Wednesday, January 19, 2022

'Access Advance' patent pool can rename itself 'Suffer Setback' after Dusseldorf court ruling: flagrant FRAND abuse concerning video codec patents

The most recent IAM Sunday Digest mentioned the Access Advance pool firm's VVC Advance announcement as well as the Dusseldorf Regional Court's conclusion that the HEVC Advance pool's terms are non-FRAND. I had mentioned the latter in a recent post on a wireless SEP dispute. While I haven't been able (yet) to obtain a copy of the Dusseldorf decision, I have meanwhile seen reports by IAM (as I mentioned) and Juve Patent, and would like to add some observations.

This is not going to be my last post on that set of cases and the underlying conduct, which I find egregious. A few organizations' greed threatens to bring the very concept of patent pools into disrepute unless we all make the proper distinction between the vast majority of pools that serve their purpose--and an abusive outlier like Access Advance. It is equally important to understand that just because a company joined that pool as a licensor doesn't necessarily mean that it has unclean hands. They aren't all wrongdoers, but the Dusseldorf disaster should give the more reasonable ones among those Access Advance licensors pause, above all with a view to the latest and greatest in video codec standards: VVC.

There's a certain complexity and some specialized terminology, so let's get an overview first.

From AVC (H.264) to HEVC (H.265) to VVC (H.266)

Wikipedia describes Advanced Video Coding (AVC), also known as H.264, as "by far the most commonly used format for the recording, compression, and distribution of video content, used by 91% of video industry developers as of September 2019." Since 2010 this blog has repeatedly mentioned patent pool administrator MPEG LA's AVC/H.264 Patent Portfolio License, which many regard as a key enabler of H.264's widespread adoption.

It's a huge problem that almost two decades after H.264's adoption as a standard, its successor--H.265 or High Efficiency Video Coding (HEVC)--is nowhere near as popular, despite its technical merits. I attribute that disappointing fact largely to patent licensing issues, as do others (for example, the Eagle Eye Networks blog says "adoption is slow" and "[t]he primary reason for this[] is that unlike [H].264 which has 1 patent pool, [H].265 has 3 patent pools with different pricing structures and terms & conditions." As a result, "many content providers have stuck with [H].264 because at least they know it will always play."

The next big thing could and should be Versatile Video Coding (VVC, H.266). It's roughly twice as efficient as HEVC.

Theoretically, you can encode videos with an 8K resolution even in H.264. But as a rule of thumb it makes sense to use at least HEVC (H.265) for 4K videos, and for 8K what you really want is VVC (H.266)--unless you don't give a damn about efficiency, of course. But with video accounting for the largest part of global data traffic (a large part of which is consumed on mobile devices), we should care.

"HEVC Advance"

I've never heard anyone complain about the terms of MPEG LA's HEVC pool license, but have repeatedly (!) been told by industry players and their counsel that they took issue with the "HEVC Advance" license. The pool administrator originally had the same name, and later renamed itself "Access Advance" so as not to be identified with a single standard. I put both names in quotes because to me "advance" means "progress"--a misnomer as that organization has actually impeded the adoption of HEVC, but I'll omit the quotes from here on out./p>

On its website, Access Advance doesn't say anything about ownership. For what I know, that entity is not the "independent licensing administrator company" it claims to be. It appears to be owned by four patent holders: General Electric (GE), Philips, Dolby, and Mitsubishi. They set up this entity in 2015 because they thought they could extract far higher royalties by not participating in MPEG LA's HEVC pool. They went on to persuade some of MPEG LA's HEVC licensors to join the HEVC Advance pool, on a basis that created all sorts of problems culminating in last month's Dusseldorf holding that the pool's terms are not FRAND.

Current Dusseldorf plaintiffs: GE, Philips, Dolby, IP Bridge

As a pool administrator, Access Advance does not hold patents, so its licensors have to enforce their IP if someone declines to take a license. In the current round of Dusseldorf cases (see this Juve Patent article), the plaintiffs are GE, Dolby, Philips, and Japan's IP Bridge (a non-practicing entity that is a public-private partnership involving the Japanese government).

Current Dusseldorf defendants: Vestel and Xiaomi

After getting a medium-sized (if not small) German company named MAS Elektronik to take an HEVC Advance license, those patent holders decided to take on Xiaomi and Vestel. The former needs no introduction, and the latter is a reasonably successful Turkish company that makes (inter alia) TV sets, though it is small compared to Xiaomi.

Overview of cases adjudicated in December

The Xiaomi cases will go to trial this spring. Six Vestel cases were adjudicated last month by the Dusseldorf Regiona Court's 4c Civil Chamber under Presiding Judge Sabine Klepsch (who in 2020 famously referred Nokia v. Daimler to the European Court of Justice, which never got to hear the case due to a settlement).

Two patents asserted by GE and two asserted by Dolby were deemed essential (and not considered very likely to be invalid). One IP Bridge and one Philips case were stayed as the patents-in-suit appeared fairly likely to be invalid.

Vestel had not only defended itself against the infringement allegations but had also counterclaimed, alleging that the electronics maker suffered damages as a result of the HEVC Advance pool terms being out of compliance with the FRAND licensing obligations of its licensors. In the cases in which GE and Dolby prevailed on the purely technical merits, the court furthermore had to reach the FRAND defense to injunctive relief--which largely overlaps with the FRAND counterclaim.

Vestel deemed willing licensee

The court applied the German Sisvel v. Haier standard, starting with an amalgamated perspective on whether the defendant was a willing licensee. The three-judge panel concluded that Vestel was a willing licensee. (We will see what happens in the Xiaomi cases, but that company does enjoy a good reputation in the licensing community.)

The hurdle for being deemed a willing licensee may be higher in Munich and Mannheim than in Dusseldorf, which is why I recently raised the question of when those courts would finally consider an implementer of a standard to be a willing licensee.

GE's and Dolby's bilateral offers disregarded

Only if and when the implementer is a willing licensee do German courts--under Sisvel v. Haier--take a closer look at the patentee's licensing offer(s).

Here, each plaintiff made a bilateral offer (i.e., a direct license that bypasses the pool) and pointed to the HEVC Advance pool license as another licensing offer. As a matter of policy, I actually believe courts should encourage that SEP holders make--wherever it is warranted--alternative offers. The more offers are on the table, the more likely it is that the implementer regards one of them as palatable. However, in Dusseldorf they have a different attitude: at least some of the judges there consider it necessary for every single one of a plurality of licensing offers to be FRAND. I believe reasonable people can disagree on this question, but in the Vestel cases the court decided not to analyze GE's and Dolby's bilateral licensing offers in detail: the court contented itself with identifying a FRAND breach just based on the HEVC Advance pool rate.

Should an appeals court (theoretically this could go up to the Federal Court of Justice) or another German court (if any cases were brought there) agree with me, it still wouldn't solve the problem for Access Advance. Its members might then win some cases (unless their bilateral royalty demands are insanely prohibitive, as may very well be the case here), but HEVC Advance would remain stigmatized regardless.

Dispositive part of FRAND analysis: HEVC Advance terms

I've seen cases (such as Microsoft v. Motorola) where pool rates served as important points of reference aiding judges as they were making a FRAND determination, but to the best of my knowledge, it's unprecedented for a court to find a patent pool's terms non-FRAND. It's a result that will live in infamy--to quote FDR. (For Vestel and its counsel, Gruenecker's Dr. Ulrich Blumenroeder, it's a remarkable success though.)

So why did it happen in this set of cases?

There's a problem of (a) double-dipping and (b) reneging on a prior FRAND licensing commitment. Vestel already had an MPEG LA HEVC license, and was therefore licensed to the portfolios owned by those former MPEG LA licensors who succumbed to Access Advance's lure of easy money. So, just as a rough estimate, Vestel was already licensed to about half of the patents in the HEVC Advance pool--and at a far more reasonable rate than the HEVC Advance rate, that is.

The court easily figured out what the HEVC Advance scheme is about. The pool charges what in my estimate is about four times the MPEG LA rate (relative to the patents in the pool). Some thought it was a simple way to make more money. But implementers who already have an MPEG LA license obviously take the position that they shouldn't have to pay again for patents they had licensed before--and especially not a far higher rate.

The court touch on multiple issues in its ruling, but needed to reach a definitive conclusion on only one in order to decide the cases in Vestel's favor. Once I've obtained a copy of the ruling, I'll go into more detail on that--and later this week or early next week I'll talk some more about why HEVC Advance is doing what it is doing. For now, suffice it to explain the following problem:

HEVC Advance refuses to lower its rate for those who have already licensed many of those patents through MPEG LA. Instead, HEVC Advance tells them to go to MPEG LA and seek a refund. What HEVC Advance should do is apply a credit. Maybe it would also be acceptable if HEVC Advance's licensors offered a refund on FRAND terms--though I have serious doubts about whether that would be the appropriate vehicle under any circumstances.

The key lesson from the Dusseldorf Vestel decisions is that you can't just join a first pool, make your patents available on one set of terms, and later join a second pool, ask for several times more money, but refuse to take into consideration that some implementers already got a less expensive license. It's a scheme to renege on a prior FRAND licensing commitment.

There are reasons for which HEVC Advance and its licensors don't treat licensees fairly. Those reasons have a lot to do with its owners' conflict of interests (with a few licensors owning and controlling the pool administrator), and with the deal structures they apparently offered to companies like Samsung and possibly also Microsoft (which just became an HEVC Advance licensor recently, so its patents were not at issue in the Dusseldorf Vestel cases) in order to get them to join. I've done some research on that, I'm still working on it, and I'll talk about this topic again in a matter of days. Stay tuned. The story must be told. While Access Advance will avoid the double-dipping problem through the terms of its new VVC pool (which will not allow licensors to participate in other pools at the same time), the root causes of the Dusseldorf disaster still pose a serious risk to the adoption of VVC.

Share with other professionals via LinkedIn: