Compress Video

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 lossy algorithms can compress up to 100 to 1. Video compression algorithms such as MPEG ‑ 2 and H.264 are codecs that are based on discrete conversion. In order to reduce the amount of video signal data between series of frames, these codecs use an inter-prediction algorithm. The essence of this algorithm lies in the method of differential coding, when the next frame is compared with the original one, after which only those pixels that have changed are encoded.

There is a common misconception that codecs play media files. In fact, media players are involved in playback, and the main purpose of codecs is to save space. Films or music in the basic version are of great weight, it is difficult to burn them to discs, upload them to the Internet. Compression with codecs can reduce the size tenfold or even more. At the same time, the quality practically does not suffer. The main advantage of the H.264 codec is the high possible compression ratio of the video stream without significant visual changes in the picture. This is achieved by analyzing not only each frame separately, but also their sequence. In a typical video clip, where the image in the frame changes quickly only occasionally, the methods of predicting several subsequent frames at once are used, which gives a significant gain when coding various kinds of motion.