Summer 2018 Update (3 years later):
HEVC Advance is heeding our advice. And *blush*—this blog has played a role in their new strategy. Three years after we published the criticism below, we noticed that 70% of visitors to this article originate from within the Boston organization or from IP addresses associated with patent pool members. The patent licensing changes have a striking resemblance to our recommendations.
Manufacturers of high-def Blu-ray players, 4K TVs, video streaming devices, cloud media vendors, software developers, and virtually all mobile gadgets are receiving an invoice, and it is a shocker. It’s not just the amount of the bill that is staggering, it’s what is demanded of them—facts and calculations that could be almost as costly to compile. The reason why this is such a big business is because no-one else can compress the file as they can. If we didn’t have this compression, there would be no Amazon Prime, Netflix or any TV guides to learn more about hbo showtime?! So we thank the inventors, please read on to learn more about H.265 HEVC compression.
The question asked in the title is rhetorical, of course. Lawyers run HEVC Advance, which is the licensing consortium for the H.265 patent pool. But according to one industry analyst, the lawyers have not consulted with streaming services, and their licensing model is completely unworkable. I agree.
H.265 is a new international standard for high-efficiency video compression. Backed by two major standards bodies and a plenitude of brand-name tech companies, the technology also goes by the name x265 or HEVC (high efficiency video encoding). Compared to H.264 (the standard for DVDs and streaming services such as Netflix and YouTube,[1] it reduces file size and mobile bandwidth by 35~50% depending on content. Adherents claim that this the coding efficiency will have a floor of at least 50% once the full feature set is baked into codecs.
H.265 is open source, but it is not free. It is the product of slew of inventors and other intellectual talent. Therefore, software developers and gadget manufacturers (especially companies that make TVs and video streaming devices) have been awaiting word on the royalty fee and licensing terms.
There are four recognized players in the market to supersede H.264, but most of manufacturers are backing either HEVC or a free encoder from Google called VP9. Ever since HEVC Advance published their licensing requirements last week, VP9 is looking a lot more attractive.
This past week, HEVC Advance published the royalty structure for anyone that wants to bake the growing market for HEVC into their products and services. The problem is that instead of sending their bill to the device manufacturer, they want a piece of every single film, game, video or graphic that is sold. They want 0.5%, and they want this royalty backdated to the first use of a standard that was ratified 2 years ago.
0.5% may not sound like much, but here’s the rub: They want it on content rather than technology.
First, it is a staggering amount when you consider that the market is set to explode—even more than it did for the previous standard H.264. That’s because few people plan to purchase video on DVDs. They want it streamed, much of it wireless, and with the very smallest bandwidth possible. The market is measured in the billions of dollars, even within the US alone.
But the structure of their licensing demands is worse than the land grab. Rather than attach to the device or user application, they seek to tax content hosts, forcing them to determine the fraction of video sales, rental or streaming services that use the new encoding technology. And they want a share of every video that is processed or streamed with a consumer standard. They fail to realize that taxing the flow rather than the vendor or pipe layer is not the way to ignite the adoption of a universally accepted consumer standard. In fact, it is stifling.
Dan Rayburn at streamingmedia.com discusses the nuances and futility of imposing an attorney-crafted bureaucracy on a burgeoning pivot point for consumer technology. His damning observations are clear and articulate. But ultimately, it boils down to a simple children’s story. If you kill a goose that lays golden eggs, you cannot harvest eggs. Lawyers fill a vital role. I am not against lawyers. But they are incapable of thinking in simple terms. [continue below image]…
If the HEVC licensing authority holds its position, it will be a pity for everyone: artists & content owners, streamers & other content providers, hardware manufacturers and consumers. H.265 is a great standard backed by an outstanding collaborative effort. Just a week ago, there was terrific potential for market unification, bandwidth reduction, storage savings, and an array of high-quality video services. BD-BR measurements give HEVC the edge in coding efficiency and the playback load is very light.
Although I could live with a half-percent delivery tax (or consumption tax, depending upon your view), it would result in a fragmented market. It hints at a bureaucracy that rivals the IRS. Ultimately the inventors will collect a tiny fraction of the potential. But the deal stopper is the effort to collect with a requisite fractional-calculation of the sales, rental and share-allocated revenue.
The market for next-generation video encoding is just starting to approach that critical adoption knee in which a 2-sided network floats all boats. We are on the precipice of a multi-billion dollar market. But now, Google—the visionary in this gambit—has the upper hand. If I could talk with lawyers at HEVC Advance, I would help them to navigate deployment dynamics and fundamental network economics…
But like Dan Rayburn, I would prefer a conversation directly with the technology stakeholders. They stand to enjoy a stunning return with a more reasonable model (reasonable on the Laffer curve and reasonable in implementation architecture). My suggestion: Take a royalty from the sales of hardware and application codecs, but not content volume. A fair model that rewards innovation while freeing explosive growth would be crafted like this:
Wild Duck Royalty Proposal
• 0.1% of software codecs within revenue-software applications or processes
• 0.15% of playback codecs built into hardware devices, [2]
• $2 or 0.15% for each encoding process or hardware device, whichever is higher [3]
(#2 and #3 are additive, for any device that encodes)
• 0.1% of paid apps that are extensible (those that support plug-ins or open source codecs) [4]
• Free – Personal reference app (published by patent consortium). It supports stand-alone video playback
• Free – Hosting or distribution Content encoded with H.265
I challenge HEVC Advance to run the numbers for any believable growth curve. This formula is not only more palatable, auditable, enforceable, and reasonable, it also delivers higher lifetime revenue to patent stakeholders. More importantly, it makes them the good guys.
Hey, HEVC Advance! I want you to succeed. Reach out to me. Use the contact link at the top of this page.
[Ellery Davies is editor of AWildDuck.com. He is also CEO &
Co-Chair of CRYPSA, a recognized standards organization]
[1] Until recently, most YouTube content was encoded in FLV. Currently, YouTube plans to support it’s own open-source VP9, but is leaving open the possibility of supporting H.265 which is licensed by HEVC Advance.
[2] The proposed royalty applies to any hardware device or app that advises or suggests the download of HEVC plug-in or compatible routine.
[3] $2 or 0.15% of sales, whichever is higher, but the fixed fee is reduced to no more than 6% of the hardware value. Up to 16 cores in a single gadget, PC, or device less than 1 cu feet not including external display.
This stacks on top of playback royalty. So, for example, if an video player application can open an x264 video and save as x265, it would pay a royalty of 0.1 or 0.15% (for the playback codec) and $2 or 0.15% for the encoding feature, whichever is more.
[4] For apps that do not include HEVC, it would be difficult to collect a royalty, of course. But developers of x265 plug-ins and helper apps would be liable.
Great article, agree. The burden of accounting for and administrating the 0.1% will almost certainly exceed 0.1%. That is except for those companies large enough for economies of scale and special audit arrangements with HEVCAdvance,
Another issue is the long tail effect. No-one has mentioned this but do HEVCAdvance really want every vlogger/minor content owner/amateur rock band/startup using video in some form to have to calculate their 0.1% of attributable income to see whether they need to mail a check for pennies of royalties? If these players move to VP9/10, they’ll take a huge part of the ecosystem with them.
You don’t talk about it but the Region 1 / Region 2 thing is just as unworkable as the 0.1% content fee. Once devices leave the factory for Region 2, how to prevent them being resold to supposed wealthier customers in Region 1? Manufacturers won’t always know how and where product will be consumed. This requirement also kills flexibility and adds cost to the device supply chain.
Other points on the wild duck proposal: percentage of device value is difficult to account: who is liable and do they know the actual retail price? Making HEVC plugin publishers liable for a % of a revenue stream they do not see or touch is also a tricky one.
When you distil everything, it’s difficult to come up with a fee structure that is very different from MPEGLA’s! And given the fair/reasonable requirement, it would be difficult to justify charging more than MPEGLA.
So I don’t think HEVCAdvance’s announcement is meant to be taken seriously. It’s an opening offer, analogous to a speculative $0.01 bid on ebay.
Note: x264 and x265 are specific encoder implementations of H.264 (aka AVC) and H.265 (aka HEVC) respectively. You shouldn’t use “x265” as the name of the format – it’s like saying “a car, also known as a Ford”. Many other encoders exist.
Great point about the potential for region-resale. In fact, it is not just a potential, it is an incentive for cross-boarder sales.
Even back in the early DVD days, I never accepted the security or economic justification of regionalization. The goals and mechanism constitute a non-sequitur.
Regarding your correction to my x265 nomenclature, I appreciate the correction Good catch, John! I am correcting the article.
~Ellery