Paradise Road: Ken’s Blog

It seems as though most of the initial furor surrounding Ogg Theora has died down quite a bit since the release of Firefox 3.5, which includes Theora support along with a heaping spoonful of political grandstanding in Theora’s favor, out of the box.  Some things have helped to derail the Theora momentum.  Microsoft has announced Silverlight 3 which will play superior-competing-standard H.264 out of the box, Adobe has open sourced the media framework surrounding its Flash player, and, most significantly, Apple has submitted a new standard for MPEG4/H.264 HTTP Live Streaming to the W3C which will provide huge benefits to video viewers, especially on a mobile network.  Lastly, the author and maintainer of the HTML5 spec that defines the <video> tag itself has capitulated and said that there will be no official codec in HTML5 for video for the foreseeable future.

So why does this interest me?  I’m a pretty big advocate of open source software, after all.  And I agree with the two biggest points that open source folks are using to triumph Theora’s benefits (that it’s “good enough” to be used at Youtube quality on the Internet, and that it’s patent- and royalty-free).  But I’m also a fan of reality, and the reality is that Ogg Theora isn’t better than H.264.  Being better, not being good enough, is what ultimately drives adoption.  Internet Explorer began losing ground to Firefox for the fact that Firefox was demonstrably better.  There was a benefit for people to use Firefox instead.  Not so with Theora.

What I don’t understand, and what irks me so badly, is why H.264 is demonized so badly by the FLOSS community.  It’s not a single-vendor solution trying to lock in a particular market to a particular player.  It’s a very good video codec with massive market adoption and widespread implementations created by a collaboration of some 250 academic institutions that have invested a lot of money into it.  And yet, FLOSS advocates, particularly those from Mozilla, are saying that H.264 is extremely expensive, that it locks away innovation, and that it puts control of the web into the hands of a minority.  The expense argument is what irritates me the most.  I won’t use the argument that it’s free until 2011, but I will use the argument that MPEG-LA hasn’t even decided on the pricing for the licensing.  Nobody knows how expensive it will be, or who will have to pay the cost.  MPEG-LA has a lot of big players in the form of Apple and Google who stand to gain from HTML5’s acceptance likely lobbying them at this very moment to make licensing reasonable for H.264, lest we delve into another GIF fiasco.  But even where things are currently headed, I think the cost to stream video is free up until something like 100,000 views, which is pretty generous in most cases.  And the expense of an encoder for the bleeding hearts and the artists?  It’s like half the price of that PS3 game they went out and bought last weekend.  Let’s talk about innovation.  Well, I’ll let the web explain that one for me, which has also been proposed to the W3C for inclusion in HTML5.  Come on people, get with the program.

I won’t really say anything that hasn’t been said about the risk of patents on Theora, other than to note that while yes, H.264 does suffer from a degree of patent risk, it is much more complete and better documented, from my understanding, making it easier to find such risks.

Mozilla is fighting a massive uphill battle by trying to convince users that H.264 carries with great licensing burden when the codec already ships for free with most operating systems or is freely downloadable in the form of Flash or Silverlight.  And fight they are: Mozilla is engaging in political grandstanding to get its own dark horse to win the race by attempting to force Ogg into the HTML5 spec instead of letting the market decide.  This serves to create an unstable, unwanted variable in getting everybody on board behind a common set of codecs.  What’s more, we’re dealing with hardware.  A hardware industry that for a decade now has been getting behind H.264 as the next-generation video standard. There’s no incentive for hardware makers to build support for a 10 year old, obsolete codec onto their chips when they can just pay the minimal per-decoder fee to have H.264 installed.  The result of which will not be a win for Mozilla, open source, or HTML5, but for Flash, Quicktime, and Silverlight.

It’s frustrating.  Somehow I think even Mozilla realizes this, stating that they’d have to “make compromises” should the web adopt H.264 instead of its dark horse, which is probably closer than anyone realizes considering that Google is playing with using H.264 videos (that they’ve already invested in) in an HTML5 video implementation.

What I think will actually happen is that there will not be a base codec specified for HTML5, nor should there be… after all, we didn’t need base image formats and we don’t need them for video.  The market will decide which one is actually better–which I believe will be H.264–and the world will move on.  I’m sure that Theora and H.264 will coexist for different purposes just as JPEG and PNG do.  But Mozilla will eventually be forced to capitulate and either use some of that Google money to license the decoder or build in backend support to play what’s allowed on the host platform.  Everybody else is already on board with H.264, folks just have to wait out Mozilla’s grandstanding.

This entry was posted on Saturday, July 25th, 2009 at 9:26 pm and is filed under Uncategorized. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.