The Shift to 4K Video: What Does it Mean for the Network?
Consumers are demanding 4K, so now it's up to operators to provide support. But how long will it take for everyone in the pipeline to get 4K-ready?
To really understand the emergence of 4K we need to look at it from several angles.
Objectively, and recursively, 4K is an inexorable inevitability; where it succeeds, it will be a success. We can already see 4K in production in a number of integrated vertical stacks, from broadcast to OTT. It is also a badge of honour for a consumer-demanded upsell on video technology. It will soon be “too expensive” to produce a non-4K TV or smartphone, so its ubiquity is inevitable. And that’s before we even ask, “Why bother?” as a network operator, publisher, or technology provider. It just is.
OK, so 4K is on a wave where it sells itself. Let us now look at the effect of 4K’s “inexorable inevitability” from the industry perspective. This should include production, broadcast distribution, IP distribution, macrotrends in infrastructure and software, and the commercial impacts.
First, let’s look at production. Capture technology, while abundant, tends to be depreciated relatively quickly, and it has a relatively rapid tech refresh cycle. Even the smartphone has 4K capability. Because news networks cannot be seen to have inferior technology to citizen journalists (or else what’s the point?), this pushes broadcasters to update their source acquisition to 4K. Even the production workflows are now concentrating on at least replacing their broadcast cabling with “data cable,” such as NewTek’s NDI.
Software like the kind used in video encoders and edit suites usually requires nominal changes to handle 4K. On top of this, a number of vendors have implemented proprietary “transmission control” protocols to ensure that IP networks can deliver high-throughput video with low latency and high reliability, ensuring that contribution from the production desk to the distribution aggregation point is of high quality.
Given the rapid technology refresh cycle and low incremental overhead, 4K brings additional benefits over time (cropping into 4K to make 1080p while maintaining 1080p resolution is a good example).
Now let’s jump to the aspect of “traditional” broadcast distribution.
Again, there is a lot of legacy equipment. Along with the equipment, there is also a ton of SD and HD content that needs to be made available. There is no debate over dropping one cost and migrating focus entirely to 4K— 4K is an incremental cost of business. Yet, the value proposition is a relatively small incremental one that makes for a fourfold (roughly) increase in network cost.
Another complexity for broadcasters is learning how to run networks that can evolve quickly and keep up with the market. The traditional networks used to have closed networks, and they bore the sole responsibility for dictating what happened and how they met their subscribers’ demand. The traditional broadcast industry has been accustomed to a combination of monopolies of technical and rights relationships. Today’s technology market is increasingly market-led. Consumers are demanding 4K by default.
Broadcasters have little or no concept of service velocity beyond, “Let’s start with a whiteboard.” Their ability to roll out network upgrades or add entirely new services to their network is measured in years, if not decades, and involves rolling a lot of trucks. 4K is, at the very least, complex for the traditional broadcasters to consider.
In contrast, our next perspective should be from our world of streaming.
As we all know, legacy technology in the IP space churns ridiculously quickly. The tech-refresh cycle is arguably measured in months, and certainly single digits of years. With IP as the lingua franca, upgrading parts of the network has become a swap-out job. This has created a massively competitive market, so pricing and vendor lock-in are things of the past.
As they refresh, telcos and hybrid fibre-coaxial network operators have quickly moved to IP. They’re focused on providing stable layer 3 IP on top of their own layer 2 (and lower) network infrastructures. They have been watching closely as different types of application layer 4 service providers become valuable users of their networks. Obviously, ISP services have been at the heart of this.
Additionally, telcos have often developed IPTV services on-net, where a premium is leveraged against the guarantee of service that on-net delivery brings. Deployment of this type of service has taken some learning for the telcos. However, they have understood how to deliver in the application space down to a production level, and now they are evolving the deployment of those services into multitenant operator class services.
The telco operators’ combination of subscribers and the ability to control the quality of service on the last mile network are of equal value to rightsholders as are the cable and satellite broadcasters’ networks. Some offer IPTV and subscriber management as a combined service. Others are exploring repositioning to offer “operator CDN” services. Just look at Netflix.
As for 4K, the great thing about an IP network is that increasing the bitrate to support 4K is really just a matter of dialling it up. Yes, of course there is some layer 2/layer 1 network provisioning to manage. But as techniques such as software-defined networking enable operators to do this on-the-fly, there is no doubt that IP distribution has a service velocity that traditional broadcast networks can only dream of.
From a technology point of view, graphics processing units (GPUs) are now appearing as a default part of the commercial off-the-shelf (COTS) and IP physical infrastructure environment. As COTS increasingly includes GPU, high-bandwidth video processing applications can increasingly be run anywhere. This ability to sometimes move your transcoding capability to the content can mean that the need to design workflows around locations where we can transport the big video content is changing.
At scale, such changes can make huge differences to network architectures. This means that software trends are also favourable for 4K’s increased adoption.
Before we finish, we should look at the commercial perspective. Yes, 4K technologies are going to incur some cost. No, we can’t really say that this super-high definition brings a massive amount of value to the consumer. However, the consumer is gradually adopting it by default.
So we have to bear the cost even if it’s just to stay in the game. For that reason, we want it to be quick and cheap to deploy and, ideally, to deploy it responsively where and when the market share matters.
IP- and COTS-based technology providers are able to meet this demand from a competitive/innovative supplier base, and they can do so increasingly quickly. Those telcos that can deploy such services are well positioned and often come with subscriber relationships too, indicating the effort can be monetised over time.
Of all the telcos, mobile networks arguably have the greatest agility and enough competition to drive innovation (unlike the incumbent telcos that are more complex and less dynamic in this regard). Mobile networks even have capabilities to multicast on their last mile (see the evolved multimedia broadcast multicast services on Google). Therefore, it is very possible that mobile network operators may evolve premium content offerings in much the same way that the wired networks have. Yet, we are typically not watching them while we think about 4K, because we think of small form factor phones and we ask ourselves, “Why would a mobile operator offering 4K be interesting on a small screen?”
But mobile operators are more focused on selling SIM cards and subscriptions than phones. Therefore, there is no reason (for example) that the SIM card wouldn’t appear in the home gateway or the smart TV, or that the smartphone itself wouldn’t increasingly become a home gateway that services the smart TV (think Apple’s AirPlay). It is possible that the mobile network operators will end up bringing 4K premium content to the consumer in a more innovative way than the fixed line operators have had to. All of this may happen long before the broadcast networks have even worked out that there is change going on, resulting in their eventual decline.
Stranger things have happened.
Finally, it is also interesting that throughout my musing here, you could broadly replace “4K” with “DASH” or “UHD” or any other such capability ... but that is just an afterthought.
[This column appears in the Winter 2016 issue of Streaming Media European Edition as "4K From All the Angles."]
While the production pieces are all in place for 4K sport and other content, delivery networks are going to have to embrace UDP, multicast, and client-assisted delivery before viewing becomes commonplace
The future of UHD is happening right now. Here's how one production company streamed low-bandwidth 4K live from a South by Southwest music stage.
At least that's what both TV and OTT video service providers believe, according to a study from SNL Kagan and Irdeto