Strutt

Strutt Help



Auditory Masking    1/1, 1/3

Strutt|Electroacoustics|Auditory Masking allows the user to calculate the masked threshold for auditory masking from a noise spectrum.

Two methods are available in Strutt:

The ISO7731 method assumes a constant slope of masking of -7.5 dB/octave and calculates the masked threshold for each frequency band as follows:

For the lowest frequency band: `L_(T,i)=L(N,i)`
For all other frequency bands: `L_(T,i)=Max(L_(N,i),L_(T,i-1)+K)`

where:
`L_(T,i)` is the masked threshold in frequency band in question
`L_(T,i-1)` is the masked threshold of the previous frequency band
`L_(N,i)` is the noise spectrum level in the frequency band in question
`K` is the slope of masking:

Note that ISO7731 allows for masking to "cascade" across multiple frequency bands - i.e.low-frequency noise can, if sufficiently loud, cause masking effects to extend over more than one adjacent frequency band.

Strutt implements a modified version of the IEC 60268-16 masking method. The key differences are:

In effect, Strutt is implementing the ISO7731 masking model except using IEC 60268-16 to calculate the slope of masking.
Hence, the equations used are the same as for ISO7731 (above) , except that the slope of masking K is level-dependent as follows:
For 1/3 octave analysis, the slope `K_(1//3" oct")=K_(oct)/3`

Note that the slope of masking in IEC 60268-16 is steeper (i.e. masking is less strong) than ISO7731.

References:


Comments or suggestions to
strutt@arup.com