Algorithm of Using ERB-spectrum to Calculate Moore's Loudness

Moore's loudness model is presented in ANSI S3.4 in 2005,which uses the signal spectrum as its input,but the method to get the spectrum is not given.It is suggested to use FFT spectrum of waveform directly for loudness' calculation.But Moore recommendes calculating six FFTs in parallel,each FFT is used to calculate spectral magnitudes over a specific frequency range,then the synthetic spectrum is used to calculate loudness.The deficiency of the two approaches is analyzed firstly,and then a new approach is proposed to improve the efficiency and precision.ERB scale is used to divide the auditory frequency into a series of bands.At last,the loudness of pure tones and car's noise is calculated using the three methods and the results are comparatively analyzed.