"We’re approaching the tail-end of 2018 and we’ve made some impressive strides this year and indeed this decade. Back in the late 00’s the industry made a huge yet sometimes overlooked shift from dedicated streaming protocols to repurposing the hypertext transfer protocol for video." -Bill Wishon
2018 has been a breakthrough year in the video streaming world. From huge mergers and acquisitions turning the media industry on its head, to record-breaking streaming viewership numbers for major sporting events, to the ripple effect of legalized in-game betting, to the consideration of esports as a form of mainstream entertainment, we’ve seen an incredible progression.
However, with any form of peaks come inevitable valleys. Consumers worldwide are impacted by the growing pains of rapid innovation, experiencing suboptimal streams of live content. It happens even on the most advanced platforms – just look at Amazon, which was put in the hot seat for its U.S. Open feed, falling victim to poor picture quality and functional issues like failure to record and rewind full matches.
Regardless, the conversation should not be a platform blame game – it’s just inevitably what garners the most media attention. Most of the time, streaming issues stem from a less exciting reason – an outdated device or a slow internet connection that are unable to support massive amounts of data flowing through. This frequently occurs when there’s more than one person trying to access the same content from the same location, network or data plan.
We’re approaching the tail-end of 2018 and we’ve made some impressive strides this year and indeed this decade. Back in the late 00’s the industry made a huge yet sometimes overlooked shift from dedicated streaming protocols to repurposing the hypertext transfer protocol for video. This allowed the massive growth we’ve seen in the course of the last decade relying on the scalability of systems built for website delivery to be adapted to streaming VOD and then adapted again for Live. Now, in late 2018, the industry and viewers are remembering a forgotten, but fundamental issue… latency. Today’s “live” video streaming is actually far from the action on the field with a typical stream stuck 30-90 seconds in the past. We’re seeing tangible proof that streaming TV platforms stand equal to cable, but when will live content have that same defining moment and get past this gate? One of the keys to unlocking this gate lies in one piece of technology: adaptive bitrate for real-time streaming.
Democratizing live-streaming for all
Adaptive bitrate provides the best possible quality to each viewer according to their specific device capabilities and network connection conditions. Without it, everyone is forced to watch the same quality regardless of what their device or network would otherwise allow. This means some viewers aren’t able to the video at all while others get a quality level far below what their network and device is capable of. This results in a sub-optimal viewing experience for a large part of the viewing audience.
Adaptive bitrate is like the “Goldilocks” scenario where it fits the individual user’s needs just right for their specific device and network connection. It’s especially important in industries where lag time is not an option, like sports, esports, sports betting and auctions. Right now, adaptive bitrate is common in the video-on-demand world, but in the sub-second real-time space, it’s rare and non-existent at scale. However, the industry is slowly but surely moving away from the single bitrate model, and there are signs that it will eventually become the norm.
How it works
Adaptive bitrate requires multiple components to work seamlessly together. While most of the industry already uses ABR for Live and VOD, the shift from “live” to “real-time” is huge and has a ripple effect through the entire video workflow. It starts with encoding and transcoding where fine-tuning is required to achieve both quality and real-time latency. Seemingly settled areas like how many layers to create and where in the architecture to create them are re-examined under the scrutiny of milliseconds. The impact on video players is perhaps most stark where Live ABR algorithms work at a leisurely pace, having at most one opportunity every four to 10 seconds to make an adaptation decision compared with the frenetic pace of real-time, where the quality of video and network is continuously monitored and decisions can be made every 16-160 milliseconds.
Esports, sports and in-game sports gambling are ushering in the future of real-time streaming. Esports is projected to reach $1.7 billion in revenue by 2021, and research firm Newzoo predicts U.S. media conglomerates like Disney, Comcast, and AT&T will invest in content rights for esports. The YES Network reported 116 percent growth in per-game average for minutes streamed compared to 2017, and a 52 percent jump in the unique per-game audience. The U.S. is on track to become the biggest betting market in the world despite being late to the party, and the NFL could gain over $2B in revenue from in-game betting.
As these areas take the industry by storm and become more mainstream, viewers are becoming more demanding and aware of the level of latency and quality they should be receiving. Especially with real money on the line, no viewer is going to make a wager on the winning touchdown if the quality is so bad they can’t see the ball, or if it turns out the player actually scored 40 seconds ago but the live-stream hasn’t caught up yet. This is where adaptive bitrate comes in, leveling the playing field for all and giving each the best possible experience. It’s a requirement in the real-time market that so many are trying to disrupt yet it’s still surprisingly uncommon, but we’re on track to turn it around.