Sound masking systems are used to improve acoustics in a variety of settings, including offices, call centres, banks, libraries and hospitals.

This technology consists of a series of loudspeakers, typically installed in a grid-like pattern above the ceiling, as well as a method for controlling their output. The sound the loudspeakers distribute raises the ambient level in a facility to a comfortable 40 to 48 decibels, which covers up conversations and noises that would otherwise disrupt occupants.

To ensure a sound masking system works as expected, a clear set of performance standards should be established prior to procurement. If not, the system may not provide the desired level of speech privacy or noise control. Occupant comfort may be sacrificed as well as the ability to easily make changes.

Six elements are vital to every project’s success: adjustment zone size; masking sound generation; volume adjustment capabilities; frequency adjustment capabilities; loudspeaker requirements; and measured results.

A specification that focuses on these qualities will allow competitive bids and, if its terms are upheld, ensure a high-level of performance from the system that’s ultimately selected.

Adjustment zone size
The single most important factor is to place an upper limit on the size of the adjustment zones (for example, groups of individually controllable loudspeakers).

Small zones of one to three loudspeakers (225 to 675 square feet) allow the masking sound’s volume and frequency to be adjusted to ensure effective levels and seamless application across the space.

If the system features larger zones (1,800 to 22,500 square feet or more), dozens or even hundreds of loudspeakers are set to the same volume and frequency. As a result, compromises must be made. For example, if the volume needs to be increased to improve the system’s effectiveness in one area, it might be too loud in another. Large zones also limit the ability to reconfigure the sound masking system without moving loudspeakers and/or rewiring parts of the system.

Masking sound generation
Each small zone should have a dedicated masking sound generation to avoid phasing (for example, uncontrollable variations in masking levels), which occurs when a number of adjacent loudspeakers emit the same masking signal.

Each masking generator should provide a sound that occupants perceive as random (for example, no noticeable repeat cycle). Its actual generation can be pseudo-random but the repeat cycle should be as long as possible. The sound must cover the entire masking spectrum of 100 to 8000 hertz.

Volume adjustment capabilities
Regardless of how the system’s loudspeakers are installed (for example, upward-facing, above a suspended ceiling, downward-facing or cut through a ceiling), the masking sound will be affected by the workplace design, including the materials used in the space, the location on the floor, items above the ceiling and furnishings.

If the zones are large, many loudspeakers are set to the same volume setting but the masking fluctuates across the space as it interacts with local conditions. Some large-zoned designs try to mitigate this problem by providing audio transformers on each loudspeaker; however, they only offer coarse adjustments in three decibel steps. When the masking volume cannot be finely adjusted in small areas, the user needs to set it to a level that is best ‘on average,’ compromising comfort or effectiveness at various unpredictable points across the space. Users can typically expect a 10 per cent reduction in performance for each decibel below the target masking volume. A poorly designed system can allow as much as four to six decibel variation, halving the system’s effectiveness in some areas. Further, greater variations call occupants’ attention to the sound.

Therefore, the specification should call for fine controls of 0.5 or even one decibel for each small zone, enabling the user to adjust the masking volume to accommodate local conditions, ensuring it is both effective and unobtrusive across the entire space.

Frequency adjustment capabilities
The sound masking system should provide fine frequency control within each small zone. The range should be specified to be between 100 to 8,000 hertz. The system should also provide control over these frequencies via third-octave adjustment, which is the industry standard and basis for masking targets set by acousticians.

Loudspeaker requirements
As long as the masking system can meet the volume and frequency targets established by the specification, it is not essential to specify the loudspeaker’s size, wattage rating or other parameters.

However, very small loudspeaker drivers (less than three-inches) are unlikely to generate sufficient levels below several hundred hertz. These low frequencies are necessary to create the full masking spectrum. While they play a relatively small role in reducing speech intelligibility, they are vital to occupant comfort. Most masking loudspeakers are four to eight inches in diameter and rated from 10 to 25 watts.

Measured results
The process should not end as soon as the system is selected. The true gauge of whether it is performing as expected is gained from measurements done after installation and initial tuning.

The specification should require a test in each 1,000-square-foot area and the vendor should adjust the masking sound within that area as needs dictate. Some systems may be able to outperform this requirement but it is a good baseline.

Measurements should include:

• Overall volume and variation tolerances: The masking volume is typically 40 to 48 decibels, depending on the type of space (open or closed) and the user’s performance requirements. Forty-eight decibels is usually the maximum for comfort and should only be exceeded in special cases. The result should be consistent within a range of one to 1.5 decibels or less to ensure comfort and dependable performance across the space.
• Masking frequency curve: There is a general curve that the acoustical community considers effective and comfortable, which is defined in third-octave bands. The specification should set maximum variations for each frequency band. Plus or minus two decibels variation is a reasonable expectation.

While temporal uniformity (for example, the consistency of the masking volume over time) can also be assessed, it is usually not an issue and is less frequently specified and evaluated.

Additional considerations
Depending on the significance to the project at hand, the specification may need to include secondary characteristics such as timer functions, zoning and control methods, security features, paging functions and aesthetics.

It is also wise to learn what services the vendor offers in support of their system.

Enforce the spec
Even if the specification is well-written, the user can end up with a nonconforming sound masking system unless someone is appointed as  responsible for ensuring that bids – and the system ultimately selected – meet the criteria.

Requiring drawings as part of the bid submission process can help because they make it easier to spot design shortcuts. Asking vendors to complete a compliance form that indicates their adherence to each aspect of the specification (and explains any deviations) is also useful.

Niklas Moeller is vice-president of K.R. Moeller Associates Ltd., a global developer and manufacturer of the sound masking system LogiSon Acoustic Network. Niklas can be reached at nmoeller@logison.com.