Educate me on WebM vs. h.264

[I4AT]

I wasn't sure where to post this, but Mobile seems like the most relevant place.

So just now for shits and giggles I decided to disable the Flash plug-in in Firefox (10.0.2) and see if I could watch a video on Youtube.

That brought me to the HTML5 info page which basically explained to me that my browser doesn't support h.264 natively, only WebM. Until this moment I've been completely ignorant to the existence of WebM so I started reading up on it.

My understanding is h.264 is royalty free (at least for the time being, until 2015?) so if Mozilla (and Opera) wanted to support it right now they could, and then just drop support for it should they start charging licensing fees? So I'm guessing they must be consciously shunning h.264 for whatever reason, h.264 is open source though isn't it, just patent-encumbered?

So my question is, who exactly owns h.264 and stands to make money off of it in the future? Apple or Microsoft, and others?

I've also read that Microsoft has developed an h.264 plug-in for Firefox, why would they do this? Firefox is the number 3 browser and for a time was Internet Explorer's biggest competition, so the only reason I can see for Microsoft themselves to develop an h.264 plug-in for Firefox users is if they somehow stand to make money off of its eventual licensing? Or maybe they see Google Chrome/Android as the bigger threat so they want to keep WebM from gaining any ground?

Also, why would Google even purchase/develop WebM to begin with when h.264 already has so much of the market? What's their long term plan? h.264 currently is better in regards to quality as well as encoding/decoding performance and it's widely hardware accelerated, but obviously WebM has plenty of potential for improvement in all of those areas. It's already being hardware accelerated by some Broadcom chips, and Intel has announced they'll hardware accelerate it on their Atom platforms if it takes off. Although that's not really a big deal in my opinion considering GPU acceleration is much better for video, and currently AMD and Nvidia are needed as key supporters with their APU's and Tegra devices.

I'm always suspicious of Google because their business is data mining. I avoid using their search engine and I'll never touch their browser basically because I have a fear of them tracking every single thing I do, and I see a future of constant pop-up ads tailored to my online activity. A paranoid part of me wants to believe they're somehow trying to crush h.264 and monopolize the codec world with WebM so they can work their data mining/advertisement business into all streaming video.

But I am pro-open source/platform, so if I can be convinced their intentions are good I'll support the WebM movement. I dream of a day where OpenGL overtakes DirectX, and Linux and Mac OS both become just as viable as Windows as a gaming platform. I have a personal distaste for Adobe because everything they put out is bloated crap, and Flash I think is the worst thing to happen to the internet. So sick of annoying obtrusive Flash ads shooting my CPU utilization through the roof, and it took them way too long to enable GPU hardware acceleration for video. I'm kind of glad Apple killed Flash in the mobile sector, even though I'm not a fan of Apple either.

Lastly, I'm wondering how an eventual surge in WebM might affect us the end users in regards to captured video. Pretty much every camcorder/phone out there today captures in h.264, right? So being able to upload video without any re-encoding and the wide hardware accelerated support for h.264 by current mobile devices is a plus for the consumer. For WebM to succeed it'll take more than just browser support, but hardware manufacturers will have to start enabling native VP8 capture for camcorders/mobile phones won't they? But then again, with the ever changing nature of an open codec in its infant stage like VP8, isn't it impossible for manufacturers to consistently offer firmware updates every time the codec is altered? It seems like Google, Mozilla, and Opera are in a lose-only scenario.

 

[MrX8503]

I won't be able to answer all of your questions as I don't know everything about the subject, but Mozilla supports webm and theora. h.264 does in fact have some form of licensing fees that Mozilla has to pay. I don't know all of the people that make money of h.264, but Apple is one of them.

I prefer h.264 because its already a format that can be shot at, edited, and viewed easily in most video editing software/players. Both Safari and Google Chrome support h.264, I'm not sure what the roadmap is for mozilla when it comes to support.

 

[I4AT]

I believe h.264 licensing currently only applies to production level entities, and streaming video is royalty free until the end of 2015. And even if they were to start charging royalties then, I think there's a $5 million annual cap. Mozilla makes something like $70 million a year from Google alone in search engine referral royalties, so they can certainly afford it. They also fully support Flash which is proprietary closed software, so both the money and the principle of openness don't seem to be the answer.

I also think h.264 is the most convenient option for the general consumer, at least today, but apparently there are end user liability issues, where not only can software distributors be held accountable for not paying royalties, but also any person that uses their software becomes liable as well. That's not to say MPEG LA would actually go after anybody, but it does remind of the whole RIAA/Napster saga, and they did actually hit a few average joes with some ridiculous lawsuits to set an example.

Google announced they'd be dropping support for h.264 in Chrome a while back, but they've yet to actually do it.

 

[mavere]

H.264 was formally ratified in 2003, and the specs themselves were finalized some time before that. Therefore, all h.264 patents will expire in about ten years.

Considering how entrenched h.264 is in gadgets and mobile hardware, it doesn't make sense to have any significant push for technically inferior alternatives in the time we have before this discussion becomes pointless.

However, h.265 is getting finalized very soon, I believe. I'd love to see the efficiency improvements offered there.

 

[Mopetar]

I wasn't sure where to post this, but Mobile seems like the most relevant place.

So just now for shits and giggles I decided to disable the Flash plug-in in Firefox (10.0.2) and see if I could watch a video on Youtube.

That brought me to the HTML5 info page which basically explained to me that my browser doesn't support h.264 natively, only WebM. Until this moment I've been completely ignorant to the existence of WebM so I started reading up on it.

My understanding is h.264 is royalty free (at least for the time being, until 2015?) so if Mozilla (and Opera) wanted to support it right now they could, and then just drop support for it should they start charging licensing fees? So I'm guessing they must be consciously shunning h.264 for whatever reason, h.264 is open source though isn't it, just patent-encumbered?

Actually the MPEG-LA has already stated the h.264 would be royalty free forever for a certain uses.

h.264 is an open standard, such that anyone can create encoders and decoders for it, but doing so requires a license from MPEG-LA. There are certainly ways around that, such as a browser not including a decoder and relying on the underlying OS to supply one, which both Windows and OSX do.

So my question is, who exactly owns h.264 and stands to make money off of it in the future? Apple or Microsoft, and others?

Microsoft has a large number of patents in the h.264 standard, and Apple has a few, but there are several other companies that also have patents used in the standard.

I've also read that Microsoft has developed an h.264 plug-in for Firefox, why would they do this? Firefox is the number 3 browser and for a time was Internet Explorer's biggest competition, so the only reason I can see for Microsoft themselves to develop an h.264 plug-in for Firefox users is if they somehow stand to make money off of its eventual licensing? Or maybe they see Google Chrome/Android as the bigger threat so they want to keep WebM from gaining any ground?

If I had to guess it's because a large number of Windows users use Firefox. If they were unable to view h.264 content, they might blame it on Windows rather than Firefox. It's probably also to snub WebM and Google, but they want to provide a good user experience.

While Microsoft does have a number of patents in the h.264 standard, it's unlikely for them to make more money licensing those patents, when they have to pay out even more money in licensing fees for every copy of Windows that they sell.

Also, why would Google even purchase/develop WebM to begin with when h.264 already has so much of the market? What's their long term plan? h.264 currently is better in regards to quality as well as encoding/decoding performance and it's widely hardware accelerated, but obviously WebM has plenty of potential for improvement in all of those areas.

My best guess is that Google didn't want to take any chances that the MPEG-LA would stick them with a large bill down the road. Eventually, YouTube would move entirely to h.264 and you can imagine that with the amount of content they serve that the bill could be quite expensive.

Lastly, I'm wondering how an eventual surge in WebM might affect us the end users in regards to captured video. Pretty much every camcorder/phone out there today captures in h.264, right? So being able to upload video without any re-encoding and the wide hardware accelerated support for h.264 by current mobile devices is a plus for the consumer. For WebM to succeed it'll take more than just browser support, but hardware manufacturers will have to start enabling native VP8 capture for camcorders/mobile phones won't they?

The h.265 standard should be established in early 2013 and is targeted to provide better image quality at half the data rate as h.264. Even if WebM were as good as h.264, it would need to build up a lot of additional support in a short amount of time to even remain relevant. That probably won't happen. Google doesn't provide any kind of indemnification with WebM so if there are any submarine patents, companies that build WebM into their products could be incredibly vulnerable to lawsuits. The MPEG-LA itself could also move against WebM if they ever felt it got big enough as they're sure to hold a few patents that WebM would need to implement.

 

[I4AT]

I'm a bit lost as to what exactly Google is hoping to achieve then...

They should know even if by some miracle WebM is adopted as the HTML5 <video> standard Apple isn't suddenly gonna fold and accept it over h.264 or they would've done it with Flash.

And with the size of the Youtube database it would probably be cheaper for them to just pay the $5 million a year in licensing fees rather than have 2-3 versions of every single video uploaded when it could potentially be one.

 

[ViRGE]

While Microsoft does have a number of patents in the h.264 standard, it's unlikely for them to make more money licensing those patents, when they have to pay out even more money in licensing fees for every copy of Windows that they sell.

Agreed. The H.264 standardization process was very results driven. No one stands to make a great deal on this (although they'll certainly make back their research costs and then some), it's more about what kinds of new products they can sell as a result of there being a common, high quality video codec in the market.

Even if WebM were as good as h.264, it would need to build up a lot of additional support in a short amount of time to even remain relevant.

At this point none of the SoC GPU guys have announced support for WebM for their next-generation SoC GPUs, which would take us through 2013. Without hardware decode support WebM is going nowhere; the costs of decoding in software are much too great for mobile devices. It's an easier sale on desktop PCs, but even then there needs to be a compelling reason to deviate from status quo.

I'm a bit lost as to what exactly Google is hoping to achieve then...

They should know even if by some miracle WebM is adopted as the HTML5 <video> standard Apple isn't suddenly gonna fold and accept it over h.264 or they would've done it with Flash.

And with the size of the Youtube database it would probably be cheaper for them to just pay the $5 million a year in licensing fees rather than have 2-3 versions of every single video uploaded when it could potentially be one.

As Mopetar so thoughtfully points out, it's clear where Google was going to go, but it's not at all clear where they're trying to go now. If nothing else the success of Android has locked them to H.264 - could you imagine an Android phone without the ability to record and play H.264? They haven't even removed H.264 from Chrome, which should give you an idea of how reliant they actually are on it.

At this point in time there's no winning scenario I can think of for Google. They're basically playing to not lose, or at least to fight the good fight. In terms of devices sold H.264 will likely be the most successful MPEG video codec ever, so there's not much of an impact Google can have other than perhaps playing hardball and driving down their royalty payments.

 

[ChAoTiCpInOy]

The problem also with WebM is that it may infringe on patents held by MPEG-LA. I think they were collecting patents in preparation to sue WebM.

 

[Mopetar]

It's a bit of a coincidence, but Ars just ran an article about Mozilla planning to add functionality for h.264 decoding into Firefox and Boot2Gecko. Pretty much boils down to Google not going anywhere with WebM.

 

[I4AT]

I just read that Google acquired On2 for over $100 million in stock. I think at that price the best they could've hoped for even with a full WebM takeover is to break even. Maybe their goal all along was to simply bluff MPEG LA into a royalty free extension, and the $100+ million dollar WebM buyout was them absorbing the cost for all the little guys out there?

Link, I've been reading old news so though they previously announced the royalty free extension until the end of 2015, apparently they later extended it indefinitely.

 

[Mopetar]

I just read that Google acquired On2 for over $100 million in stock. I think at that price the best they could've hoped for even with a full WebM takeover is to break even. Maybe their goal all along was to simply bluff MPEG LA into a royalty free extension, and the $100+ million dollar WebM buyout was them absorbing the cost for all the little guys out there?

Link, I've been reading old news so though they previously announced the royalty free extension until the end of 2015, apparently they later extended it indefinitely.

I think that Google falls into a class of use that still requires a license for some of the things that they do. If the build a software decoder into their browser they'd need to pay for every copy distributed. Also, every Android device has h.264 hardware decoders, but Google doesn't pay that cost since they don't make the hardware.

I don't know exactly what all of that would cost them, but over a long enough period of time, the $100 million spent on a codec that they didn't need to license would have covered it. Having WebM take off would have also made Android cheaper for devices makers if they didn't need to include h.264 support in their SoCs. It would have also been a benefit for ChromeOS, although that never really took off either.