Audio Coding

Audio data reduction by means of perceptual audio coding, as it is used for example in the well-known “mp3” algorithm, has become a widely deployed technology that can be found in each portable music player mobile phone and PC, as a format for music on the Internet, on digital media and in digital broadcasting.

Research in the field of perceptual audio coding, under the leadership of Dr. Bernhard Grill and Prof. Dr. Jürgen Herre, covers some specific aspects of audio data reduction for high quality audio signals, such as bandwidth extension of audio signals or joint coding of spatial (multi-channel) audio signals. Audio bandwidth extension is the re-synthesis of the audio signal’s high frequency range in the audio decoder from a transmitted low bandwidth base band signal plus some side information that describes the properties of the original signal in the high frequency range. Bandwidth extension is a very attractive approach to achieve good audio quality also at extremely low bit rates.

Joint coding of spatial audio signals (e.g. 5.1 surround sound, sound from 10.2 or even 22.2 3D loudspeaker setups) aims at how to efficiently encode many loudspeaker signals compactly with a very high quality that preserves the original beauty of the sound scene (sense of localization, envelopment, ambience etc.).

These and many more contributions to improved quality and higher efficiency of audio coding were made at the “home of mp3”, at the Audio and Multimedia division of Fraunhofer IIS. Through close collaboration between AudioLabs and Fraunhofer IIS, the next generation of successful audio codecs could be born in Erlangen again.

Read more on the history of mp3.