Digital video compression

Continuing with our articles on compression, we are now going to focus on digital video. There are no major differences compared with audio, since the underlying approach is the same, that is, based on codecs. As we did with sound, let's consider what the size of a video implies, and the need for it to be compressed. An uncompressed video frame can take up 1 megabyte (MB) of space on our hard drive. At a rate of 25 fps (frames per second), each second would take up 25 MB, so we would need a lot of space. Under normal video editing and processing circumstances, video has to be compressed. Digitization is carried out at the highest quality, to avoid losses as far as possible.

There are two different types of compression for video: intra-frame and inter-frame. In the first type, also known as spatial compression, each frame is compressed separately, so the images are complete (keyframes). This is the highest-quality system, which means it takes up more space. An example of this type is the AVI format, which we all know.

Inter-frame compression works in a completely different way, based on similarities between frames. As a result, it has lower quality than the previous type, but in turn takes up less space. Among the formats of this type are the MPEG formats. MPEG was developed by the Moving Picture Experts Group.

The compression method it uses is based on content similarity: if it detects a part common to the whole, it stores one instance and discards the rest. This achieves a reduction in space. This standard includes MPEG1, MPEG2, MPEG3 and MPEG4. We are going to discuss each of them in more detail, since the choice between them (in DTT, to look no further) is currently generating some controversy and is a matter of the greatest interest:

MPEG1 is a digital video compression format that emerged in 1991. It was created so that video could be stored on a medium very familiar to all of us, the CD-ROM. Its data rate is 1.5 megabits per second and it came at a resolution of 352 x 240 pixels in NTSC or 352 x 288 in PAL. Today, it is widely used for viewing videos on the Internet. It includes the popular Layer 3 (or, as we all know it, MP3).

MPEG2 appeared in 1994 and is one of the formats offering the highest picture quality. It achieves data transmission rates of 3 to 10 Megabits of bandwidth. It allows full-screen or broadcast digital video files to be transferred. It also introduces and defines transport streams, which are designed to carry digital video and audio over unpredictable and unstable media, and are used in television broadcasting. With some improvements, MPEG2 is also the current standard for HD broadcasts. A decoder that complies with the MPEG2 standard must also be able to play MPEG1.

It is currently the format used by digital television and for encoding video DVDs.

It works with resolutions from 352 x 480 to 1920 x 1080 pixels, or 720x576 (PAL) and 720x480 (NTSC).

MPEG3 was developed for HD digital television, although the MPEG2 format was also capable of performing this function. MPEG3 has greater bandwidth, but in the end MPEG2 was chosen. For this reason, the project based on MPEG3 was abandoned.

MPEG4, introduced in 1998, is best known thanks to its codecs, the very famous DivX and XviD. One of its great advantages is very good quality, similar to that of DVD, in exchange for a much higher compression factor than other formats, resulting in more compressed files that are ideal for transmitting data over the Internet. It uses a resolution of 176 x 144 pixels.

At this point, the projects ahead of us are based on MPEG7 as the standard that will be most widely used on the Internet and in interactive television. In addition to images and sound, this format will encode data in XML. MPEG7 will be of great help in the development of the new interactive television, with the introduction of content and audiovisual search engines and many more features. It consists of a standard representation of audiovisual information that allows the description of metadata) for keywords, semantic meaning (the famous "5 Ws": who, when, where, how and why) and structure (shapes, colors, textures, movements, sounds).

It is a standard of the International Organization for Standardization ISO/IEC, developed by the MPEG group. Its formal name is Multimedia Content Description Interface. The first version was approved in July 2001 (ISO/IEC 15938), and the latest version published and approved by ISO dates from October 2004.

The future brings us MPEG21. The fundamental basis on which it rests is the definition of the term digital item. These will be the files that will be used within the marketplace established on the MPEG21 network. In addition, various information is specified, such as the intellectual property and usage rights that each user has over the available digital items, a rather controversial issue.

As we anticipated in our Cutting Edge section devoted to the new digital video formats, the future lies in, or more strictly speaking will stem from, the information provided by metadata.

As for intra-frame formats, the big star is AVI. The initials stand for Audio Video Interleave. It was created for Windows and developed by Microsoft in 1992. The audio and video tracks are recorded consecutively in several layers. Recording alternates between picture and sound, but so quickly that we humans can only perceive them as simultaneous. AVI is a file format that can store data encoded in various ways inside it and that uses different codecs capable of applying different compression factors. As for playback, fragments of the file are read consecutively. The player then separates each of the audio and video streams interleaved in the file. Each of these streams, once separated, is stored in a memory buffer and passed to the corresponding codec. The video codec returns another buffer containing each of the frames to be displayed. The audio codec returns another with the digital sound sample to be played. With this information, the player only has to synchronize the frames and the sound and output them at the right speed.

In short, and by way of conclusion, you only need to look at the launch years we have given for the different formats to see that digital video compression is moving at breakneck speed and that the television of the future will be shaped by these standards.

You only have to realize that formats come first and media follow, to see that these will have a decisive influence on television or, rather, on the broadcasts of the future. Welcome to the real world.