The most recent non-proprietary video compression standard, High-Efficiency Video Codec (HEVC), also known as H.265, is expected to become the video standard of choice for the next decade.
This article focuses on the technical and market implications of HEVC’s adoption in the content creation and delivery market.
HEVC: Architected for the Most Efficient Video Compression Techniques
Without sacrificing video quality, HEVC can reduce the size of a video file or bit stream by as much as 50% compared to AVC/H.264 or as much as 75% compared to MPEG-2 standards. This results in reduced video storage and transmission costs and also paves the way for higher definition content to be delivered for high-quality-of-experience consumer consumption. Designed to evolve the video compression industry, HEVC intends to:
- Deliver an average bit rate reduction of 50% for a fixed video quality compared to H.264
- Deliver higher quality at same bit rate
- Define a standard syntax to simplify implementation and maximize interoperability
- Remain network friendly—i.e. wrapped in MPEG Transport Streams
While H.264 featured seven profiles (defined sets of coding tools used to create a compliant bit stream), the HEVC spec currently supports three: Main, Main 10, and Main Still Picture. Future profile extensions for HEVC will likely include increased bit depth, 4:2:2 and 4:4:4 chroma sampling, Multiview Video Coding (MVC) and Scalable Video Coding (SVC).
HEVC’s Main profile supports a bit depth of 8 bits per color, while Main 10 supports 8 bits or 10 bits per color. Because of the additional bit depth option, Main 10 has the potential to provide better video quality than Main. Finally, Main Still Picture profile allows for a single still picture to be encoded with the same constraints as Main profile.
The HEVC spec also defines 13 levels, which are sets of constraints that indicate the required decoder performance to playback a bit stream of the specified profile. The levels are split into two tiers: Main, which includes levels 1 - 3.1, and High, which includes levels 4 - 6.2 and is designed for highly demanding applications. HEVC levels share a number of similarities with the levels of H.264, with a key difference being the addition of levels 6, 6.1 and 6.2, which define requirements to support 8K resolution video.
How HEVC is Different
The primary goal of the new HEVC standard is to provide the tools necessary to transmit the smallest amount of information necessary for a given level of video quality.
While there are a number of differences between H.264 and HEVC, two stand out: increased modes for intra prediction and refined partitioning for inter prediction.
Intra Prediction and Coding
In the H.264 standard, nine modes of prediction exist in a 4 x 4 block for intra prediction within a given frame and nine modes of prediction exist at the 8 x 8 level. It’s even fewer at the 16 x 16 block level, dropping down to only four modes of prediction. Intra prediction attempts to estimate the state of adjacent blocks in a direction that minimizes the error of the estimate. In HEVC, a similar technique exists, but the number of possible modes is 35—in line with the additional complexity of the codec. This creates a dramatically higher number of decision points involved in the analysis, as there are nearly two times the number of spatial intra-prediction sizes in HEVC as compared to H.264 and nearly four times the number of spatial intra-prediction directions.
Inter Prediction and Coding
H.264 uses block-based motion compensation with adjustable block size and shape to look for temporal redundancy across frames in a video. Motion compensation is often noted as the most demanding portion of the encoding process. The degree to which it can be implemented intelligently within the decision space has a major impact on the efficiency of the codec. HEVC takes this to a new level.
HEVC replaces the H.264 macroblock structure with a more efficient, but also complex, set of treeblocks. Each treeblock can be larger (up to 64x64) than the standard 16x16 macroblock, and can be efficiently partitioned using a quadtree. This system affords the encoder a large amount of flexibility to use large partitions when they predict well and small partitions when more detailed predictions are needed. This leads to higher coding efficiency, since large prediction units (up to and including the size of the treeblock) can be cheaply coded when they fit the content. By the same token, when some parts of the treeblock need more detailed predictions, these can also be efficiently described.
Market Implications of Adoption
There are several cases where the improved quality to bit rate ratio of HEVC will impact the industry at an application level. As high quality video distribution consumes enormous network capacity, the benefit of these efficiency gains include:
In the mobile streaming market, the HEVC bit rate reduction of 30 – 50% to achieve comparable quality to H.264 is realized in the cost savings of delivery across networks. Mobile operators will not need to deliver as much data for a given quality level, making for lower costs and more reliable playback, assuming the device hardware can smoothly decode HEVC.
Media companies with significantly-sized content libraries will also feel the positive impact of bit rate savings. With HEVC halving file sizes, transitioning to the new codec will stretch storage capacity twice as far going forward.
HEVC also aligns with the push towards high-resolution Ultra HD 4K and 8K video in the mainstream market. With 4K resolutionfeaturing four times the number of pixels as 1080p, the efficiencies provided by HEVC make broadcasting 4K much more feasible.
Elemental software-defined video processing solutions have enabled a number of high-profile 4K HEVC technology achievements over the past year, signaling one area of market potential for the codec. Highlights include:
- CES 2014: world’s first real-time 4Kp60 HEVC showcase and Samsung showcase with DIRECTV (January 2014)
- Sochi Games: world’s first real-time 4K HEVC over satellite with NTV+ and on-demand to Samsung TVs with Comcast (February 2014)
- NAB 2014: world’s first real-time 4K HEVC DASH over a CDN (April 2014)
IP & Satellite:
oSky Deutschland Football and Vienna State Opera in 4K with HEVC DASH (May 2014)
oBrazil World Cup: NTV’s 4K UHD transmission of the 2014 Sochi Games over satellite and the BBC’s 4K UHD World Cup coverage of the Brazil World Cup over IP
Graphics Processing Units (GPUs): Ideally Suited for HEVC
The techniques and algorithms used in HEVC are significantly more complex than those of H.264 and MPEG-2. There are more decisions to make when encoding a given video stream or file and as a result, more calculations need to be made in compressing video assets.
The computational intensity of HEVC lends itself to the processing performance advantage available with graphics processing units (GPUs). While HEVC tools are designed to improve parallel processing capabilities, the sheer number of tools with increased complexity is very large. Elemental has estimated that HEVC encoding will require up to ten times more processing power than H.264 encoding. For example, with 500 different ways to encode each macro block, the processing power requirements are significantly higher for HEVC when compared with H.264 encoding. A doubling of spatial intra-prediction sizes, doubling of the number of transform sizes, almost four times the number of spatial intra-prediction directions, and the increase in the inter prediction search space will strain many existing hardware platforms and highlights the value of powerful GPU-based video processing in media companies’ endeavors to support the new codec.
The introduction of ultra-high resolution 4K and 8K content to the consumer market coupled with the focus on HEVC to support these resolutions will cause additional issues for most existing encoding solutions. GPUs hold an advantage again in this case, as these powerful processors are optimized to handle increased resolutions.
With a flexible software-based architecture, video processing solutions from Elemental offer support for HEVC via a seamless software upgrade. Elemental has deep experience developing video codecs from open specifications to full implementation using general-purpose programmable architectures (GPUs and CPUs).
Easing the transition to HEVC within legacy MPEG-2 and H.264 infrastructures, upgradeable solutions like those from Elemental can incorporate new compression approaches much more quickly than existing fixed hardware encoding and decoding platforms, such as ASICs and DSPs.
Elemental expects that HEVC will find widespread adoption in streaming, broadcast, satellite, cable, IPTV, surveillance, corporate video and gaming applications before the end of the decade.