FORMAT



Autor: Redaktion


H.264

Since 2003 a recommended standard for audiovisual applications.

H.264 is one of numerous "H. standards", which are published as recommended standards for audiovisual applications by the International Telecommunication Union (ITU).

This compression format was released in 2003 and is largely based on its predecessors MPEG-1, MPEG-2 (H.262), and MPEG-4, but achieves much greater efficiency than, for example, MPEG-2. H.264 is often mistakenly referred to as X264. However, X264 is actually an open source software library used to encode the H.264 format.

MP4 is the most common container type for H.264-encoded media. Occasionally MP4 is used synonymously with H.264, which is not technically correct. MP4 is a file format (based on Apple's QuickTime MOV) and not a data reduction method. It can include much more than H.264 media, including other MPEG variants, H.265 and audio and image files.

Due to its efficiency, H.264 is designed for high-resolution image content. It is used to compress  video in the Blu-Ray standard as well as for coding TV signals for HDTV and satellite broadcasts (DVB-S2). The visual quality is comparable to that of MPEG-2, although it only uses around a 1/3 of the data rate. Compressing to H.264 is more processor intensive than MPEG-2, although most modern computers can handle this without major challenges.

Apart from its usage in HDTV, this compression method has experienced a large uptake in the multimedia world. For example, the codec has been distributed by Apple since version 7 of QuickTime. H.264 has also been implemented in Flash Player v9 from Adobe.

A third, large field of application for H.264 is with DSLRs and prosumer video cameras. Although H.264 was specifically developed as a distribution codec, it is also used for acquisition. Since the compression method is lossy, its use as an acquisition codec must be treated with caution. In addition to extremely reduced colour information, restricting colour correction opportunities, and loss of fine image-detail (eg landscape shots and close ups are threatened by compression artefacts), inefficient compression strengthens any deficiencies from one generation to another. Mixing H.264 acquired footage with higher quality acquired material is only advised as a creative choice, as the difference in picture quality may be jarring to the viewer.

This codec has achieved its data reduction through a series of different compression methods applied equally to the image content. Complex movements within an image may be better encoded by so-called Macro Blocking. The image is divided into pixel blocks of a defined size, each of which is stored with its own motion vector. The "Intra Prediction" method is a form of prediction, estimating the pixel values of a block based on surrounding, already decoded pixels. Only the difference from the actual image is coded. Another form of data reduction is the "Long Term Prediction". In this method, image contents are analysed to (theoretically) as far back as there are frames to encode, which then allows the codec to encode the video by understanding the recurring movement patterns efficiently over many changing frames and only adding additional data where there are changes. It is customary to use the previous 16 frames.

The H.264 codec has been standardised in various levels. The highest level is denoted by 5.2 and supports an image grid of 4096x2160 pixels at 60 frames.

Related Articles: