compression

1. (Or "compaction") The coding of data to save storage space or transmission time. Although data is already coded in digital form for computer processing, it can often be coded more efficiently (using fewer bits). For example, run-length encoding replaces strings of repeated characters (or other units of data) with a single character and a count. There are many compression algorithms and utilities. Compressed data must be decompressed before it can be used.

The standard Unix compression utilty is called compress though GNU's superior gzip has largely replaced it. Other compression utilties include pack, zip and PKZIP.

When compressing several similar files, it is usually better to join the files together into an archive of some kind (using tar for example) and then compress them, rather than to join together individually compressed files. This is because some common compression algorithms build up tables based on the data from their current input which they have already compressed. They then use this table to compress subsequent data more efficiently.

See also TIFF, JPEG, MPEG, Lempel-Ziv Welch, "lossy", "lossless".

Compression FAQ.

Web Content Compression FAQ.

Usenet newsgroups: news:comp.compression, news:comp.compression.research.

2. Reducing the dynamic range of an audio signal, making quiet sounds louder and loud sounds quieter. Thus, when discussing digital audio, the preferred term for reducing the total amount of data is "compaction". Some advocate this term in all contexts.

(2004-04-26)