μ-law

The µ-law procedure (often written u-law ) is a digitization procedure for analog audio signals that is standardized in Recommendation G.711 of the International Telecommunication Union (ITU). In a similar way to the A-law method, the µ-law method works with a logarithmic quantization characteristic in order to achieve a better signal-to-noise ratio .

A-law and µ-law characteristics and linear quantization in comparison

As with this method, 8-bit values ​​are generated in each case. However, the quantization characteristic is steeper at low levels. In addition, the coding is designed not to generate continuous 0 sequences, but rather constantly changing bit states. This facilitates a certain method for clock recovery at the receiver of the digital signal. The µ-law method is used by PCM technology in North America and Japan.

${\displaystyle f_{\mu }(x)=\operatorname {sgn}(x){\frac {\ln(1+\mu \cdot |x|)}{\ln(1+\mu )}}}$with , where is the sign function and represents the analog signal from −1 to 1. ${\displaystyle \mu =255}$ ${\displaystyle \operatorname {sgn}(x)}$ ${\displaystyle x}$

The quantization curve is divided in this method in 15 segments.

Web links

• ITU - International Telecommunication Union