If video codecs were presidential nominees, then it wouldn’t be too much of a stretch to characterize HEVC as the “establishment” candidate of the codec coterie. The presumptive successor to the hugely successful and widely implemented H.264, HEVC’s position as the next generation codec seemed assured when it was first solidified.
But just as establishment candidates in the Democratic and Republican primaries have seen their smooth path to the nomination disrupted and even derailed by insurgent dark horses, HEVC is being challenged by a number of upstart video codecs promising a “royalty-free” alternative. And as with presidential nominee candidates, some of those promises may be hard to fulfill.
Since we chronicled the rise of rival next-generation codecs last year, several members of the insurgency have consolidated their efforts around a new organization: the Alliance for Open Media (AOM). Launched in September 2015 with a mission to “develop next-generation media formats, codecs and technologies in the public interest,” the Alliance’s founding members include some major heavyweights: Amazon, Cisco, Google, Intel Corporation, Microsoft, Mozilla and Netflix. Just in April, the Alliance added chipmakers AMD and NVIDIA and the semiconductor IP company ARM to its roster of supporters, shoring up critical hardware support.
All of these members bring significant expertise to bear on the problem of video delivery and several members—Mozilla, Cisco and Google—are using the alliance to consolidate independent efforts at developing a next generation video codec that is open source and royalty free. Specifically, Cisco’s Thor, Mozilla’s Daala and Google’s VP10 (a planned successor to the still nascent VP9) will all be thrown into the AOM mix and are expected to play a foundational role in whatever codec emerges.
ARM’s Mark Dickinson, general manager of the company’s media processing group, said in a statement that AOM was “working to ensure that the new codec can and will be implemented in hardware once the software specification is finalized. ARM’s focus on mobile multimedia will enable hardware implementations of the AOM Video codec to quickly become available to all OEMs operating in the mobile market.”
This is, of course, easier said than done. Like the other rivals we canvassed last year, AOM faces steep hurdles if it wants to unseat HEVC. It’s had a comparatively late start and will likely lose even more time as it attempts to recruit more members and vet their technologies to ensure they don’t violate existing video patents (no small feat).
Meanwhile, HEVC is already deployed in the wild—in processing chips from major supplies such as Broadcom and Sigma, in smart TVs from all of the leading TV vendors, in broadcast encoders, in set-top boxes being deployed to customer’s homes, and in mobile phones. More importantly, those content owners preparing UHD content are already encoding and distributing with HEVC. The exception is YouTube, which not surprisingly has encoded 4K/UHD content in Google’s VP9. It did the same in the last codec upgrade cycle by encoding content with VP8, but as AVC/H.264 became the defacto standard, YouTube switched to AVC/H.264. Each passing day brings still more products, entrenching HEVC’s status.
That’s not to say that the Alliance faces bleak prospects, particularly outside of the broadcast industry. Web companies, Google especially, have been complaining that the 10-year cycle of codec development and implementation that marked the evolution from MPEG-2 to AVC and now to HEVC is far too slow to innovate in a rapidly evolving tech marketplace. With Google, Microsoft and Mozilla on board, three of the four major web browsers are in place to support the AOM codec—and browsers can be updated rather quickly relative to other pieces of the video delivery ecosystem.
The inclusion of Google’s VP10 and Cisco’s Thor technology in the AOM software pool indicates that the members are looking perhaps not at a replacement for HEVC but at a successor to the codec, which means they’d have to deliver technology that doesn’t infringe on existing patents and technology that offers compression efficiencies and other improvements on HEVC—a tall order for sure.
Still, it’s wrong to view the emergence of AOM alongside HEVC as a zero-sum contest. Unlike presidential elections, we can have more than one winner. The world can live with two (or more) codecs governing video delivery, even if that’s a less-than-elegant solution. Google’s VP9 and its support among SoC suppliers who also support HEVC is a good example of how co-existence is likely to work should AOM succeed in delivering a viable alternative.