Waveform Analysis for High-Quality Loop-Based Audio Distribution

Existing compression techniques for waveform audio still rely heavily on the removal of data from the stored representation – lossy compression. This reduces the quality and is particularly an issue when the audio file is to be distributed over low bandwidth communication media. Small file sizes can be achieved using compression procedures, but the quality of the audio suffers, to the detriment of the reproduced audio, and thus, the listening experience. Though acceptable audio quality can be achieved using these compression methods, the file sizes increase dramatically at acceptable levels and become unsuited for low bandwidth transmission purposes. A possible solution to this problem is to use a compression procedure which maintains the original quality of the audio, whilst exploiting perceptual coding techniques. Using such a method would provide a high quality audio representation, and attempt to reduce the size of the file by making use of waveform analysis and comparison techniques. By performing such analysis, and producing a representation of the audio in a small distributable file a high quality of audio could be maintained, whilst the only factor to be lost would be information that is deemed to be unperceivable to a degree. Sufficient audio chunks can be built up so that the entire original audio can be reproduced by using as few chunks as possible, to attain the minimum file size an accurate representation. This document represents initial theories and research conducted into the possible development and application of such a compression method.