Lossy Audio Compression
General Terms - Lossy Audio Compression
Lossy audio compression refers to a method of audio compression in which the compressed audio file loses some of its original data. Lossy audio compression algorithms offer higher compression ratios, resulting in only 5 percent to 20 percent of the original stream, by discarding less critical data.
Applications
Lossy audio compression finds a wide range of applications. Some of the direct applications include computers and mp3 players. Digitally compressed audio streams are used in video DVDs, digital television, streaming media on the Internet, satellite and cable radio, and terrestrial radio broadcasts.
Most of the lossy compression algorithms reduce perceptual redundancy, i.e., they identify the sounds which are not perceived by human auditory system. Examples of such sounds include the higher frequency sounds and sounds that occur simultaneously with louder sounds. These audio data are coded with decreased accuracy, or they are not coded at all.
A major part of bits saved in lossy compression, however, comes from noise shaping. This process includes reducing the number of bits used to code a signal. However, this method increases the amount of noise in that signal. To hide this noise, the algorithm uses very small numbers of bits to code the high frequencies of most signals.
Usability
Usability of a lossy audio compression is determined by factors such as perceived audio quality, compression factor, speed of compression and decompression, inherent latency of algorithm, and support for software and hardware.
Lossy audio compression results in a perceived reduction of the audio quality, which may range from little to severe.
