FAST love: Lawo on the year of absolute elasticity

By Phil Myers, Lawo CTO.

While the broadcast industry may not quite be in turmoil, established outfits are rushing to adapt to new business and consumption models. Fickle viewers, rather than content distributors, rule, and manufacturers need to support them.

For a while, it looked like subscriptions to streaming services were the new norm until it became clear that viewers are subscribing to multiple services simultaneously – but only while it suits them. One-year subscriptions, let alone automatic renewals, seem to be dwindling everywhere. Some are already hinting at the maturity of the streaming subscription model as new customers are offset by viewers that unsubscribe in droves.

The new magic word seems to be FAST, as in Free, Ad-supported Streaming Television, because most viewers under the age of 50 don’t seem to mind commercials if that means the content is free.

There is, in short, a lot going on, on several levels. Sure, public and private broadcasters have ceased to bet on their traditional linear service models and are diversifying their offering. Competing with big tech or global streaming services on content looks trickier by the day, as this requires deep pockets that need to be filled with advertising dollars. Plus, the ones that will likely benefit most from digital advertising may use content production and distribution only as one means to a much larger end.

What still attracts a lot of eyeballs, though, are sporting events, provided the rights situation is kept in check. Luckily, there is a growing market for tier 2 or 3 events that can be served using pop-up or special-interest channels.

Be prepared

While the production of fictional content is increasingly farmed out by private and public broadcasters, the production tools for news shows and live events are still very much needed. Yet, in the face of stagnant or shrinking budgets, predicting how much of each functionality will be required in the foreseeable future is becoming tricky.
Purchasing too much production hardware dents your spending power and will be frowned upon by your financial controlling department when a substantial part of what your company purchased sits idle for most of the time.

That is precisely why everybody is rushing to the public cloud, right? Are they really? According to the EBU and other trusted sources, the public cloud may have lost some of its lustre. There are several reasons for this: latency issues still turn live coverage workflows into mild nightmares; you need a (reliable) network connection, which may not be available where the action is; and while getting content into the public cloud is comparatively affordable, getting it out again tends to be cost-prohibitive.

Mix and match for lift-off

Broadcasters and content producers need the flexibility that a software-based approach provides. Software does away with heavy investments in bespoke, single-function hardware for both cruise-speed and peak-load requirements. Yet, real flexibility should mean that the software-based processing functionality runs exactly where operators need it, at short notice, on demand.

This may be on-premise where everything is in place to keep the content securely in-house, in a central data centre (a so-called private cloud), or in the public cloud. If operators want to combine two or all of these usage models, this should also be possible. Anything outside the public cloud can run on standard servers that users can purchase directly from a trusted manufacturer. Dynamically mixing and matching compute capacity between usage models sits at the very heart of the proposed approach.

Is this the same old emperor (processing capability) in new clothes (software running on generic rather than bespoke machines)? No, because server-based compute power is function-agnostic. If the infrastructure management system allows for elastic allocation of processing tasks to pooled compute resources, the net result is a significantly higher utilisation of the hardware. Now couple that with a large number of choices for the licences that operators can purchase.

Someone who only needs two multiviewer heads with four PiPs each can use them without paying for bundled processing capability that they simply do not need. Larger operations, on the other hand, can select incarnations of the same multiviewer functionality with 16 or even more PiPs, and they can instantiate as many multiviewers as they need.

Although already flexible, this is still not quite the dynamic solution operators expect for a swift response to their changing requirements. Support for different protocols, such as ST2110, SRT and NDI, as well as compression formats (JPEG XS, H.26x, NDI, etc) needs to be baked into the software offering, as does the ability to go from one to the other at any processing stage, or to serve a variety of versions based on a single incoming stream.

We hear you

The functional elasticity is complemented with a commercial model that puts users first: those who wish to know what they’re in for are welcome to purchase perpetual licences. Users who need to remain agile in the face of changing consumption models, on the other hand, will prefer the ability to take out function-agnostic subscriptions based on credits that can be freely allocated to any – present and future – app functionality at any time and in any way they see fit. They can use a UDX processing app for conversion tasks, for instance, followed by a stream transcoder, then a multiviewer, and so on, within a single day.

The processing toolkit thus becomes as elastic as current viewing demand. The ability to include add-ons for the various video and audio microservices means that users can tailor their virtual infrastructure to their real-life needs at short notice, on a budget that can remain stable despite the almost endless combinations afforded by the credit-based subscription model.

The bottom line is that broadcasters no longer need to plan for the next five or ten years. Whenever they need more processing capability, they can take out additional subscriptions, perhaps only for the duration of an upcoming event. Everything can be charged to the operating expense account. How’s that for something to look forward to in 2024?