That’s Heavy, Man … 4K Over OTT

By now, two trends are clear when it comes to the momentum around 4K/Ultra HD video. One is that Ultra HD is far more of an Internet-protocol (IP) thing than traditional broadcast video methods.

Samsung and Comcast, for instance, etched a 4K/UHD deal at the 2014 International CES. Those sets are in production, and the Comcast 4K app on them will carry a mix of NBC and Hulu titles (“more to come”), over the broadband input — not the cable input.

The industrial transition to “all-IP” — meaning distribution networks that carry stuff over spectrum apportioned for cable modem/broadband delivery, as opposed to traditional set-top box/broadcast delivery — is a continuum, with different service providers at different points.

(It wasn’t all that long ago that any cable chief technology officer, when queried about when the network will be “all-IP,” answered with some variation of “not in my lifetime.” Nowadays, actual date targets exist: By 2020, by 2016 and so on.)


The second clear trend, as it relates to 4K/Ultra HD, is that it’s the undisputed darling of the over-the-top (OTT) video community. Amazon Studios is shooting its all of its original content in 4K; Netflix is out with an $11.99-per-month package for Ultra HD access. YouTube, UltraFlix, Sony and M-Go, among others, are dabbling in 4K.

The OTT community is hurtling forward with 4K content, ahead of their pay TV competitors, because — well, because they can. Meaning they don’t have to deal with anything beyond making the content, and receiving money from consumers to view it.

They don’t have to do any of the work of getting video to people, for instance — compressing, modulating, multiplexing, combining and then reversing the whole process, at the receiving end, in the home.

That’s why they’re called “over the top” in the first place, for heaven’s sake. They plunk their wares on top of the broadband networks built by others.

High-volume harrumphing tends to accompany this conversation. Why is raw broadband consumption up 50% or more, year over year, since 2009? Ahem. When did Netflix start streaming, and stop mailing DVDs? If you guessed “2009,” you’re a winner! Take the rest of the day off!

Here’s why the advancements of the OTT video community into 4K/Ultra HD video make broadband network engineers want to smash their foreheads into their desks: It’s huge. Video is already the biggest and bulkiest passenger on the Internet, but 4K is to HD as Fat Albert is to Stuart Little.

Here’s what Netflix recommends: Bandwidth of 3 Megabits per second for standard-definition (SD) television, and 5 Mbps for high-definition TV.

But for Ultra HD, Netflix recommends a beefy 25 Mbps. As in five times more than the “weight” of HD, which is already pinching broadband carrying capacities, especially during prime viewing hours. (On any given evening, on average, Netflix represents about 60% of all Internet traffic.)

It’s worth pointing out that earlier this year, the Federal Communications Commission changed the definition of “broadband” from 4 Mbps to 25 Mbps. Never mind that the move essentially booted digital subscriber line (DSL) from the ranks of “broadband.” Never mind that the exclusion of DSL as “broadband” upends the aerial view of the marketplace.

It’s also worth pointing out that 25 Mbps is but a fraction of a Gigabit per second (Gbps) — a Gig of anything is 1,000 Megs of the same thing. Yet, getting to a Gig in delivery speeds is the new target for service providers, whether or not it’s needed.

What’s a network engineer to do? Two things, for starters. Happily, both are already underway in the engine rooms of the industry we used to call “cable.”

One is more IP capacity, which is coming (and likely just in the nick of time) with the next chapter in cable modems. It’s known in tech circles as the acutely nerdy “DOCSIS 3.1,” where “DOCSIS” is pronounced like a word — “dock-sis” — and stands for Data Over Cable Service Interface Specification.

DOCSIS 3.1 promises to double downstream and upstream broadband speeds, without the buzzkill that is the “forklift upgrade” — which tends to make finance and operations people batty.


Two is the latest in video compression, known as HEVC, for “High Efficiency Video Coding.” HEVC improves upon H.264 (which also goes by “AVC” and “MPEG-4”, which improved on MPEG-2, the granddaddy of digital video compression, dating back to the first digital set-tops, in the mid-’90s.

Each new compression chapter, HEVC included, doubles the efficiency of its prior version. HEVC is two times better than H.264/MPEG-4. H.264 is two times better than MPEG-2. That makes HEVC four times better at squishing than what’s inside millions of early-generation set-tops, installed in homes.

Bottom line: We’ll likely see more (or at least hear more buzz) about Ultra HD from the OTT camp than from traditional broadcasters, for now. That it’s big and bulky matters little to an industry that doesn’t pay for freight.

The good news is that innovation is ubiquitous these days, and the double balm of DOCSIS 3.1, for capacity, and HEVC, for compression, will make the hurt more like a stubbed toe than a broken ankle.

Our advice? Find and be kind to all network engineers responsible for keeping up with capacity. They need some serious love. Being the poor soul constantly asking for budget to buy more capacity stuff — when the only answer to “why do we need it?” is “to keep up with OTT” — is hard for everyone involved.

This article first appeared in the Next TV section of the July 6 issue of Multichannel News. For more translations, visit Leslie Ellis at