The process of removing redundant bits of information from a digital file (especially a graphics, audio, or video file) as a means of reducing file size and/or expediting processing and transmission over networks. Essentially, data compression works utilizing one of a variety of algorithms which remove large chunks of information, but include codes for its later restoration. When the file is reopened or recreated from the data file, the system decodes the compression algorithm, replacing the “removed” data. (Video compression algorithms are thus also known as codecs, standing for “compress/decompress.”) However, the greater the compression, the less detail can be encoded, and the more detail is lost when the file is recovered.
As algorithms become more sophisticated, the compression ratios will increase while resulting in decreasing amounts of data loss. Currently, 10:1 is a compression ratio (in other words, the compressed file is ten times smaller than the original) that is performed without significant loss of information. Compression rations of 30:1 are also common, but they tend to result in some degree of detail loss. Compression algorithms that result in no data loss when the files are subsequently decompressed are called lossless, while those that do result in some degree of data loss are called lossy.
Some common data compression algorithms in use include JPEG and MPEG, to name two. See also File Compression.