“I need not fear my enemies because the most they can do is attack me. I need not fear my friends because the most they can do is betray me. But I have much to fear from people who are indifferent.” – Russian Proverb
Now I know that most people don’t really care about the mechanics behind playing video files and I can’t say that I blame you for caring more about your content than the technology behind it, but while this post will get into some of the more mundane mechanics of the codec industry, I ask that you stick with me because behind the scenes a war is being fought for control of your very television.
This particular battle has been going on for over 10 years now and centers around something called a codec.
When J.D. Rockefeller set out to monopolize the oil industry, there were several crucial areas where he attacked. He knew that he couldn’t control all of the oil fields because it was literally bubbling out of the ground, but what he could control was the distribution method for getting oil to the end customer.
In building his monopoly he seized assets used to transport oil from raw material to the end consumer. Whether it was owning all of the oil pipelines, so that he could control what oil cost him, owning the railroads so he could dictate how far his competitors could reach or owning the distribution points where consumers bought kerosene to light their homes, he made sure that he had control over every aspect of it. This was good for Standard Oil investors, but wasn’t very good for competitors or consumers.
Online video may not seem like it has a lot to do with the oil industry, but if you look at it’s early development, there are many similarities. So much content is bubbling up that the real challenge isn’t finding video oil, it’s getting it to consumers. Instead of pipes, now we have internet access, instead of railroads there are CDN networks, instead of gas stations, there are operating systems ready to serve us 24 hours a day.
In all of these industries, competition has been limited to a handful of big companies, but the industry that I’m most interested is much smaller than any of these. In the grand scheme of things, codecs (and the filters that go along with them) are the refineries of the video world. They take digital signals and convert them into the flickering magic that appears on our screens. Consumers may not understand the technical details behind it, but they are a crucial chokepoint in your digital video experience.
This battle has been fought on many fronts, but in the end it always comes down to one issue. Those who think consumers should have a choice and those who think they know better. It’s about control over your entertainment experience. Who, What, Where, When, and How you are allowed to consume YOUR media. On one side, well funded corporations with huge financial stakes, on the other, an unorganized patchwork of misfit companies and an army of guerrilla volunteers desperately fighting for a better entertainment experience for all of us.
The war over how video is transmitted may not make it to the front pages, but how it turns out will be important for the success of digital video. In order to better understand how this battle is going, I reached out to interview one of the Colonels in this digital revolution.
Dan Marlin is the CEO and Co-Founder of CoreCodec. His company has built many of the tools necessary to play video files. Before starting his company, he worked for DivX and over the years has contributed extensively to the open source codec movement. He also sits on the board of the Matroska Foundation, an organization dedicated to enabling high definition digital video support for as many consumers as they can.
In our interview, we discussed the growing momentum behind the MKV format, his thoughts on DivX and the competitive landscape of the codec industry and had a passionate discussion around a controversial decision by Microsoft to prevent outside developers from using alternative filters in Window’s Media Player.
In regards to MKV, Marlin had many positives things to say about the momentum that they are seeing. When I asked him about interest in the format, he said that over the last 8 months, they’ve seen a “20 fold increase in the inquiries in regards to more details, about usage about enhancing the current feature set.”
This interest should mean good news for consumers. As more and more customers ask “where’s the MKV?“, hardware companies are starting to respond. When I asked Marlin about how long it would take before we see MKV reach critical mass he said,
“If you look at the adoption scale, you’d probably have to say that we’re at the Ubber Geek stage right now. It will probably take 2 – 3 years. We’re just starting to see the penetration now and it’s been three years since our last release. I would probably have to say two years. Not this Christmas, but the following Christmas you’ll probably start to see more devices.”
One of the more interesting things that came up during our conversations was some of the trends that Marlin is seeing in the MKV adoption curve. It’s no surprise that the anime community was one of the first ones to start using the technology, but I was surprised to learn that countries in Asia and Europe have been more enthusiastic in adopting MKV then in North America. In fact, the trends for MKV adoption mirror the original DivX adoption curve exactly. It’s almost as if the people who’ve been long time DivX users are the first ones to upgrade to an HD experience.
“Absolutely, as a matter of fact it’s mirrored exactly. You could look at DivX in the early days when I was there going back to 2001 and you can actually see the same adoption happening, the anime, the ripped releases from the AV heads, it’s mirroring it, but you have to ask why they are doing it? They are doing it because of the flexibility that it brings to what they’re doing. They can add, especially when it comes to some of the guys that rip DVD and the like and Blu-Ray, they kind of make it their own. They can add menus, there are menus out there that even though they are text, they do very basic things, but there can also be a ton of files inside the container itself, there are info files and pictures you can group.”
While Matroska was technically created by CoreCodec, Marlin told me that he has plans to spin it off into a foundation similar to Mozilla. They plan to offer sponsorships to companies that want to tap into their early adopter customer base. One of the things that I found fascinating throughout the interview was the openness behind such a transformative piece of technology. Instead of monetizing their creation, CoreCodec is building a business around the open source eco-system. Big media companies that believe you can’t build a business around “free”, would be well served in looking at how Core Codec has been able to position themselves by giving a good portion of their technology away.
“we looked at it not looking to make money and that wasn’t really the intention, but even what has been proven now and maybe not so much back then, open source and the ecosystem around open source, there can be profit. Even in a non-profit foundation or a not for profit foundation I should say, which the Matroska Foundation will eventually become, you know is pretty much the same thing. You still can be profitable and make money to support what you developed.”
When I ask Marlin about his thoughts on DivX and how they are positioned in the codec industry, his thoughts were bittersweet, “it’s a love-hate thing.” On one hand, having DivX adopt the MKV container does a lot towards making it a standard. It also helps to speed up the amount of time it will take to get into hardware devices. On the other hand, not a lot has changed since DivX and XviD split paths and now that the open source movement has taken the upper hand, he doesn’t like to see confusion between X.264/MKV and DivXHD.
“Obviously they’ve rethought what they had to do with H.264 which is a migration, but they’re not providing anything of value to what’s already out there. As a matter of fact, it brings more confusion than anything else and that’s the frustrating part because they have their own eco-system with certification and us as a solution provider like with CorePlayer or the CorePlayer platform itself is working with third party OEMs and they are asking questions in regards to DivX and DivXHD and we say the same thing we’ve been saying all along. DivX is Mpeg video and DivXHD is AVC video.”
Of all the topics that we discussed though, the most controversial was the decision by Microsoft to restrict how third party filters work within Windows media player.
To fully understand the issue, you need to know how your computer reads media files. When you click on your file, filters take a look at that data and tries to figure out what to do with it. If it’s audio, they’ll send the data to an audio decoder so your soundboard can play it. If it’s video, then it gets sent to a splitter where the audio stream and video stream are separated. From there a decoder looks at the video data, decodes it and sends it to a renderer for display on your screen.
The controversy revolves around how Microsoft prioritizes filters when you play back content. Currently, if you have several filters installed that can all handle the same job, WMP will look at the merit value of each filter and give preference to the highest one. Since you have the ability to prioritize which filters you want your computer to use, it allows you to create the ideal settings based on your hardware.
This comes in handy if you’re trying to play H.264 video in WMP and it happens to conflict with your video card. Since the user has control over the priorities, you’re able to create a better (more credible) configuration.
With the Windows 7 RC, Microsoft has taken away your ability to prioritize which filter you can use. From their perspective, they get a ton of complaints about filter problems and by making it a closed system it improves the experience for their customers. For the codec industry though, it will reduce the incentive for engineers to continue to work on filters because Microsoft has just essentially seized the entire filter market.
Microsoft will argue that because they allow people to install whatever filters they want on their own media players, that this restriction is somehow reasonable. After all, they’re not preventing customers from downloading another media player and configuring the settings anyway you like, they’re controlling their own product.
The problem with this argument though, is that while consumers have shown that they’re willing to download a codec, by and large, they’ve been very reluctant to download an entire media player. It’s a big commitment to mess with the default settings on Windows and because Microsoft bundles a copy of Window’s media player into every operating system they sell, it drastically minimizes the potential market that companies like CoreCodec, DivX and Nero can serve. This ultimately leads to less investment in codec technology and lower quality video for consumers in the long run.
Take a look for yourself at a real life comparison between video played using Media Foundation’s preferred filters and an open source combination. While the differences may be subtle, there is clearly better focus and definition in the open source solution. It might not be much, but it makes a huge difference when you put it on a 60″ screen. Today, you’d have the option of recreating the ideal settings in WMP, but with Windows 7 Microsoft is now in control.
While Marlin wouldn’t go as far as to accuse Microsoft of using their dominance over the operating system as a way of stifle third party codec competition, he did agree with me when I suggested that this may have more to do with preventing competition then securing their media player for consumers.
“You said it I didn’t, but essentially when it comes down to it, that’s what it is. It’s just frustrating that we all have to go through what we have to do and they could have provided an integrated solution without having to lock out third parties. Period.”
Now we can argue over whether or not Microsoft had an evil intent when they choose to shut down part of the codec industry, but regardless of the motives, competition is hurt by their decision to close media player to third party vendors. When I asked Marlin whether this would hurt his company or whether it was a dam in the river that would fork around the issue, he had conflicting thoughts.
““I think it’s going to be both. Microsoft will probably tell you that there is no problem and then the Core people will fork around it, but you’ve got to question the value of it though. You could still have embedded DirectShow filters, why have them under media foundation?”
Later on in the interview he extrapolates,
“I would say that as long as the default decoders are not set as the default and can be overwritten, I think we’re OK. The question is what steps will you have to go through and will Microsoft allow those steps. Right now you can edit it, they posted the solution online, but Microsoft could bypass that solution with the next RC. So that’s kind of like a wait and see thing. It does affect our business though, it does affect DivX’s business, it affects everyone’s business. “
Now Microsoft is free to run their business in anyway that they see fit and while the issue over filter compatibility within WMP may be an inch in the grand scheme of things, with each inch consumers lose a little bit more control. What’s so surprising to me about Microsoft’s behavior though, is how bold their actions are given the current regulatory climate.
Someone should nominate them for Alpha Dog of the Week because it takes giant brass balls to use your ability to bundle software, in order to shut down an entire industry, while you’re being accused accused of abusing your monopoly by bundling software within the operating system. If the EU understands even a little bit about codecs, I would expect them to be up in arms over this issue because it essentially proves the argument that they’ve been trying to make. Microsoft’s dominance in the operating system is having a detrimental effect on competition in other areas of the software industry.
It could very well be that Microsoft has good intentions here, but given their long history of doing whatever it takes to gain control of the codec industry, I can’t believe that this is by happy accident. This is a company that just spent a ton of money to exclusively webcast the Olympics in their Silverlight codec. The lack of MKV support in Windows 7 prompted the Hack 7 MC blog to write that “Microsoft’s support of the format is borderline neglectful.”
The decision to interfere with the priority filter settings is so Machiavellian I still don’t know what to make of it. My cold banker heart says yes! yes! yes!, but the consumer in me says dear God no. While I understand that these issues are hard to figure out and that there are many ways to look at them, I hope, for the sake of the entire codec community, that Microsoft will rethink their decision to exclude third parties from Windows media player.
For a complete transcript of my interview with Dan Marlin, please click here.]]>
What exactly is CoreCodec and how does it relate to what you are doing with Matroska?
“Matroska is a container that our engineers developed some years ago. It is part of CoreCodec, I mean CoreCodec technically owns the rights to the trademarks and the like for Matroska. We have begun to form a separate Matroska foundation which will pretty much takeover from what we’ve begun to more of an independent, something along the lines of the Mozilla foundation, where they independently control the source code, but for right now CoreCodec is maintaining and helping to startup the Matroska foundation itself.”
CoreCodec also does for-profit work as well is that right?
Why did you go with such a radical open source license for Matroska then? Was that the only way you saw it coming together and competing or is it a reflection of your core fundamentals?
“Really it goes back to what our roots were. If you go back to even from when I started with multimedia when I was originally with DivX, I was one of the original people in DivX and we had what was called the OpenDivX decore, that was pretty much under an open source type of derivative license that I think was called the DivX open license or something along those lines.
It didn’t come out, it didn’t really pan out very well for them because I thought that the eco-system that they were trying, which they eventually came to realize, that the eco-system they were trying to go for just wouldn’t work with that license.
so in a lot of ways at Matroska we saw, well we saw a lot of opportunities in that AVI itself was an outdated technology. At the time, MP4 the container wasn’t very well defined. You’re going back again to about six or seven years ago now, so we looked at it not looking to make money and that wasn’t really the intention, but even what has been proven now and maybe not so much back then, open source and the ecosystem around open source, there can be profit. Even in a non-profit foundation or a not for profit foundation I should say, which the Matroska Foundation will eventually become, you know is pretty much the same thing. You still can be profitable and make money to support what you developed. It’s no different than Redhat, though they are more at the corporate level trying to support their customers or their derivative works of the Red Hat OS on Linux.
With the CoreCodec side of the business how are you helping other companies? Who would be an ideal customer for you and how are you helping them?
“With regards to Matroska, it would have to be implementation. While Matroska is open source, we also have a free BSD style license version of our library for playback. What that means is that you can still playback the content even though you may have a closed eco-system. We still allow you to integrate the technology into your closed eco-system, so that’s the difference there. We help third party solution providers integrate Matroska into their solutions. Some examples that are currently out there implementing would be LG, Panasonic, Toshiba and the like, more recently Western Digital with their hardware that they just released for multi-media playback.”
Over the last 12 months it seems like we’ve seen awareness of MKV really start to take off. Even some of the more mainstream technology writers are starting to make it a check list feature, has that surprised you? What are you seeing in terms of the interest around the container? Has there been a ramp up over the last 12 months?
“Oh absolutely, as a matter of fact we’ve probably seen a 20 fold increase in the inquiries in regards to more details, about usage about enhancing the current feature set. It was really designed from the very beginning and if I can go back to the conversation that some of us had here in the early days, it was designed for the next 25 years, in that no matter what the MPEG consortium came up with in regards to the spec for whatever might be the next gen including Mp4 as a container, it was really designed to go beyond that, it was designed to let the spec grow as the technology kind of grew at the same time and again that’s the whole point of it.
Part of the early days, going back to DivX, one of the biggest things that held DivX back was the AVI container. It’s really why DivX right now is adopting Matroska in what their efforts are, and we’re all for that, even though, I’m no longer with DivX, I wish any third party that wishes to adopt a technology like Matroska the best of luck.
I really think it’s only going to enhance what has been in the past, a frustrating experience for end users. In that Matroska, once adoption has come and the tools are there, people start to understand more about the difference between a codec and a container, the experience will be much smoother because there won’t be hacked flavors of AVI or this or that, they’ll just be standards. No difference than what MP4 is, but the difference I think is about control.
People feel that they have no input whereas we listen to all third parties. We adjust and adapt as needed, but the core code itself has not needed to change itself in three and a half years, so we have not needed to touch the container itself in well over three years now and again that’s a testament to the flexibility of what is in place. Now in the future, that’s going to grow obviously with 2.0, but we’re actually going to slow our development for 2.0 because of the adoption curve of 1.0 right now. Which will allow us to be a little more meticulous and listen to the customers, listen to the consumers, the OEMs and the integration providers and see what their needs are and then we’ll kind of adapt to it, so we’ve kind of slowed down a little bit in what will be the future version of Matroska just so that people can start to enjoy what’s already out there.”
With Matroska, because it’s a container it allows you to put various codecs in it. Does that create problems when you go to try and certify MKV with a CES company, where maybe it could play some MKV files, but not all of them?
It’s setup so that the structure within it is the codebase with the exception of only one or two formats which are only recent formats and I think one of them wouldn’t actually be seen at all. It’s pretty much setup, I don’t think you’re really going to see problems.
Now certification, that’s something that we’re just getting into now. We’ve put out there I think in our specs, we actually put out levels of certification what a mobile and desktop platform are going to need. We put that out there for comment and later on we will have some form of certification, but it’s more so for sponsors and things down that line where we’ll certify their hardware against a certain profile and allow them to use logos and the like for distribution on the device itself or on the box and things like that, so it’s more so the flexibility and the updates hasn’t needed it for awhile and I don’t think any existing formats, even the ones I briefly mentioned, their not pressing. I think one is E-AC3, but I think even the guys that discuss their current code will still support it.”
Given the momentum that you’ve been gathering, what do you feel the next step is? What needs to happen for the next stage of MKV’s growth?
“Well probably, more hardware. That’s where we see it. The growth in the past 8 months, I won’t even say 6 months, prior to Christmas 2008, we saw big growth and we see this year that the growth even will be much more explosive going into this year’s holiday season. So I think that would probably be the biggest thing that we would want and we’re also expanding out from the services we’re providing within CoreCodec. People are starting to look more closely at what we have with our other technologies with CoreAVC, CorePlayer and even next year with CoreMVC which is based on H.264 and CoreAVC. So we’re excited about doing that, helping others as an alternative to what is currently out there right now. Again, no different from what we’ve done in the past.
How do Core services work within an open source eco-system? You’re working on creating these filters for private customers right?
Yes, it depends, both You’ve got to take it two different ways or at least we look at it two different ways. One as a consumer and one more from the OEM/consumer electronics side. Each one is supported in different facets. On the consumer side you have our codecs themselves outside of the container, the Matroska container, we have all of our decoders CoreAVC, CoreAAC, we support Mpeg2, Mp3, etc. We pretty much have our own codecs very similar to Ffmpeg, except it’s closed source.
Now later on at some point, we do plan to open source pretty much our entire eco-system, if the business warrants it and right now it looks like does and that is because of the popularity of the CoreAVC as it is. We can still open source it and monetize it and also release our encoder as well, but at the same time we’re very cautious about what we do. We take steps appropriately. We do, again going back to the core of what we’ve always done, we open source a lot of our technology. Like Matroska, the Haali media splitter may not be open source, but it is free and we put it out there for everybody to use. Most of our other directional filters from CoreVorbis to our original CoreAAC is all open source and available on our website.”
Help me understand some of the terms out there that can be a bit confusing, for me at least. What’s the difference between a codec, a splitter and a filter?
“You can throw renderer in there as well. Each one plays a certain role in playback. Consider a container like Matroska a ball. Inside of that ball you can put anything. You can put pictures, you can put video, you can put audio, you can put text files, you can put pretty much, even executibles in there if you support it with your parser. It gives you unlimited possibilities to put stuff in that ball. Then you have the codec which allows playback. They are the physical means for playing back the audio or video. Now in between the playback and the file itself, you’re going to have to have a player that supports it and to be able to do so, you need a splitter and a renderer and that’s where the Haali media splitter comes in.
Technically what it does, it puts pin outputs to the audio and video and connects the playback to the physical software itself within the operating system and that’s on Windows, Linux is something different and CorePlayer is something different as well. CorePlayer has it’s own built in filters. It’s not DirectShow, it’s pretty much has it’s own internal decoder, renderer and splitter.
Each playback means is going to have it’s own and Mplayer does it by mimicing DirectShow where you physically take the Windows .dll and you’re able to take it over to Linux and kind of emulate DirectShow within Linux. It’s hard to understand, but once you start to use it a little bit more and you start hearing the terminology you become much more familiar with it, but just know that each part plays a very key role in processing the playback of audio and video.”
Looking at the step forward to Window’s 7 that is coming out, there’s been a lot of positive press in terms of native codec support and the h.264 and AAC support. A lot of people would say this is a good thing, that this is going to make the codec experience more seemless for consumers. Why should anyone care about Windows locking down their filters and is Window’s 7 really bad for the codec industry or is this something where the good outweighs the bad?
“It really depends on how you look at it. If you’re a consumer, absolutely. It’s a controlled eco-system, it’s no different than what Apple does with the iPod or even the iPhone where they’re trying to control the experience for the end consumer. For technology companies, however that has a very large investment in what they’re doing as a service provider for their individual solutions, it’s bad.
Going back to our case with CoreAVC, our H.264 filter, we have people that have purchased it, consumers, OEMs that have purchased it and to have to shut out what is something that people have come to expect, ie. the flexibility of being able to add filters or add playback filters on their own, I think is bad.
You should always give them the choice.
I understand the point of trying to control the environment. I’m sure that one of the largest problems that Microsoft deals with when it comes to Window’s media player is people complaining that their multimedia doesn’t play. They’re trying to integrate that, it does make sense from a business case, but to have to shut out, under the guise, maybe of saying that it is a matter of controlling it, I think we can pretty much say that, or guess here, I don’t want to overstep myself, that it may be a guise and that may be in lieu of control.
It has to do with DRM and if you think about the formats that have been shut out, it’s AVC or H.264 and it has to do with AC3, this is what’s in Media Foundation itself. Now Media Foundation is a higher class replacement for DirectShow, in all honesty, it’s DirectShow, that’s all it is.
So they basically used their own filter and are now saying that we’re going to control the codec at the filter level and anybody who wants to use a different filter, you can do that, but you can’t do it on our player, you have to build your own application?”
“Right and there are some work arounds where people are hacking the registry and are trying to come up with bypasses where it overwrites the priorities for media foundation so when somebody clicks an h.264 file, that it actually will open up with a third party filter, DirectShow. The point is, there’s no documentation, or very limited documentation and there’s not even clarification.
Even Microsoft is confused in this case.
If you take Silverlight for example, Ok lets do a comparison with Adobe. Adobe has the Adobe media player and the like. They all share the same architecture essentially. Whether it’s Adobe Air, they all interangle amongst each other, whereas Microsoft’s Silverlight has it’s own collection of decoders within it and does not share the embedded ones that are on the OS. It’s kind of confusing and goes against what they should be doing.
If they’re truly trying to control it, meaning yeah that’s OK that Silverlight is a stand alone product, but if they’re going for the largest control then why aren’t they sharing the Media Foundation codec within Windows 7, within Silverlight? Maybe I could see stuff like XP, Vista making sense, but it doesn’t in Window’s 7, they should be sharing the same architecture and again, I don’t know I would probably say that the argument is going to come for control for their users experience, but I would probably have to lean towards it being more in line with DRM.
It’s about control for secured content.
Do you think DRM is the reason why Silverlight hasn’t been made portable? I mean isn’t the solution that Matroska is solving partly because there isn’t a high def portable format out there?
“You know, I’ve heard people talk about that, but I don’t really know about that. I just think that it’s almost about staying the course, but trying to migrate to something new at the same time and I just don’t know if the flexibility of being able to get things like Matroska into what they have, would work. How would you even get Matroska into Silverlight?
I don’t know
“I don’t know either and we have some of the top guys over at some of the forums like Doom9 and you have some of the engineers from Microsoft kind of chatting about it and as long as we’re putting the thought in their head, I’m hoping that it’s going to play a role in their future decisions. Not just for Matroska, but even for third party decoders like CoreAVC or DivX or PowerDVD or Nero or any of those other third parties.”
Looking at that piece of the eco-system, it seems like there are a lot of MKV files out there and now that we’re beginning to see more hardware devices support it, how far away are we from software solutions where you can start editing files or is that not even related to your goals with MKV?
No, it absolutely is, in fact that was one of the keys of what we’ve always done. You have AVIMux or MKVtoolnix which, is probably one of the most advanced tools, it is THE authority when it comes to mixing and creating Matroska files. That was a very key part of Matroska from the very beginning. We wanted to unify, or mirror some of the stuff that we’re doing with CorePlayer. Even CoreAVC is a single unified core for every operating system and Mosu with what’s he’s done and even Steve or some of our lead engineers, what they designed from the very beginning, is to be able to do that for adoption purposes. That way there were no restrictions, Linux OSS, Windows Pocket PC, Windows Mobile, even though it doesn’t really matter, it will create outputted formats supported on all operating systems and the tools for creating Matroska and editing Matroska is no different.
They’ve evolved over the last five years as much as the format has not evolved. The tools are the complete opposite I should say. They’ve evolved with at least a dozen versions every single year whereas the container hasn’t really needed to change. That’s a key to our success. Again, all under the same type BSD license where they can take the source code and integrate into the product, whatever that product might be.
They’re not required to pay royalties or licensing, all we want is for people to adopt the format itself, that’s the key to the success of it.
It seems like there are going to be three standards for downloadable media. The Silverlight solution, The Apple solution and the open solution that people can go towards. Recently, it seems like a lot of the hardware providers are partnering with studios to do streaming deals and the hardware is more about incorporating YouTube and Netflix into the devices then full portability for consumers. How long do you think it will take for mass adoption of MKV and having a serious penetration level where most of the devices out there have to support to?
If you look at the adoption scale, you’d probably have to say that we’re at the Ubber Geek stage right now. It will probably take 2 – 3 years. We’re just starting to see the penetration now and it’s been three years since our last release. I would probably have to say two years. Not this Christmas, but the following Christmas you’ll probably start to see more devices.
Even with companies like DivX putting DivXHD and the like, but we’re not convinced that DivXHD is going to succeed at all. It really is just a rehack of H.264. It’s no longer DivX. DivX was a file format and they’ve done nothing different with this, they just have the eco-system with their hardware partners for when they upgrade. I don’t know, I’d probably have to say within the next year and a half, we’ll be at a more critical mass.”
DivX adoption of MKV helped to solidify support in creating it as a standard, but as far as what DivX is trying to do with it, do you see a solution there or do you think that they have to reinvent their business as DivX SD becomes less popular? Where do you feel that they are positioned in the codec space?
“Well they have. Obviously they’ve rethought what they had to do with H.264 which is a migration, but they’re not providing anything of value to what’s already out there. As a matter of fact, it brings more confusion than anything else and that’s the frustrating part because they have their own eco-system with certification and us as a solution provider like with CorePlayer or the CorePlayer platform itself is working with third party OEMs and they are asking questions in regards to DivX and DivXHD and we say the same thing we’ve been saying all along.
DivX is Mpeg video and DivXHD is AVC video.
You can still say DivX name, just put a disclaimer saying blah blah blah, there’s no difference. Now DivX would probably argue with that and say they have file specific format stuff, and they do if you talk about the DMX menus, which they do have or they have Sub-X, which is their subtitle format and yeah that’s proprietary to their decoder and for them to be able to playback those file formats, but do they really bring anything of value to the table? No.
We support sub-titles in Matroska in .KF, so any kind of SSA or other sub-title formats we already support and the menu system is already out there for the 1.0 menus. You can create menu overlays right now within Matroska.
In a lot of ways, it almost seems like they’re back at the Xvid/DivX split where you have something where there is now this open source solution and correct me if I’m wrong, but if you’re Toshiba and want to support MKV in your products does that cost them anything at all?
“With DivX sure”
No, no, no, recently JVC had an MKV box without DivX, did it cost them anything at all to include Matroska support?
So that’s very disruptive when you look at the eco-system and it goes back to the question of you’ve got these closed sourced solutions where you’re building a business and making money, but then you’ve created this open source solution. Is that because the only way to take on Microsoft and Apple is to keep it open and you need that help or is it more that you wanted to future proof your own container?”
“It wasn’t so much container it was more or less a need. The limitations on AVI as a container were very frustrating for engineers to be able to do things that they wanted very easily, you could not. Lets say you wanted to add or remove elements from within the container, for example an audio track where you wanted to be able to add an additional audio track. You couldn’t do that with AVI easily, you could hack it, but there weren’t tools to be able to extract.
With our tools you can extract or remove or replace very easily, so it wasn’t a move against Microsoft by any means, it was more of a necessity of what we wanted and then once we found out what we wanted, we tried to look at the future and see what we can do to extend the adoption for that 20 – 25 year period and we feel right now considering the feedback that everybody’s provided over the past few years and the fact that we haven’t changed for the past few years that we’ve kind of accomplished that.
Now to monetize it? The whole point wasn’t really to monetize it. We will with some of the stuff we’re doing, we have contractors that go out and help out and integrate within CoreCodec by helping third parties and that’s great and with the Foundation that will change a little bit because of the sponsors, but it’ll be no different than XMPP or Mozilla where you have sponsorship levels.
It gives you funding to do marketing, so that you can get people interested
Yeah, you want to be able to advertise. We also want to be able to further our eco-system, not only with Matroska, but also with CoreCodec and some of the stuff around other technology that we’re doing and we want to be able to provide that so having that will help us continue to stay in business and continue to further evolve the technology.
Providing the logos and the certification for those levels will definitely help to that point and cause less confusion in the marketplace, which again I feel that DivX is and will continue to cause. Again we’re not sold, they’re going against everything that we have tried to put out there, but again I understand it.
What’s to differentiate an MKV file that was created with DivX tools vs. Us? The only thing we can do is warn the user. We can detect it in the header of the file that was created and we can say you’re trying to play back a DivXHD file and this file may not be supported. We’ll continue to play it back in our eco-system with CorePlayer. With the filter we can do the same thing, we can warn the user that it was created with the DivX tools and it may not be supported because it may not be and DivX will try to tell you that everything is compliant with Matroska and I’m sure that it is, but that’s just what they’re saying. We’ve already identified stuff that’s not compliant and we’re reporting back to them and it’s good to have a type of relationship with them.
There’s a bit of a dichotomy there. It sounds like one hand you’re competitors, but on the other hand they’re ability to get this out into their ecosystem could really benefit the container as well?
Yeah, it’s a love-hate thing, same thing with Microsoft really. I wish them luck. I think that what they’re doing is great for the container. It shows that we were right all along. Hacking AVI was not good and Matroska was built for the future and I’m glad for that, but at the same time I don’t want them to ruin something for the consumers or confuse consumers anymore.
That’s my only concern. I wish them luck. When I left DivX, I left them wishing that they could have done better and I’m glad to see that they made something of a good move, but the question is how will it hurt in the long run.”
They made a splash, but where they’re going is what everyone is trying to figure out right now?
“What they have going for them is the current partnerships with the companies that they have and that’s a great great thing and again I just feel that under the guise of saying that it’s DivX and DivX only is what the third parties need to realize. Most of them aren’t stupid please, the third parties whatever companies you mention, they all know.
They have to see some value in the certification in order to spend the money so if it’s not there it won’t shake out that way.
“Absolutely and again you have to question the true certification and the cost for that certification. I’m not going to sit here and spit out numbers that they charge to certify, but the value is about control. They’re looking to control the experience for the customer, for that third party. That’s going to become less and less of an issue over time, especially with faster hardware coming into play and even TVs and the like, I’m not sure if you saw the latest TV reviews from Google?”
I saw that you have an LG TV coming out, but that’s thru USB stick right, not thru an ethernet connection?
“Yeah, but the speed of the menu system is horrible, it’s bad.”
So there is this trade off where you need the chips to be a certain speed in order to support your format?
“Yeah, but like with anything that will become less of an issue. In the mobile market CorePlayer excels because of the speed of the player and that’s because it’s written for 386, we write at a very low level, so when you start taking bloated code and start applying that to something that doesn’t have a lot of CPUs or dedicated CPU cycles, that’s when you start to realize the same performance. Going in it’s favor though, you do have hardware catching up especially when they have to upgrade to H.264. With TVs, it’s no different from some of the stuff they’re doing. Even with CorePlayer it’s already integrated with some third parties for TV usage, so that’s great for us and what we’re doing, but all we’re doing is providing the multimedia framework for playback on the TV itself and the user interface with CoreUI.
Going back to the Window’s 7 issue real quickly, do you think the decision to freeze out the outside filters, is that going to hurt the codec industry, is that going to result in less competition or will this be another case where there’s a dam in the river and the river forks around it?
“I think it’s going to be both. Microsoft will probably tell you that there is no problem and then the Core people will fork around it, but you got to question the value of it though. You could still have embedded DirectShow filters, why have them under media foundation?
Because you can lock competition out, people won’t download other media players they will download codecs.
“You said it I didn’t, but essentially when it comes down to it, that’s what it is. It’s just frustrating that we all have to go through what we have to do and they could have provided an integrated solution without having to lock out third parties. Period.
That’s the most frustrating thing. Then to not provide the information needed in order to overwrite, obviously we can now the hacks are out there, guys have already posted the registry key that we can kind of use, but what we’re doing here is we’re waiting. They’re not providing us with any of the information that we need now, so we’re in a wait and see mode. We’re in RC-1 right now with Windows 7 we’ll have to see with RC. Maybe they’ll post more information.
It sounds like this isn’t so much a threat to your ability to do business, but rather a lot more man hours, a lot more work to get this worked out?
“Yeah, I would say that as long as the default decoders are not set as the default and can be overwritten, I think we’re OK. The question is what steps will you have to go through and will Microsoft allow those steps. Right now you can edit it, they posted the solution online, but Microsoft could bypass that solution with the next RC. So that’s kind of like a wait and see thing.
It does affect our business though, it does affect DivX’s business, it affects everyone’s business.
When you can’t have filters within your application work properly . . . whether anybody realizes it or not, there’s mass adoption going for Windows 7. I can tell you that the demographics from our website, the ubbergeeks that are early adopters, I can tell you right now that the people running XP and Vista jumped to Windows 7. A huge percentage of them, probably 75% of them are all on Windows 7 and this is just since the RC came out.
It’s going to become an issue very fast, much more so than anybody thought. We need some kind of answers and I guess you’re not really going to get them right now from Microsoft. They are on our forums talking a little bit chatting back and forth with Haali, Steve and myself, so it’s good to see some dialog, but we’re still not getting any answers and I don’t expect that to change.
It seems like it’s an issue that affects your industry quite a bit, but it’s hard for people to see past that just because it’s not as visible.
“It’s definitely not as visible, but when people start screaming . . . . we’ll probably see that, if they are planning on releasing Windows 7 by Christmas . . . It’s going to be ‘great’ fall.”
I know that when DivX first came out, it was released in Europe and there was big adoption in France and to a certain extent spread geographically from there. Have you seen similar trends with MKV?
“Absolutely, as a matter of fact it’s mirrored exactly. You could look at DivX in the early days when I was there going back to 2001 and you can actually see the same adoption happening, the anime, the ripped releases from the AV heads, it’s mirroring it, but you have to ask why they are doing it?
They are doing it because of the flexibility that it brings to what they’re doing. They can add, especially when it comes to some of the guys that rip DVD and the like and Blu-Ray, they kind of make it their own. They can add menus, there are menus out there that even though they are text, they do very basic things, but there can also be a ton of files inside the container itself, there are info files and pictures you can group.
It’s creates a more interactive experience for consumers?
“Yeah and is that different than .divx? it is, but again, it is what it is. They’re looking for something that has a hacker edge to it and not that Matroska was really created for that, but we’re glad to see adoption no matter where it occurs and that kind of pushes hardware companies to do the same. They’re getting their customers asking why isn’t MKV support in there? Look at Sony, why isn’t there MKV support in the PS3? It’s free, they should add it, same thing with the Xbox 360.
Why do you think that the hacker community choose MKV for the Blu-Ray rips?
“Probably because of rips support, we started to see it right off the bat really. A huge Japanese following with Matroska back in probably 2004, somewhere right around there. We saw that right off the bat, the anime people jumped right on it.”
Do you have any sense as to why anime seems to be so far ahead of the curve when it comes to online video?
“We’ve actually had this discussion internally. I wish I could answer that, you could say it’s geographic and that the Asian market is faster to adopt cutting edge technologies, but at the same time why is China still using Real video? It could be argued both ways.
I would tend to think that they just look for something that’s flexible to their needs and I think that’s what Matroska is providing just like DivX did back in the early days with quality video. It was night and day vs. anything that was prior to that. I think Matroska brings the same thing to the table, but we do it with a little bit more flexibility than what they can do with the files themselves, being able to attach multiple files within the container.
What do Matroska fans have to look forward to next? In the next 3 – 6 months, are we going to see more hardware coming out?
Is 2.0 going to be rolling out or is that further down the road?
2.0 like I said, we’ve definitely pushed it out for awhile because were going to hang back and let 1.0 get the adoption. You’ll probably see things finalized with MKS/MKV/MKA so there’s support within the web browser itself by default. Within all web browsers, Safari and the like. The application type will be set. That will be coming in setting us up for maybe tweaking it for some more menus.
We’re also talking about adding a bit more menu work into 1.0 and we’ve already done some work prior to 2.0 which is a major part of Matroska technology called EBML. We’ve already completed EBML 2.0, as a matter of fact EBML 2.0 is a complete rewrite of 1.0 which is in Matroska 1.0. It really provides us greater flexibility for what we will be doing for Matroska 2.0. At the same time, it uses the technologies that we created here at CoreCodec and utilizes those technologies within EBML which is CoreC which is our AVC library language. Going along those lines is what we’re planning for 2.0.
It’s exciting, you guys are at the heart of something pretty big here I think and to be part of that is neat. Since this is your 2nd go around is there anything that you think you’ll do differently from when you were at DivX?
“Yeah, it has to do with playback. While Matroska for us is great, we’re obsessed with fluid playback no matter what it is, including Matroska, so right now our main goals are to extend the eco-system, the CorePlayer eco-system even further and then be cutting edge with Matroska as we roll out our 1.x version and also 2.0 and that’s pretty much it. We’re excited to be able to provide that for the consumers and our OEM customers.”