Back to Insights
Blog

Beyond Standards: Making Hybrid Broadcast Work

Ian WagdinVice President of Technology and Innovation9 min read
Making Hybrid Broadcast Work
Key takeaways
  • Standards compliance alone doesn’t guarantee interoperability in hybrid broadcast. Two manufacturers can implement the same specification correctly while making different decisions about optional functionality, such as metadata handling.
  • Hardware still wins for deterministic workloads. It offers predictable latency, high processing density, and efficient power consumption when a broadcaster knows its fixed channel count.
  • Software and cloud suit unpredictable demand. Capacity for a major event’s supplementary feeds can be deployed when needed and released afterwards.
  • DMF and MXL are moving the industry from transport to compute. DMF provides a reference architecture for software-based media environments; MXL defines how applications exchange media efficiently within it.
  • Timing can’t be assumed to survive every network boundary. Media may need reclocking as it enters a new compute environment, with timing preserved through the rest of the processing chain.
  • Shared compute introduces resource contention that hardware avoids. Applications from different manufacturers on the same node compete for CPU, GPU, memory, and network capacity.
  • Technical partnerships make the foundations dependable. Standards only work reliably once manufacturers implement and validate them together.
Key terms

DMF (Dynamic Media Facility) — A reference architecture, developed through the EBU, for software-based media environments covering infrastructure, orchestration, workflow management, and media applications.

MXL (Media eXchange Layer) — A common approach to exchanging media between software applications within a compute environment, without repeatedly converting content into traditional transport formats.

JT-DMF — The Joint Task Force on Dynamic Media Facilities, examining requirements including timing, resource management, orchestration, and the commercial implications of software-based workflows.

PTP (Precision Time Protocol) — The protocol that provides a common timing reference across connected infrastructure within a conventional ST 2110 facility.

ST 2110 — The SMPTE standard for transporting uncompressed media over managed IP networks.

SRT — Secure Reliable Transport, a protocol that helps address less predictable networks, with buffering and latency trade-offs for protecting against packet loss.

RIST — Reliable Internet Stream Transport, a protocol addressing similar challenges to SRT on unpredictable networks.

Hybrid production has changed the way broadcasters design and operate their infrastructure. Dedicated hardware, software applications, and cloud resources increasingly contribute to the same live workflow, often with technology from several manufacturers. The expectation is that these environments will interoperate and the architecture be invisible to the end user, regardless of location..

That sounds straightforward until we examine what happens at the boundaries.

A signal might originate in a hardware encoder, move across an IP network into a cloud-based processing environment, pass between applications from different vendors, and eventually return to dedicated hardware for distribution. Each transition introduces questions about timing, control, security, and resource management.

Making those transitions work reliably requires more than connecting systems that support the same standards. It demands a common engineering approach to how media moves, how applications communicate, and how the complete workflow behaves under pressure.

There will also be different demands on any system during different phases of a production, Planning will need to look at compute availability and licence models, The build will require flexibility during production itself the system must be reliable and robust.

Choosing the Right Technology for the Workload

One important development in recent years has been a more pragmatic understanding of where hardware and software deliver the greatest value. Not long ago, the industry appeared convinced that almost everything would eventually move into the cloud. Broadcasters have since recognised that the economics and operational requirements are more complicated.

Dedicated hardware, like Appear’s X Platform, continues to offer advantages for deterministic workloads. When a broadcaster knows how many channels it needs to encode or decode, hardware provides predictable latency, high processing density, and efficient power consumption.

Software and cloud infrastructure become attractive when requirements are less predictable. A major sporting event might need additional processing capacity for supplementary programming, social media content, or temporary distribution feeds. Those resources can be deployed when required and released afterwards.

Broadcasters can now establish a permanent foundation and extend into the public cloud wherever flexibility creates operational value. However, combining these approaches introduces engineering challenges that cannot be addressed simply by moving functions between platforms.

Matching technology to the workload

Approach Best suited to What it delivers
Dedicated hardware Deterministic workloads — a known, fixed number of channels to encode or decode Predictable latency, high processing density, efficient power consumption
Software and cloud infrastructure Unpredictable or occasional demand — supplementary feeds, social content, a major event’s temporary distribution Capacity deployed when required and released afterwards

Where Standards Meet Reality

Broadcast interoperability has always depended on standards, but compliance alone does not guarantee compatibility. Two manufacturers may implement the same specification correctly while making different decisions about optional functionality.

Metadata handling offers a useful example. Information that appears unnecessary to one application may be essential to another further downstream. A vendor might legitimately omit an optional capability, only for that decision to create an unexpected problem elsewhere.

We have encountered similar issues throughout broadcast technology, from SDI Level A and Level B implementations to MXF files and different codec configurations. A tightly integrated system from one manufacturer may offer predictable behaviour, but customers increasingly want the freedom to combine specialist technologies. Open standards provide the foundation for that choice, and engineering collaboration makes it practical.

Moving from Transport to Compute

Two emerging industry initiatives are particularly relevant to this transition: the Dynamic Media Facility (DMF) and the Media eXchange Layer (MXL).

DMF, developed through work within the EBU, provides a reference architecture for software-based media environments. It establishes a common understanding of how infrastructure, orchestration, workflow management, and media applications fit together.

Much of the underlying compute infrastructure is already well understood, but how media functions from different manufacturers interact within that infrastructure remains fairly unknown.

MXL addresses an important part of this challenge by establishing a common approach to exchanging media between software applications. Instead of repeatedly converting content into traditional transport formats as it passes between functions, applications can exchange media more efficiently within the compute environment. The Joint Task Force on Dynamic Media Facilities (JT-DMF) is examining additional requirements, including timing, resource management, orchestration, and the commercial implications of software-based workflows. Together, these efforts will help determine how media behaves when processing takes place within shared compute environments.

Timing Cannot Be an Afterthought

Timing is arguably one of the most difficult engineering challenges in a hybrid broadcast.

Within a conventional ST 2110 facility, Precision Time Protocol (PTP) provides a common timing reference across connected infrastructure. Difficulties arise when media moves beyond that controlled environment, particularly across wide area networks or into cloud infrastructure.

Timing references cannot simply be assumed to remain consistent across every network boundary. Media may need to be reclocked as it enters a new compute environment, with timing information maintained throughout subsequent processing. Protocols such as SRT and RIST help address less predictable networks, but protection against packet loss and variable network conditions introduces buffering and latency considerations.

For live production, these compromises matter. Multiple contribution feeds arriving with different timing characteristics can create operational complexity, particularly when low-latency turnaround is essential or synchronous arrive via different routes.

Ideally, accurate timing information should be established as close to the original media source as possible and preserved through the processing chain. Achieving that consistently across hybrid broadcast infrastructure remains an important area of industry development.

Sharing Compute Without Compromising Performance

Software interoperability introduces another challenge that dedicated broadcast hardware largely avoids: competition for shared processing resources. When applications from several manufacturers operate on the same compute node, each consumes CPU, GPU, memory, and network capacity. A change in one application can potentially affect another.

Broadcasters therefore need reliable information about resource requirements, appropriate reservations and isolation, and consistent monitoring across the workflow.

This includes understanding how applications behave under peak loads, how resources are allocated, and how operators receive warnings before performance deteriorates.

The JT-DMF’s work on compute resource management is particularly relevant to monitoring resource consumption, which will help broadcasters build multi-vendor systems with greater confidence.

The same principle extends to control and orchestration. Open APIs are essential and increasingly provisions for MCPs to allow agentic automation, but applications must also be discoverable, manageable, and observable within the wider production environment.

Common misconceptions about hybrid broadcast interoperability

Myth: if two vendors support the same standard, their systems will interoperate.
Fact: compliance alone does not guarantee compatibility. Two manufacturers can implement the same specification correctly while making different decisions about optional functionality, such as metadata handling — and that difference can create an unexpected problem further downstream.

Myth: timing references stay consistent automatically once media crosses into the cloud or a wide area network.
Fact: timing cannot be assumed to remain consistent across every network boundary. Media may need to be reclocked as it enters a new compute environment, with timing information maintained throughout subsequent processing.

Myth: a shared compute environment behaves like dedicated hardware, where one application’s load doesn’t affect another.
Fact: applications from several manufacturers operating on the same compute node each consume CPU, GPU, memory, and network capacity — a change in one application can potentially affect another.

Engineering for the Complete Workflow

Hybrid production offers broadcasters greater freedom to choose where processing takes place and which technologies best support their operations. That freedom also changes the engineering responsibility. We can no longer evaluate a system solely by examining individual components, and must understand how those components interact at the boundaries between hardware, networks, and software.

The industry’s work on DMF, MXL, timing, and orchestration is helping establish the foundations for this transition. Yet those foundations only become dependable when manufacturers implement them consistently and validate their behaviour together.

That is where technical partnerships become essential.

In the next article, I’ll explore what those partnerships look like from an engineering perspective, and why proving interoperability requires considerably more than confirming support for a standard.

Rules of thumb for engineering hybrid broadcast workflows

  1. Match the technology to the workload — dedicated hardware for deterministic, known-capacity needs; software and cloud for unpredictable or occasional demand.
  2. Don’t assume standards compliance equals compatibility — test how vendors handle optional functionality, such as metadata, in practice.
  3. Reclock media as it enters a new compute environment, and preserve timing information through the rest of the processing chain.
  4. Demand reliable information on resource requirements, reservations, isolation, and monitoring before combining multi-vendor software on shared compute.
  5. Treat technical partnerships as essential — standards only become dependable once manufacturers implement and validate them together.

Solution brief
One REMI workflow, every event

How to deliver low-latency, resilient contribution across managed and unmanaged networks—while reducing on-site infrastructure.

Download the brief
Remote Production solution brief cover
About the author
Ian WagdinVice President of Technology and Innovation

Ian Wagdin collaborates with industry partners to push the boundaries of live production technology. With over 30 years of experience in broadcast and production, Ian has been at the forefront of major industry transformations — from the transition between SD and HD, to the evolution from tape-based to file-based workflows.