NEW YORK—With video traffic on pace to make up more than 80 percent of all internet traffic within two years, the tech industry is working on ways to move the bits more efficiently and increase viewer engagement in the process.
At the Streaming Media East conference here this week, Facebook said it is starting to see how much its Live video, first rolled out in 2016, is increasing engagement with viewers.
Facebook Live has enabled more user interactions—clicks, likes or comments—and two times more time spent with content than non-Live content, said Erin Connolly, Facebook Live product manager, who was the keynote speaker on Tuesday.
The Facebook Live API, which allows publishers and content providers to embed their streams into Live, resulted in a 1.5 times increase in views on publisher pages, Connolly said. "Live is a tool to further our mission of building community and bringing the world closer together," she added.
Facebook is working to further that interaction with three new features. Facebook is adding a persistent stream key, which will enable users to post video with fewer steps than before, and is introducing Live Crossposting, which will enable publishers to stream to multiple Facebook pages at the same time.
The social media network is also testing a feature called Live Rewind, which will give users the ability to rewind a Live video while it is streaming.
Other new things in the works include Watch Party, which enables users to invite other users to view the same video and share comments in a chat. Facebook Premieres will allow users to show or schedule prerecorded content as a "Live" moment, she said.
Moving the Bits
The newly minted AV1 codec for compressing video from the Alliance for Open Media was also on display at Streaming Media East. After two years of development and anticipation, in March the Alliance released the AOMedia Video Codec 1.0 (AV1) specification, which sets a new bar in the ongoing effort to squeeze more data into video streams.
The new codec is backed by the biggest names in media, including Amazon, Apple, Google, Facebook, Microsoft and Netflix. Despite its promise, however, AV1 comes with high overhead in technology, time and cost, said consultant Jan Ozer, who spoke here. It's also not "quite" ready, he said, and it faces another two years of ramp time before it gets baked into hardware.
While AV1 promises to enable 4K streams over low-bandwidth connections, it comes at a cost. AV1 is "thousands" of times slower than the current benchmark codec, known as H.264, and AV1 will cost more to implement on a per-stream basis, Ozer explained in a presentation. That means that it's really only for cloud-scale providers with lots of compute power, such as YouTube and Netflix.
Some vendors are already working on their own implementations of AV1 and seeing improvements over the base specification. Bitmovin, based in Austria, has been working on AV1 from its early stages and has been able to get that performance hit down to only five times, said Christopher Mueller, CTO and co-founder of Bitmovin, in an interview.
The real benefit, he explained, is that AV1 is open source and free to use, compared with other specs or proprietary technology, which saves money for content providers and creates an ecosystem of ISVs who can continue to hone the platform. "Every bit counts," Mueller said.
The Next Generation
Advanced compression technology and field-programmable gate array (FPGA) hardware are enabling ubiquitous compression of data streams, and FPGA vendor Xilinx, working with partners, demonstrated ways to get the most out of codec acceleration at the event. The company showed a scenario using V-Nova's PERSEUS 2 codec, which can stream a 4K file with the same quality at half the bandwidth of High Efficiency Video Coding (HEVC).
Also, in something out of "Minority Report" or "Iron Man," Xilinx partner Skreens demonstrated its software-as-a-service (SaaS)-based product that functions as a "visual engine" that can turn any connected screen into a personalized viewing space.
Using a phone app or touch screen, users can open, close, expand or zoom in on any video window, as well as also run multiple video streams on different parts of the screen. The technology has use cases beyond the cool features, officials said here, including the internet of things and smart home design.
On top of all this, the use of artificial intelligence is starting to enter the picture among providers, which is helping them manage and categorize the mass of video being created every day. Said Jason Hofmann, vice president of architecture at video hosting company Limelight Networks: “You need to be 20 to 30 percent more efficient every year just to keep your head above water.”
Scot Petersen is a technology analyst at Ziff Brothers Investments, a private investment firm. He has an extensive background in the technology field. Prior to joining Ziff Brothers, Scot was the editorial director, Business Applications & Architecture, at TechTarget. Before that, he was the director, Editorial Operations, at Ziff Davis Enterprise. While at Ziff Davis Media, he was a writer and editor at eWEEK. No investment advice is offered in his blog. All duties are disclaimed. Scot works for a private investment firm, which may at any time invest in companies whose products are discussed in this blog, and no disclosure of securities transactions will be made.