Video with minimal loss of quality
Why do I need video on my computer?
In fact, what does storing video information give us on a computer and not on a normal videotape? Basically, digital video storage and processing is attractive for two reasons: First, digital video is not “stuck” from multiple views; secondly, it is essential to obtain a copy of the video material, absolutely identical to the original. These two reasons are enough to save your home video digitally.
What is the principle behind implementing video on a computer?
Just like the rest of the video and the movie. The frames are successively replaced 25 times per second (for Pal / Secam), which our brain perceives as a continuous movement. From this principle, we conclude: long video sequences are huge. So let’s say we save a video with a frame size of 352×288, 24 bits per pixel. It takes approximately 435 MB to save at least one minute of video. This is the main problem with digital video: an incredible gluttony in terms of file size.
How to deal with huge sizes?
There are two main ways of dealing with it: by reducing the basic parameters of video capture and compression. The main parameters of video capture are: frame size, color coding and frame rate.
The term “bit rate” is often introduced. Average bitrate is the size of a video stream in bits, relative to its length in seconds. The unit of bit rate is 1 bit / s – 1 bit per second (1bps – 1 bit (s) per second). Since 1 bit / s is very small for digital video, Kilobits / s (Kbps) and Megabits / s (Mbps) are also entered. For the video in the previous question, the bit rate is 58 Mbps. The compressed video bit rate in VideoCD, which has the same frame size and frame rate, is 1.1 Mbps.
What are the main types of compression?
Compression is divided into two types: lossless (often called “lossless” for short) and lossy (“lossy”). The difference between these types is clearly apparent from their name. Most lossless compression methods do not take into account the visual similarity of adjacent frames in a video stream. In contrast, lossy compression techniques use this similarity in most cases. Because of this, the maximum compression ratio of an average video clip achieved by lossless algorithms does not exceed 3 to 1, while algorithms that operate with loss of quality can compress up to 100 to 1.
Very often, methods that take into account the similarity of adjacent frames in a video sequence are called “recursive.” Only individual frames are fully saved to them, called keyframes (sometimes intra). All other frames contain only differences from the previous ones (sometimes they also contain links to information contained in the following frame).
Techniques that compress each frame separately from the others are called “separate”. In video sequences compressed by such methods, all frames are keyframes.
Lossless compression often uses algorithms similar to those used in archive archives (ZIP, RAR, LZH). Algorithms that use the discrete Fourier transform with the preservation of such a number of coefficients, which is sufficient for the complete restoration of the original information, are compressed more strongly.
In lossy compression methods, algorithms based on the discrete Fourier transform are also used more frequently, but the number of stored coefficients is almost always much smaller than algorithms that work without loss of quality. Lossy compression generally encodes the amount of quality reduction for each frame or the average or instantaneous bit rate of the output file.
What are the problems with video compression?
Lead the so-called. artifacts. Artifacts are visible violations of the video stream quality that occur during the lossy compression process. The most common compression artifact is dividing an image into square blocks. “Recursive” compression algorithms also often have “garbage” near contrasting edges and moving objects, as well as “wavy” and “fuzzy” colors. The number and intensity of artifacts depend on the compression algorithm, the ratio of the output bit rate to the original, and the nature of the compressed image.
