Noise Archives | Campbell Associates

Background Noise and Audiometric Investigations – Acoustic Calibration

Noise

Are you Measuring Background or your Sound Level Meters Noise Floor?

Audiometric testing requires environments with very low background noise to avoid two major problems:

  1. Patient distraction, and
  2. Masking or corrupting the test signals.

Because human perception of sound is strongly contextual and subjective, clinicians cannot reliably judge an environment’s suitability without objective sound level measurements.

The strictest requirements are for measurements down to 0 dB HL with open ears, whereas screening with enclosed headsets at thresholds above – 30 dB HL allows much more background noise. It’s prudent, however, to start by examining the most demanding case.


Sound Level Meter Requirements and Self-Generated Noise

Every sound level meter has a self-generated noise floor that limits the lowest measurable level. It comes from:

  • Electrical noise (thermal noise in electronics), typically at low frequencies.
  • Acoustic noise (Brownian motion acting on the microphone diaphragm), dominating higher frequencies.

Since noise adds energetically, the error becomes:

  • Negligible if background noise is >20 dB above the noise floor
  • +0.4 dB error when the difference is 10 dB
  • +3 dB error when the background is equal to the noise floor

Thus, the meter’s noise floor should be at least 10 dB below the lowest background noise level you need to verify.


Calibration and Interpretation

Accredited calibration reports normally show two self-noise measurements:

1. Complete instrument with microphone attached (dB(A))

This includes both electromagnetic and acoustic self-noise plus unknown residual noise in the “quiet” calibration room.

2. Electrical self-noise using a dummy microphone

Reported for A-, C-, and Z-weightings, allowing assessment of frequency-dependent behaviour.
Typically:

  • A-weighted noise is lowest, because A-weighting attenuates low frequencies
  • Z or C-weighted values are higher, because electrical noise is mainly low-frequency

Calibration standards do not impose pass/fail criteria for self-noise; they only require reporting. Meters naturally get noisier with age and environmental stress. The technician must therefore:

  • Know the instrument’s noise floor
  • Understand that the calibration figure includes ambient noise from the test room
  • Recognise that the real noise floor may be higher than the reported one

Audiology-Specific Requirements: Third-Octave Bands

Standard calibration reports use broadband frequency weightings, but audiology requires third-octave band criteria, as specified in:

  • BS 8253-2:2009, and
  • BSA Guidance

Therefore:

  • A third-octave band SLM is essential, and
  • Self-generated noise values must also be given per third-octave band and calibrated to verify compliance.

Enhanced calibration services can measure and report the SLM’s true third-octave noise floor, enabling direct comparison with audiometric room criteria. Typical results shown below:

Recommended background noise level for 0 dB HL audiometry and the verified LASmax of a reference class 1 sound level meter showing the -10 dB signal to noise can easily be obtained.

Campbell Associates supply sound level meters and audiometer calibration hardware, software and have an independent UKAS accredited calibration laboratory (0789) for acoustic and audiometer test equipment.

A procedure for the assessment of low frequency noise complaints – Salford University Guidance 2009

Noise

Application Note – Norsonic Measurements for A procedure for the assessment of low frequency noise complaints – Salford University Guidance 2009

Make sure your instrument is setup to make a 1/3 octave measurement each second.

Load your measurement into NorReview in the normal way.

Click the level versus time/ frequency display – Lf +Lt

You then see a split display of average levels for each second in top part of the screen and the frequency in the bottom.

Identify the recordings where the complainant has noted the low frequency noise is present. Listen to this recording and make sure it does not include people moving around the properties or transportation noise – The concept of the standard is for investigating mechanical/ industrial type noise sources. If the recording meets this criteria click on it to display the frequencies for this recording. For longer recordings which contain other noises you can highlight part of the recording and just display the frequencies for this.

The numbers are displayed on the right side of the screen – Click the Num Table tab. These can be copied and pasted into Excel. We are only interested in frequencies from 10Hz to 160Hz for this standard.

Table V from the standard:

In the example above the highest 1/3 octave is at 100Hz. The Average value at 100Hz is 49.2dB which is above the limit for the day and the night. Note if the noise under investigation is impulsive and not constant you can apply a 5dB penalty to the measured values.

The next stage of the assessment is to view this frequency over time to see if it has correlation with the complainant’s log/ diary sheet.

The example (from the office) shows repetitive presence of 100Hz tone which correlates with the recordings.

If there is a correlation you may wish to try and find the source of the noise. This is not always easy as low frequency noise is harder to pinpoint and identify. The third octave measurements can help you identify/ locate the source and it is sometimes useful to have the FFT module in your meter which gives a more detailed ‘acoustic fingerprint’ of the noise making it easier to locate sources where the acoustic environment is more complex.

Please contact Campbell Associates support@campbell-associates.co.uk with any questions.

Salford University Low Frequency Guide