Steps in algorithm:

1.

Use convolution filters to divide the audio signal (e.g., 48 kHz sound) into frequency subbands that approximate the 32 critical bands -> sub-band filtering.

2.

Determine amount of masking for each band caused by nearby band using the results shown above (this is called the psychoacoustic model).

3.

If the power in a band is below the masking threshold, don't encode it.

4.

Otherwise, determine number of bits needed to represent the coefficient such that noise introduced by quantization is below the masking effect (Recall that 1 bit of quantization introduces about 6 dB of noise).

5.

Format bitstream