Reed Hastings, Netflix, and the Future of Streaming Video
First reported by Multichannel News, a recent interview at the Copenhagen Future of TV Conference had Netflix CEO Reed Hastings promising that consumers won’t need more than about 15 Mbps to stream 4K video files once the content is up and running on the site, which is slated to happen within the next two years. That means those who get regular data speeds of around 50 Mbps already have the means to stream 4K content into their homes. You can view the entire interview below.
NscreenMedia curated some highlights from the talk:
On 4K Video
In this interview, Mr. Hasting states (23 minutes) that he views 4K video in the home as an evolutionary process, not revolutionary. He thinks that 4K will show up first on PCs and tablets, as these devices will acquire screens at 4K resolution and find their way into consumers’ hands before 4K flat panel TVs. Today, Apple’s Retina displays come in resolutions at about half the horizontal resolution of 4K, so it seems likely in a year or so that tablets and PCs will commonly have displays at the required resolution. Further into the future, Hastings sees the price of 4K TVs coming down sufficiently that people will buy them as part of the normal TV replacement process.
Mr. Hastings also restated that Netflix will start streaming in 4K resolution and said it would likely happen next year. The company expects bandwidth requirements for 4K streams to be about 15 Mbps. This is considerably above the average bandwidth of broadband connections in the US today, though Verizon FiOS and Google fiber customers likely will be able to stream consistently at that speed. That being said, the penetration of 4K displays by this time next year is liable to be restricted to the videophile market.
No change in direction
Those looking for Netflix to break out of its current business model will find no solace here. Mr. Hastings restated the company’s commitment to the current business of all you-can-eat TV and movie streaming with no commercials (16:50 minutes.) He also says the company is focused on improving content in all the existing regions (0:30 minutes.) The company will continue to plow money from a growing subscriber base back into licensing more content (15:40 minutes), which, in turn, will help attract more subscribers — the classic virtuous circle.
The Future of TV
On the future of TV (8 minutes), Mr. Hastings restated his belief that TV would become more like an app. He went further, predicting:
“In the next 10 or 20 years, almost all video watched is gonna be Internet video and it’s gonna be on-demand.”
He also held the BBC up as model for the evolution of broadcasters (10:30 minutes) saying that all broadcasters will go over the Internet. He did hedge a little, though, saying broadcast would still be around for a “very long time.”
As someone who has spent 25 years at the bleeding edge of technology with the last 12 dedicated solely to video compression, my honest initial reaction was: Wait, what did he say? To be clear, I am not doubting nor debating that 4K is coming or that Hastings (who is rumored to be a candidate for the Microsoft CEO job) is a visionary. I simply want some someone to bring the 4K conversation back to reality. For those who want to understand what 4K is in terms of size I suggest you read a great piece from Tim Siglin called What is 2K and 4K video?
In the end, no media headlines can change some simple realities about the current state of video compression:
1) In order to achieve the compression necessary to stream 4K, either the picture quality will be degraded or a new compression standard will be needed. HEVC is still new, and the results are not fully settled yet on how it will perform on general video. It is claimed that HEVC is twice as efficient as H.264 (the best currently-fielded video codec), but 4K video frames are four times as large as 1080p frames. Even granting the claimed HEVC performance (no sure thing), this makes Hastings’ claim that 4K videos will be able to be streamed at 15 Mbps (a bitrate where H.264 struggles with some current 1080p videos) dubious at best. Additionally, a new encoding standard means new decoders on the back-end.
2) As Netflix is one of the big users of the Internet, adding the 4K ability will clog the Internet even more. Although the infrastructure is in place to handle a few 4K video streams, what about thousands or millions streams? Will this be the last straw that brings the Internet to its knees?
3) As device manufacturers create the ability to shoot and view in Ultra 4K, they still will need to address how to handle the size of these videos. More storage on devices is needed, and as you can see from this article, storage isn’t the only thing affected by the increase in resolution; battery life will be limited to 2 hours.
The average consumer will not be happy with the short battery life of their tablet when streaming 4K videos, and using today’s typical data plan to stream 4K videos is probably out of the question. Technology may develop in the future to enable smooth streaming of 4K video (e.g., adoption and high performance of HEVC, more capable devices, higher-capacity broadband networks), but I question whether the technology of today is ready for it. When it comes to 4K video, my belief is that video deliverytechnology is going to be challenged to deliver on the vision of this videoviewing technology. I look forward to seeing what innovations will come next to help make it a mass market reality.