NAB 2019: LiveU Explains What 5G Will Mean for the Future of Cellular Bonding
Jan Ozer: Jan Ozer here from NAB. I'm with Dan Pisarski who is the VP of engineering here at LiveU. We're talking about two trends in contribution delivery, 5G and HEVC. Thank you for joining us. So you guys came out with some interesting HEVC-related statistics this year, tell us about those.
Daniel Pisarski: Yes, since LiveU's launch of our HEVC-based products, we've seen a huge uptick of HEVC use. I mean, for contribution, it'll only make sense, right? You're getting—even if you're still doing HD, not UHD—you're getting a much better-looking stream at half the bits and so in a cellular-bonded environment where every bit counts, and if your bandwidth takes a little dive, you still want the greatest quality you can get, that's a huge boost. Plus, on this cellular-bonding side, everybody's mindful of how much data they use. We all want to do more production, more video, for the same price that we're doing. And so being able to use half the bits to get that has been a huge boost for broadcasters. So we saw just a gigantic use of HEVC in our first year of really having HEVC on the market.
Jan Ozer: Okay, and you say half the bits. I mean, is that what you've been experiencing? Is it really cutting the data rate in half?
Daniel Pisarski: We've experienced with the hardware that we deployed it's really about half, right. You get the same quality, 720p or 1080i or 1080p stream, for half the bitrate that you would have used before.
Jan Ozer: So how's that work because when I tried the product, I had to—my ultimate destination was YouTube Live—and I had to route the signal through your cloud system to kind of convert the... So if you're sending to a system that doesn't support HEVC, how does that work?
Daniel Pisarski: So we're a first-mile solution, right? We use HEVC from the actual portable device up to the cloud and then we would convert it to H.264 'cause you're right, most of the online platforms right now are not taking HEVC yet. And we've spoken to a lot of the online platforms about their plans and it seems like, for a lot of them, HEVC ingress isn't on the roadmap for various reasons. Their relationship with the HEVC licensing bodies and things like that. So it's probably going to involve that transcoded in the cloud back to H.264 for some time. But all the benefits that you're looking for in the first mile, less bits used on your data plans, better-looking video in those tough conditions of actually out in the field and broadcasting over cellular are all solved. And then that H.264 step is what we call cloud-to-cloud. So it's going from wherever you're hosting, either our decode server or where we're hosting our cloud software, into your ingress point, that's usually going directly from some giant cloud data center to another cloud data center. So the bandwidth is rarely a challenge there. You have plenty of bandwidth there.
Jan Ozer: Okay, there's one more conversion, I guess, from HEVC to H.264 and then to the service.
Daniel Pisarski: Right.
Jan Ozer: Are you seeing any degradation in quality with that additional transcode?
Daniel Pisarski: No, you can tune the quality there. So you've got a great looking image coming from the HEVC decode and then you're feeding that into a high-quality H.264 software encoder and you're able to pick a very high bitrate there because, again, you're going cloud-to-cloud. And it's not something that's costing you on your cellular data plan, for instance. So if you're doing four megs of HEVC out of the field, you're doing eight megs of H.264 into YouTube or higher and you still get a great-looking image.
Jan Ozer: Okay, what's the cloud cost? Are you familiar with their pricing structure? How much does that add versus point-to-point H.264?
Daniel Pisarski: So, you know the way LiveU does it, we really have kind of dedicated cloud resources and we have a pricing on that, but it's not a pricing that's isolated to just the HEVC-to-H.264 transcode. You are paying that cloud cost with LiveU anyway to get our receive bonding, to get all the functionality that that cloud server provides to you, and that price hasn't changed to add that transcode. We already had enough CPU dedicated to that instance to do that transcode for you. I know some other companies have kind of broken it off into a free-standing service and come up with pricing for that. So for the average user just across the industry, I'm sure there's a cost factor there. But for us, you're already paying us for that CPU power, and that CPU power was already enough to get done what you needed to get done.
Jan Ozer: And that's a great point. Even if you're doing multi-modem bonding, you need to receive in your cloud anyway because Facebook isn't going to know what the signal's coming in from one of those if you've got four modems doing it. So it shouldn't, how much latency does it introduce?
Daniel Pisarski: It's right around 500 milliseconds latency for that transcode in the cloud.
Jan Ozer: Okay, latency is or is not a big deal for most of your customers? And if it is a big deal, what do you guys doing in that respect?
Daniel Pisarski: So it is a big deal for a lot of our customers. And so, what we've done in our bonding algorithm is drive down the latency in each new version we've put out. So, probably just about a year ago, probably at the show last year, we were at 1.2 seconds latency point-to-point. So really, glass-to-glass, if you were going to base ban, but from the unit in the field to that transcode stat when we're talking about going streaming to online. So 1.2 a year ago, then the next version came out, drove it down to 1. And now we have a version out that does 800 milliseconds. That's all over modems. If you happen to be on LAN or fiber or something like that, you can go down to 600 milliseconds on our units.
Jan Ozer: Okay another completely different topic relates to 5G and I've just heard some pretty outrageous things about it in terms of health scares and so when is 5G going to be a thing, when are you guys going to have a product for it, and what's the reality as it relates to that health-related concerns we've been hearing?
Daniel Pisarski: So, 5G to reach the numbers that really kind of get talked about a lot and advertised about 5G, gigabit through the air, or even higher, uses what's called millimeter wave, which are frequencies at 30 gigahertz and above. That's a high-frequency transmission compared to 5 gigahertz at the highest that you use right now for either LTE or Wi-Fi. But it's all the same RF energy in the air and there's been a lot of studies done about RF energy in the air that show that you really have to be very, very close to the transmitting unit and it has to be at very high power to actually impact the cells of the body. So when you're actually just talking about a unit that's in a backpack or sitting on a table or even a phone held to the side of your head, it's not close enough, right. Only the use case of literally holding the device to the side your head for an extremely long amount of time is really close enough to have that risk. The strength of the signal from a point of view of impact to cells of the body falls off so rapidly, even just an inch away, that there's almost no impact. That doesn't change with millimeter waves just because the high frequency. All that properties of RF frequencies are still true. So, millimeter wave is exciting. It's got some challenges from a technical standpoint of it bounces off of all solid objects. It can even bounce off of wood or in the air. So it doesn't have as long distances that can go as some of the sub 6 gigahertz frequencies. But that's all the challenge that the carriers are taking on to make 5G work. And they've got solutions for it, like directional antennas and beam forming antennas and things like that, to have that signal work. And we've seen tests where they've gotten great mobile, moving millimeter wave speeds, gig through the air, in tests in the field. So, it's very exciting stuff for us. We know it's gonna lead to units that have higher bandwidth, lower latency, better-looking pictures, more video transmissions over a single unit, more video transmissions over a single modem, all sent through it. We're really excited to kind of start launching features like that. Realistically, people get a little gung-ho on the kind of launch dates we think. So, right now you're starting to hear about, oh, we've got 5G deployed here or now it's deployed in these cities, but the deployments are very, very limited. And we think that, really, it'll take more towards the middle of 2020 before you see true, real true city-wide deployments that are usable by many devices.
Jan Ozer: So when do you get out in front of that? When would you expect to see your first 5G product out?
Daniel Pisarski: Definitely before then. I mean, we're in the lab right now with the carriers trying it with the larger antennas and the larger modems that are available right now, so that we can see how the whole system operates. We're talking to the modem vendors that make the smaller units and the smaller antennas and figuring out how they're going to fit in the units. So, I think you'll see a LiveU unit that's got true millimeter wave 5G integrated in it, really, maybe early 2020, but it's just my guess.
Jan Ozer: And that's not really you're thing. Basically, you push data out the serial board and you've got to decode it and...
Daniel Pisarski: Well, we get involved because we've often, a lot of our products use internal antennas. We work with antenna designers. It makes us a little bit unique in the market compared to some other products and so we get involved at that level. Right now, it requires brand new antennas that work in different frequencies, parts like that. You're right, the actual modem itself comes from another manufacturer and you talk to it through a serial API and it broadcasts data. But we definitely get involved in, alright, how much power does that have to draw? Does it need more cooling than the other ones? Is it the same physical size as the other modem? Now it's got a plug into an antenna, how big's that antenna? Where does the antenna have to sit in the unit? So, there's a lot of engineering there to do that we're already started on.
Jan Ozer: Alright, is it going to show up in a unit that size or is it gonna be a lot bigger or what's your sense of that?
Daniel Pisarski: So, we're kind of working with millimeter wave antennas right now to get a sense of that, but we've seen some examples of millimeter wave antennas that would fit in a unit like this. So we have got high hopes that it's not going to make units get bigger all of a sudden to support millimeter waves.
Jan Ozer: Okay, listen, I appreciate you taking the time and have a great show.
Daniel Pisarski: Thank you very much.
Per-title encoding is on the way out as Brightcove and others demonstrate the value of a more holistic approach. Streaming Media's Jan Ozer interviews Brightcove's Yuriy Reznic at NAB 2019.
When Telestream Vantage customers need a large video library encoded in a hurry, they can now turn to a hybrid cloud option. Streaming Media's Jan Ozer and Telestream's Ken Haren discuss Vantage Cloud Port on the show floor at NAB 2019.
Twitch uses a head-and-tail encoding strategy where popular content (the head) is encoded one way and less popular content (the tail) another. In this interview, a Twitch engineer explains what's on the video gaming powerhouse's five-year roadmap.
Streaming Media's Jan Ozer and NGCodec's Oliver Gunasekara discuss NGCodec's live HEVC 4k60 encoder, and why the company was wrong about the future of H.264.
Streaming Media's Jan Ozer interviews Phenix Technologies' Kyle Bank on the show floor at NAB 2019, and hears why low-latency CMAF isn't good enough for Phenix's demanding customers.
Streaming Media's Jan Ozer and Encoding.com's Greg Heil discuss findings from the 2019 Global Media Format report, such as why better codecs don't always find wide adoption.
Streaming Media's Jan Ozer and NETINT's Ray Adensamer discuss NETINT's Codensity T400, which is aimed at companies that need to do large live video encoding jobs at scale.
At NAB, NPAW showed off its new smart ads service, which not only measures whether or not an ad is delivered but looks at its streaming quality and how it impacts overall viewer engagement.
See you later, per-title encoding. There's a new system in town and it's called context-aware encoding. Learn which companies unveiled CAE solutions at NAB, as well as other streaming breakthroughs.