1992 年 16 巻 61 号 p. 49-54
The algorithms standardized by ISO/IEC JTC1/SC29/WG11 for audio data compression (MPEG/Audio Coding Algorithms for short) are overviewed. The standard has a layered structure with Layer I, II and III algorithms. Layers I/II provide satisfactory coding quality with reasonable complexity, while Layer III dedicates additional complexity to the best possible quality. Two typical compression techniques for audio signals, the subband coding and the transform coding, are introduced. Key technologies inevitable for the coding algorithms, such as psychoacoustic weighting, block-size selection, MDCT, are covered one after another. Algorithms of Layers I through III including optional joint stereo coding are discussed with their subjective qualities. Fast algorithms for less computation and compensation for distortion caused by errors are also presented. A future ISO standardization schedule with important topics concludes this article.