How Does Compression Work?

By Author: Jack Church
Total Articles: 158
Comment this article

Most likely you have been using compression even if you did not realize it. For example when archiving files into a Zip file or when saving digital photos in a compressed JPEG format you are using compression. So how does this compression magic work?

The idea behind data compression is very simple. Data compression converts data from one coding to another. It converts data from an inefficient representation format to a very efficient representation format that requires less storage. In computer words data compression is an algorithmic way to represent the same data in less bits and bytes. Although the concept behind data compression is very simple the implementation can be very complicated and involve sophisticated algorithms and mathematics.

One might wonder why isn't the data always represented in an optimized compressed format? After all if there is a more efficient way to represent data why not always use it? The reason is that there are many domains of efficiency. Data compression optimized file storage space. But when files are compressed it is harder for computer software to process them or in other words ...
... working with compressed files is less efficient in terms of speed. For that reason files are usually compressed for storage purposes and later on uncompressed for processing.

There are two types of data compression known as lossless and lossy. Lossless compression uses inefficiencies in data representation or in other words it removes redundancy in the data to save storage space. For example if a text file includes a long list of the same letter such as a thousand times the letter Z in a row it would require a thousand characters of storage. A data compression algorithm would represent this file in a better way by simply writing something like Write A Thousand Z taking only about 20 characters of storage.

Another example for data compression by removing inefficiency is character encoding. In normal text files each character in the alphabet is represented by one byte or eight bits. In English however not all characters are as likely to appear or in other words not all character are as popular as others. A better way to encode character is to use less bits for the most popular characters and more for the less popular.

Lossy compression on the other hand converts the original data into a more efficient data that is a very close representation of the original data but not exactly the same. In other words as opposed to lossless compression if you convert a file using lossy compression algorithm and then revert it back to the uncompressed format you will not get the exact same data.

So why use lossy compression? Because by allowing the data compression algorithm to create a close enough representation but not exactly the same as the original data file the algorithm can be more efficient and create much smaller compressed files. Lossy compression is commonly used with multimedia files such as photos and videos.

JPEG is a very know lossy compression file format. If you take two photos an original pixel raster raw photo and a JPEG version of the same photo the JPEG version would be extremely smaller than the raw file. If you look at the photos themselves with a viewer the JPEG photo would look the same to the human eye but in reality there will be tiny differences between it and the raw file. In fact once converted to JPEG format it is impossible to recreate the original raw file again. Jack Church explains more on interface command line interface