HEVC IP Owners Are Killing the Golden Goose Over Royalties
Recent input at Streaming Media East reveals HEVC to be a diverse and powerful technology that many publishers have permanently eschewed due to unclear royalty policies and excessive royalty demands. I don’t mean this to be (yet another) bash-the-greedy-HEVC-IP-owners post, but a continuation on the cautionary tale on how HEVC IP owners have potentially killed the goose that could have laid the golden egg. With VVC coming down the pike, it’s a story that can’t be told too often.
HEVC Is a Powerful Technology…
Here’s what I heard and saw at Streaming Media East. First, on the show floor, Teradek’s John Landman showed a 500 kbps 1080p HEVC-encoded stream produced by a Cube 755 that retained good detail and color considering the data rate, and exhibited higher quality than many news videos we see on TV every day.
A few hours later, Beamr VP of strategy Tom Vaughan showed me a 1920x800 version of Tears of Steel encoded at 1 Mbps that you can download and view on the Doom9 forum. Vaughan encoded the clips with a pre-release (development) version of Beamr 5 (v4.5) in response to an HEVC encoding challenge from Amazon’s Ben Waggoner to see if any encodes could beat his x265 encodes.
Figure 1. Beamr 4.5, 1920x800@ 1 Mbps.
Vaughan’s video looked great to me, but this was a quick look on a notebook screen under fluorescent lights in a busy conference setting. On the board, commenter Emulgator stated (presumably after longer study in a more conducive environment), “Remarkably kept/simulated texture, sometimes glued onto the underlying surface, but nicely done. The expected artifacts at such low bitrate are unexpectedly small, and well shifted to domains where it doesn't hurt...Nice research, never believed that would be possible!”
This and the Teradek clip shows that HEVC can be an exceptionally useful codec.
...Whose Ship Has Sailed
Contrast this with the discussion after Christian Feldman’s excellent Streaming Media East session on AV1 and VVC, wherein the Bitmovin codec engineer contrasted and compared the two codecs technically and from a licensing perspective. Inevitably, the HEVC royalty situation arose during the Q&A.
Several publishers in attendance proclaimed that they would never deliver HEVC-encoded video because of the risk of content royalties, which the MPEG LA and HEVC Advance pools have long ruled out, but Velos refuses to do. Since roughly 25% of HEVC owners aren’t in a pool, there’s also a risk that these owners might attempt to impose royalties at some point in the future.
What I found remarkable was how angry these publishers were when discussing these policies, perhaps a reaction to the sheer unreasonableness of HEVC royalty policies. When one attendee mentioned that the Media Coding Industry Forum (MC-IF) had been formed to promote a more cohesive royalty package for VVC, another correctly observed that the MC-IF had no teeth and that it would be up to the VVC IP owners to formulate a cohesive strategy.
Note that it’s not the mere presence of royalties that doom a technology; MPEG-2 and H.264 both were royalty-encumbered and were tremendously successful for licensees and licensors. Rather, it’s dysfunctional royalty policies that kill the goose. Let’s hope VVC IP owners understand since it’s a lesson that many HEVC IP owners still haven’t learned.
The old realities that used to dictate codec adoption no longer apply. Opening up new markets now matters more than reducing operating expenses. How are HEVC, AV1, and VVC positioned for the future?
Live sports streaming and an insistence on professional quality IP video are helping cement HEVC's role as the standard.
Is AV1 all that people expect it to be? How much better would HEVC be doing with a fair royalty policy? Look to these charts for the answers to tomorrow's codec questions.
Early testing of HEVC playback in HLS showed positive but inconsistent results. We need StreamingMedia.com readers to step up and help provide more data.