound you play can be quieter than a room-wide masking level would need to be. The catch is that you cannot verify your level the way a building system can: a slider position is a gain multiplier, not a decibel value, so the 40-to-45-decibel window that the research identifies is not something you can dial in. The practical substitute is to raise the sound only until nearby speech stops being followable, and leave it there.
There is a second difference worth noticing. An installed system raises the floor uniformly, which is why it shrinks the radius for everybody and why it has to respect the 45-decibel ceiling: the people whose voices it covers are also sitting in it. Headphones carry no such constraint on the room, only on your own ears, so the ceiling that applies to you is the hearing-safety one rather than the speech-effort one. That makes a personal setup freer and less accountable at the same time, since nothing stops you from running it louder than any building system would be allowed to.
For masking speech specifically, white noise has the advantage of a flat spectrum — equal power in every equally wide band, so there is energy wherever the speech is. Pink noise rolls off at about three decibels per octave and is easier to sit under for a full day, giving up less speech coverage than that slope suggests. Avoid anything with recognisable content; the whole mechanism is about removing information from the background, and birdsong puts some back.
Taking It to Whoever Owns the Room
There is also a hearing-safety floor under all of this. The National Institute on Deafness and Other Communication Disorders reports that sound at or below 70 A-weighted decibels is unlikely to cause hearing loss even after long exposure, while long or repeated exposure at or above 85 dBA can cause it, with normal conversation at 60 to 70 dBA. A masking level in the 40s is far below that, which is another reason to treat any urge to go much louder as a sign that the sound is the wrong tool for the problem.
One structural note that the interaction study makes unavoidable: if the open layout already suppressed face-to-face conversation by seventy per cent, a floor of headphones is not what broke the collaboration. It was broken before anyone put them on.
None of this makes an open-plan office a private one. The measured range across twenty-one real offices spans more than a factor of seven in distraction distance, and the difference between the good end and the bad end is absorption, screens and an installed masking system — building decisions, not listener decisions. If your room sits at the bad end, a personal setup is a partial compensation and the published target values are the thing to raise with whoever owns the space.
References
- Bernstein, E. S. & Turban, S. (2018). The impact of the 'open' workspace on human collaboration. Philosophical Transactions of the Royal Society B 373(1753), 20170239 (two field studies using wearable sensors and electronic message logs; face-to-face interaction fell by approximately 70% after the move to open space, with electronic interaction rising)
- Haapakangas, A., Hongisto, V., Eerola, M. & Kuusisto, T. (2017). Distraction distance and perceived disturbance by noise — An analysis of 21 open-plan offices. Journal of the Acoustical Society of America 141(1), 127-136 (883 respondents; distraction distance predicts disturbance, odds ratios 1.10 to 1.15 per metre; measured range 2.5 to 18.0 m; background level correlated with distraction distance at -0.83)
- Hongisto, V. & Haapakangas, A. (2017). Distraction distance predicts noise disturbance in open-plan offices. 12th ICBEN Congress on Noise as a Public Health Problem (the published target values: under 30% highly disturbed where distraction distance is under 7 m, speech level at 4 m under 46 dB and background level above 40 dB; masking above 45 dB LAeq not recommended because speech effort rises)
- Jahncke, H., Hygge, S., Halin, N., Green, A. M. & Dimberg, K. (2011). Open-plan office noise: Cognitive performance and restoration. Journal of Environmental Psychology 31(4), 373-382 (participants in the high-noise condition remembered fewer words, rated themselves more tired and were less motivated than in low noise)
- Bradley, J. S. & Gover, B. N. (2004). Criteria for acoustic comfort in open-plan offices. Inter-noise 2004, Prague — National Research Council Canada (a 45 dBA background level, giving a Speech Intelligibility Index near 0.20, as the criterion for acoustic comfort)
- Vasilev, M. R., Kirkby, J. A. & Angele, B. (2018). Auditory Distraction During Reading: A Bayesian Meta-Analysis of a Continuing Controversy. Perspectives on Psychological Science 13(5), 567-597 (sixty-five studies; noise, speech and music all carry a small but reliable cost, with intelligible speech and lyrical music the most distracting)
- White noise — ASA/ANSI Standard Acoustical Terminology, term 2.37 (equal power in every equally wide frequency band)
- Noise-Induced Hearing Loss — National Institute on Deafness and Other Communication Disorders (at or below 70 dBA unlikely to cause hearing loss even after long exposure; at or above 85 dBA can with long or repeated exposure; normal conversation 60-70 dBA)
Frequently Asked Questions
What volume should white noise be in an open office?
Published target values for installed masking put the useful band at roughly 40 to 45 A-weighted decibels, and above 45 the recommendation is against going further because people raise their voices to compensate. Through headphones you cannot read your level off a slider, since the slider sets a gain multiplier rather than a decibel value, so the workable substitute is to raise the sound only until nearby speech stops being followable.
Are noise-cancelling headphones necessary for open offices?
They are not necessary. Their strength is steady low-frequency sound such as ventilation, which is not the part of office noise that drives the complaints: speech intelligibility is. Closed-back headphones attenuate before you add any sound of your own, which lets you play it quieter, and that is the main benefit either type provides.
Does masking make speech inaudible?
No, and it is not meant to. The measured driver of disruption is intelligibility rather than level, so the goal is speech you can tell is happening but cannot follow. The standard quantity for this is the Speech Transmission Index, and the distraction threshold sits at 0.50.
Will white noise through headphones help colleagues too?
No. Distraction distance is a property of the room, set by absorption, screens and the installed background level, and headphones change only what reaches one person's ears. Improving conditions for everyone means changing the room, which is why the published target values are worth taking to whoever manages the space.
Is all-day listening in an office safe?
At masking levels it is far below the hearing-risk thresholds: at or below 70 A-weighted decibels is unlikely to cause hearing loss even after long exposure, and the risk figure is 85 dBA with long or repeated exposure. The honest caveat is that the reading evidence shows any added sound carries a small cost, so the case for it rests on the distraction it replaces being worse.