For more than 30 years, digitalisation in the media industry has enabled new services, increased reach of conventional services and brought a continuously improving TV experience to consumers all over the world. One of the most central technology components in digital media distribution and consumption is compression and in particular video compression, since video data represents by far the largest amount of data in most media services. Over the years, multiple different video coding standards and proprietary codecs have emerged and the compression performance of new generations of codecs is constantly improving. However, the compression efficiency is not the only factor that determines how well a codec is suited for usage in different application areas – and eventually – how widely deployed the codec will become. The MPEG-5 Essential Video Coding (EVC) standard, presented in this paper, is being developed by MPEG using a novel process targeted at addressing business requirements, including licensing, as well as technical requirements, so as to facilitate rapid deployment and widespread use of the standard throughout the media industry.
Development of a standard typically starts with an analysis of the requirements: What are the problems that will be addressed by this standard? What are the desirable technical properties of this standard? How will this standard interact with other components of the eco-system? What are the target applications and use cases?
For video coding standards, the core problem has remained the same over the years: reduce the size of transmitted video data as much as possible while keeping the visual quality as close as possible to the original video. The convention when it comes to video coding standards has been to only define the bitstream format and the decoder (not the encoder). This allows for cross-industry compatibility of the most critical component (decoder) and at the same time, it allows for flexibility in the design of the encoding process, for example, to meet requirements of latency and availability of computational resources.
Even though the input format to an encoder and the output format from a decoder (uncompressed samples of video data) has remained the same over the years, the applications of different generations of video coding standards has been quite different. MPEG-1 Video , completed in 1993, saw great deployment with Video CD. MPEG-2 Video , completed in 1995, became immensely popular with the dawn of DVD and digital TV broadcasting. For MPEG-4 Advanced Video Coding (AVC/H.264) , completed in 2003, the primary applications were HD TV and IP based video services such as Over The Top (OTT) Video On Demand (VOD). MPEG-H High-Efficiency Video Coding (HEVC/H.265) , completed in 2013, has mainly seen deployment for 4K and HDR/WCG video applications.
Since April 2018, MPEG (ISO/IEC JTC 1/SC 29/WG 11) has together with VCEG (ITU-T SG 16/Q 6), worked on a new project; Versatile Video Coding (VVC) which is expected to see deployment in applications providing an immersive video experience such as Virtual Reality streaming, 8K broadcasting and High Dynamic Range (HDR) video.
It should be noted that from a technical perspective, and from a standards point of view, there is nothing preventing old standards being used for new applications. Even though there might be no decoders to support it and there is no level explicitly defined for it, you could very well apply the compression methods from MPEG-1 to 4K video. However, the bandwidth requirements for such a system would go through the roof and it would not be cost-efficient to run such a service. By using the latest and most advanced video coding standard it is possible to minimise the resources needed for distributing encoded videos and provide as high quality as possible at all different operating points. The net result is a win-win situation for both service providers and consumers – provided that the cost of deploying and licensing the standard itself is reasonable.
No comments yet