Slashdot Mirror
AV1 is Well On Its Way To Becoming a Viable Alternative To Patented Video Codecs, Mozilla Says (mozilla.org)
Here's a surprising fact: It costs money to watch video online, even on free sites like YouTube. That's because about 4 in 5 videos on the web today rely on a patented technology called the H.264 video codec. From a report: It took years for companies to put this complex, global set of legal and business agreements in place, so H.264 web video works everywhere. Now, as the industry shifts to using more efficient video codecs, those businesses are picking and choosing which next-generation technologies they will support. The fragmentation in the market is raising concerns about whether our favorite web past-time, watching videos, will continue to be accessible and affordable to all.
Over the last decade, several companies started building viable alternatives to patented video codecs. Mozilla worked on the Daala Project, Google released VP9, and Cisco created Thor for low-complexity videoconferencing. All these efforts had the same goal: to create a next-generation video compression technology that would make sharing high-quality video over the internet faster, more reliable, and less expensive. In 2015, Mozilla, Google, Cisco, and others joined with Amazon and Netflix and hardware vendors AMD, ARM, Intel, and NVIDIA to form AOMedia. As AOMedia grew, efforts to create an open video format coalesced around a new codec: AV1. AV1 is based largely on Google's VP9 code and incorporates tools and technologies from Daala, Thor, and VP10.
Mozilla loves AV1 for two reasons: AV1 is royalty-free, so anyone can use it free of charge. Software companies can use it to build video streaming into their applications. Web developers can build their own video players for their sites. The second reason we love AV1 is that it delivers better compression technology than even high-efficiency codecs -- about 30% better, according to a Moscow State University study.
Over the last decade, several companies started building viable alternatives to patented video codecs. Mozilla worked on the Daala Project, Google released VP9, and Cisco created Thor for low-complexity videoconferencing. All these efforts had the same goal: to create a next-generation video compression technology that would make sharing high-quality video over the internet faster, more reliable, and less expensive. In 2015, Mozilla, Google, Cisco, and others joined with Amazon and Netflix and hardware vendors AMD, ARM, Intel, and NVIDIA to form AOMedia. As AOMedia grew, efforts to create an open video format coalesced around a new codec: AV1. AV1 is based largely on Google's VP9 code and incorporates tools and technologies from Daala, Thor, and VP10.
Mozilla loves AV1 for two reasons: AV1 is royalty-free, so anyone can use it free of charge. Software companies can use it to build video streaming into their applications. Web developers can build their own video players for their sites. The second reason we love AV1 is that it delivers better compression technology than even high-efficiency codecs -- about 30% better, according to a Moscow State University study.
At the outset, I just want to say how happy I am that AV1 has taken off, and how seriously it is viewed by so many technology companies as a way around H.264 and (even worse) HEVC. Particularly with respect to HEVC, there are three separate patent pools with different participants. HEVC is, in many ways, already set up to fail due to a large number of participants that participate in either none or one of the pools (see https://streaminglearningcente... for how chaotic it is). There are some other proprietary technologies such as Perseus that are out there that claim better performance than HEVC from a PSNR/SSIM perspective, but they will likely remain fringe.
What is of more concern to me is how carefully AV1 has been constructed in terms of its coding tools to avoid patent trolling and patent submarining (e.g. Rambus at JEDEC with DDR). This is a very serious and very technically complex issue, as any company could easily assert patents on AV1 if they feel there is infringement on their claims as pertains to any of the coding tools. There are increasingly limited ways of dealing with spatiotemporal entropy in non-infringing ways that do not involve exponential increases in gates or CPU cycles.
A recent and simple example of this is the MPEG-LA claiming they license patents related to the MPEG-DASH streaming framework. MPEG-DASH is, essentially, an XML schema for a streaming manifest combined with either MPEG-4 Part 12 (the MP4 container originally specified by Apple as the MOV format), or MPEG-2 Transport Streams encapsulating H.264 video. Nobody on the DASH Industry Forum really thought that MPEG-DASH would be subject to this type of activity, yet magically MPEG-LA began waiving it agreement around about two years ago.
As a result, many in the industry have held onto the virtually universally-supported HTTP Live Streaming, which is an M3U playlist with tag extensions and MPEG-2 Transport Stream container for the codecs. Even that standard developed by Apple has never become a fully ratified within the IETF, and nobody knows if the same thing will happen there either.
Incidentally, any time Google has presented VP8 or VP9 at previous conferences and is asked about patents, they avoid answering questions and the audience usually laughs. I've seen it personally, and I think it's the industry's cynicism for the various patent holders and some of their past actions. Where it becomes critical is for silicon suppliers, whose front-loaded costs are now in the neighborhood of nine figures to launch some SoCs, and for content distributors, who invest a tremendous amount of time and money encoding all of the required profiles for streaming to new codecs. Commitment to efficient hardware acceleration by them for the codec is risky, as they could easily be legally enjoined from selling their products if they didn't get their patent licenses in order, and this would also leave content holders scrambling to fall back to already-established codecs.
I will admit I'm cynical here too. While I'd love to see a patent-free open standard, I'm not optimistic that someone will not come out of the woodwork claiming infringement on a key coding tool. I wish Google and the rest of the AV1 participants luck. They'll need it.
Forgotten? No. VP8 lead to VP9, which is used by YouTube and Netflix. The work on what was to be VP10 was merged into AV1, which also includes contributions from Cisco's Thor and Mozilla's Daala.
VP8 hasn't been forgotten so much as its development has been continued.
The other issue is when will they start taking alpha channels (transparency) seriously. Been promised since vp8.
If they did, it could quickly become the standard in production and post production, which had to be Good thing right?
But they just don't seem to realise that. It's not (very) difficult.
Armchair lawyers are hilarious morons.
That was an interesting post by the founder of MPEG. He assumes that the rise of AOM will mean the end of video codec advancement because no on will be making money on codecs. He's completely and horribly wrong on that assumption. There is no longer a need to make money on the codec, the major content providers that provide video to the public have a massive incentive to continue to improve codecs because it literally costs them money. Google, Netflix, Apple, Facebook etc all save money if the Codec improves, and those savings can be multiples more than the licensing fees MPEG-LA collects.
The need for MPEG and MPEG-LA is over. HEVC should be a dead standard. The rise of AOM and AVC1 is a blessing to all of humanity. A free codec, developed and supported by the very people broadcasting and producing all the video. The very people with the largest incentive to continue to improve the codec because every byte saved saves them money.
MPEG and MPEG-LA should wander off into the night, they simply aren't needed anymore and have been destroyed by the same patents they sought to exploit.
What's really sad is that the patent pools are so packed with greed that they'd rather crap their pants and die an ignoble death than offer a better deal. They will not be missed.
Why not? There is a reason why MPEG 1 layer 3 still is used today, even 'though there are far better audio codex in existence.
An average consumer only needs 'good enough'. 'Exceptional' and a pain in the a** to use freely (beer) will always lose out. This applies to AV codex, UHD Blu-rays, etc. Content creators and distributors will only pay for licenses if the bandwidth they save, really save them that much more money. And the largest content distributors use their 'patent unencumbered' 'free' codex, whether patent holders like it or not. Also, don't forget, software patents are still not valid and/or not enforced in large parts of the world (including a number of very large modern economies). Then, what's left is hardware support. But generic processors and other re-programmable hardware nowadays is fast enough for most real-time video decoding. A truly dedicated ASIC style hardware decoder often is not needed at all.
So, you're left with consumers that either pirate - so they don't care about patent fees at all, use less efficient codex because they have bandwidth to spare anyway, use 'free (beer) patent unencumbered' codex because their main streaming site 'tells them to', use the patented codex because the patents don't apply in their jurisdiction...
MPEG-LA et al are dinosaurs and will die out soon. At least for AV de/encoding.
Can anyone explain why all the players in broadcast TV aren't pushing to have the next-generation broadcast TV standards (ATSC and DVB-T and stuff) using AV1 instead of H.265/HEVC as the replacement for MPEG2 and MPEG4/H.264?
Because the bitstream was codified in the last couple of months.
Maybe future standards will have it, but it was unavailable until 2018.
It's only just reached 1.0 and the encoding time of the codec is mind blowingly slow. It makes encoding HEVC look extremely fast.
I'm praying that AV1 takes off in a big big way, I like the idea of a superior codec, saving me disk space and being open source and free, my inner PC hippie is into that.
I don't know if it does every single feature HEVC or 264 does mind you, it might be crap at 10bit or 12bit or something, I just don't know, but my understanding is, it's fairly good.
None the less, it's not going to be replacing anything for several years. You need to wait multiple generations for smartphones, tablets, laptops, PCs, TVs and god knows what else to have new AV1 capable chips in them. Plus the encoder needs obvious, intense optimisation. Honestly the litmus test is when the piracy teams (or at least a few hardcore anime groups) start using the codec.
When I can replace some of my stuff on my NAS, with something at least 33% smaller and identical or better quality, I'm much more interested.
I do wish them well and I hope these hype articles continue, but patience will be a virtue here.
So, I provide a link that literally shows how to encode alpha channel in VP8. Which I personally use quite a bit for lower thirds overlays. And all you do is push your head a little further up your ass... I can see why this whole codec adoption issue is such an uphill battle.
Mozilla employs people from Xiph such as Chris Montgomery, Timothy Terriberry, Jean-Marc Valin, and Thomas Daede. I don't think paying the bills is laughable. Mozilla has funded development of Opus, Daala, and AV1.
If it helps, here's a recent blog post from Chris Montgomery on AV1's contstrainted directional enhancement filter.
As mentioned by others, it's not finalized and tested enough for the new ATSC 3 standard.
Also, remember that ATSC is the same organization that famously rejected the superior COFDM modulation when designing ATSC 1. ATSC has only now conceded that COFDM was the right choice all along.
That poor decision to use 8VSB instead of COFDM is why ATSC 1 suffers from multipath and it's also why multiple transmitters can't share frequencies. Sharing frequencies is something that the competing DVB-T standard has always done on its COFDM platform. As an aside, both Sirius and XM Satellite Radio have thousands of terrestrial repeaters that share the identical frequency nationwide because they use COFDM.
ATSC 1 is also why you can't view HDTV in a moving vehicle because of the Doppler Effect. A half-assed ATSC-M/H specification was hastily approved that almost nobody ended up using because mobile data ended up being the de-facto choice for mobile video consumers.
Kriston
So, I point out that ideas were thrown around and nothing become a standard, and it is not generally accepted and implemented throughout the VP codec family, and you repeat your link with the one method of encoding via the one codec that implements the hack and you think *I* have my head up my arse?
Perhaps you should actually read the link your own link points to, from 2012, then you may realize that this hack involved using a supplemental frame of content for Alpha, by substituting it for Y and zeroing U,V, and encoding in the same lossy way as the primary frame (and only in yuva420). I really hope you dont want any well defined alpha edges in your lower thirds..
Here is a good reference for how 'clean' this solution is:
https://lists.ffmpeg.org/pipermail/ffmpeg-user/2014-January/019283.html
I see in 2017 this VP8 hack was pushed in to an adobe export path, the discussion is enlightening as to how nicely official this has always been:
https://github.com/fnordware/AdobeWebM/issues/17
ffmpeg forced you to sample in yuva420p, son you are in subsampled hell, perhaps ok for final pass for the web, not useful for production editing.
and again, it uses the double-frame hack, without specific rate control for alpha - which is just broken. it took them years to even bother to support
the same hack in vp9..
AV1 is seems is playing a similar and foolish/broken game, trying to treat alpha channel as a separate image.
This inevitably leads to artifacts, and incompatible codec implementations.
To be clear, what I was is thought out, first class alpha support, in the official spec, that doesnt just exist as a hack though up in 2012 on a whim and half
implemented in some of the codecs.
Does that not make sense?
AVI supports alpha to about the same level of quality - actually slightly better, as RGBA is supported, it is just not standardised as to how the specify the A
doing in and out, so different applications handle it or not, randomly.
That is why people also dont use AVI for alpha content - quicktime is usually used, unless something else is forced, because it just works (tm) for codecs that claim they support it.
So, perhaps instead of being an insulting douche, you could perhaps consider that *proper* alpha support would be a good thing.
Once more open source proves to be effective for innovation. I'm surprised that Chiariglione still seams to believe that proprietary software is the only model for progress. I think reality has shown us, for quite a while now, that this is not the case. On the contrary. He seems stuck in the past.