BT

Google Explains Chrome Dropping H264

by Alex Blewitt on Jan 21, 2011 |

After last week's announcement that the Chrome team was dropping support for H264, Mike Jazayeri has posted a more detailed explanation of the rationale behind the decision.

The initial announcement generated highly polarised discussion, and the explanatory note achieved much the same. In it, Mike argued that it was necessary to make a stance for the future of the video tag by having an "open" codec in order to view videos in the browser. This stance is seen as positive by those who develop open-source software including the other browser teams behind Opera and Firefox, who only support webm.

Producers of content are not as happy with the decision, however. Outside of the desktop browser, H264 is widely used for video content. Since it is used in Blu-Ray, many devices have hardware accelerated decoding and even Intel's latest chips have processor-built functions to permit the CPU to assist where necessary. Not only that, but the smartphone market is saturated with devices that can decode H264; none have hardware assisted webm decoding.

Critics argue that Safari's browser share is a rounding error in the scope of computer browsers (currently, around 5%); but the real explosion in technology is happening in the mobile space, not the traditional desktop. Especially for handheld products, battery life is of key importance so video without hardware acceleration is likely to be a non-starter. For example, in Apple's most recent earnings report, iPads outnumbered Macs 2-1; iPhones/iPods outnumbered them 4-1. For each Safari browser on the desktop, there's six on mobile devices. Putting them into perspective, the 2010 global computer market is estimated at 350m in 2010; compare that with Blu-Ray devices (25m) and iPhone/iPod/iPad devices (100m). 25% of these devices have H264 support built in; and a majority of the 350m computers will be Windows with software or hardware H264 support.

The irony is that the Chrome team are arguing for "open", and since "open" is good, everyone must like it. However, critics point our that not only do Windows and OSX natively support H264 decoding as part of the OS, but Flash also decodes video with H264. Flash is famously not open, yet is still firmly supported by Chrome, as are other patented technologies such as MP3, AAC and even the humble GIF; yet these too are still supported.

Supporters argue that the switch will force content producers to encode video in webm as well as H264, though the comments suggest this is unlikely to happen. Regardless of Chrome's position, Firefox and Opera only support webm currently, and instead of dual encoding, content producers create a Flash player that's capable of decoding the same H264 stream that other browsers can handle natively. Since the technology is already in place, all that will happen is browser detection, which is used currently for other browsers, will be extended to Chrome as well.

Lastly, webm is seen as open and H264 is seen as closed. In fact, only two companies have ever developed webm - On2 and Google - and much of that was behind closed doors. Instead, the webm codec and specification was presented as a fait accompli, with the spec often delegating to the reference implementation in case of inconsistencies. H264, on the other hand, was developed by a consortium of companies who had a vested interest in getting a stable specification in order to implement in hardware. Although covered by patents, the H264 specification was arguably developed in a more open model than webm was.

Recently, the Free Software Foundation has come out in support of Google's webm proposal, arguing that only a non-patent encumbered video codec can be truly free; (although this view requires that the webm codec be not infringing unknowingly on existing patents).

Some reaction to Google's move has suggested that it represents a step back for standards on the Web, because H.264 is supported by more hardware and software. Those comments represent a fundamental misunderstanding of the vision of the Web as free and unencumbered. We can only be free if we reject data formats that are restricted by patents.

In summary, the decision to drop H264 from Chrome is unlikely to make any practical difference to content production of videos either for the web or for hardware devices. Content producers will continue to deliver H264 video and deliver natively or via Flash wrappers for those browsers that don't support it, as they already do today. Further, although Chrome is pandering towards open-source by claiming to be open, the reality is there are several other closed and patent encumbered parts of Chrome which are not being removed. Whether Chrome supports H264 in HTML5 becomes irrelevant, since all browsers support H264, either natively, or with a Flash wrapper.

Hello stranger!

You need to Register an InfoQ account or to post comments. But there's so much more behind being registered.

Get the most out of the InfoQ experience.

Tell us what you think

Allowed html: a,b,br,blockquote,i,li,pre,u,ul,p

Email me replies to any of my messages in this thread

Google owns YouTube by Leandro A

You're forgetting that Google defines the standard because they own YouTube. If YouTube serves only WebM, then everyone will have to support it.

Re: Google owns YouTube by Werner Schuster

The Webm format doesn't make sense.
All clients out there, from TVs, game consoles, to smartphones, to PCs can view h264 video, in special applications or in browsers (either natively or with Flash).
If YouTube were to switch to WebM today (or in one year or whenever), most of those clients would stop working, in particular the ones that are hard to upgrade... Andoid devices, for instance, many of which aren't upgraded to newer OS versions by the manufacturers (2.3 is the first Android with initial webm support) Not to mention all the hardware and entertainment systems that are not usually updated by users most of whom probably don't even know how.

That means: no matter what happens, YouTube will have to keep encoding all its content as h264... which means there's no incentive for anyone else to jump on the webm bandwagon. Sure, YouTube's big, but there are tons of video providers out there with tons of video; just look at the content provided by the BBC every day or any of the other content providers, from entertainment and news video to people hosting screencasts they've created.

None of those providers have any incentive to a) go and re-encode _all_ their back catalog as webm (CPU cycles and employee time to make it happen ain't free and re-encoding something for ideological reasons is hard to justify unless you're Google or the FSF) and b) don't care about webm enough to lose customers and viewers by switching to webm and dropping h264.

It follows: everyone (video providers, client platform vendors) will stick to h264. Clients will either view it natively (with special apps or inside the video element) as it is now or have to fall back to playing the video using Flash. Frankly, I'd rather use anything besides Flash to watch video, which'll probably mean staying clear of Chrome and Android (well, at least once Google decides to remove h264 from Android... which has neither happened nor been announced yet).
Encoding files in webm won't make sense unless you know you'll only be serving Chrome, Firefox or Opera users; or once Flash gets webm playback capabilities... at which point we're back to relying on closed source Flash. Yay......

WebM and Ogg-Theora might be inferior by Adam Nemeth

I don't believe openness is the only benefit for a user.

What we should understand is, that H.264 is the official successor of the MPEG technologies.

It was designed by an open scientific and engineering community, people who do know the most on the field, and - not accidentally - people who were paid by the largest companies related to the field.

Google wasn't them, nor was Theora.

Therefore, what I see here, is that an effort of the best minds of mankind at a given field is thrown out on a political / business decision.

While some people may argue, that some parts of H.264 costs money, many engineering hours were invested by really senior staff to make H.264 possible. Think of your own salary... somehow it has to be earned by the companies who employed them.

Re: Google owns YouTube by Kra Larivain

I don't think google believes it can pull this trick.
This move looks like a big FUD move against h264 and an attempt to corner apple on the mobile market.
Google owns youtube, but youtube is only the tip of the iceberg of digital video. The rest of the industry is h264 based, and the includes all the professional video solutions: hardware encoders, autohring tools etc.
Like you said, there are is too much h264 content and too many devices out there with no webm support, and it's not even a matter of software support since they don't have the webm chip.
Typically, any internet enabled tv, apple tvs, rokus and all those fancy devices.
Phones aren't much of an issue since people usually change them every couple of year. So 18 months from now, a significant portion of smartphones could be webm enabled.
H264 won the standards battle 4-5 years ago, by lack of competitor.

This is not true for iDevices and preventing them from seeing youtube videos could be a major annoyance for them. Apple could provide software, but so long for battery life, it's not any better than flash support.
As for revenue, i don't believe ios youtube has any advertisement, so it might not be that big of an issue for google if they loose ios customers.
Which brings me to another point, i'm not sure google has that much interest in html5 video, as the technology does not allow for any advertisment at this point. Google mentionned embedding the ads in the video on the fly, but that would require a lot of development effort.
And i'm not sure mp4 container is flexible enough to let this happen. Matroska has better multi track support and might be more suitable for this kind of tricks.
Not even to mention the fact that no browser can play html5 videos in full screen. As of today, flash remains the best technology for google's requirement.

Where I Google is double faced, is when they're saying that they'rer doing this for openness, standards and to workaround patents. It's more likely to be about money and contrrol, openness will just be a side effect, and might not be at all if google controls the format (and they most likely will, it's too dangous for them to let the community control something as critical as that).
Webm is a spinoff h264, only not as good, so most likely covered by some patents, there is no spec, only poorly commented code, coder and decoder share the same code base, hiding potential bugs in the implementation.
Time will tell!

My two cents by Nikolay Tsankov

I find it hard to grasp how patent-encumbered, royalty-payment-requiring codec is better than a free one (for anyone else but the people who receive those royalties).
The only argument is that everyone already uses H.264, therefore we should continue to use it...
Everyone used steam engines once, they were built by the best engineers and scientists of the time, but then we found something better and we moved on.

Allowed html: a,b,br,blockquote,i,li,pre,u,ul,p

Email me replies to any of my messages in this thread

Allowed html: a,b,br,blockquote,i,li,pre,u,ul,p

Email me replies to any of my messages in this thread

5 Discuss

Educational Content

General Feedback
Bugs
Advertising
Editorial
InfoQ.com and all content copyright © 2006-2013 C4Media Inc. InfoQ.com hosted at Contegix, the best ISP we've ever worked with.
Privacy policy
BT