Sound Level Meter

Overview

This analyser implements the features of a basic sound level meter. The algorithm is the same as that of a sound level meter: the audio is spectrally weighted (A, B, C, Z), rectified, and integrated (using a ‘fast’ and/or ‘slow’ exponential integrator) and then transformed to sound pressure level in dB.

User Controls

Weighting

Weighting is a process of filtering the signal such that frequencies are emphasised or de-emphasised. A-weighting is based on the 40 phon equal loudness contour, and as such is thought to be useful for estimating the loudness of quiet sounds. However, A-weighting is very widely used in environmental acoustics, building acoustics and occupational acoustics for a wide range of sound pressure levels (not just quiet sound).
B-weighting is based on the 70 phon equal loudness contour, and as such should be useful for estimating the loudness of moderately loud sounds. Despite its apparent appropriateness for measuring such sound, B-weighting is not very widely used. The RLB analyser provides an alternative spectral weighting that includes the low frequency attenuation of B-weighting, but not the high frequency weighting. Currently RLB is a separate analyser because a different method of temporal integration is used.
C-weighting is based on the 100 phon equal loudness contour, and as such should be useful for estimating the loudness of loud sounds.

Integration Time

Ignore Initial Filter Delay

Checking this box makes the analyser produce no output for the first part of the anaysis, equal to the duration of the time constant used in the exponential integrator. This can be helpful in avoiding the output of very low sound pressure levels at the start of a file. It is most appropriately used for sounds in which a relatively steady state situation is assumed (i.e., sound would have occured before the start of the sound file had it been recorded). It is inappropriate to use this for the measurement of 'energy' signals (isolated sound events that have a defined duration and have no context of background noise).

Outputs

Note that if the synchronisation option is not selected, each output time series of this analyser has a data rate equal to the audio sampling rate of the file being analysed.

Code Authors

This analyser was written by the PsySound3 team.

Key References

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License