HEVC Advance License Fees

[x265_Project]

Well, this is disappointing... http://blog.streamingmedia.com/2015/07/new-patent-pool-wants-share-of-revenue-from-content-owners.html

[nevcairiel]

That was that for HEVC then, guess we'll wait for the next video codec to see wide adoption.
Greedy lawyers ruining the industry..

[Jamaika]

So at this moment a chance to other software vendors, ie. VP9, Daala. I'm just not sure if streaming TV is possible.

[foxyshadis]

Who is this HEVC Advance consortium? The MPEG LA already has its own license for HEVC that covers the entire standard for similar rates as AVC. These guys don't list their patents, appear out of nowhere to charge exorbitant fees to people who've already paid for patent protection from MPEG. If this is an example of some minor patent holders holding out in order to carve a bigger slice of pie all for themselves, then all they've done is assure that they end up in international litigation for years (decades?).

[x265_Project]

Who is this HEVC Advance consortium? The MPEG LA already has its own license for HEVC that covers the entire standard for similar rates as AVC. These guys don't list their patents, appear out of nowhere to charge exorbitant fees to people who've already paid for patent protection from MPEG. If this is an example of some minor patent holders holding out in order to carve a bigger slice of pie all for themselves, then all they've done is assure that they end up in international litigation for years (decades?).

"The initial list of Licensors is expected to include GE, Technicolor, Dolby, Philips and Mitsubishi Electric and will offer more than 500 essential HEVC/H.265 patents through the license HEVC patent pool. The number of patents is expected to grow significantly as the additional HEVC patent owners become Licensors." - http://www.hevcadvance.com/#company

Obviously, the companies that have already signed up to license their patents through MPEG-LA would be getting substantially less (20 cents/unit after the first 100,000 units in a contract year, capped at $25 million annually, with no royalty on content). MPEG-LA licensors include Apple, British Broadcasting Corporation, Fujitsu, Hitachi Maxell, JVC KENWOOD, NEC Corporation, NTT DOCOMO, Samsung, and Vidyo.

Who's still missing from both lists? Google (owns Motorola's HEVC patents), Microsoft, Sony, Panasonic, Fraunhofer...

Companies like Netflix, Amazon, Hulu, Apple, Facebook, Twitter, Comcast, DirecTV, etc. will never agree to pay 0.5% of their "applicable (top-line) revenue" to this patent administrator. They've badly overplayed their hand. This won't stand.

[Parabola]

Agree, this won't fly: it's unworkable. To me it seems like an opening bid and I suspect in a year's time the deal will be very different. Anyone with an alternative codec is going to be very happy (and busy right now).

[nevcairiel]

As the article in the OP mentioned, MPEG-4 Part 2 (ie. MPEG-4 ASP) had similar licensing, and what was the result? Noone simply used it for anything. They stuck to MPEG-2 until H.264 came to be.
I sure hope that this isn't the fate of HEVC because some patent holders wanted more money.

[BadFrame]

Bummer. When I first read about this I was convinced it was simply 'saber rattling' from the companies in this 'breakaway group' to negotiate a bigger slice of the MPEGLA royalty pie.

Now it seems they are fully intent on going their own way, thus destroying the advantage of having one entity to negotiate with (MPEGLA), as well as increasing the cost of using HEVC.

Hope this won't end up hurting x265's bottom line too much :/

[Parabola]

As the article in the OP mentioned, MPEG-4 Part 2 (ie. MPEG-4 ASP) had similar licensing, and what was the result? Noone simply used it for anything. They stuck to MPEG-2 until H.264 came to be.
I sure hope that this isn't the fate of HEVC because some patent holders wanted more money.

Going off topic slightly but not entire industry bypassed MPEG-4 Part 2. DivX established a successful ecosystem around it. A problem with that standard was that the new deblocking/deringing filters were out-of-loop and non-normative. Consumer electronics decoders generally did not bother with this post filtering and the resultant performance on these devices was no better than MPEG-2. H.264 brought the filtering in-loop, making it mandatory.

[pieter3d]

So at this moment a chance to other software vendors, ie. VP9, Daala. I'm just not sure if streaming TV is possible.

Most of the new higher-end connected 4k TVs have VP9 decoders integrated. A hardware VP9 decoder is also in the Galaxy S6.

[roo1234]

I guess h265 is fated for most home use, because of the patent hell. Congrats!

[foxyshadis]

Ars just wrote up an article about this. (http://arstechnica.com/tech-policy/2015/07/new-patent-group-threatens-to-derail-4k-hevc-video-streaming/)

Of course this is as much a negotiating tactic as it is an actual money grab, but this announcement really was the nuclear option. (Well, aside from flat-out suing some users.)

[schweinsz]

"The initial list of Licensors is expected to include GE, Technicolor, Dolby, Philips and Mitsubishi Electric and will offer more than 500 essential HEVC/H.265 patents through the license HEVC patent pool. The number of patents is expected to grow significantly as the additional HEVC patent owners become Licensors." - http://www.hevcadvance.com/#company

Obviously, the companies that have already signed up to license their patents through MPEG-LA would be getting substantially less (20 cents/unit after the first 100,000 units in a contract year, capped at $25 million annually, with no royalty on content). MPEG-LA licensors include Apple, British Broadcasting Corporation, Fujitsu, Hitachi Maxell, JVC KENWOOD, NEC Corporation, NTT DOCOMO, Samsung, and Vidyo.

Who's still missing from both lists? Google (owns Motorola's HEVC patents), Microsoft, Sony, Panasonic, Fraunhofer...

Companies like Netflix, Amazon, Hulu, Apple, Facebook, Twitter, Comcast, DirecTV, etc. will never agree to pay 0.5% of their "applicable (top-line) revenue" to this patent administrator. They've badly overplayed their hand. This won't stand.
A real very big player of the video coding technology patents is NOKIA.

[Jamaika]

Nokia's probably bankrupt. In Poland, I don't see advertising. I haven't heard anything about the company for five years. I read that it is Microsoft.

[x265_Project]

The proposed device royalties are up to 12.5 times what MPEG-LA charges for their HEVC license. But we go from insane to completely ludicrous with the proposed fee on "applicable revenue" of content distributors.

I'm confused by the legality of this proposed licensing scheme. What is an HEVC bitstream? With respect to patent law, it's a signal. Signals, it turns out, are not patentable (https://en.wikipedia.org/wiki/List_of_patent_claim_types#Signal) in the US.

[CruNcher]

It's a very good thing that the smartest independents moved away from MPEG and their Research Patent Domination as soon as possible understanding the Risks very early on ;)

i feel that x265 wont be as successful anymore as DivX ;-)/XviD/x264 that thing ended right here and it was very predictable (seeing also x264s commercialization move by CoreCodec) that everyone is so surprised now by this move is kinda strange ;)

Now economical demands are gonna push alternatives and alternative Research more then ever before good work MPEG showing your real Face ;)


@x265_Project

They of course will pay but now they gonna start to take the knife and stab in their back (already started) ;)
Though a thing MPEG Members should know very well how that works stabbing each other, lot of experience there ;)

In the End it all depends if the End user will accept higher Service Fees of course for a MPEG driven solution ;)

[pieter3d]

VP9 is free, well-developed and on-par with HEVC. Also there is good HW support for it in 4K TVs, even some phones, with more coming. Also Chrome and Firefox support it, so playback support is actually much better than HEVC.

[mandarinka]

Well, format might be on par, but is that the case for libvpx encoding? It is still horribly slow (and not properly threaded, even) and has rate control that is absolutely unusable according to most people who tested it.
Being stuck with VP9 as such might not be that bad, but having to watch stuff encoded by libvpx? Damn (http://compare.bakashots.me/compare.php?setId=1086&comparisonId=7375&imageNum=1).

[x265_Project]

Well, format might be on par, but is that the case for libvpx encoding? It is still horribly slow (and not properly threaded, even) and has rate control that is absolutely unusable according to most people who tested it.
Being stuck with VP9 as such might not be that bad, but having to watch stuff encoded by libvpx? Damn (http://compare.bakashots.me/compare.php?setId=1086&comparisonId=7375&imageNum=1).

I should make it clear that we have no problem with intellectual property holders being able to protect and license their work. In our market (codec implementations), there is nearly perfect competition, and so the price of our IP is dictated by the laws of supply and demand. When worldwide technical standards are developed, however, the owners of the IP in these standards gain more of a monopoly position, as everyone naturally wants to adopt a leading standard, but any IP holder in the standard can act as a gatekeeper. This is why standards bodies require contributing companies to promise to license their IP on Fair, Reasonable and Non-Discriminatory (FRAND) terms. But what is reasonable? MPEG-LA's terms are reasonable. What HEVC Advance proposes (which would be in addition to MPEG-LA license fees), is clearly not reasonable.

The combined market power and political influence of the world's web video services, broadcasters, movie studios, semiconductor companies and device OEMs far outweighs the influence of HEVC Advance. If the HEVC Advance companies aren't persuaded to drop this insane proposal and join the MPEG-LA license pool, the world's content companies will find a better alternative.

VP9 is one option, but it has its own issues. You've noticed that if you want to encode VP9 you have one encoder to choose from. This isn't the same ecosystem that MPEG standards enjoy, where you have dozens of companies building encoder implementations. Despite the fact that VP9 is free and open source, it's controlled by Google, and that makes many potential adopters uncomfortable. Content companies and broadcasters feel better about adopting a standard that has a large and competitive ecosystem creating the implementations they need. It's also difficult for us and others to build and sell an encoder that competes with a free option offered by Google, the standard owner.

The world's content companies will never pay a content royalty on a new format, and the world's video hardware and software developers will not pay unreasonably high device royalties.
The ITU and ISO need to do better next time. The Joint Task Force on Video Coding should never have been allowed to develop the HEVC standard until the participating companies had first agreed to a better framework for licensing the technology. Had this been decided up-front, we wouldn't be dealing with uncertainty today. Going forward, it's reasonable to envision a new standard that includes all of the HEVC coding techniques developed by reasonable patent holders, dropping all of the methods proposed by GE, Technicolor, Dolby, Philips and Mitsubishi Electric. This would take years to develop, test, implement and deploy, but the world's adopters will do whatever is necessary to avoid unreasonable, impractical license fees.

Another good step would be to change and/or clarify patent laws. We need to make it clear that while the methods and systems used to create or decode the bitstream can be patented, the bitstream (data, or signal) cannot. Video makes up 70% of the bits crossing the Internet, and orders of magnitude more bits flowing through dedicated video distribution systems (cable, satellite) and distributed on physical media (DVD, Blu-ray). No one should be allowed to tax content.

Personally, I'm optimistic that the HEVC Advance patent holders will recognize that it's not in their interest to delay or kill worldwide adoption of HEVC. The longer this cloud hangs over our heads, the more that the world's implementers and adopters will seek and develop better alternatives. A reasonable share of a very good revenue stream is better than a large share of nothing. HEVC is poised to really take off, so this needs to be figured out quickly.

[pieter3d]

Well, format might be on par, but is that the case for libvpx encoding? It is still horribly slow (and not properly threaded, even) and has rate control that is absolutely unusable according to most people who tested it.
Being stuck with VP9 as such might not be that bad, but having to watch stuff encoded by libvpx? Damn (http://compare.bakashots.me/compare.php?setId=1086&comparisonId=7375&imageNum=1).

These kinds of artifacts have been addressed in recent builds. Same goes for the speed and threading. You should really try out the top of the git repository, not the release tags.

[CruNcher]

VP(X)10/9 situation is iffy so much buildup on MPEG only some changes it carried on since True Motion/On2 and it can not be assured that someday someone knocks on the Door an calls out his patent Google never can assure this wont happen wee need something to buildup from the Old days where most patents expired and Research again in a different direction for me this is Daala (IETF NetVC) the right way to get free from MPEGs Domination once and for all :)

They never gonna Adapt this fast they are slow and driven by so many rules and stabbing each other after their Meetings about which patent gets in over and over again we are more agile, faster, focused and have 1 common goal ;)
They not gonna understand what overrun them in the End ;)

User Content is much more in quantity then Hollywood will ever produce we have a chance to break free if not now when then ? ;)

We just have to set the rules and im sure they gonna follow if they under enough pressure, that's practically proven it's their weakest point :)

We need to produce Content and flood the System with it a lot of content in this 1 Goal format asap :)

I hope many will join in they don't need to show themselves they can do it hidden but preparing to Stab back is the only Chance also they (content distributors) have :)

Viva la Revolution ;)

Neo Wakeup !

[foxyshadis]

HEVC Advance was supposed to be meeting for the first time in Tokyo this week, beginning yesterday. Anyone know if that happened and have any details?

[x265_Project]

http://www.hevcadvance.com/pdf/HEVC_Advance_press_release_FINAL_9.22.15.pdf

HEVC Advance is also actively soliciting input from market participants and considering adjustments to arrive at a royalty structure that enables continued and rapid adoption of HEVC and brings the associated benefits to stakeholders within the media and technology industries.
“We have received significant market feedback, particularly on content fees, and will adjust fees to support widespread use of HEVC,” Moller said."

[x265_Project]

Article - http://www.streamingmedia.com/Articles/Editorial/Featured-Articles/HEVC-Advance-Releases-Revised-Licensing-Terms-108230.aspx

Fee Structure - http://www.hevcadvance.com/pdf/RoyaltyRatesSummary.pdf

[BadFrame]

This is good news, however as I understand it, this still means you will have to pay royalties to both MPEGLA and HEVC Advance in order to use HEVC ?

[foxyshadis]

The most important part of this to many people here is that all content licensing fees are waived for any content that the user doesn't pay for, whether it's given away, ad-supported, or available in any other way without charge. The other fees are more reasonable than before but still significantly higher than MPEG-LA's. I smell further challenges and renegotiations coming up next year.

[x265_Project]

This is good news, however as I understand it, this still means you will have to pay royalties to both MPEGLA and HEVC Advance in order to use HEVC ?
As things stand now, that's accurate. Of course this would be more expensive and difficult to manage for licensees. And it would certainly result in an unfair collection and distribution of HEVC license royalties. Let's hope that further negotiations result in a single pool with much more reasonable terms.

[smok3]

If I assume that this pool is kinda legit, how is this unfair to content providers? (They can either charge costumers for better HEVC stream and keep vp9/whatever cheaper/free or invent whatever streaming method they want). And the article is mentioning "poor" apple and I'am pretty sure apple won't pay a cent to anyone, misleading. And what has CDN to do with anything?

As for the subjective view, most of the time the encoding tech is way better that the content, so why not give tech people, the actual patent holders some moneys (Tom Hardy is not poor I imagine).

[Washka]

More about HEVC Advance consortium and their fees.

http://www.guru3d.com/news-story/licenses-for-the-hevc-videcodec-become-cheaper-yet-its-too-late.html

[BadFrame]

More about HEVC Advance consortium and their fees.


The numbers were certainly interesting, it's not hard to see why companies like Google, Netflix, Amazon and the like (leading video content providers over the web) is cooperating on a royalty free codec, there are HUGE amounts of money to be saved in the long run.

Of course it helps when the HEVC licensing goes haywire with new patent pools emerging with their own royalty demands, unlike the article I think there's still a chance that this can be resolved, preferably with one single patent pool again, but it doesn't look very promising.

The article also stated that VP10 is almost ready (!), I thought it was still in heavy development and/or in transition to become the official codec of the 'Alliance for Open Media'.

[iwod]

For the first time ever I hope Apple will join the Open Media Alliance and use VP9. Despite I think VP9 is inferior to HEVC, or heck even a decent H.264 encode.

The terms for HEVC advance still sucks. It allows them to get a 20% increase every five years. So the combined total revenue to HEVC patents, MPEG-LA and HEVA Advanced together, will properly exceed 10B over its lifetime.

10B+ for these patents? Are they freaking nuts?

P.S- I am not against patents, I think they provide additional revenue to universities research. But as far as I am concern, those money aren't even going there. Cooperate Companies are simply being a patient trolls themselves.

[Procrastinating]

Patents are fine and dandy, but are held for way too long in this day and age, especially software patents.
It is unreasonable for a start, that companies can sue for "previous infringement" of patents, over 5 years after the patent expires.

And then on the software side, entirely new markets, codecs and algorithms can live and die in the span of 5 years. For a company or institute to hold an algorithm ransom for 20 years is crazy, and almost certainly stifles more innovation than it encourages.

[iwod]

A question pops up on my head, are you require to pay licensing fees if you use free implementation of HEVC such as x265, purely a software implementation and play on a devices which uses software decode only?

[x265_Project]

A question pops up on my head, are you require to pay licensing fees if you use free implementation of HEVC such as x265, purely a software implementation and play on a devices which uses software decode only?
To patent licensors, a "device" includes software implementations. When you use x265, whether it is under the GPL v2 license or our commercial license, you are getting the right to use our copyrighted software, but these licenses don't include a license to any patents. So, distributors of software applications or systems powered by x265 are also responsible for licensing the necessary patents in their part of the world.

[x265_Project]

Technicolor has decided to withdraw from HEVC Advance.
http://www.econotimes.com/TECHNICOLOR-TECHNICOLOR-WITHDRAWS-FROM-THE-HEVC-ADVANCE-POOL-TO-ENABLE-DIRECT-LICENSING-OF-ITS-HEVC-IP-PORTFOLIO-155090

There is no word on their proposed device royalties, or other terms of their license, other than the fact that they will not charge for content.

They claim they are doing it "in order to accelerate adoption of the standard". At the same time they admit that this will "generate value more effectively than would have been possible through participation in a patent pool".

So, as it stands now, HEVC adopters will now have 3 organizations to pay license fees to - MPEG-LA, HEVC Advance and Technicolor.

[iwod]

Technicolor has decided to withdraw from HEVC Advance.
http://www.econotimes.com/TECHNICOLOR-TECHNICOLOR-WITHDRAWS-FROM-THE-HEVC-ADVANCE-POOL-TO-ENABLE-DIRECT-LICENSING-OF-ITS-HEVC-IP-PORTFOLIO-155090

There is no word on their proposed device royalties, or other terms of their license, other than the fact that they will not charge for content.

They claim they are doing it "in order to accelerate adoption of the standard". At the same time they admit that this will "generate value more effectively than would have been possible through participation in a patent pool".

So, as it stands now, HEVC adopters will now have 3 organizations to pay license fees to - MPEG-LA, HEVC Advance and Technicolor.

This sucks.....

So, could a company theoretically, license all the patents for HEVC, and then create their own codec, which is essentially the same as HEVC and call it something like XXX.

And then relicense XXX under a different terms to recoup for their expenses?

( I think this is similar to what happen to Divx in the Mpeg4 era )

[x265_Project]

This sucks.....

So, could a company theoretically, license all the patents for HEVC, and then create their own codec, which is essentially the same as HEVC and call it something like XXX.

And then relicense XXX under a different terms to recoup for their expenses?

( I think this is similar to what happen to Divx in the Mpeg4 era )
You took a left turn there; way off topic. Technicolor leaving the HEVC Advance patent portfolio won't lead to the kind of outcomes that you're envisioning.

The original DivX codec was a reverse engineered hack of a Microsoft MPEG-4, Part 2 codec. The guys who developed DivX didn't develop new coding standard, they just developed their own implementation of an existing coding standard.

Patent licenses usually apply to specific implementations of the patented technique (products) sold by the licensee to end customers. They usually cannot be sublicensed in the way that you described.

[iwod]

You took a left turn there; way off topic. Technicolor leaving the HEVC Advance patent portfolio won't lead to the kind of outcomes that you're envisioning.

The original DivX codec was a reverse engineered hack of a Microsoft MPEG-4, Part 2 codec. The guys who developed DivX didn't develop new coding standard, they just developed their own implementation of an existing coding standard.

Patent licenses usually apply to specific implementations of the patented technique (products) sold by the licensee to end customers. They usually cannot be sublicensed in the way that you described.

Not really off topic, nothing more then my fantasy mind to solve the issues.

So this is it then. No more HEVC coming out ( apart from China which doesn't care about Software patents anyway )

We could have had very nice 1080P 1Mbps streaming. Which help those of us who can't get a faster internet connection.

We could have had BPG, finally a graphics format that beat jpeg in everyway.

We could have had 4K in our Set top Box.....

Thanks to patents and greed this is now all gone.

[x265_Project]

Not really off topic, nothing more then my fantasy mind to solve the issues.

So this is it then. No more HEVC coming out ( apart from China which doesn't care about Software patents anyway )

We could have had very nice 1080P 1Mbps streaming. Which help those of us who can't get a faster internet connection.

We could have had BPG, finally a graphics format that beat jpeg in everyway.

We could have had 4K in our Set top Box.....

Thanks to patents and greed this is now all gone.
Dude... take it easy. It's going to be OK. All is not lost.

Technicolor pulling out of the HEVC Advance patent portfolio makes a messy situation a bit messier (3 places to license from instead of 2, and slightly more expensive), but certainly not impossible. We don't even know the details of what Technicolor will attempt to charge for their patents.

[birdie]

We could have had BPG, finally a graphics format that beat jpeg in everyway.

Aside from heavy blurring even at the highest compression ratios and by that time BGP is not really more effective than JPG.

BGP shines at mid to lower compression ratios when JPEG start looking like pixellated s***, and BGP looks like big blurry gradients.

In fact I cannot find a single reason to use BGP. Oh, and it doesn't support EXIF/IPC so it's useless for photo professionals. So, again, what's its use case?

[foxyshadis]

Aside from heavy blurring even at the highest compression ratios and by that time BGP is not really more effective than JPG.

BGP shines at mid to lower compression ratios when JPEG start looking like pixellated s***, and BGP looks like big blurry gradients.

In fact I cannot find a single reason to use BGP. Oh, and it doesn't support EXIF/IPC so it's useless for photo professionals. So, again, what's its use case?

Not sure where your info on BPG came from, but it fully supports EXIF/XMP/IPTC metadata. We're all entitled to our opinion, but I'll say that my experience regarding its quality is that it looks quite a bit better at any given size. The intra prediction alone changes everything.

Compared to JPEG-XR, however, it doesn't always win, and the lack of royalty-free licensing makes it a bit of a bummer.