YouTube announced yesterday that it would now be supporting live streaming in 4K at up to 60 frames per second (fps). This is aimed at the growing market for live streaming on social networks. (See Social Video Could Be the Next Great Challenge for Network Operators.)
YouTube Inc. has supported 4K video for several years, but has now extended that to live streaming, for both standard videos and 360-degree video. Other social streaming platforms such as Facebook and Twitter's Periscope only support streams up to 720p at 30fps.
YouTube's "Live" channel competes directly with these services but was launched following the initial success of apps like Periscope, Facebook Live and Snapchat Live Stories. However, live streaming on social networks is expected to grow rapidly, with Facebook CEO Mark Zuckerberg famously predicting most of Facebook would be video before 2020.
Given the initial success of user-generated content (UGC) -- largely driven by YouTube -- it seems likely that live streaming will take off. I'm not sure, however, that rolling out 4K support is going to have much of an impact today.
Experts tell us that to really see the benefits of 4K, you need streams of about 20 Mbit/s to 25 Mbit/s, especially for live streams. There is more time, and therefore more potential for tweaking the various tools and settings, to further improve video compression for on-demand video. But for live, going under 20 Mbit/s will usually affect video quality. And obviously there's no point streaming low-quality, compromised video and calling it 4K.
Just as a comparison, according to Netflix's ISP Sped index, most streams are being delivered at between 3 Mbit/s to 4 Mbit/s. So while some homes in developed markets may have 50Mbit/s or 100Mbit/s connections, the average connection speed is far closer to 4 Mbit/s or less, even in those countries.
Also, most tests have found that 4K quality is difficult to discern on screens under 55-60 inches. Given that the bulk of demand for social live streaming is likely to come from smartphones, 4K streams then may not be particularly meaningful. I should note that this is for standard video; the 360-degree video experience will likely benefit from higher bit-rates.
There's also the issue of network capacity. Bandwidth is limited on telecom networks, despite regular network upgrades. How will the network handle high-bandwidth video streams, and what will that mean for end users? In previous Heavy Reading research, buffering delays and rebuffering interruptions were rated the greatest drawbacks to watching Internet video.
An additional challenge is that video will be shot and uploaded mostly via mobile devices, often on mobile networks. That will affect not just the downstream bandwidth, but also upstream -- which is still usually more limited than downstream capacity.
So why do it? I can come up with four reasons, but they are all a bit speculative.
Firstly, YouTube is a little late to the market and needs to catch up before the other players run away with what could be a huge market. Supporting 4K streams gives it a selling point, a differentiator.
Secondly, Google (Nasdaq: GOOG) is developing technology on a number of fronts, and if it has developed the capability to offer 4K live streaming, why not roll it out?
Third, YouTube also offers live streamed coverage of events, and partners with various content providers. So it's not just aimed purely at UGC, and offering 4K streams could be more significant in these other use cases.
Lastly, live streaming is often better suited to compression because it's mostly "talking heads," i.e., a person talking into the camera without a lot of visual information or high-speed action in the frame. This can be compressed more effectively than sports or action movies. So YouTube should be able to bring bit-rates down lower than providers of other 4K services.
Fundamentally though, I think it's primarily an opportunity to differentiate its product and underscore the company's technology expertise in a potentially important market.
— Aditya Kishore, Practice Leader, Video Transformation, Telco Transformation