A Method to Reduce Audio Data without an Evident Loss of Information. Perzeptual audiocoding procedures, especially those of 0 MPEG 1 Layer III standard (known as MP3), have found a constantly increasing recognition the last years. These algorithms are used in mobile phones, digital walkmen, on DVD as well as to the file transfer or as real time streaming on the internet. With a compression of about 10 to 15 the Layer III en-ables the implementation of audio data transfer over communication networks (Internet, ISDN) for video conference and internet radio applications as well as the transfer of audio data by wireless modems or radio telephones.

Procdure: In a first step an audio signal is divided in a filter bank into 32 subbands. With an addi-tional frequency transformation, the dissolution of frequency is increased to 576 bands. The data reduction is carried out by reducing the resolution in every subband in such a manner that it is inaudible for the human ear. This becomes possible due to the covering effect of the ear. An effect, which everyone knows, who tried to order a beverage in a disco: Loud sounds cover weaker signal components so that it does not matter if these components are quantified with fewer bits. The finished scanning values are embedded now into a bit stream, or written in a standardised file and sent to the decoder.

The target of this thesis (diploma) is to explain this complex technology in a simple way to students. So the focus lies further on comprehension then functionally aspects. As pro-gramming language we used MATLAB and the pertinent graphic surface. For the user we developed thereby an intuitively usable surface, from which all important steps can be re-constructed. It is also possible to represent the effect of special signals and adjustments on the coding.