Video coding format

Videos use a special way to save space. This helps the video play on your screen. It makes the files small.

One way to save space is to look for things that stay the same. If a part of the picture does not move, the computer does not need to save it again. This makes the file much smaller.

Computers use a plan to do this. This plan is called a format. A special tool called a codec uses the plan.

Most video files also hold sound. The video and the sound are put together in one big container. This lets you see and hear the show at once.

This is how we watch shows online. It is very cool!

Have you ever wondered how movies play on your screen? Video files are often very big. To make them fit, we use a video coding format. This is a plan for how to save video data. A special tool called a codec follows this plan. The codec makes the video small so it can travel through the internet.

One way to save space is a method called DCT. This stands for discrete cosine transform. It is a way to change how the image data is stored. Another way is motion compensation. This looks for parts of a picture that move. Instead of saving every new frame, the computer just saves the changes. This makes the file much smaller.

Videos also need sound. The video and the sound are put in a container. An MP4 is a common type of container. It holds the video and the audio together.

Many formats exist today. H.264 is a very famous one. It is used for YouTube and Netflix. A newer format is HEVC, also called H.265. It was made in 2013 to help video look even better.

Have you ever wondered how huge movies fit on your phone? Video files are often too large to send easily. To fix this, we use a video coding format. This is a set of rules for how to save video data. A special tool called a codec follows these rules. The codec compresses the video to make it much smaller.

How does this shrinking work? One common way is using a method called discrete cosine transform, or DCT. This changes how image data is stored to save space. Another way is called motion compensation. This looks for parts of a picture that move between frames. Instead of saving every single detail, the computer only saves the changes. This clever trick makes the file much smaller and easier to play.

People have been working on this for a long time. In 1929, R.D. Kell suggested sending only parts of a scene that change. Later, in 1972, Nasir Ahmed proposed the DCT method for images. In 1974, Ali Habibi introduced hybrid coding to combine different techniques. By 1975, researchers John A. Roese and Guner S. Robinson made this work for moving video. These ideas helped create the video technology we use every day.

Many different standards have been made over the years. The first practical standard was H.261. Later, MPEG-2 became the standard for DVDs and digital TV. One of the most famous formats is H.264, also called MPEG-4 AVC. It was developed in 2003 and is used by YouTube and Netflix. A newer format called HEVC, or H.265, arrived in 2013. This newer version helps videos look even better.

You can think of a video file like a lunch box. The video and the sound are both inside. The lunch box itself is called a container format. Common containers include MP4, AVI, and Matroska. An MP4 container might hold H.264 video and AAC audio. This keeps all the parts together in one neat package. Just like your lunch box, the container holds everything you need to enjoy your meal.

A video coding format is a set of rules used to encode digital video content. This content can exist as a data file or a bitstream. Because raw video data is massive, these formats use compression algorithms to shrink the files. Most modern formats rely on two main techniques: discrete cosine transform (DCT) coding and motion compensation. A computer component, such as software or hardware, that performs these compression or decompression tasks is called a video codec. While people often use the terms "format" and "codec" interchangeably, they are different. The format is the technical specification or the layout plan. The codec is the actual implementation of that plan, like a specific piece of software.

To understand how video is compressed, we must look at the mechanism of motion-compensated DCT. This is a hybrid coding algorithm. It combines spatial compression with temporal compression. Spatial compression uses DCT coding to shrink data within a single frame. This was first proposed by Nasir Ahmed in 1972 for image compression. Temporal compression uses predictive motion compensation to find similarities between different frames. Instead of saving every pixel for every frame, the codec finds blocks that move. It then copies those blocks and only records the small differences. This process is often called block-matching. Finding the absolute best way to match these blocks is an NP-hard problem, which means finding a perfect solution is practically impossible in a reasonable time.

Because the specification does not mandate exactly how to find these matches, codecs have freedom to innovate. This allows for different trade-offs in speed and space. A live video feed might use a fast algorithm that uses more space. In contrast, a DVD might use a slow, highly efficient algorithm to save more space. This flexibility is why different codecs can implement the same format in different ways. For example, the H.264 specification can be implemented by many different codecs like x264 or OpenH264. Each codec might use different mathematical paths to reach the same goal defined by the standard.

The history of video compression began much earlier than the digital age. In 1929, R.D. Kell proposed transmitting only the parts of a scene that change. In 1952, researchers at Bell Labs proposed differential pulse-code modulation (DPCM). By 1959, NHK researchers introduced the concept of inter-frame motion compensation. In 1967, researchers at the University of London proposed run-length encoding (RLE) to reduce bandwidth for analog signals. Early digital video in the 1970s was very inefficient. Uncompressed standard-definition video required bitrates around 45,200 Mbit/s. This was 2,000 times greater than the 100kbit/s bandwidth available at the time. Even modern 1080p high-definition video requires over 1Gbit/s if left uncompressed.

Practical video compression truly emerged with the development of motion-compensated DCT. Ali Habibi introduced hybrid coding in 1974. Later, in 1975, John A. Roese and Guner S. Robinson extended this to the temporal dimension. They found that DCT was more efficient than other transforms like the Fourier transform. They showed that DCT could compress data down to 0.25-bit per pixel. This was much better than the 2-bit per pixel required by typical intra-frame coders. By 1981, Wen-Hsiung Chen developed a practical algorithm called adaptive scene coding. This motion-compensated DCT became the standard technique for video from the late 1980s onward.

There have been many important video coding standards over the years. The first digital standard was H.120 in 1984, but it was too inefficient for practical use. The first truly practical standard was H.261. This was followed by MPEG-1 in 1991 for VHS-quality video. In 1994, MPEG-2/H.262 arrived and became the standard for DVDs and digital television. MPEG-2 could achieve a compression ratio of up to 100:1. This allowed for new technologies like high-definition television. In 2003, H.264/MPEG-4 AVC was released. It is currently one of the most widely used formats in the world. It is used by streaming services like YouTube and Netflix, and is required for Blu-ray Discs. The current generation is HEVC (H.265), introduced in 2013.

It is important to distinguish between the video format and the container format. A video coding format like H.264 is usually bundled with an audio stream. These two streams are placed inside a multimedia container format. Common containers include MP4, AVI, FLV, RealMedia, and Matroska. A user usually does not have a raw H.264 file. Instead, they have an MP4 file that contains H.264-encoded video and AAC-encoded audio. A single container can hold different types of video. For example, an MP4 container can hold either MPEG-2 or H.264 video. The WebM format is another example. It uses the Matroska container to hold VP8 or VP9 video and Vorbis or Opus audio.