FIELD: physics.
SUBSTANCE: invention relates to processing of audio signals, in particular to devices and methods of classification and processing of audio signals, in particular to control of dialogue amplifier, virtualizer of ambient sound, volume leveller and equalizer. Disclosed is a loudness normalization method and a sound processing device for loudness normalization based on a target loudness value. In one embodiment, dynamic amplification parameters applied to audio segments of an audio signal are determined based on a target loudness value: for the first audio segment - depending on the short-term characteristic of the audio signal, for the second audio segment - depending on the long-term characteristic of the audio signal. Long-term characteristic is determined differently from the short-term characteristic, and the loudness of the audio signal during reproduction is consistent with the target loudness value for the sound segments. Further, the volume of the first and second audio segments is modified by applying dynamic amplification parameters.
EFFECT: providing automatic adjustment of a device for improving sound quality in continuous mode depending on the reproduced audio content to prevent audible distortions at switching points.
11 cl, 41 dwg, 2 tbl
