The Reality of Virtual Reality: Despite the Hype, VR Isn't Viable

It seems like you can’t browse the web or read a magazine without something popping up about VR. Sure, there’s a lot of hype around the technology (probably as much as, or maybe even more than, the hype around mobile in the mid-1990s), but consumer adoption is relatively low. Sales of VR headsets number in the low millions (according to UploadVR, 6.3M sold in 2016), barely a drop in the bucket compared with the proliferation and market penetration of smartphones and their ilk.

Gaming is probably the most viable use case for VR, but consuming video content is right up there. The promise of immersive 360° video experiences is crucial to VR’s future growth. But what’s the reality of providing those VR 360° videos right now?

Before I answer that question, let’s get a few things out of the way. First, there’s a difference between run-of-the-mill 360° video and its immersive, VR cousin. Lots of the former is showing up in places like Facebook, where users can pan around a video using a mouse and keyboard. It’s the latter that holds all the promise, delivered using head-mounted displays (HMDs) and, oftentimes, a mobile phone.

Second is the quality required to deliver an immersive VR video experience. According to Josh Courtney from Sky VR (in a keynote presentation he gave at Streaming Media East in 2016), delivering a quality VR experience with 360° video requires 4K per eye. That’s 8K content. Third is the bandwidth needed. To stream 8K content encoded in H.264 (still the predominant encoding format), a viewer would probably need about 50 Mbps of throughput. Finally, there is latency. As viewers move, their heads to look at different perspectives in a VR video experience, any lag in rendering that content can potentially induce nausea.

As you can probably already surmise, the current outlook for delivering a top-quality VR video isn’t that great today. Not only do we not have the base of high-efficiency encoding (through codecs like HEVC, VP9, or AV1), but the bandwidth for delivering it isn’t there either, especially considering that the most popular consumption means would be through a mobile phone and an HMD like the Samsung Gear VR. Imagine trying to stream 50 Mbps over an LTE connection! The lack of bandwidth coupled with high throughput requirements lead to only one end—latency in the viewing experience.

Thankfully, lots of companies are working to solve the problems. Not only is Facebook (which acquired Oculus several years ago) pioneering new methods for immersive VR rendering, but Apple recently announced support of HEVC for its devices (perhaps a precursor to a new iPhone with VR capabilities). Other companies, like Beamr and Harmonic, are making it easier to provide high-quality bitrates with lower throughput requirements. The confluence of a number of different industries working on solutions to some of the critical problems associated with VR video today—quality, bandwidth, and latency—promises to accelerate the adoption and availability of VR video. As those challenges are overcome, and as the cost to produce immersive VR video content continues to drop, it will only spur premium content production and, as a result of there being more content available, consumer adoption as well.

So, to answer my initial question, VR video isn’t really viable today. The reality is that premium content owners and other content creators are experimenting with the medium, but there can’t be an expectation of ROI at this time. Poor video quality is already a leading contributor to subscriber churn. Imagine how consumers would feel if they were paying for VR video content and the experience was laggy and pixilated, and sucked up all their bandwidth?

There’s no doubt that VR video experiences will improve over the next few years—bandwidth requirements will lessen, content will be more immersive, and computational rendering power (especially on smartphones) will continue to climb. But if I were a content owner thinking about producing VR video? I might experiment a little to make sure my workflows were capable of supporting it, but I’d probably spend my money elsewhere.

[This article appears in the September 2017 issue of Streaming Media Magazine as "The Reality of VR and Video."]