The Mozilla Foundation is putting its weight behind an effort to create an open video format on the Web. It is doing this by giving $100,000 in grant money, to be administered by the Wikimedia Foundation, towards the development and support of Theora, an open-source video codec. More importantly, it is also building support into the Firefox Web browser for both Theora and Vorbis, an open source audio codec.
Many other video codecs and encoders require licensing fees or come with restrictions. Mozilla hopes to change this over time. Although I suspect the Theora video codec is inferior to other technologies, as long as it can improve over time, it could eventually become a serious contender to MPEG-4 or Windows Media Video (WMV).
Evangelist Christopher Blizzard explains why Mozilla is backing open video in a long and windy post:
Although videos are available on the web via sites like youtube, they don’t share the same democratized characteristics that have made the web vibrant and distributed. And it shows. That centralization has created some interesting problems that have symptoms like censorship via abuse of the DMCA and an overly-concentrated audience on a few sites that have the resources and technology to host video. I believe that problems like the ones we see with youtube are a symptom of the larger problem of the lack of decentralization and competition in video technology – very different than where the rest of the web is today.In my mind there are two things that help drive that kind of decentralization:
* You should be able to easily understand how something moves from a computer-readable format to something that is presented to a user. For example, turning HTML into a document, turning a JPEG file into a picture on the screen or using HTTP to download a file.
* You must be able to implement and deliver that technology without requiring anyone’s permission or license. In reality this means that it should be available on a royalty-free basis and without encumbered documentation.In the video world, there are some formats that fit the first quality: Some formats are documented, understood and even widely deployed. But more often than not they are subject to to per-unit royalties, large up-front fees and creating content in those formats (the encoders) are often so expensive as to be prohibitive to all but only the deepest-pocketed corporations or well-funded startups. And there are very few video formats that meet the second. This is not the kind of decentralization that made the web thrive. It is quite the opposite.
Beyond the compression algorithms, what is especially exciting about Theora is that as an open-source project it might be easier for it to eventually evolve into a format that can more easily interact directly with other documents and data types on the Web. Videos should include more hyperlinks, for instance, and become part of the very fabric of the Web rather than an exception, which is still how it is treated today.
This is a good thing
It’s not really THAT big a deal. Xvid is already open source (and has been around for 8 years), and it hasn’t been widely used for commercial streamed encoding.
See the Wikipedia page on Xvid, section “patent issues”, for information on why it is not suitable as an open video standard for the web.
Xvid is an open source library for encoding/decoding of the proprietary MPEG-4 codec.
Surprising how many well-known codecs are based on the Mpeg-4 video specification. Want to read something that will keep you up all night? Read about the upcoming license fees for Mpeg-4 use that begins in 2010.
May be a good thing, but good luck trying to build a competitive codec, sans patent issues, for $100K.
The best possible way to solve this problem at this point is for MSFT and everyone else to compete intensely driving the price down.
WMV is a very inexpensive codec to use and is not at all encumbered with patent issues like the ISO codec.
Seems like a very long time investment because it would take a long time to make that ball start rolling.
This is VERY VERY good. Not only would it free video from its proprietary shackles it would eliminate the unnecessary license restrictions and costs currently keeping borders shut.
Throw in its effect on Ad $’s and you have the makings for a revolution.
This is good in theory, but could be a lot more difficult than it sounds. Codecs are patent minefields and unless the open standard is scrubbed clean, chances are most of that 100k will go to lawyers.
That depends on where the foundation is located. The Us is almost entirely alone in its views on intellectual property. So, if for example this Org were setup in the Netherlands. An idea patent in the US would not effect it… there are actually many codec’s out there that work like this at the moment. Its “we the people” that keep electing special interest groups that keep “we the people” from having things like global standards and freedom of innovation. While the rest of the world marches on.
Not entirely true. I actually had to work on the licensing for AAC and H.264. In the case of H.264 for example, MPEG LA handles the patent pool, which is comprised of myriad patent holders from all over the world, including US, Europe, and Japan. Before these codecs get licensed, everyone has to agree on what versions of the codec get licensed for what scenarios and for how much. Further complicating matters is valuing one entities patents over another. So, imagine having 20-30 patent holders from all over the world sorting this out and you can see why this is a mess. To top it off, enforcing patents varies across all of these geographies.
Now, the point I was making earlier though is a bit more subtle. I once asked two top codec engineers if it would be possible to create a new codec from scratch without violating someone else’s IP and got a resounding “NO!” The reason for this is that there are so many different methods involved that one would eventually have to use some or many patented methods to process the video. I think that the best way around this would be to piggy-back off of something like VC-9 which is an open version of WMV that MSFT got approved as an open standard in 2006.
What Geoffrey said. Plus the whole world is already using MPEG-4 (AVC/H.264 and AAC) — Flash, all mobile phones, all personal media players including the iPhone/iPod and PSP, all game consoles, Blu-Ray, etc. There are already great open source implementations.
This is as silly and pathetic as tilting at the MP3 windmill with Vorbis was a decade ago.
Christopher Blizzard’s points are mostly either uninformed or not well thought-out. Sadly, it doesn’t surprise me that Mozilla is wasting its time here, because they really don’t seem to know that much about any of this stuff (as witnessed by their pointless Ogg support in Firefox).
If Mozilla wanted to add MPEG-4 or MP3 or some other format to Firefox, they would have to pay for a distribution license. That’s the problem with these proprietary codecs. How many installs of Firefox are there…let’s say a few million. The license would probably be per download or some sort of blanket license and it would definitely cost more than the $100 000 that Mozilla is giving up for work on Theora.
>How many installs of Firefox are
>there…let’s say a few million.
Let’s say a few hundred million.
Although I certainly approve of their supporting of open source codecs, this a s slightly odd choice.
Theora is more of a competitor to XViD/DivX than the current video codec of choice (h.264). Although I haven’t personally tested them all thoroughly, from what I have seen, h.264 is on a completely different level.
On the other hand, Ogg Vorbis (audio) is roughly similar to its proprietary alternatives such as AAC (and both notably better than MP3). It would have been nice to see FLAC getting some support as well for its open source lossless audio compression.
In my opinion, this money would have been better spent supporting something like Dirac (which is currently supported and in use by the BBC).
As long as someone supports open source formats, that is good news for all…
I think that this money should have been spent on the BBC’s Dirac codec, which is technologically superior, having been written from scratch to encode using wavelets.
Charles, Tim: H.264 has open-source implementations but that doesn’t mean much because anyone using that code for anything serious without a license from the MPEG-LA will get their pants sued off. We couldn’t ship it in Firefox, for example.
The codec quality argument really only matters for people pumping out a lot of HD video. For someone who just wants to publish some video on the Web — or use video in their app — Theora works great, and doesn’t carry the risk of license fees down the road that H.264 has.
As for Dirac — we’re definitely interested in it, but Dirac playback is still more immature than we’d like it to be. If Dirac lives up to its promises we’d love to see it in Firefox.
“H.264 has open-source implementations but that doesn’t mean much because anyone using that code for anything serious without a license from the MPEG-LA will get their pants sued off.”
First, Mozilla shouldn’t be creating and distributing their own media playback architecture any more than they should be distributing printer drivers — it’s no only another distraction of the kind that seems to be impacting what should be Mozilla’s mission, but there are people and companies doing this better than Mozilla will ever be able to. The browser should simply use what’s available.
Second, your comment presumes that Mozilla is stupid enough to do it without a license, and I’m sure that’s not the case. Look to Apple as a model of a company who’s distributed AVC decoders for free for many years with QuickTime and iTunes installs. Surely Mozilla can work something out with MPEG-LA if the standard license agreement doesn’t work for Mozilla’s business model.
” We couldn’t ship it in Firefox, for example.”
Good, because you shouldn’t. You should focus on making Firefox suck less on the PC before Chrome eats your lunch there, and before Opera completely pre-empts any mobile opportunities.
“The codec quality argument really only matters for people pumping out a lot of HD video.”
That’s just crazy-talk. AVC/H.264’s greater efficiency is nice for HD, but where it’s crucial is mobile.
“…Theora works great, and doesn’t carry the risk of license fees down the road that H.264 has.”
That’s hilarious. Nobody’s gone after Theora yet because nobody of any consequence knows about it. If hell froze over and Theora did become popular, prepare to be sued out of existence by the companies who own the patents it steps on.
> prepare to be sued out of existence by the companies who own the patents it steps on
This threat has been charged at free software companies before, most famously at the Linux operating system. All that can be said is what Xiph has stated, that an exhaustive legal study has been done, no patents appear to be violated, and anyone disputing the claim should step forward. The source code is all available, so there should be little difficulty for patent holders to find problems, if any exist.
Yet, for decades, they have received no claim of infringement!
Ogg Vorbis and Ogg Theora were built for the express purpose of being free from patent restrictions. You’ll need some evidence, if you are going to seriously claim their entire mission is groundless.
That said, your idea that Firefox and Theora can somehow be “killed” makes me wonder if you know much about how these open source projects work.
I can’t find any video codec that has licensing fees or restrictions. r u saying if i want to put a flash video player and video on my website that i have to pay adobe?
Ryan, as a consumer, you typically do NOT have to pay (at least on the web). Application providers, however, typically have to pay for encode/decode or both. For this encoding content to publish on the web the licensing fees vary. On one of our video sites, we pay ~$3k/year for 1 encoding server license from On2 so that we can encode our content. Another complication is mobile devices, the licensing behind these is costly and completely inconsistent with web and applications.
We’ve been here before. I find it odd that nobody has mentioned the PNG image format. It was created for much the same reason that this effort is being pursued. Back in the early 90’s there were only proprietary formats for compressed images. So, a bunch of folk got together and created PNG to give us an open source alternative to GIF, etc.. Let’s hope that this video format is at least as successful as PNG and hopefully more successful.
bob wyman
@Geoffrey, @Charles: the projects that the $100K go towards are already fixed. They are all technical and none of this money will go towards lawyers. If you are concerned about patents, you might want to help this organisation http://openmedianow.org/ which is indeed trying to make sense of the patent and therefore the license space in media.
“the projects that the $100K go towards are already fixed. They are all technical and none of this money will go towards lawyers.”
That just shows how clueless Mozilla is about patent issues intrinsic to video codecs.
If Mozilla is going to try to forge their own path with open-source twaddle rather than adopt standards with several open-source implementations and an established ecosystem, they deserve whatever they get.
who are you you strange man? you seem very upset at the very existence of open source media!
@Robert – That’s good to hear. What all the open formats need is getting hardware developers onboard. If all those DVD players that support DivX also start supporting Theora or Dirac or both, things would really start moving forward. (Well that and some of the more ahem interesting film and TV distributors offering these formats of course). I haven’t looked into the processor power needed for decoding of course which is one of the main reasons stand alone DVD players don’t read h.264 at the moment along with license fees.
It has long been the case that if iPods started being able to play Ogg Vorbis audio it would start taking off in a much bigger way. As DRM (at least in music) falls to the side, people are downloading MP3s from iTunes/Amazon/eMusic/Play/whoever so they can play them back on any music device. As Ogg is open with no royalties due, I’m sure the stores without ulterior motives (ie. most stores except iTunes) would prefer that to be the format of choice if only it would also play everywhere.
Either way, keep up the good work investing in these. Hopefully they can get some traction.
It’s theoretically possible that major websites and/or manufacturers would drop the current market leading formats en masse, but it would take one hell of a business case in order to become a reality.
“But it’s OPEN!!!” is not a business case.
“But it’s FREE!!!” is sometimes a good business case.
Anyway, not everyone is entirely motivated by business cases. Some groups, including Mozilla, Wikimedia and many others, care about unencumbered standards and open source and factor that into their decisions. And we are less concerned with taking over the world than the companies you’re probably familiar with; enabling unencumbered video production and consumption is the goal here, not world domination.
What about the codec that BBC was developing ?
OMG a kick in the balls for flash .
Twitter = 500 millions dollars
Open Video Standard = 100.000 dollar
Sometimes i just don’t understand…
100K sign me up
That’s Good!
Dailymotion already started to encode some videos in Theora for the OLPC project: http://olpc.dailymotion.com
Check out a possible solution to linking video to the web: http://profile....endID=342277570
It’s a myspace app under development. More at blipd.com.
It’s good to see two successful non-profits working together. I think it’s equally important that the big players in the corporate world embrace this format (Google, Yahoo! & [dare I say] Microsoft).
I’m sure the Webkit crew will also be on board with this standard.
The BBC and others did a lot of work on DIRAC a few years ago, thats an open codec spec.
Dirac is a true open codec like Theora. Theora and Dirac have very different targets right now; Dirac for high rate and HD (and HD+) and Theora for mid to ultralow rate. That’s a result of where the two projects are coming from (Theora from the web video world, Dirac from highest-end production and broadcasting). The thought is that they will eventually meet and overlap in the middle.
The Theora codec is very lightweight compared to Dirac and should fit the niche of ‘easy, fast, simple, general purpose vide codec for anything’ for a good while. Dirac is still finding its legs, but the format has been designed for a higher-rate sweet spot. Dirac is most efficient coding high-rate video, Theora is most efficient at mid- and low-rate (especially low).
Dirac must be very poorly designed then. The fact is, any well designed codec, e.g. h264 or even the crappy VC-1 are more efficient then competitors and older codecs at a very wide range of bitrates (there’s no bitrate I’m aware of where MPEG4 ASP beats h264 for example). Hence the reason why most people are moving to h264.
And given that Theora is, sad to say, a rather crappy codec (most relatively recent blind comparison tests I’ve seen have shown that Theora pales in comparison to h264 at most bitrates), it looks like there’s no sign of any good truly open source codec emerging any time soon. The simple fact is Theora is never going to succeed when they are actively competing against h264 (Apple and Flash) or even VC-1 (WMV).
I had high hopes for Dirac, but if it is so crap, I guess we’re still going to have to wait.
Surf the Internet on your television . . .
Congrats.
Firefox 4 will be
x264“theora + vorbis” powered one.( tri licence to be quattro licence?
I like the foundation did Right investment.
Lords of FORGE cheers
I hope they seriously consider Dirac, it’s currently not supported enough to spread, and it has much better quality then first-gen Theora, which frankly looks like poo.
But something is better then nothing…
PS: to all the “Dirac is for high rate and HD” crowd, I thought HD and high-bandwidth are the “wave of the future” anyway!
There have been quite a number of improvements to Ogg Theora over the last couple years. Sounds like you’ve not had a chance to check them out yet.
Also, this announcement does not mean Dirac will be left in the cold. In fact, Dirac can be used within an Ogg format, so it is an easy next step after the framework is put into place.
I think the question has to be asked whether supporting Theora is worth it. The world is moving to h264 (and to a less extent VC-1). From all tests I’ve seen h264 is a load more efficient then Theora. If Mozilla thinks that offering people a crappy inefficient codec is going to be enough they’re dreaming. The sad, and simple fact is 99% of the world doesn’t give a flying flip about open source and free software. If site 1 uses h264 (e.g. FLV) and has better quality videos then site 2, guess which site they’re going to use? As long as these problems remain, Theora will be used by those who truly care about free content e.g. wikipedia while the rest of the world stays with the ugly state we are in now.
Theora has had ample time to try and beat h264, they haven’t and I personally doubt it’s possible (there are just too many patents they have to avoid). Theora has the advantage of being newer and being in an area with fewer patents and should hopefully be able to provide the efficiency of h264 at the very least. It’s really the only way if Mozilla wants to h ave any chance of making a difference.
Leaving it in the sidelines while wasting money and time on Theora is useless.
Sorry that should be “Dirac has the advantage”
Eat it on your tv