Google Outlines Crucial HTML5 Video Development During Keynote
In an opening day keynote to the Streaming Media East conference in New York City, Matt Frost, senior business product manager for Google Chrome, guided the audience through the past and the possible future of online video.
"Video at Google started as an experiment," said Frost, noting how a single fulltime Google employee worked on video in 2004, along with a few others giving 20 percent of their time. Only two years later, Google was a video giant when it acquired YouTube.
"There is no single standard for video on the web," said Frost, explaining the challenges for online content distributors. While a single format would be beneficial, there currently isn't one that works well with all devices, he said.
What's needed is a solution that lets content owners encode once for all destinations. YouTube is popular even with enterprises, Frost said, because it removes the burden of formats, handling all transcoding in the background.
Google's approach to HTML5 solves many of the problems facing content owners today, Frost asserted, explaining how the Chrome team is devoted not only to making sure HTML5 video works well in Chrome, but also in other browsers. Its goal is to improve HTML5 video so that it can fulfill the same role that only Flash plays today.
Frost listed key areas where HTML5 needs improving: content protection, adaptive streaming, and captioning. The first two are especially important to premium content owners, he said, and Google is making progress on all three.
For content protection, Google, Microsoft, and Netflix recently made a proposal to W3C suggesting encrypted media extensions in HTML5. The proposed solution could be achieved with only a minor change to the HTML5 spec, Frost said. "It is a very promising initiative that should result in content protection appearing in HTML5 browsers soon," he added.
For adaptive streaming, Google has evaluated all the possible formats and has thrown its weight behind MPEG DASH, Frost said. Google has published a media source extension proposal at W3C.
Looking at the work that remains to be done with HTML5 video, Frost said that native browser support for video and audio isn't enough; the core technologies must be modern and keep pace with the web. The goal isn't just media playback, but turning the browser into an application platform and social platform. The developer community can't build cutting-edge experiences with ten-year-old technology, he added.
Today's viewers put up with current online video playback because they remember the "bad old days" of postage stamp-sized video and proprietary streams, Frost said, but there's much yet to be done.
"Web video can be so much better," Frost said.
Frost finished by detailing four key areas of investment for web video: authoring (the ability to easily and quickly encode once for all devices), distribution (flexible for any user's environment, with monetization options), playback (fast access with low power consumption), and innovation (an eye to future compatibility).