Open-plan offices are often criticized for being noisy and distracting. Coworkers’ conversations in particular can make it difficult to focus. Until now, most assessments of office noise have relied mainly on sound pressure level (LAeq). But sound level alone doesn’t tell the whole story.
At Forum Acusticum/ Euronoise in Málaga, Valtteri Hongisto, from Turku University of Applied Sciences introduced a new method that helps predict how distracting office noise really is.
Aim of the study
The goal was to create an objective prediction model that can reliably estimate how distracting office sounds are- especially speech- using standard acoustic metrics available from common sound level meters.
Study design
The research team conducted a psychoacoustic experiment with 39 Finnish-speaking participants.
- Stimuli: 111 office noise scenarios, including different numbers of simultaneous speakers (1–12), background masking sounds, and overall levels between 40–60 dB LAeq.
- Procedure: Participants rated “How much the sound would distract your work?” on a scale from 0 (not at all) to 10 (extremely).
- Metrics: For each sound, the researchers measured equivalent sound level (LAeq) and sound level variability (SLV, LA5–LA95).
Sound Level Variability – since office noise is rarely steady
Sound Level Variability (SLV) describes how much the sound level fluctuates over time. In an office, noise is rarely steady: people start and stop speaking, voices rise and fall, and background masking interacts with speech. These changes create a pattern of peaks and pauses that can make noise more distracting than a constant hum at the same average level.
SLV is measured using standard sound level meters that can record statistical sound pressure levels:
- LA5: the sound level exceeded 5% of the time (represents peaks, like speech syllables).
- LA95: the sound level exceeded 95% of the time (represents background or quieter moments).
The difference between these two values gives the Sound Level Variability. This simple calculation captures how “dynamic” or “fluctuating” a sound environment is. Examples:
Low SLV (~1–3 dB): Steady background noise or masking (like ventilation). This is usually less distracting.
High SLV (>10 dB): Clear speech with noticeable peaks and pauses. Even at moderate loudness (e.g. 50 dB LAeq), this can be very distracting.
Results
The experiment confirmed that distraction increases with overall sound level (LAeq) but also depends strongly on sound level variability (SLV).
- Single- or dual-speaker conditions were typically more distracting than multi-speaker babble, even at the same LAeq.
- Using mathematical optimization, the team developed the following prediction model:
The model showed an exceptionally high correlation (r = 0.96) with subjective distraction ratings.
Practical implications
For professionals working with offices- architects, acousticians, interior designers, and workplace consultants- this model is a valuable tool:
- It allows distraction to be measured objectively, not just based on complaints.
- The required data (LAeq and SLV) can be collected with standard sound level meters.
- The model can help identify areas in an office where distraction risk is higher, supporting better layout and design decisions.
- It can also be used in renovations or workplace health assessments, giving objective benchmarks for improvements.
Conclusions
This study marks a significant step toward bridging subjective annoyance and objective measurement in office acoustics. For professionals designing or assessing open-plan environments, this model provides a much-needed tool to evaluate and mitigate noise distraction—helping to create workplaces that support concentration and well-being.
