High Efficiency Video Coding (HEVC), also referred to as H.265, is a video standard that is being developed through an ISO/IEC collaboration. HEVC planning was begun way back in 2004, shortly after H.264 was finalized and the topic has been getting a lot of exposure in the industry over the past few months. Many questions remain about HEVC including how quickly it can be implemented into the current video ecosystem and when content owners will adopt it.
The Digital Media group at Frost & Sullivan, which I am part of as a Principal Analyst, has been doing a lot of coverage on HEVC as of late. Our lead analyst on the transcoding side, Avni Rambhia, has published three reports that include details on HEVC including the Global Broadcast and DTT Video Encoders Market, Global Pay TV Video Encoders Market and Global Media and Entertainment Video Transcoding Market. Based on what we have seen in the market and data we have collected from suppliers, here’s our take on why HEVC adoption for consumer services is at least five years away.
The current HEVC draft was put out in July 2012 and the standard is expected to be ratified shortly. MPEG-LA, the licensing coordinator for all MPEG technologies, put out a call for applicable patents in July of last year. Several companies have early demos already available including Mitsubishi, NHK, Cyberlink, Broadcom, ATEME, Ericsson and Elemental Technologies, among others. Integration with chipsets is unlikely to begin until the standard is finalized and even after that, production will probably be initiated until critical mass of demand is achieved. However, although decoding will eventually reach the point of chipset integration, software based playback is not unrealistic given growing smart phone horsepower.
Video is a push industry, the community is constantly pushing the boundaries on resolution, compression efficiency and user experience to drive the media and entertainment ecosystem forward at ever-increasing speed. Two apparently disconnected use cases are the focus of market innovation this year, the explosion of video consumption on portable and personal devices, and rising investment in Ultra HD video at resolutions of 4K and even 8K. Each has its own challenges and technology requirements, but both share one crucial stumbling block – they cannot become global phenomena without a significant leap forward in video compression efficiency. MPEG-4 or AVC is the de-facto video compression standard today, and has played a key role in recent years in enabling Internet video, OTT services, IPTV and HD across all Pay TV services.
However, the technology has matured to the point where any advances are incremental, and prices are correspondingly seeing tremendous downward pressure. MPEG-4 is also not well positioned to enable Ultra HD transmissions in an economical fashion. Furthermore, with video accounting for as much as 90% of total bandwidth usage in North America during primetime – with less than a quarter of the Pay TV subscriber population watching on-demand OTT video – it is clear that lower-end video services cannot be served by AVC in the long-term. This is especially true because growth in HD-capable devices like tablets and the rising trend of watching OTT content on HD and Ultra HD connected TVs is further exacerbating an already challenging bandwidth situation.
HEVC or H.265 has been heralded as the solution to this problem. Offering up to 50% compression efficiency improvement over state of the art AVC codecs, H.265 is poised to disrupt the video ecosystem – both for M&E and for enterprise applications – yet again. Vendor excitement around the technology is high, with a number of announcements for HEVC-enabled products at CES 2013 and energetic R&D efforts underway to develop encoder and decoder cores (both software and hardware) now that the standard is finalized and development of a patent licensing program is underway. Amidst the fervent hype, it is easy to believe that HEVC is an immediate technology whose adoption curve will be soaring upwards in 2013 and 2014. But that’s not reality.
One can draw a parallel between the adoption curve of MPEG-4 as it gradually encroached into the supremacy of MPEG-2. We believe that while token adoptions – such as incorporation into DVB standards for terrestrial broadcasting – will occur in the short-term, and a few channels may also be launched by 2015, a critical mass of adoption will not begin to occur until at least 2016. History indicates this – even a decade after the launch of AVC, MPEG-2 remains a formidable force in Pay TV (particularly cable), owing to the massive footprint of legacy equipment such as set top boxes and transmission infrastructure that is all designed to work with MPEG-2 video.
Cost also remains an issue – many Pay TV operators in regions like Africa, Asia and Latin America are choosing MPEG-2 rather than AVC because of the significantly lower cost of consumer premise equipment (CPE) and video encoders. Considering the massive wave of investment in AVC equipment that we have seen in the last two years, we expect at least 5 more years of equipment life before economically stressed broadcasters and service providers will consider systemic upgrades. Any video technology touches many components as it travels from glass to glass, such as cameras, NLE systems, video indexing systems, statistical multiplexers, satellite transponders, head-ends and (perhaps most importantly) CPEs.
Similarly on the OTT side, transcoders, file formats, streaming protocols, streaming servers, content protection systems, network optimization platforms and end devices all need to support HEVC before an end to end solution becomes broadly viable. In their continual endeavor to fight commoditization and drive demand through continued technological disruption, vendors of video technology and consumer electronics devices alike are engaged in fast and furious product development around HEVC, with many announcements made already and several more significant milestones expected throughout 2013.
Silicon vendors are also looking towards the technology but at a somewhat more sedate pace, both to maintain profitability levels on existing AVC chipsets and also given the considerable challenges of achieving real-time power-efficient encoding and decoding of HEVC content (particularly at higher resolutions). But any large scale migration from AVC to HEVC will take time, much as the transition from MPEG-2 to AVC is still very much an ongoing process.
One can also draw parallels between 3D and HEVC, as technologies that were aggressively marketed before a robust content pipeline was in place. Video technology without adequate, compelling content is like a painting without adequate illumination – it’s very hard for a viewer to see the value. 3D has by wide consensus failed to realize its expected potential, partly because too much was expected too soon, but also because the content community was not able to find a viable business model to justify the expense and disruption, and users failed to see value in the technology – even as television set prices dropped dramatically as vendors struggled to establish demand at expected levels.
From a service provider perspective, the industry has only just overcome the alphabet soup of fragmented video formats to converge around AVC. Interoperability standards like Ultraviolet and MPEG-DASH, streaming technologies like HLS and Smooth Streaming, and pretty much every digital terrestrial transmission and cable standard today all embrace AVC. AVC has been the technology to break the walled garden mentality that pervaded OTT video during much of the 90s, with vendors like Adobe, Apple, Google, Microsoft, Rovi and Sorenson Media all embracing AVC in the interest of optimizing OTT video, despite the fact that each owns their own proprietary video compression technology.
Many of these vendors – particularly those who provide core codec components that power other vendors – are in the process of developing HEVC software cores to power the inevitable innovation that the new technology will catalyze. At the same time, they acknowledge that their service provider customers are loath to disrupt a delivery ecosystem that has only just settled down around converged solutions focused more on delivery and quality of experience than with just connecting the dots.
In terms of services rollouts, Frost & Sullivan expects closed-loop solutions such as enterprise video conferencing (from vendors like Cisco Systems and Vidyo) and Ultra HD broadcast services in the far east (powered by vendors like Cyberlink, Rovi and Samsung) to be among the earliest HEVC-based services to be rolled out. Video on demand services for low bandwidth, such as HD video delivery over cellular networks, are also likely to be early adopters of HEVC due to the significant potential for operating expense savings and growing demand for higher quality of experience over increasingly stressed mobile networks. These are expected to rely on software decoders in the short-term, with more power efficient hardware decoders or video co-processors expected to become available in the 2015 timeframe.
Satellite DTH service providers are also expected to leverage HEVC to roll out Ultra HD channels in the 2014-2015 timeframe, although it is hard to predict the level of uptake they will see in the early years of the technology. Similarly, some pilot DTT channels are expected to roll out in the 2015 timeframe, but the level of uptake remains to be seen. High-end encoding vendors such as ATEME, Ericsson, Fujitsu and NTT, and high-end CPE vendors such as Technicolor are all beginning to add HEVC to their product portfolios. All things considered, while certain applications will embrace HEVC much sooner than the norm and HEVC encoding and decoding cores should mature by 2014, we expect it will be around 2017 before a comprehensive ecosystem of first-generation HEVC-enabled products will come to market by 2017. Furthermore, we expect AVC to remain in widespread use even in 2018, although it will definitely be considered a commodity technology at that point – much as MPEG-2 is today.
The key takeaway from all of this is that HEVC won’t be adopted as quickly as some may think and if they bet big on HEVC too early, that’s one bet they are going to lose. We’ve done a lot of work at Frost & Sullivan on the topic of HEVC and in addition to the three reports we’ve already publish, we’ve done a lot of private research on HEVC for clients. If you’re looking to get more details on HEVC technology, get copies of the reports I mentioned, or need any custom research on the HEVC market, please feel free to reach out to me for more details.