The Future of Video Technology

The initial focus of MC-IF is on the Versatile Video Coding Standard, an international standard developed in a joint effort of VCEG (ITU-T Q.6/SG 16) and MPEG (ISO/IEC JTC 1/SC 29). The standard, published in July 2020, delivered a substantial increase in compression efficiency over its predecessor, HEVC/H.265, achieving around 50% bitrate reduction for the same video quality. VVC’s unparalleled compression performance stems from a combination of efficient coding tools designed from the start to support a wide range of video content properties including High Dynamic Range (HDR), Wide Color Gamut (WGC), and computer-generated imagery for gaming and remote screen content sharing. Additional versatility can be achieved by metadata signaled in supplemental enhancement information (SEI) messages specified in VVC itself and in VSEI (Versatile SEI messages for coded video bitstreams) (ITU-T Recommendation H.274 | ISO/IEC 23002-7). As a whole, VVC’s coding tools, functionality, and metadata signaling capabilities make it the leading state-of-the-art, feature-complete video codec available.

Applications and Use Cases

The VVC standard is designed to handle a wide variety of applications that involve video compression. Video streaming, digital TV and video conferencing are all important use cases handled by VVC. But VVC has also been developed and tested for more advanced and emerging use cases such as 360° video, multilayer video and High Dynamic Range (HDR) video. The experts involved in the VVC development have ensured that the final standard will truly be versatile by including test sequences and test conditions for such emerging applications throughout the standardization process.

Compression Methods

Although VVC is not a direct extension of HEVC, the basic principles of the VVC standard are the same as HEVC, as well as other previous standards produced by MPEG and VCEG. Pictures of a video – represented by sample values of up to three color components – are partitioned into blocks of different sizes and predicted from previously coded samples in the same picture or in previously coded pictures.

There are a lot of refinements and novel methods
included in VVC that together result in superior
compression performance. A few examples include:
  • Binary and ternary block splitting
  • Geometric partitioning
  • Affine motion compensation
  • Matrix-based intra prediction
  • Combined inter and intra
  • prediction
  • Dependent quantization
  • Adaptive loop filtering
Some coding tools provide benefit for specific use
case and/or under certain circumstances.
Such tools include:
  • Reference picture wraparound
  • Reference picture resampling
  • Independent subpictures
  • Palette mode

In combination, these and many more coding tools make it possible to compress video more efficiently and thereby provide superior quality, reduce bandwidth requirements and enable services that could not be realized with previous generations of video coding standards.

VVC Profiles

Profiles are used in video coding standards to define interoperability points for bitstreams and decoders. The first version of VVC includes six different profiles targeted at different use cases and different applications. All profiles of VVC include support for pictures with sample bitdepth of up to 10 bits, and all profiles include support for subpictures. The profiles can be grouped into three categories:

Main 10 4:4:4 profile and Multilayer 10 4:4:4 profile

– supports video of four different chroma formats: Monochrome, 4:2:0, 4:2:2 and 4:4:4. The difference between Main 10 4:4:4 profile and Multilayer 10 4:4:4 profile is that Multilayer 10 4:4:4 profile supports multiple layers that can be used for example for spatial scalability and multi-view video while Main 10 4:4:4 profile only supports temporal scalability (through a concept called sub-layers).

Main 10 profile and Multilayer 10 profile

– are subsets of Main 10 4:4:4 profile and Multilayer 10 4:4:4 profile, respectively, supporting only video of two different chroma formats: Monochrome and 4:2:0.

Main 10 4:4:4 Still Picture profile and Main 10 Still Picture profile

– are subsets of Main 10 4:4:4 profile and Main 10 profile, respectively, supporting only a single coded picture.

Join the Video Revolution Today

With significant advancements in every aspect of media coding, VVC will be the the new "it" for years to come.

Get Started with VVC

Discover how VVC can help your organization advance your media coding.

Join MC-IF Today

Engage with us and become a part of the Media Coding Industry Forum