Scale factor (audio coding)

Scale factors are used in audio data compression (e.g. MPEG-2 ) to reduce the amount of data in an audio file .

A common dynamic range in audio coding is 16 bits per sample with linear PCM . This dynamic range is sufficient to prevent the quantization noise from being perceived. However, since these 16 bits are seldom fully used, data can be reduced by first using coarser with z. B. 6 bits is quantized. The scale factor now indicates the level of this 6-bit quantization in which the audio signal is located. This step value is further quantized with so many bits that the quantization noise is below the hearing threshold of the respective frequency band (see subband coding ), which happens with less than 16 bits depending on the signal amplitude. The sum of the bits for the scale factor and the bits for the amplitude is often less than 16 bits. Empirical studies show that the scale factor hardly changes for immediately consecutive samples. As a result, a further data reduction can be brought about by adding a scale factor z. B. applies to eight consecutive samples.

literature

• Udo Zölzer: Digital audio signal processing. 3rd, revised and enlarged edition, reprint. Vieweg + Teubner, Wiesbaden 2008, ISBN 978-3-519-26180-3 .
• Roland Enders: The home recording manual. The way to optimal recordings. 3rd, revised edition, revised by Andreas Schulz. Carstensen, Munich 2003, ISBN 3-910098-25-8 .