TranSPORT: Consumer Demand Drives Streaming Platforms, Format Adoption
If this year’s Super Bowl and FIFA World Cup are any indication, the live streaming of marquee sports events is integral to broadcasters’ workflow. In fact, for an increasing number of fans — particularly the younger demographic — streaming sports events to desktops, tablets, and mobile devices supersedes watching on a traditional television.
The second screen, in many cases, has become the first screen, but streaming to this screen is hardly one-size-fits-all. Content creators must contend with a range of device capabilities, streaming formats, and consumer demands.
“One of the biggest challenges that people have is consistently delivering to things that are a couple years old,” said Rick Wisher, Adobe product manager, Primetime, at last week’s TranSPORT event in New York City. “If you take older Android models or older iPad models, they tend to not be able to handle the same bitrates as some of the later stuff.” He noted the challenge of “putting together the profile sets that actually work well across all those different devices [and] tweaking them as things move forward [so that] you can actually start to deliver higher bitrates.”
Building out the streaming infrastructure presents another challenge. Marquee events like the Super Bowl and FIFA World Cup drew streaming audiences in the millions who expected a broadcast-quality experience. And, in some cases, the streaming product did not live up to expectations.
“The Super Bowl … was the most widely watched sporting event, at least the first hour. The reality is that it also had the highest abandonment rate of any given event,” said Allan Konar, sales engineer, Qwilt. “Why? As soon as [they] get buffer, people decide that they’re going to go elsewhere. They’re going to do something else. … There’s a certain level of service that we’re used to.”
However, an audience exceeding 1 million is hardly the norm. Networks must decide how best to build their streaming infrastructure: whether to accommodate the maximum audience and lie dormant during the rest of the year or to accommodate the average audience and scale up for the marquee events (and hopefully avoid buffering).
“A lot of it just really has to do with scale, particularly when you’re talking about streaming live events where you go from nearly zero end points to hundreds of thousands and millions, [and] back down to nearly zero,” said Jason Warman, senior sales engineer, Aspera. “I think part of designing these types of environments is being able to have some modularity around that so you can plug in and have transcode if you’re not set on a particular universal format. Or you can plug in a big transcode farm that’s halfway around the world if you happen to have a single event that requires that kind of scale.”
The variety of streaming formats factored heavily into the discussion, with panelists choosing sides on MPEG-DASH, an adaptive-bitrate format used for streaming multimedia over the Internet. For Wisher, DASH represents an opportunity to provide customers with a universal format.
“It allows you to form better business relationships because the ecosystem is much easier to compete in,” said Wisher. “By and large, there’s a spec that anybody can develop against and nobody controls.”
While DASH may be a less proprietary, more accessible streaming format and an alternative to HLS, several vendors reported that they are not in any rush to adopt it. According to Verizon, the company’s Universal Streaming platform addresses the multiplatform issue and saves customers from having to worry about underlying formats, making the need to adopt MPEG-DASH less pressing.
“We haven’t seen any demand from our live-event customers for it,” said Paul Heitlinger, product manager, Live Streaming, Verizon Digital Media Services. “In my opinion, we solved that multiformat problem. DASH is part of the LTE multicast spec, so we’ll support it because we have to for the wireless side. But, from a universal streaming or video-unicast-streaming side, if it becomes so popular that everybody’s demanding it, then, sure, we’d support it at some stage. But, as of today, nobody’s come to us and said we want MPEG-DASH.”