AV1’s open, royalty-free promise in question as Dolby sues Snapchat over codec

Big Tech declaring AV1 royalty-free “doesn't mean that it is."

See full article...

 

[Gnafu the Great]

OK so I guess Dolby is going on The List but really now its a difficult list to still administer, now it's harder to find a vendor I'll still want to buy from.

Not that I've paid for anything media related for decades, but still.

I've never willingly bought anything from Dolby, but they've sure managed to weasel their way into all kinds of products I enjoy. Screw Dolby!

 

[Gnafu the Great]

Regarding "clean room": they didn't reverse engineer anything, so clean room practices don't apply. However, I can say the folks who made the AV1 standard worked closely with lawyers every step of the way to do everything they could to avoid infringing existing patents. They didn't just pool what they had and hope for the best: they intentionally avoided infringing other patents. That's what makes these suits so laughable to me. I watched the development of AV1 closely, and some of the biggest companies in the world probably spent just as much money on lawyers as they did engineers.

 

[Gnafu the Great]

To this point I wonder if AV2 is basically just waiting for the patent expiration of h.264 to happen around 2030 before they fully implement the spec and take advantage of some of those techniques to lower either bandwidth or processing power required.

AV2 is nearly done, set to be finalized Any Day Now™:

https://av2.aomedia.org/

https://gitlab.com/AOMediaCodec/avm/-/milestones/18 (I'll be very surprised if this isn't the last milestone before the final 1.0 spec, and it's 95% complete)

I'm guessing AV2 has incorporated techniques from expired H.264 patents, but the full portfolio may have to wait for AV3 (or the remaining patented techniques won't be useful by then).

 

[Gnafu the Great]

Sad to interject, but:

1st, Dolby has been churning out usefull things for audio since the 70s, video was an evident next step.

2nd, Dolby does things much better than the "more open" competition years in advance. Then the competition slowly catches up.

In Vision vs HDR10+ you are watching now in 2026 the level plaing field, but go back to 2014 and see where the alternatives stood. Then do the same excercise in 2016 (one year after HDR10 [snas the plus] was introduced) and see where was dolby vis a vis the alternatives (HDR10, Log Gamma et al)

Ditto for Atmos vs DTS:X vs MPEG-H 3D. Nowadays they are level, but at time of introduction ~10 years ago, Atmos was ahead in some key areas.

Yes, the may be dicks, but they are very competent dicks.

I've long been under the impression that the advancement of open standards is time and time again hindered by patent trolling and organizations like Dolby. I have a hard time believing we couldn't have had things like HDR10+ a lot sooner if we didn't have to work around patents on math. These advancements don't require this kind of rent seeking.

 

[Gnafu the Great]

Do we actually need Dolby (c) any more? We have DTS now. Nobody is interested in making Dolby surround music, despite a lot of people trying. Dolby hasn't been a leading supplier of anything for 30 years.

Dolby Vision and Dolby Atmos are big in UHD Blu-ray and higher end streaming. I agree that we have sufficient alternatives technologically, but Dolby has their marketing claws in a lot of pies. AOM is working on other formats like IAMF to produce open standards for more of these use cases.

 

[Gnafu the Great]

Considering the innovative work that Dolby has done over the years I would never consider them a patent troll. If the patents are infringed, then Dolby should be paid. However, this idea of charging a percentage of the total device value needs to be stopped. Patents should be licensed at a flat rate independent of the use.

In my comment, patent trolls and organizations like Dolby are separate but both problematic. I don't believe we need patents or patent licensing organizations for these types of advancements. AOM and other organizations have shown that there is plenty of interest in advancing the state of the art just because it helps in other places (in the case of AOM, it's reducing operational costs and increasing subscriber retention; but I'm sure there are examples of publicly-funded research that is beneficial to society in other ways and where patents wouldn't serve anyone but rent seekers).

 

[Gnafu the Great]

As a side note, it is also possible that someone may hold a newer patent on a technique that would be beneficial to encoding that isn't a normative part of the spec. It's possible a near-magic encoder like x264 uses techniques that are covered by patents newer than the H.264 specification itself such that decoding H.264 may be safe sooner than it would be safe to deploy a device that uses x264 for encoding. The same could apply to a specific implementation of AV1.

(I have no special knowledge; this is hypothetical.)

 

[Gnafu the Great]

I would like to see a lossless format with PAR file support, or some other kind of checksum/repair tech integrated.

I'm not sure, but there might be some of that in the work the CELLAR working group has done to standardize FFV1, FLAC, and Matroska. I at least know FLAC has checksums. I'm not sure if FFV1 or the Matroska work includes anything in that area.