Lossless data compression
The term "Lossless data compression" is used to describe a method where processing is applied to data to reduce the total amount of data. In lossless data compression none of the original information is discarded. The result is that all of the information contained in the original file is encoded in the output file and the total amount of data is reduced (thus the term "compression").
Lossless data compression works to varying degrees depending on the nature of the information contained in the file. For example- documents containing only text can be compressed to a greater degree than audio files. This is due to factors such as the total number of possible "symbols" and the number of times each symbol repeats.
Alpha-numeric files contain a fairly limited number of possible symbols whereas 16 bit audio files contain 1 + 32,768 possible symbols ("sign" bit plus one-half of 65,536). The increased number of possible symbols in combination with no reasonable method of predicting how often they will repeat requires the entire audio file to be analyzed to generate the coding required to achieve a useful level of compression. Increasing the wordlength to 24 bits makes this process take even longer.
The Meridian Lossless Packing system was developed to address the need for lossless data compression on audio DVD formats. There are also other formats such as FLAC and ALAC which are more commonly used for computer digital audio files.
Please see data compression for details.