| Subcribe via RSS

NAB Perspectives: ATEME’s Antonovich Explains How HEVC Will Drive 4K, At-Home–Production Models

April 27th, 2015 Posted in Headlines By Jason Dachman

ATEME was focused squarely on the burgeoning HEVC compression standard at its NAB 2015 booth. With a 4K HEVC high-dynamic-range (HDR) demo that drew plenty of interest, the company teased the Kyrion HEVC encoder and decoder scheduled to ship late this year and early 2016. In addition, the company showcased its ultra-low-latency encoder/decoder, as well as virtualized compression for the converged headend.

SVG sat down with Mike Antonovich, GM/SVP, Americas, to discuss ATEME’s HEVC offerings, the evolution and promise of the standard, and how it will drive both HD and 4K production in the future.

Can you tell me a bit about the 4K HEVC HDR demonstration?
This is clearly targeted, at least today, to the cinema industry. It’s all about creating a unique 4K experience. More pixels is really not compelling enough. It’s making better pixels for better-quality TV. So a wider color gamut and high dynamic range are really big drivers for what makes 4K truly interesting and immersive, and that’s what we’re demonstrating. We’ve got some really good Hollywood-studio content that has been transcoded and optimized for full-HDR display. The whites really pop, and the colors are saturated, and the blacks are deep enough to set off the colors and give you a visually better experience. So, clearly, there will be a time for that for the sports market, but, right now, it’s really good for the cinema marketplace.

What else is ATEME highlighting at its booth?
We have a lot going on this year. For our Kyrion product line for contribution, the two most exciting things are the software path to HEVC and both our encoder and decoder. Because we are really a software solution, we have the unique ability to upgrade our software with regular, recurring releases, including a roadmap that will deliver HEVC decode in the fourth quarter and HEVC encode in Q1 of next year. That is going to provide our broadcast customers who want high-quality 4:2:2 10bit sports contribution to save bandwidth by 40%-45%. So that makes a huge difference for anybody who’s paying for bandwidth on satellite or on broadband or merely needs to send these signals home or distribute them.

Have you seen a growing demand for HEVC compression in real-world deployments?
We are starting to have clear, concrete requests [for HEVC]. While HEVC is essential for 4K, it actually has a lot more applicability for anybody doing SD and HD because of its ability to work with interlace. Most broadcast-sports production is still heavily reliant on interlace processes, so it’s important that we were able to provide the same kind of bandwidth efficiency for them as we get with progressive, which is what the standard was really optimized for.

The standard has been there, but we’ve been waiting on what’s downstream in terms of distribution. But the good news is for the contribution industry: as soon as it’s available in this kind of form factor, it’s immediately useful for sports contribution. To save bandwidth and to capture higher data rates at better quality within the same channel bandwidth or [to reduce] bandwidth with the same relative quality: these are big, big drivers for both cost-conscious and quality-conscious broadcasters and sports producers.

Downstream, chipsets are being developed. You can certainly see smart TVs already with HEVC, but the addressable universe is still growing. You are going to see demand initially through OTT or second screen for sports, because that’s where the biggest addressable audience is today. It’s not yet aggregated in the way we classically saw with HD the first time with DTH, where you created an audience. We’re still waiting for the sports community to step it up and produce more in 4K, but, in the meantime, we’re going to find applications.

How soon before we start seeing HEVC deployed for contribution applications?
You can have it for Christmas. Between our encoder being available in Q1, we’re hoping to pull that earlier, and, certainly on the decoder side, it’ll be there. There are ways to get there outside of our Kyrion product. I can provide HEVC encode now all the way to 4K on our Titan product, which is on an Intel Blade server. It’s not fully optimized for remote productions, but that’s what we used at Roland Garros and the Olympics. Certainly, on a fixed-location basis, we can provide 4K HEVC today.

Where are you seeing the most interest in 4K contribution and delivery in these early days?
What’s interesting is, we’re actually seeing a great deal more 4K demand for proof of concept and product launch in Latin America than we are in North America. That’s very exciting. Certainly, in Mexico for high-end football as a driver. Brazil will not be far behind with the Olympics coming next year. While [Olympic Broadcast Services] hasn’t announced any intentions to do 4K just yet, I would watch the space; I think that could change and move pretty quickly.

Thus far, we have seen 4K content transmitted via a quad-split method using four H.264 encoders. Is this a sustainable model long-term?
There’s a great deal of use of H.264 quad-tile for contribution already, and, of course, that’s off-the-shelf for us. We can do that six ways to Sunday and just rack and stack. This is here, and it’s acceptable. It’s certainly clunky, and it’s not the best solution for long form, but it meets the requirement for now. There’s the issue of stitching the tiles, but, of course, any one encoder failure can destroy the signal, and you’re tying up four encoders, so it’s less than optimal. Obviously, with redundancy, it means tying up eight encoders.

But the good news is, because it’s off-the-shelf, rack-and-stack hardware, you can use it for a 4K event and then return it for normal HD/SD requirements. It’s not dedicated 4K hardware, which is very hard to justify in the economics of the truck business and the remote-production business. A standalone 4K encoder is going to get only a fistful of dates this calendar year, right? But [it’s useful] having stuff that you can return to normal inventory on one truck or multiples when the event’s over. For the sake of argument, if the Super Bowl or World Series needed 4K, tomorrow, it goes back into regular service. In that respect, it’s economical and flexible.

Have you seen an increase in the use of “at-home–production” models, and how is ATEME enabling this?
The trend is towards reducing the cost of remote production because, if the cost of transport is cheaper than [traveling crew and equipment to the remote], people are going to pull camera feeds back and use dedicated resources at the studio, where they’ve got the expertise without having per diems, hotels, and everything else. So I think it can get cheaper. Wherever there’s good, reliable transport — and, clearly, in North America, we generally have that — where you can pull all the camera feeds, all the angles, even construct the slo-mos, either in the field or back home, the economics are going to be very interesting to a lot of people. We’ve seen pilots of that at the last World Cup and Olympics; we are going to see more of that as a trend, and, clearly, we can support that with high-quality encoding and if we can reduce the bandwidth for that by 40%. A seven-camera shoot can be transmitted inside the bandwidth of, essentially, a four-camera shoot, or a 35-camera shoot can be transmitted inside the bandwidth of a 20-camera shoot. That is when all of this starts to get very compelling to folks.