MPEG-5 EVC: The new standard in video coding
When digitization of TV and media began, it quickly became apparent that these files were much too large for conventional formats. This prompted the need for efficient video compression. The increased screen resolutions and streaming services demanded new technology.
Video codecs provide this technology. By efficiently using available resources (such as network bandwidth), video can easily be displayed through different service providers and over a variety of devices.
What is MPEG-5 EVC?
MPEG-5 Essential Video Coding (EVC) is the latest video compression standard from the Moving Picture Experts Group (MPEG), a standardization group that has been very active and successful in developing and publishing standards for video coding for more than 30 years.
The goal of MPEG-5 EVC is to provide a standardized video coding solution that addresses the needs of businesses, such as video streaming, where existing ISO video coding standards have not been as widely adopted as might be expected from their purely technical characteristics.
MPEG-1 video, completed in 1993, saw wide distribution with video CDs. MPEG-2 video, completed in 1995, became immensely popular with the creation of DVD and digital TV broadcasting. MPEG-4 advanced video coding, completed in 2003, was primarily used for high-dimensional television (HDTV) and Internet Protocol (IP) video services such as over-the-top (OTT) video on demand (VOD).
The MPEG-H high-efficiency video coding (HEVC), completed in 2013, has been less widely adopted. It is an ultrahigh-definition (UHD) and high dynamic range/wide color gamut (HDR)/(WCG) that is only useful for some cinema and video applications. Encoding.com reported in their Global Media Formats Report for 2019 (six years after the HEVC standard was completed) that only 12% of their encodes use HEVC, while 82% still use the AVC standard from 2003.
The new MPEG-5 EVC is expected to be deployed in a wide variety of applications, including streaming TV and cinema. The requirements for the project specifically emphasize the importance of real-time encoding for live OTT streaming and offline encoding for streaming VOD, but other applications such as video conferencing and traditional broadcasting are also expected to be efficiently supported. The standard will support video resolutions of up to 7680 × 4320 and high frame rate (HFR) of at least 120 frames/sec.
To provide the highest possible quality on current and future displays, the standard will support HDR and WCG with 10-bit precision.
The development of HVEC
MPEG-H HEVC was developed by a joint standardization project of MPEG and VCEG with participants from across the entire media technology industry. The standardization phase for the first version of HEVC was conducted between 2010 and 2013. This phase included four face-to-face meetings per year, with approximately 500 participants and up to 1000 contributions. The exact number of organizations involved in some way is difficult to tell, but it's safe to say that more than 50 companies and research institutes contributed to the development of the HEVC standard.
A wide range of participants has advantages:
- more innovative technology is brought forward for consideration
- more resources are available to review all the different parts of the standard
- many organizations backing the standard in the industry and supporting its deployment in different applications
However, the downside of so many participants is that the final standard can contain patented technology from a large number of holders. For earlier generations of MPEG video-coding standards, a single licensing instance (patent pool) covering a vast majority of the technology in the standard, and licensing terms publicly was available.
The situation changed for HEVC when the large number of contributors resulted in highly fragmented HEVC patent ownership. Three different patent pools, MPEG LA, HEVC Advance and Velos Media, combined to provide alternative licenses for HEVC. But each held a subset of the various patent holder's portfolios, so it is not always clear if the content distributed in HEVC format is subject to royalties.
Such patent ownership fragmentation and observed uncertainty on the licensing terms availability have had a negative impact on the commercial deployment of the standard.
The EVC main profile outperforms the HEVC anchor by around 26% in bitrate reduction, whereas the EVC baseline profile shows around 31% bitrate reduction versus AVC. For the complexity estimates, the EVC test model in the main profile shows around 5× and 1.75× run time relative to the HEVC anchor for encoder and decoder, respectively.
The EVC baseline profile shows 60% encoding time reduction with approximately the same decoding time as the AVC anchor. It should be noted, however, that the EVC test model software is currently
under development and that further improvements in encoding speed and decoding speed are expected.
In performance tests, the EVC main profile outperforms the HEVC anchor by around 23% in bitrate reduction.
Will EVC become the new standard?
EVC was developed using a novel standard design principle, designed to provide increased insight into the ownership of the technology included in the standard. This provides a better foundation for licensing of the codec.
The draft standard is built on technology submitted by Samsung, Huawei, Qualcomm, and Divideon, and includes a baseline profile only consisting of technology available under royalty-free terms. All the technology components included in the main profile can be turned off individually, thereby providing users the codec with the ability to avoid using technology that they are unable to license.
The results show that the EVC standard can reach the same video quality as HEVC at 26% lower bitrate on average, over a set of UHD and HD sequences. Combined with the licensing aspects, these results put EVC in a good position for widespread adoption, particularly for applications and services that are still based on the AVC standard.