Comprimering

Comprimering, or data compression, is the process of encoding information using fewer bits than the original representation. The aim is to reduce storage requirements or transmission bandwidth while preserving as much of the original information as possible. In some cases, exact recovery of the original data is required (lossless compression); in others, some information is discarded to achieve higher reductions (lossy compression).

There are two main categories. Lossless compression preserves every bit of information and allows exact reconstruction. Techniques include entropy coding (for example Huffman coding and arithmetic coding), dictionary coding (such as LZ77 and LZ78), and combinations used in formats like ZIP, PNG, and FLAC. Lossy compression deliberately loses some data to achieve greater reductions and is common for multimedia. It relies on perceptual models and quantization to remove information considered less important to human perception, as used in JPEG images, MP3 audio, and various video codecs.

Key concepts include compression ratio (original size divided by compressed size), bit rate (for transmitted or streamed data), and the trade-off between speed, memory usage, and quality. For evaluation, objective metrics (PSNR, SSIM) and subjective quality assessments are used.

Applications span data archiving, file transmission, streaming media, and storage systems. Practical considerations involve algorithm complexity, latency, error resilience, and compatibility with existing formats. While many data types are highly compressible, some data yield limited gains, and excessive compression can degrade quality in lossy schemes.