Examining the HTML 5 Video Codec Debate

Ars Technica has a great breakdown of the codec debate for the HTML 5 video element. Support for the new video element seems to be split into two main camps, Ogg Theora and H.264, and the inability to find a solution has HTML 5 spec editor Ian Hickson throwing in the towel. "Hickson outlined the positions of each major browser vendor and explained how the present impasse will influence the HTML 5 standard. Apple and Google favor H.264 while Mozilla and Opera favor Ogg Theora. Google intends to ship its browser with support for both codecs, which means that Apple is the only vendor that will not be supporting Ogg. 'After an inordinate amount of discussions, both in public and privately, on the situation regarding codecs for and in HTML5, I have reluctantly come to the conclusion that there is no suitable codec that all vendors are willing to implement and ship,' Hickson wrote. 'I have therefore removed the two subsections in the HTML5 spec in which codecs would have been required, and have instead left the matter undefined.'"

Comments

It's a toughy

[91degrees]

Do we use an inferior standard or a closed standard?

Maybe "implementation dependent" is the term we're after.

Re:It's a toughy

[nine-times]

I think Microsoft has lost the media wars, and they pretty well know it. (admittedly, just a guess) Expect their products to support H264 and AAC. The bigger fly in their ointment is probably improved web standards in general. They've been gearing up to fight Adobe (Silverlight vs. Flash) for the proprietary "rich web" market, and if HTML/CSS gets rich enough that we don't need a proprietary plugin, that might not end up being a market worth winning.

Re:It's a toughy

[Draek]

Inferior standard. Judging from HTML4, by the time we could safely drop HTML5 support from our web browsers there'll be at least a dozen codecs that perform far, *far* better than H.264 does today so alleged superiority buys us very little, there'll still be a time where people interested in performance ignore the standard altogether. On the other hand, H.264's patent concerns will be with us for the next ~20 years, so Theora's advantage in ease of implementation will likely hold up for a much longer time.

Re:It's a toughy

[maxume]

It matters very little. If Microsoft and Apple fail to implement Theora, the fact that the standard calls for it will not matter (because it will not be practical as a universal fallback).

Mozilla can't license H.264 in a way that lets downstream packagers use it, so they don't want to put it in the standard either.

The previous /. story discussing the email Hickson sent out covered this stuff pretty well.

It isn't particularly hard to do things like put a flash fallback inside of a video tag, so people that want to use the standard but still have wide reach have lots of options (flash is the de facto way to play 'web' video today, so I don't think it is unreasonable to assume that this may continue).

Seriously? Lolcats?

[Anonymous Coward]

Ars Technica has a great breakdown

Oh, I totally agree. The best articles always insert two lolcats into their page so that we get a better idea of what's going on.

Did I miss something or is it still 2006?

why does the codec have to be in the spec?

[ibookdb]

and let the content providers decide.

Re:why does the codec have to be in the spec?

[Serious Callers Only]

The whole point of the element is to allow content providers to choose one of the always supported formats and therefore know a-priori that it will work in the user's browser. A "choose one from this list" strategy, or creating a new plugin-hell for codecs doesn't accomplish this end.

I disagree - the video element explicitly allows for several source files, so the whole point is not to allow only for one codec, or to mandate several codecs which are supported by everyone. That would have been nice, but hasn't been possible. As it is the video element is now being treated more like the image one - different browsers will support different image formats, but most will support a few core ones.

The whole point of the video element is to allow pages to easily embed video files (as opposed to the messy complicated method using object elements). The video element allows for several encodings in order, so the process of choosing a codec is transparent to the user, so long as you can give them something they can play, and is painless for the provider, given that there are free options for converting to ogg.

So it's quite possible right now, in theory at least, to serve video that every browser on every device can play (h.264/ogg/flash) - here's an example.

Life would be great if there was one clear unencumbered codec with no drawbacks, or at least a choice of a few (as there are for image formats), but there isn't one clear winner (ogg theora has definite disadvantages, the most important being lack of hardware support and quality issues). I think Apple should support Ogg, and see why Mozilla resist h.264 - there are strong arguments for both sides.

In the meantime the video element makes presenting video possible without a plugin with any sane browser (i.e. not IE), and is a step toward native browser support when people converge on a codec (or several) as they did with image formats.

Re:why does the codec have to be in the spec?

[Java Pimp]

Because people shouldn't have to be prompted to install codecs in order to view in-browser videos.

Yeah really! Every time I go to certain sites I get a prompt to install a new codec to see this hot video... And each time I do that my homepage changes to something not quite safe for work... I wouldn't mind so much if they'd quit messin with my homepage... sheesh!

Apple and Xiph

[_Hiro_]

It seems like Apple has something against implementing any Xiph codec... FLAC and Vorbis support in iTunes is nonexistent, and even with the QuickTime plugin, iTunes still doesn't have proper tagging support. And now refusing to add Theora support in Safari?

Perhaps someone on the Xiph board did something to one of Apple's Media guys when they were kids or something?

Re:Apple and Xiph

[Anonymous Coward]

Regardless of why they have some hatred for Xiph who cares what Apple's doing? Just specify Ogg. Apple will either lose market share as people switch to a browser that doesn't suck or they'll cave and use Ogg. If you can get 3 of them to agree I'd say that's pretty good. Are we just going to stop bothering to innovate because Apple won't give us its blessing? Let's just rename Apple to "Microsoft" and call it a day.

We (developers) are the ones that determine who wins the browser battles. We make the sites and we tell people what browser to use. FireFox didn't install itself on grandma's computer - that was us.

Re:Apple and Xiph

[99BottlesOfBeerInMyF]

Regardless of why they have some hatred for Xiph who cares what Apple's doing?

Ipod and iPhone owners care. Content providers looking to target iPod and iPhone owners care.

Apple will either lose market share as people switch to a browser that doesn't suck or they'll cave and use Ogg.

You're oversimplifying. This about more than just Web browsers. It is also about content services. When you don't have Google's Youtube on board with Ogg and you don't have iTunes on board with Ogg and it won't play on iPhones or iPods and you have little likelihood of that changing, specifying Ogg in the spec results in the spec not gaining widespread implementation and failing.

Are we just going to stop bothering to innovate because Apple won't give us its blessing?

Apple is one of the companies pushing HTML5 and already implements it in Safari. They aren't holding back progress so much as trying to push it in a different way than what Mozilla and Opera want.

We (developers) are the ones that determine who wins the browser battles.

I'd say the content providers have as much or more influence than browser developers. If the video element is implemented in a way content providers like iTunes and YouTube are not happy with, then it will be ignored by them and we''ll be stuck without any progress and a Web still locked into a fragmented mix and dominated by Flash video and Silverlight.

Re:Apple and Xiph

[truthsearch]

Apple uses open standards in their MobileMe / .Mac implementation. They also write standards-based server components, like CalDAV. Their platforms' preferred 3d library is OpenGL, another open standard.

Clearly they support many open standards, so it's not just about control over their customers.

Re:Apple and Xiph

[shutdown -p now]

You misunderstand the nature of HTML5 standardization process. Unlike previous HTML iterations, which were designed by W3C committee which largely did not intersect with people who actually implemented it, HTML5 is a vendor-driven effort that had only recently came under the aegis of W3C (after the latter's XHTML 2.0 died a quick and painless death). Since it's vendor-driven, it's going to be exactly what the vendors can agree upon - no more, and no less.

Re:Apple and Xiph

[Evanisincontrol]

We (developers) are the ones that determine who wins the browser battles. We make the sites and we tell people what browser to use.

Woah woah woah. That's a huge misconception that needs to be squashed right now: We, the content providers, do not tell the customer what browser to use; rather, the customer tells us what browser they're willing to use to view our content.

Why do you think so many "IE6 approved" sites still exist? Because those website's operators desperately want people to continue using IE6? No, they do it because a very large number of people are still using IE6 and are going to continue using IE6 regardless of what browser we mighty developers to try "force" others to use.

As someone else pointed out above, the problem with trying to hardball Apple into playing nice is that Apple will just sit and wait. When website developers go to create their sites and try to ensure cross-browser compatibility, their response to the problem will NOT be "Oh, Apple is just being douchebags. I'll just not bother supporting Safari until they support Theora." Instead, what they'll probably say is, "Hey, flash videos work in every browser. Why should I bother using this stupid VIDEO tag?"

Re:Apple and Xiph

[radtea]

The fact that it's open source or royalty free doesn't mean there are no patent trolls ready to file a lawsuit once Apple or Microsoft use it.

Likewise, simply because the MPEG LA controls the licensing of KNOWN patents for H.264 doesn't mean there are no patent trolls ready to file a lawsuit once it gets adopted as a standard.

There is also no assurance that the MPEG LA won't try to monetize their position as the sole licensing authority for H.264 if it were to be adopted into the standard. Unisys anyone?

So Apple's case would only be plausible if they can show that there is any reason to believe that the software-patent-related risk is higher for Theora than H.264, and they have not done that.

irrelevant

[Anonymous Coward]

"Apple is the only vendor that will not be supporting Ogg"

Except IE, which doesn't support, and has not announced plans to support, anything. Until they decide what they're going to do, it really doesn't matter what everyone else is doing.

Re:irrelevant

[nine-times]

Well it does matter, it's just that the matter is far from settled.

Honestly, I think it is possible to overestimate the power of Microsoft's vendor lock-in. If they don't get in gear and really compete in the browser market, it's only a matter of time before it bites them in the ass. They've already lost of decent chunk of the market to these other browsers.

If these browsers get to the point where they're all offering a clearly superior experience on the web, and Microsoft is still dragging their feet, they will eventually become irrelevant themselves.

Major browser vendors

[glwtta]

Apple and Google favor H.264 while Mozilla and Opera favor Ogg Theora.

Right, while convenient, that doesn't strike me as a very comprehensive list of "major browser vendors".

Re:Major browser vendors

[sam31415]

If you click through to Hickson's actual summary, you can see why Microsoft is being largely omitted from the discussion:

"Microsoft has not commented on their intent to support <video> at all."

There was a simple solution...

[Guspaz]

They could have simply specified that a browser must support ONE of the two options, h.264 or Theora. This would have at least provided a reference to websites, such that they can guarantee that they need support no more than two codecs. Without a standard, they can't necessarily guarantee that a browser will support either. A third party browser may come by and decide to implement nothing but MJPEG since it isn't specified.

I mean, there are legitimate concerns in both camps. Theora's hardware support is non-existent, and h.264 has expensive licensing fees. So why not allow browser manufactuerers to pick the one that best suits their position, rather than leaving it undefined entirely?

A guarantee of at least one of two being supported is better than no guarantee at all.

Re:There was a simple solution...

[samkass]

HTML doesn't specify what image format must be supported (PNG, GIF, JPG, etc); why is video any different? If HTML had specified GIF explicitly up-front, we'd all be in trouble when UniSys became dicks about it.

Let the market decide. If h.264 succeeds despite the extra cost, it means folks found enough value to justify the cost. If DivX or VC1 come out of nowhere to take over the web we won't be left with an out-dated standard. If a sleeper patent hits Theora hard we'll be glad we didn't lock ourselves down.

Hardware Encoders

[Nate53085]

The best reason I have seen so far as to why Apple/Google favor H.264 is because their current products have H.264 hardware encoders in them. Switching to ogg/theora would hit battery life hard in these devices since it would have to be done in software. While I agree that its a selfish reason, its a reason better then "cause we want it". I would really like to see Theora succeed though, an open standard for web would be a beautiful thing

Re:Hardware Encoders

[pembo13]

Hardware encoders/decoders would come pretty fast if Theora was made the HTML5 standard.

Why does it care?

[kindbud]

Really? Why does the HTML5 spec care what codecs are used? Why doesn't it just provide a way to specify which codec the author used to encode the media file, and let the browser prompt the user to get it if needed?

Re:Why does it care?

[shutdown -p now]

The fear is that the "good format" in this case will be H.264, and once it will stick and become de facto standard, we'll have the same mess as with GIF all over again - since FOSS browsers won't be able to support it legally (at least in U.S.), nor free content creation/editing tools.

XiphQT Components

[Anonymous Coward]

http://xiph.org/quicktime/

Adds support for Ogg Vorbis and Ogg Theora to QuickTime (which is used for nearly all media playback on OSX). Easy to install (but could be made easier easily - such as making into a .pkg), and makes Safari 4 work with <video> and Theora.

Also, can we please stop whining about this in relation to the HTML5 spec? HTML has never specified file formats for media/objects (<img>, <object>) and it should *not* start now.

MPEG-LA licensing is a fucking bitch

[Nicolas MONNET]

and a poster child against software patents. It's *very* expensive for small players, it's incompatible with free media, the terms are almost impossible to comprehend (or at least you need several "IP" lawyers on staff), plus you aren't even assured that you won't be sued in Texas by some scum sucking, syphillitic pus-drinking, rotting corpse-devouring and worm-infested defecation-eating patent troll.

Firefox will NEVER support H.264

[sadler121]

Unless there is a miracle and Software Patents are deemed illegal, Firefox will never support H.264. Being tri-licensed at least the GPL/LGPL would prevent Mozilla from licensing H.264.