For decades, building a new broadcast facility has followed the same ritual - specify the hardware, install it, wire it together, and hope it still makes sense in five years. That model is starting to break down, and for national and public broadcasters, replacing it has become one of the most urgent conversations in the industry.
The idea gaining ground is what's known as the Dynamic Media Facility, or DMF, an approach where media workflows are built and changed in software, not hardware. Instead of installing specialised equipment and physically wiring devices together every time something needs to be built, adjusted or torn down, pre-built configuration templates handle the change in software, in a fraction of the time.
Crucially, DMF does not distinguish between private, public or hybrid cloud environments. That's what solves the scaling problem, existing resources can be combined with on-demand capacity to absorb peak loads, infrastructure can be repurposed quickly for different workflows, and resources can be shared across departments, or even between organisations.
Three pressures are converging.
The first is budget. Content budgets are not growing at the same rate as audience expectations, and when broadcasters have to choose between investing in platforms or investing in content, content wins. DMF offers a more granular way to build modern facilities and get more out of existing hardware, without buying new equipment or spinning up a new server for every function.
The second is fragmentation. Audiences now expect more content, produced faster, with fewer people involved. A software-first approach lets broadcasters combine media functions from multiple vendors and orchestrate an entire production through a single, custom interface, sometimes with just one person at the controls.
The third, and the one I think gets underestimated, is resilience. New servers can be added to a cluster with minimal impact on live services. High-value workflows can be spread across clusters in different server rooms to meet the highest availability requirements. The Dynamic Media Facility reference architecture uses six abstract layers to group different functions. The first three layers are generic with the common IT and include physical infrastructure, host platform and container platform. The next three layers are media industry specific and include Media Exchange and Application & User Interface layers. Every layer of the DMF model is built around control, monitoring and security, so the service stays on air.
I was at the EBU Network Technology Seminar in Geneva recently, and there was genuine consensus that this model is needed to move past the hardware constraints broadcasters have been working around for years. It's still early, end-to-end synchronisation across workflow chains, compute resource management, Media Exchange Layer (MXL) cross host communication, there's a list of technical questions still being worked through.
Where views diverge is less about disagreement and more about vantage point. From the broadcaster side, the expectation is that open-source software becomes more important, that more gets built in-house or with partners, and that vendor licensing shifts toward genuinely pay-per-use models. From the vendor side, someone still has to build the software, maintain it, and patch it as new vulnerabilities surface, and that has a cost.
My own experience with pay-as-you-go licensing has been mixed. Some vendors price it sensibly. Others offer flexible terms that only make sense for the first two or three months, after which an annual plan would have worked out cheaper. I think both sides land somewhere in the middle eventually, but it isn't there yet.
If a broadcaster asked me today where to start, my answer wouldn't be technical. It would be: get your goals right first. Technology is the enabler, not the starting point. If the decision is driven by the technology rather than the business outcome, the result won't match expectations.
DMF isn't the right fit for every broadcaster. A media house built around thematic content, or streaming events produced by external partners, probably doesn't need it. But an organisation running newsrooms, studios, and live and remote production, one that wants to share resources across departments and repurpose workflows as needs change, is exactly who this is for. Either way, the prerequisite is well-mapped workflows: without that groundwork, there's nothing solid to build the configuration templates on.
The trade-off is complexity. Flexibility doesn't remove risk, it relocates it. Teams need to understand every layer of the system well enough to fix problems fast, and more importantly, to catch them before they happen while building and testing configuration templates. Get that wrong, and the risk isn't abstract: it's a hiccup or an outage on a live programme.
Strip away the hype, and here's where I think this goes. The DMF model matures through several rounds of refinement, and a lot of today's open questions get resolved. Open-source communities build general workflow templates that others can pick up and adapt, rather than starting from scratch.
Broadcasters end up running compute clusters across multiple, geo-redundant server rooms, with most production workflows spun up when needed and torn down after. And the major public cloud providers start natively supporting the Media Exchange Layer (MXL), making it far easier for professional audio-video production software to exchange media, whether that's on one server or across a cluster.
DMF is, today, mostly a conversation for large national and public broadcasters rather than something core to Levira's own operations. But it's exactly the kind of shift we pay close attention to, because it changes what “infrastructure” means for the broadcasters and media organisations we work alongside. Whatever shape this takes over the next few years, it's less about any single platform and more about how well the pieces, hardware, software, vendors, in-house teams, fit together. That's the conversation we're always happy to have.
To talk through what a dynamic approach to media infrastructure might mean for your organisation, get in touch.