The AV2 Video Standard Has Released (Final v1.0 Specification)

https://av2.aomedia.org

by ksec • 2 months ago

365 183 2 months ago

av2.aomedia.org

jjcm2• 2 months ago

A few things - this is one step in a long, LONG path. AV2 is currently unusable in its current state (the encoder typically runs at around 1fps on good hardware), and likely will remain so til ~2028 when the first av2 hardware accelerated chips start dropping. Even then, I wouldn't expect AV2 streams to be common til 2030.

IMO, if it were just the efficiency gains on the table (which are substantial - ~20-30% over AV1), I'd say that AV2 isn't worth it. The biggest thing it does add though is multi-stream support, which will be a big win for VR and live sports. The other fun thing is you can send an alpha channel as a separate stream, which the file will then composite for proper transparent video support.

adgjlsfhk12• 2 months ago

Based on AV1's trajectory, hardware encode isn't necessary (though it is nice). The current encoder is a reference encoder. Now that the spec is finalized, expect significant speed improvements from production encoders (realtime likely won't happen until we get it in hardware though)

Gigachad11• 2 months ago

Hardware encode is required if you want things like video calls, camera recording and such to use it.

It isn’t required for content distribution platforms which aren’t realtime and the cost of encode is easily made up by hundreds of thousands of streams.

diegocg• 2 months ago

Anything running on a battery will need hardware acceleration

jjcm2• 2 months ago

I strongly disagree with it not being required. I run a small social news site - AV1 is still prohibatively expensive both for the server and clients for software encoding/decoding. Without hardware encoding, the tradeoff for better compression ratios in exchange for massive battery use + very long processing times for encoding simply isn't worth it. In order to get AV1 out, I have to often process a h264 version of a video first anyway, just so the client isn't left waiting for their video upload to finish encoding. This means to support AV1 I'm not saving anything on the storage side. Even youtube only does AV1 encodes for extremely popular videos - it only makes sense to do at significant scale.

I love AV1, don't get me wrong, and I can't wait til I can switch over to it as a single unified format for both images and video, but for now the cost is too high until hardware acceleration becomes ubiquitous

dgreensp• 2 months ago

Where do you see information about the efficiency gains over AV1?

tom_alexander• 2 months ago

It's at the end on the conclusion slide @24:40:

https://www.youtube.com/watch?v=Se8E_SUlU3w&t=1480

shmerl11• 2 months ago

That's fine and not anything new for codecs, they always take a long time before mass adoption.

Take a look at AV1 itself, you can't even say it's really ubiquitous on all hardware. It's quite well along in adoption compared to early days, but some mobile devices are still lacking hardware acceleration for it.

BillStrong2• 2 months ago

It is still ironic to me that my Steam Deck has decode AV1 acceleration, on a really old CPU/GPU combo.

d--b• 2 months ago

Does anyone know what is so costly to calculate in AV2?

ZeroGravitas5• 2 months ago

In general they just increase the numbers on everything when they go up a generation.

e.g. if you check in 4 directions to see if you can reuse a chunk then make it check in 8 or 16.

Faster encoders will have smart heuristics on when to use these new abilities and when to skip them but the reference encoder will try everything in a dumb way to eke out a tiny win to maximize a theoretical advantage and map out the extreme best case.

xbmcuser2• 2 months ago

The way things are going, we can pretty much forget about AV2 hardware encoders in PCs anytime soon. All the newest, best chip capacity is being completely hogged by Apple and AI companies.

Unless chipmakers port the AV2 design to older, cheaper nodes, it’s just not happening for average users. We’ll probably see some Chinese TV chip makers throw in an AV2 decoder just to check a box, but as an actual encoder? I wouldn't count on it anytime soon.

Gigachad11• 2 months ago

PCs don’t need hardware encoders. There are no realtime jobs on these. You can let it encode in however long it takes.

You need hardware encoders for things like cameras because they need to encode in real time since the buffer would quickly overflow otherwise.

kryptiskt• 2 months ago

I wouldn't be so pessimistic, Intel and AMD aren't going to stop making CPUs, and if their integrated graphics adds AV2 it will be motivation enough for others to follow.

miohtama• 2 months ago

In HW accelerated chips, what part of the calculations they usually accelerate? Could it be possible to repurpose old HW?

boriskourt• 2 months ago

One of the biggest gains of having dedicated hardware is that the computation doesn’t happen on the general hardware.

This is what makes it viable on mobile devices where system responsiveness and power efficiency are high priority.

Generally these hardware decoders haven’t been retoolalble.

BillStrong2• 2 months ago

Essentially all of the processing of the video data, barring the container format which the CPU uses to know what part of the data to send to the GPU or the Audio chip or codec.

And HW acceleration is generally a preset baked in version of the encoder or decoder. These are mostly codec specific.

So, no using hardware from previous versions.

Now, you can see some software that tries to use the GPU itself, instead of the dedicated hardware acceleration, to decode, but that isn't the HW accelerated, and may not operate in real time.

At the same time, that will consume much more power, eliminating some of the advantages or the pure HW rendition, especially important for mobile.

I could see an argument being made for encoding, if it is 2x or faster than the CPU, but I haven't looked at any in a while, so don't know the speeds.

cyberax3• 2 months ago

Typically things like quantization and motion estimation.

Bombthecat• 2 months ago

I feel like in 2030 it's more likely that we send 480p and just upscale with ai on the other end

hulitu2• 2 months ago

480p ? More like 200i. There's a race to the bottom driven by those "up to" codecs.

PunchyHamster• 2 months ago

The biggest thing it does add though is multi-stream support, which will be a big win for VR and live sports.

AV1 supports it too ?

ksec6• 2 months ago

........I'd say that AV2 isn't worth it.

Unless they have hardware encoder and decoder design done in parallel, otherwise it would be 2028 before a hardware block design is done and 2030 for the earliest product to ship with it. In reality I think 2031 or 2032 is more likely.

And I have been saying the same for quite some time that 20-30% for a generational codec improvement isn't worth it. I think they originally aimed at 50%, and then 40% and then 30%.

nubinetwork3• 2 months ago

enabling high-quality video delivery at significantly lower bitrates
likely will remain so til ~2028 when the first av2 hardware accelerated chips start dropping

This might sound dumb, but whats the point if its intended for slower devices, but those slower devices don't even exist yet?

rovr1384• 2 months ago

so that new devices can adopt it

They can't adopt it if it doesn't exist.

andersa• 2 months ago

It's not for slower devices, it's for lower data transfer bills for providers like YouTube.

mrbigbob• 2 months ago

i believe you would need new hardware to ENCODE AV2 video but DECODING AV2 video should work on current and a fair bit of older technology. The folks of Videolan are developing an AV2 decoder called "Dav2d" that should be able to play (decode) AV2 on modest hardware today.

https://www-test.videolan.org/projects/dav2d/

po1nt• 2 months ago

Looking at how GPU development has been sidetracked for NPU, I worry that this is 2035 target at best. Manufacturers will push for maximising matrix operation silicon area. In the era of trillion dollar investments into datacenters, traffic cost is afterthought. The only benefitors might be YouTube, Netflix and such, but on their scale investment into ISP level caches might be cheaper.

cyberax3• 2 months ago

The biggest thing it does add though is multi-stream support

This was supported in H.264 MVC but only saw real use for 3D movies on physical BluRays. With almost no content available outside that.

kalleboo• 2 months ago

The biggest thing it does add though is multi-stream support

I would have thought this would be a part of the container format rather than the video codec?

xmichael9093• 2 months ago

Google H264 SVC

2OEH8eoCRo0• 2 months ago

I feel like these encoders that require acceleration are a big reason why good hardware obsoletes so quickly.

mmastrac2• 2 months ago

Dav2d doesn't have the same nice ring to it. I hope there's someone with a decent repo-name punning skill who'll contribute before that.

avi2ude? av2go?

toast0• 2 months ago

2av2quit?

Dwedit2• 2 months ago

At least it's not D4vd.

zamadatix7• 2 months ago

I like it - not as punny as the first but pretty straightforward.

jbk• 2 months ago

It was difficult to find a nice name, with av2 :(

It works in French d2vid (Deuvid)

jgtrosh• 2 months ago

Also in French, different similarity, av2il hav2e

Slightly worse in English, av2age mav2ick

themafia• 2 months ago

e2ed

Retr0id• 2 months ago

daviid - or trim to davii and pronounce it "davey". But tbh I quite like dav2d.

shmerl11• 2 months ago

Congrats!

How is the case of fighting off Dolby's patent racketeering going? They tried to attack Snapchat for using AV1.

mmastrac2• 2 months ago

Last update seems to be "lawsuit was filed" with zero updates since then. That stuff tends to move slowly.

shmerl11• 2 months ago

Hopefully their patents will be busted and preferably Dolby will be also forced to pay damages for filing invalid lawsuits. That's the only way to teach patent trolls proper lessons.

basilgohar• 2 months ago

This will always happen. There are just some entities that can't stand not seeking rent.

ethin3• 2 months ago

It may always happen but it would happen less if we updated patent laws to fine people who filed invalid patents or enforced some kind of similar punishment. If you file a patent, it's up to you to verify that your patent is actually valid, and the courts shouldn't have to do that legwork for you. It also doesn't help that the patent office/components of governments don't review patents as thoroughly as they used to. Same with trademarks.

cyberax3• 2 months ago

Dolby's patents are garbage, and they know it. They managed to patent entropy coding somehow. This will not stand to challenges, and they managed to poke a well-resourced company.

Dwedit2• 2 months ago

What I'm interested in is seeing how this will improve the AVIF image format. AVIF stomps the competition for low-bitrate still images (where chroma subsampling is used). For lossless images, not so much. Lossless JPEG XL and lossless WEBP make lossless AVIF look like a joke.

wmf3• 2 months ago

Honestly AVIF2 is the last thing we need now. There are way too many minority image formats already.

Gigachad11• 2 months ago

There aren’t _that_ many. HEIF has been an unusually large pain in the ass just because it’s both patent encumbered and incredibly popular since the iPhone and many cameras use it.

JPEG is woefully outdated with the lack of HDR and modern compression, HEIF can’t be used without paying a license, webp was designed just for extremely efficient small images rather than local storage, avif I’ve never seen used ever, and JPEG XL is on track to be the next major format.

I agree we don’t need an avif2, but until jpeg xl there really weren’t any decent alternatives for jpeg.

fishgoesblub• 2 months ago

It's a silly idea to not improve on things for something as "too many" formats. There's too many ARM processors out there, should companies stop development and keep using old, less efficient CPUs?

ChadNauseam5• 2 months ago

AVIF is for sure my favorite image format right now. No other format has the quadfecta of lossless, HDR, transparency, browser support. Plus as you said, for very compressed images it looks amazing. It blows my mind how small AVIF files can be. Also, unlike HEIC and Ultra HDR JPEG, it actually supports HDR natively as part of the file format rather than doing the hacky sidecar gain map trick. I know it doesn't matter to everyone, but I just love HDR and AVIF is the only format that I feel like really takes it seriously.

ksec6• 2 months ago

but I just love HDR and AVIF is the only format that I feel like really takes it seriously.

JPEG XL would like a word.

erwan577• 2 months ago

Do you think advanced users should consider AVIF/AVIF2 along JPEGXL for long term pictures archival ?

Which kind of encode settings do you suggest for conversion from high resolution RAWs or JPEGs ?

BoingBoomTschak4• 2 months ago

I really don't.

1. Lossless AVIF is a joke often beaten by WebP and even PNG. Even worse for grayscale.

2. Chroma subsampling remains a bad idea for still images unless the resolution is high enough to hide the artifacts.

3. Tooling is the worst part, AV1 encoders are basically focused 99% on video and leave a measly 1% to image; unlike JXL, of course. SVT-AV1 still doesn't do YUV444 and libaom was unusable. Fortunately, the unpaid enthusiasts were here: https://giannirosato.com/blog/post/the-multimedia-renaissanc... (and more recently https://giannirosato.com/blog/post/oavif/)

I don't see AVIF being used for lossless, which is the largest reason I'd prefer JXL to win: one codec to rule them all sure is an alluring future.

theandrewbailey• 2 months ago

Maybe libavif will pick up AV2 support, so all one needs to do is update it to v2.0.0 or something and bam! Extra efficiency!

xmichael9093• 2 months ago

Not always https://tonisagrista.com/blog/2023/jpegxl-vs-avif/

ParadisoShlee• 2 months ago

Mostly a joke... I've been waiting for the AV1 Apple TV, so now I'm just waiting for AV2 support as Apple TV as well now.

londons_explore• 2 months ago

Outside the apple ecosystem, AV1 is supported nearly everywhere.

techpression• 2 months ago

They’ve had hardware decoding since M3 and equivalent A cpus. So I’d say it’s pretty well supported.

breve3• 2 months ago

My 10 year old iPhone 7 can play 1080p AV1 video in software for more than 200 minutes with VLC. The iPhone 7 was released a year and a half before AV1 was.

So I think it's a safe bet the current Apple TV devices are capable of playing AV1 video in software. There's a VLC release for Apple TV:

https://www.videolan.org/vlc/download-appletv.html

https://apps.apple.com/us/app/vlc-media-player/id650377962?p...

jshier• 2 months ago

Not especially relevant, as the obvious use of AV1 on the AppleTV is streaming, and the OS frameworks don't request AV1 without hardware decoding. Services which provide their own video decoding (are there any?) don't seem interested providing their own software decoder for the ATV, despite the bandwidth savings.

dwaite• 2 months ago

Apple A17 Pro / A18 include AV1 hardware decode.

tysonbru• 2 months ago

I’m curious how much AV2 will actually help older hardware in practice.

I’m on a 2019 Intel MacBook Pro: 2.6 GHz 6-core i7, 64 GB RAM. The machine is still more than powerful enough for normal desktop work and software dev, but YouTube in Chrome has become borderline unusable for me. My internet is fine, Safari plays the same videos smoothly, and YouTube “Stats for nerds” shows plenty of buffer but the decoding makes youtube unusable in chrome for me.

wmf3• 2 months ago

Unfortunately for you, newer codecs use more CPU than older ones so AV2 would probably be even worse.

alephnil• 2 months ago

Newer codecs has generally introduced more ways a video can be encoded, so that the encoder need to work much harder to encode a video, so that it actually achieve the gains that the newer codec allows much more processing will be required. Decoding on the other hand, will mostly stay the same or increase only slightly. It's not likely do decrease though, so if you struggle playing av1 today, you will also struggle with av2.

For encoding, you can always write a simple encoder that use only the features that were present in mpeg2, and it will be about as efficient as mpeg2 as well. Newer codecs has more features that allows more efficient encoding, at the cost of more processing.

Telaneo3• 2 months ago

Sound like a Chrome/Youtube problem. My 2012 Macbook Pro plays 1080p AV1 just fine in VLC (pretty sure Youtube works fine too in Firefox, but I didn't check whether or not it was AV1 or H264).

Telaneo3• 2 months ago

For reference: dav1d 0.5 can decode 143 FPS of a 1080p 8-bit video on a third gen core i7.[1] I doubt there's been much in the way of regressions since then. 10-bit and 4k is obviously a lot more heavy, but not really relevant to older devices.

[1] https://www.phoronix.com/news/dav1d-0.5 (it's mislabled as a core i3, but the 3770K is a core i7).

seam_carver• 2 months ago

use the enhanced h264-ify to block the av1 stream, av1 takes a lot of cpu

ZeroGravitas5• 2 months ago

I use Firefox but YouTube has recently started giving me a pop-up occasionally telling me that they are intentionally slowing down the site because they don't like some of the browser extensions I use.

drob5183• 2 months ago

I gave up using Chrome a decade ago. It’s a power sucking pig. Safari has its own issues, but at least it’s usable. When I need something that isn’t Safari, I use Firefox.

LunaSea• 2 months ago

Safari has a terrible developer experience and has been behind in implementing the various browser API for years, including AV1 support.

lern_too_spel• 2 months ago

Chrome has used less power than Safari for a few years now. As far as I know, the only good reason to use Safari is to use FairPlay DRM for sites that need it. For all other purposes, Firefox or Chrome is superior. https://birchtree.me/blog/everyone-says-chrome-devastates-ma...

kasabali3• 2 months ago

Has nothing to do with video codec.

Download the video with yt-dlp & play it in mpv you'll see it even flies on a potato.

Play the same in browser and it'll be dropping frames left and right.

ethin3• 2 months ago

And how long will it take before someone implements this standard and gets sued because Adobe or Dolby or whoever wanted to get slapped down? My knowledge may be out of date but if this is as "open" as AV1, I'm very skeptical that the individual companies will actually allow that. Greed and all that.

zamadatix7• 2 months ago

It took 7 years for the first patent assertion claim against AV1 to go to the courts and it will probably take a while for that case to resolve. Funnily enough, it wasn't from the pool constantly putting itself in the news about it over the years. I.e. it can take quite a while before attempts are made.

dataflow• 2 months ago

On that topic: https://arstechnica.com/gadgets/2026/03/av1s-open-royalty-fr...

buran77• 2 months ago

It took 7 years for the first patent assertion claim against AV1

This might just mean that if the claim is found valid, there's seven years' worth of inertia slowing down any effort to move on. Seven years in which HW and SW manufacturers worked to build in the support, and you the user developed your processes or workflows around assumptions specifically tied to that solution. I'd rather know on day one if I should go that way or not.

throawayonthe• 2 months ago

i'll be surprised if anything comes of it because like... https://aomedia.org/about/members/ if the legal departments of all these corporations haven't made them drop av1 in all these years there's probably not much there?

markvdb• 2 months ago

Patent trolls are nasty. How long will it take for one to get the full support of those shaking the independence of the US judiciary for their own gain? Let's hope the rot stops before that.

shmerl11• 2 months ago

No one can predict patent trolls - they can surface at any time randomly. But there should be a more organized fight against them. I assume AOM should be backing that.

maxloh• 2 months ago

It takes a few years for vendors to support hardware decoding for a new standard, so we won't see it in widespread use anytime soon.

Telaneo3• 2 months ago

Looking forward to a decently speedy encoder coming around. The reference one for AV1 is really not that great, and the same is true here. But as soon as we get SVT-AV2 or whatever, I'll be a very happy camper.

thinkingQueen• 2 months ago

AV1 is being actively claim-charted by a lot of companies right now, and lawsuits are almost certainly coming. The same process is already starting for AV2, but most players are waiting for the AV1 cases to mature first.

People keep calling the AV-family codecs “royalty free,” but in practice they increasingly look like a legal and financial gamble.

ZeroGravitas5• 2 months ago

People have been saying this for decades now.

I've never understood why some people seem to cheer this on like a corporation owning some maths was their local sports team.

For a while I assumed some people had put in a lot of effort on H.264 encoders and so the digital sharecroppers were angry and jealous that someone might be advocating for messy freedom.

But some people seem to just enjoy the thought of corporations putting a tax on video distribution.

Luckily those greedy corporations have repeatedly shot themselves on the foot and so their influence is waning.

Klaus232• 2 months ago

How long has it been since AV1 was released? About eight years, and there's still no credible patent holder. The vultures are always circling around compression standards. You shouldn't take that too seriously. Even if a lawsuit is filed, there's a legal defence fund to protect against baseless claims.

Klaus232• 2 months ago

Most importantly, there is no alternative. As kasabali said, the patent situation for the other codecs is a mess. Additionally, they could be hit by the same "No FRAND" problem. If someone comes forward with a patent that was used unintentionally, the situation is the same for all codecs.

kasabali3• 2 months ago

in practice they increasingly look like a legal and financial gamble

As opposed to what, like HEVC? Where you need to pay 3 different patent pools to be sure (which all has different terms), then there's still other patent holders that aren't in any pools and come and hit you with loyalty requests any time under terms however they like to?

throw0101a2• 2 months ago

People keep calling the AV-family codecs “royalty free,” but in practice they increasingly look like a legal and financial gamble.

And the alternative is… ?

For H.265 there are two HEVC licensing pools you have to sign with plus at least two non-pool companies:

* https://en.wikipedia.org/wiki/High_Efficiency_Video_Coding#P...

Going with a non-AVx codec is no less complicated and fraught with lawsuit risk AFAICT.

amelius• 2 months ago

It should be not possible to patent communication standards. The opportunity for abuse through lock-in effects is just too big.

shmerl11• 2 months ago

There should be a stronger push to fight these trolls. AOM has enough resources to bust their garbage patents and make them pay damages for patent racketeering. At least until software patents are banned for good.

sarah-robiin• 2 months ago

AV1 already was a big leap toward efficient and open video formats. I'm awaiting AV2 since a long time.

Sure it'll take a while since it's implemented in chips and hardware so we got efficient and fast hardware encoding/decoding.

But a ~25% higher efficiency sounds very promising in times of increasing storage prices and chip crises.

angelmanuel• 2 months ago

I'm not a expert on video encoding

But i wonder if the future could depend less on fixed-function compression methods and more on AI networks that recreate the video but weight much less that a compressed video.

Neural codecs such as github.com/Orange-OpenSource/Cool-Chic

jech3• 2 months ago

It will probably depend on whether NPUs are universally available in smartphones, and whether we get a standard API for accessing NPUs. But I don't know whether AI-based codecs can have battery usage competitive with fixed-function hardware.

hgoel• 2 months ago

Agreed, unless AI progress slows down enough that it becomes reasonable to bake weights into circuitry, the conventional approach will probably remain preferable for encoding, at least on power constrained devices.

perching_aix• 2 months ago

anyone performed a encoding and decoding benchmarking with the reference codec yet? I'd expect encoding to be dreadful, but maybe the decode is already usable

BoingBoomTschak4• 2 months ago

Now the real question: will The Industry™ again need for nerds and digital privateers to actually write and graft all the psy coding tools that makes their encoders usable (and I mean the word, using x264 as benchmark) for the quality-conscious and not just CPU-efficient VoD blurred to death with marketing yelling "PSNR! PSNR! PSNR!" at the top of their lungs? Will FGS be more usable?

Take a look at https://gitlab.com/AOMediaCodec/SVT-AV1/-/work_items/2269 for details (PS: SVT-AV1 claims to be suitable for AVIF yet doesn't support YUV444, lel)

abacadaba• 2 months ago

Is there an AV1 2.0? I'm not using this codec if they can't do basic semantic versioning right.

itvision• 2 months ago

Codecs are final (the only exception being H.265 with a couple of later additions/extensions that required new codec IP and were not that widely deployed or supported). They are not software. You cannot have upgrades for fixed function HW decoders (which is a must for pretty much any established codec and AV2 will be also one of them).

xmichael9093• 2 months ago

Neat but the real world still, and for decades to come - lives on h264 https://www.wink.co/documentation/Why-H264-Is-Almost-Always-...

the-grump• 2 months ago

Did you read what you shared?

It’s talking about security cameras, low-bitrate video.

FTA: "Myth Busted: At security camera bitrates (400-800 Kbps), H.265 provides negligible compression benefits. The marketing claims of "50% savings" apply only to high-bitrate content like 4K movies at 25+ Mbps, not security cameras."

IshKebab• 2 months ago

That's also bullshit anyway. Not surprising because the article is slop.

The AV2 Video Standard Has Released (Final v1.0 Specification)

Related Stories