Easing the bandwidth crunch means making video compression more efficient, Broadcast International says
There’s no doubt about it: Video is truly becoming a reality on mobile devices, from Verizon Wireless’s V CAST mobile TV service to the iPhone and Google’s recently announced Android platform.
According to Broadcast International VP of sales and marketing, Rob Chipman, however, the increasing amount of video demanded by mobile consumers could lead to a bandwidth crisis. Aging mobile network infrastructure combined with the demand for video may make it increasingly difficult for broadcasters and operators to offer mobile video to a mass audience. Chipman says one potential solution can be found in handling video compression in a new way.
Mobile TV Update: Recently, Broadcast International (BI) has been discussing the possibility of a near-term bandwidth crisis, especially as more consumers begin to access video and other data-rich types of content on mobile devices. Could you provide more detail on this?
Rob Chipman: The reality is that we are always dealing with bandwidth issues, because there is always going to be more content to put over the pipes than can currently be handled. I think that what is exacerbating the situation right now, however, is that you have a very large influx of video content on the one hand — mostly in the online world at this point — and an aging infrastructure on the other. The major broadcast studios are now embracing online video and forming their own online video portals, and the MySpaces and the Facebooks of the world are adding video components to their offerings. You also have YouTube and the user-generated content [phenomenon].
So, the content side is one aspect. On the other side of it is an aging infrastructure that has not been kept up to date and is not sufficient for carrying large amounts of video. When you consider a typical cell phone, the conversation bandwidth may be 3 or 5KB. Video traffic may be a hundred times that, so you get a sense of the proportion of the problem as more people start to use video in a mobile environment.
Mobile TV Update: How do you think this will affect providers of mobile services, especially those looking to deploy mobile TV?
Rob Chipman: It’s a combination of the infrastructure and the network capabilities as well as the amount of video that is [truly being] offered and available. It’s not like you can just go to the mobile Web and play a video, even on a video-enabled phone. Typically, you have to subscribe to video packages. For example, Verizon has the V CAST package. The Weather Channel has a package for about $12.95 a month, with weekly and daily options. To be honest, these are almost disincentives. You have to want video bad to spend these not trivial amounts. It’s not like the Web, where you can call that video content up for free. And part of what’s driving those costs is the fact that it costs these companies so much more or uses so much more bandwidth than could be used for just regular calls. It’s not that they have necessarily maxed out yet, but if all those users start using video and again you start using a factor of one hundred — one person watching video makes it so a hundred people can’t — it’s using up the bandwidth of one hundred people making just regular voice calls. It’s almost at a point where with the current offerings, both in terms of the cost and because there are so few video options, there’s really almost a disincentive to be able to use it [or for operators to deploy it].
Mobile TV Update: What would need to change in order to spur more mobile video services and offerings?
Rob Chipman: Bringing down the cost of being able to use video in a mobile environment is what really will spur the growth, and bringing down the cost is a function of making it so that instead of using 300K of bandwidth (hypothetically) to send video down, if it’s in the 40K, 50K or 60K range, then you can make a case for not charging anything or certainly charging a smaller amount. This would be more similar to what’s going on in Asia and other parts of the world that are using video to a much higher degree.
Mobile TV Update: BI has developed the CodecSys technology, which is essentially a new way of dealing with video compression. How can handling video compression in new ways prevent or at least assuage a bandwidth crisis?
Rob Chipman: I wouldn’t go so far as to say it is ‘the’ solution, but it is certainly a key component of the solution. If you can bring the cost down, which is probably a barrier to people using it right now, then there will be more content available because the demand would certainly be increased. Frankly, where we need to get is to be able to basically mirror what’s happening in the online world today, where if you want to watch video, you go to a site that makes it available and some of it is available free of charge, some of it you may have to pay for. That’s where we ought to be in the mobile world, but the bandwidth issue is keeping the costs of doing this up.
What will kind of soften this for the major network operators is one aspect of the CodecSys technology, which is that it is essentially future-proof. One of the struggles that any kind of content delivery or broadcast entity is dealing with right now is converting 90 percent of the content delivery infrastructure around the world, which is still based upon MPEG-2. And everyone is kind of at this point now of thinking, ‘Well, we know we need to upgrade. Do we upgrade to MPEG-4 or some flavor of that?’ They can make that move, and it is not an insignificant decision because it means having to replace a lot of expensive equipment. This would be fine if you felt like MPEG-4 was going to be the answer for 10 or 15 years out. But with the pace of technology, it will be something else. If it’s a year down the road, or 18 months, or two years down the road, and you’re looking at that same decision again, that’s a frightening prospect.
Mobile TV Update: What makes the CodecSys future-proof?
Rob Chipman: The beauty of it is that it doesn’t matter what codecs or standards develop over time. In fact, we welcome development and new codecs and new standards, because it’s a software download. You can upgrade an encoder, for example, and add new codecs to the library that you switch between in terms of processing the video, and if the next standard becomes adopted and/or somebody comes up with a great new codec, we welcome it and could incorporate it as opposed to making an expensive encoder obsolete and having to buy a new one.
Mobile TV Update: BI and IBM have partnered to package and sell the CodecSys video compression software running on the IBM BladeCenter QS20 Cell Blade multicore processor. Is the hardware in this system upgradeable?
Rob Chipman: Yes. The BladeCenter is a scaleable hardware platform. One of the things that encoding video requires is a lot of processing power. And if you need to add to that, it’s simply a matter of dropping in another blade. So, the hardware platform is also scaleable as well as the CodecSys software platform is in being able to add new codecs.
Mobile TV Update: The CodecSys uses artificial intelligence to analyze a video stream and select the codec best suited to a particular video frame or sequence. How does this add to bandwidth efficiency?
Rob Chipman: The essence of the CodecSys technology is being able to switch among multiple codecs, because with any particular codec, or even a particular setting between codecs, you are to some degree setting it to optimize for a particular kind of content. By nature, any particular codec is kind of a generalized application that’s trying to do the best it can on all types of video. Our premise is that as the video changes, such as when a scene changes from a dark, indoor setting or a close up between two people to outside, in bright sunlight and/or fast motion, these are vastly different compression schemes. Rather than try and [process it] with the same codec or the same codec setting, we switch codecs. As you look at bandwidth use, just using a single codec, you’ll see it performs fine on certain content, but when that content changes, it’s almost as though it is gagging trying to process it. Our approach would be to switch the codec at this point and find the one that is doing the best job of processing the content and then changing to another one when that scene or frame changes. The concept is being able to change or switch codecs in terms of processing and encoding video.
Mobile TV Update: How could a broadcaster or operator use the CodecSys technology to its advantage?
Rob Chipman: The obvious benefit is being able to deliver high-quality video at lower bandwidth. The technology’s upgradeability is one that removes a lot of the risk from making an infrastructure upgrade. Unfortunately, there’s no way around making the initial one, but when you know that this is going to be a platform that you can live with for many years, that is a big factor.
Mobile TV Update: Could you offer a real-world example of a broadcaster or wireless provider using the CodecSys technology to more cost-effectively provide mobile video services?
Rob Chipman: As an example, take a wireless provider. To offer more video, the provider would have to add to its capacity in a major way, with cell towers and transmission hubs and other types of infrastructure. If there is the choice of being able to scale up to the tune of billions of dollars to increase your capacity and infrastructure or offset that by a platform that allows you to use much less bandwidth, the platform [is probably the more attractive option]. The reality is that you will probably have to increase your capacity regardless, but to a much lesser degree if you can decrease the bandwidth.
Mobile TV Update: As a final question, in what form will BI and IBM be making the CodecSys available for mobile video offerings?
Rob Chipman: We’ll use the IBM sales channel as the primary selling organization. They’ve got 10,000 resellers out there worldwide as well as established client relationships. They’ve never made an encoder before, but they have a lot of other hardware in the plants of wireless operators and other broadcasters, and they think this is something that ought to be added to their ecosystem as well. The Cell Blade and Blade Center with CodecSys is kind of a hand-in-glove fit, where they’re putting our multicodec video compression on their multicore processor. It will have the IBM brand name on it and will be essentially powered by CodecSys, using their sales channel. That’s the way we see this rolling out to the market.
For more information, visit www.codecsys.com.