The sound level within an attenuator diminishes along the length of the attenuator as the sound energy is absorbed. However, if the attenuator is in a relatively noisy area to attenuating sound that would otherwise break out to a quieter one, it is possible for noise around the attenuator to break into the quieter end through the attenuator’s casing, unless it provides sufficient insertion loss to prevent this happening.

Similarly, if the attenuator is in a quiet location controlling higher levels of sound from a noisy space, that sound can break out through the attenuator’s casing from the ‘noisy end’ of the attenuator.

This means that the attenuator’s casing must be designed to take account of the sound levels inside and outside the attenuator, together with the level of attenuation that it is expected to provide.

In this example it is expected that the attenuated level will be between 15dB and 20dB in the 500Hz, 1kHz and 2kHz octave bands. However the level of sound breaking back into the attenuator is similar to or higher than the expected attenuated level. This means that the overall sound level within the attenuator, which is the (logarithmic) sum of both components, is approximately 5dB higher than expected. As a result, the attenuator’s performance is correspondingly lower than expected.