Google Announces New WebM Video Standard, Open-Sources VP8

Yesterday Google announced WebM, a new web media project that combines the VP8 codec (which Google acquired in its purchase of On2) with the Vorbis audio codec and parts of the Matroska multimedia container. As part of this project Google has open-sourced the VP8 codec, which could have far-reaching consequences for the future of video in HTML5.

If you’re late to the HTML video game, here’s the play-by-play: the upcoming HTML5 specifications include a <video> tag, which is intended to provide integrated video playback in web pages without the need for a third-party plugin (the browser would provide the necessary video decoding and playback controls). However, the HTML5 spec doesn’t specify a particular codec to be used for the video content. At present, the state of browser support is varied: Google Chrome and Safari both support the proprietary H.264 codec, with both browsers’ parent companies paying licensing fees to include the codec. Chrome, Opera, and Firefox all support the Ogg Theora codec, a free and open source alternative that many say provides inadequate quality. IE currently supports neither, but has confirmed that the upcoming version 9 will support H.264.

Up until recently, this was a standoff: Apple and Microsoft won’t use Theora because they say it lacks the performance of H.264, and Opera and Mozilla refuse to use H.264 because it’s patent-encumbered and they want to avoid paying licensing fees.

Enter Google. In August of 2009 it acquired On2, the company that owned the up-and-coming VP8 codec, which was rumored to offer comparable performance to H.264. At the time there were murmurs that Google planned to open source the codec, thus potentially putting an end to the “video wars” once and for all. Well, that’s exactly what Google has done.

The open sourcing of the VP8 codec comes as part of Google’s launch of the new Open Web Media project it’s calling WebM. The details can be found on the new project’s first blog post.

The new WebM project has already received support from Mozilla (who has announced that it will be supported along with Theora for HTML5 video in Firefox 4), Opera (who has already released a “Labs” build of its browser with WebM video support), and Adobe (who has confirmed that the VP8 codec will be included in an upcoming release of the Flash player). Microsoft has taken a more nuanced approach, stating that, “In its HTML5 support, IE9 will support playback of H.264 video as well as VP8 video when the user has installed a VP8 codec on Windows.”

Of course, it goes without saying that Google Chrome will be supporting the new format in the near future. A post on the Chromium blog states that preliminary support for WebM is already available in development versions of the browser.

Of course, there’s still one major player missing from that list: Apple. The Guardian Technology Blog reports on an email reply from Steve Jobs to a reader asking about Apple’s position on the new video codec. Jobs simply links to a forum thread on the X.264 developers’ forum, pointing out current technical weaknesses of the codec as a specification. Of course, WebM as it currently stands is labeled as a “developer preview,” so of course those issues could potentially be resolved by the time an official release rolls around.

Moreover, if Google proceeds with converting the existing YouTube library of clips to VP8 and making them available in an HTML5, Apple’s position might prove to be untenable, and VP8 support could find its way into Safari. In fact, Google has already stated, on the WebM FAQ, that “all videos that are 720p or larger uploaded to YouTube after May 19th will be encoded in WebM as part of its HTML5 experiment.” This appears to mean that if you’re viewing the experimental HTML5 YouTube pages using a nightly build of Firefox or Opera, you can already view WebM-encoded video in the wild.

This is certainly an exciting development, and one which appears to bring the dream of open video on the Web one step closer to reality. What’s your take on what this holds for the future of web video?

Free book: Jump Start HTML5 Basics

Grab a free copy of one our latest ebooks! Packed with hints and tips on HTML5's most powerful new features.

  • WTFrank

    Great news for one format, but somehow I doubt that this is the end of the discussion.

  • didgy58

    all i have to say is great news!!!

  • http://www.optimalworks.net/ Craig Buckler

    Well done Google! We may finally have a codec we can depend on.

    As for Apple — who cares what they decide to do. Safari has fewer users than IE6 and iPhone/iPod/iPad users are used to missing video content! Also, how could they claim to be supporting HTML5 if every other vendor supports WebM but they don’t?

    • Louis Simoneau

      Yep. That sounds about right :-)

    • TheEmperor

      Apple? What about Microsoft? As usual they are the most likely player to be dragging their feet’s and slowing the hole process.

      You are among the many who got (partially) fooled by MS initial announcement of “support” I suspect.
      Being able to get WebM-support by installing *extra* software is of very very little use, and it’s of less use than being able to install Chrome Frame and get a complete modern

      The reason you got my reaction is because you wrote like Apple was the only bad guy not gonna play along. (My apology if I completely misread you) But Safari already, this very day, got the same kind of WebM-support that MS has gotten positive PR for the last day after they announced that very same kind of support coming to IE some time in the future.

      With Safari you can already install extra QuickTime-filters and get 3d-party codecs supported in Safari. That IE9 is gonna support 3d-party codecs (actually MS say they are gonna block everything but H264 and WebM) if you install extra filters for DirectShow is just the same thing coming much later.
      Definitely no reason to present Apple as any better in this matter than Microsoft.

      Yes, Apple might be “bad”, but MS is most likely gonna be worse, and most definitely more important.

  • http://www.davidclick.com Zygoma

    Well reserached, well written article. Video content on the web is going to be big lets hope all the big movers will make it easy to publish without the proverbial cross browser headaches :-)

  • http://inspirationsunlimited.co.in radiant_luv

    Well, Correct me if my understanding is wrong, is that mean I do not have to rely on a FLV player anymore to integrate a video on the web?

    • http://www.optimalworks.net/ Craig Buckler

      That’s correct. But don’t remove it just yet – not every browser supports the video tag and those that do won’t support WebM yet.

  • boltronics

    Finally! Video playback coming natively to a browser near me!

    I think it’s inevitable that only truly free and open general-use formats will rule in the end… it just takes time. The Word .doc format, .mp3 format, StuffIt, are all examples of non-free formats that have died or are on the way out. Web video is just another example of this occurring.

    I’m so sick of putting up with Flash video playback controls that don’t work, I always download video and play locally at any chance I can get. I personally hope this is the last nail in Flash’s coffin.

    I also agree with Craig – Safari can go and do whatever it wants. IE6 is only still around because of corporate desktops that are so locked down that the end user has no choice in what they use. I’ve never seen an office of Macs that were locked down like that. As such, end users on Apple’s platform will just replace Safari with something else if Apple doesn’t play ball.

    • arts-multimedia

      I do not know where you get that idea from that .doc and mp3 are on their way out. I don’t see any evidence of that.
      You have to realize that open source and commercial software industry need each other. It’s not one way or the other.

      • boltronics

        Last year, Microsoft introduced support for the OpenDocument format, which is the primary format used by OpenOffice.org. If you are familiar with Microsoft’s history regarding open source software and open formats, this speaks for itself in volumes. I also sincerely doubt MS would ever have tried to make Office Open XML if it wasn’t so scared of the competition.

        Ogg Vorbis is becoming increasingly common due to licensing issues. Video games use it, a ton of portable music players support it, and why not? It costs nothing to throw in as a feature and is as close to zero risk as you can get. For lossless high quality audio there’s also flac. If there’s a flac option and you’re not on dialup, who in their right mind would use mp3 over flac? I personally rip my entire audio collection to flac, and transcode it to Vorbis when flac isn’t supported by something.

        The commercial software industry contributes more code and developers to the Linux kernel (for example) than spare-time hackers. Perhaps you mean to say that proprietary software complements free software? Even so, I seldom find that to be the case. It’s always the opposite – FOSS complementing proprietary software.

        The supporters of free software that I know are able to run a computer 100% on free software, including BIOS and drivers, quite successfully (it’s actually not difficult if you choose your hardware carefully). I don’t know anyone who uses a 100% proprietary operating system. You don’t think MS has never incorporated BSD code into Windows, do you? They’ve been doing it for a long time.
        http://www.kuro5hin.org/story/2001/6/19/05641/7357

  • W2ttsy

    @Craig Also, how could they claim to be supporting HTML5 if every other vendor supports WebM but they don’t?

    Wait a sec dude. The article purports that Google invented this WebM spec and now not supporting it means you’re no longer supporting a completely different spec? Unless the official codec of HTML5 is going to be WebM, a browser can still support HTML5 video playback using a codec supported by your browser (and in Apples case, thats x.264 for the moment).

    Little less Apple bashing and lot more accurate comments would be good.

    • http://www.optimalworks.net/ Craig Buckler

      Good point. But why is Apple the only vendor hesitating? The HTML5 spec may never specify a codec.

      Apple will do whatever they want but, in this case, I really don’t think it matters. iPhone users can’t view many videos now and they still won’t be able to if Apple decide not to implement WebM.

  • http://fvsch.com Florent V.

    Great news, but:
    - The hardware acceleration part will take time, may never happen. A good thing is that they have several chip makers on board, but will they deliver VP8 support in their products?
    - According to the author of the x264 encoder, the VP8 spec is not a real spec that you can use to build an implementation. If you want to build a different implementation, you basically have to reverse-engineer the “spec” and the open-source implementation; not impossible as the reference implementation is open-source, but still. Oh, and he says the quality is not as good as H264, but since it’s not a big gap like H264 v. Theora it could be ok.
    - MPEG LA (the company managing the patent pool for H264) says it will probably create a patent pool for VP8. Which means that companies with video compression patents which they think apply to VP8 would get together and setup a license and license fees, just like H264. It could be some bluff or FUD, but that’s far from certain. http://i5.be/aPZ

    This could end badly. It could be that, barring a ban on software patents (yes, please!), there will not be a non-patent-encumbered codec in the foreseeable future.

    • Louis Simoneau

      A number of good points there.

      Re: hardware acceleration: yes, but non-hardware-accelerated native video decoding will still use less CPU than hardware-accelerated video in a flash wrapper ;-)

      Re: the spec, that’s valid in terms of his criticism of it as a spec, I definitely think there’s room for cleaning up the code and improving documentation, but again this is a developer preview. And his criticism of the quality has to be taken with a grain of salt, as there are a number of comparisons out there that show comparable performance, and he has a considerable stake in the 264 platform.

      Re: patent pools: I can’t help but think that Google would have done its due dilligence before making a move like this, so we can hope they’ll be willing to fight the patent trolls. Of course, only time will tell.

  • Wolf_22

    I think in the end, this will prove to be a great step!

    Standardization seems to be the primary necessity in everything “web” and while it may cripple a few things along the way, I fail to believe that the gain will fall short of its intended benefits. These things move in baby steps but in my eyes, this move on Google’s part seems to be a HUGE stride. I think they did a good thing here.

    I can only imagine the amount of politics involved with all this…

  • hairybob

    In the end, I think standardization will do nothing more than concentrate power in the hands of the large media companies. Rather than opening up opportunities, html5 will facilitate the likes of News wrestle back the control that had been handed out to the unwashed masses over the past 10 years.

    • Wolf_22

      “Power” is a somewhat relative and or subjective way to describe any business entity. It can mean different things to different people on different days! Woa… Talk about factorial perspectives, right?

      Assuming you use it to indicate “all of the above,” one could argue that “large media companies” will always be powerful, regardless of general standardization. Just look at Microsoft and Internet Explorer (given, of course, we’re talking about the browser market)…

      [QUOTE]Rather than opening up opportunities, html5 will facilitate the likes of News wrestle back the control that had been handed out to the unwashed masses over the past 10 years.[/QUOTE]

      What do you mean by the above remark? To me, that sentence makes no sense.

  • http://www.2webvideo.com/index.html orebaba

    Many things are still cloudy. But you have explained well, Louis. Thanks.
    Since a month or so, the video quality displayed in YouTube has changed for better. Prior to this the quality of video in other video sharing sites was much better. I, for one, have stopped embedding videos from Dailymotion, Metacafe, etc., instead using YouTube videos only for my blog.

  • arts-multimedia

    Interesting development. I do not know yet how this is going to support subtitles and closed captions but it is definitely a good development.

    For the Flash-haters among you: Flash can do a lot more then showing videos, so your “enemy” is certainly not going to die out because of this. Blame the dumbos doing stupid things with Flash, instead of cursing the application itself.

    Sad news for Apple and MS who would have loved to dominate the market with their stuff. H246 is nice for lots of action in a video but talking heads show quite a few artifacts, which proves that there is no perfect codec and Steve Jobs would do well to remember that before critisizing others.
    Sorry for the Apple bashing, I do have a mac as well as a pc, but mister jobs has been a pain in the neck of the internet community for quite some time now, so he deserves a slapping.

  • arts-multimedia

    An interesting article from Jeroen Wijering on the subject:
    http://www.longtailvideo.com/support/blog/11887/html5-video-not-quite-there-yet#comment-76513
    Jeroen Wijering is the author of the video player for YouTube. His JW player is the most versatile player around, based on Flash. He too is working on a html5 version and in doing so, he discovered quite a few flaws.

  • SICK

    i hope it wont take too much time for the page to load.