Live Media’s Mise en Place Problem

Why broadcast workflows still feel like cooking without your knives — even in the age of automated kitchens.

In my last post, I talked about how much I love to cook — and how much that love is tested when I can’t find the tools or ingredients I need. Instead of getting lost in the joy of creating, I find myself hunting through drawers, questioning my memory, and breaking my own rhythm.

That same friction shows up in my professional world too — live media production.

We have powerful tools. Incredible tech. Cloud elasticity. But when it’s time to actually do the thing — to get the show live — something always gets in the way.

Broadcast Without Mise en Place

If you’ve worked in live production — especially cloud-based REMI workflows — you know this story:

• Standing up cloud infrastructure using tools like Terraform or CloudFormation
  
• Watching EC2s spin up in minutes — which is genuinely impressive
  
• And then spending the next 4 to 8 hours manually:
  
  • Modifying configuration files
    
  • Connecting contribution signals
    
  • Mapping IP addresses and ports
    
  • Diagnosing app mismatches and encoder quirks
    
  • Double-checking if your monitoring tools are actually pointing at the right stream
    

It’s like having a professional kitchen delivered to your driveway — but all the drawers are locked, and nothing is labeled.

The mise en place is… missing.

The Real Work Begins After Boot

Let’s be clear: the infrastructure is there. We’ve nailed that part. Our teams can, and often do, provision complex live environments at the push of a button.

But what comes next still feels far too manual, too brittle, too opaque. Every production team I talk to says the same thing:

“We can spin up the machines in minutes. But wiring the signals and getting the system actually ready to produce takes too long.”

They’re not building infrastructure. They’re wrestling with integration.

It’s not the kitchen itself that’s slow to build — it’s the process of finding the cutting boards, figuring out where the knives went, and fixing the broken walk-in fridge that someone forgot to plug in.

A Glimpse of a Better Way

That’s why there’s growing excitement around a deceptively simple but transformative idea:

What if you didn’t have to move the media at all? What if you could just reference it — like memory?

This is the core promise of shared memory in live production workflows. Rather than building out complex media transport layers, the media is made available in place — and applications simply access it directly, zero copy, without unnecessary re-encapsulation or re-transmission.

In a kitchen, it’s like this: instead of walking to the fridge or sending a runner to the pantry, you simply think “lemon zest” — and it appears, already measured, right next to your cutting board.

The concept itself isn’t new. Shared memory has been used for years in various production systems, but what’s changing now is how the industry is beginning to rally around shared memory as a standard interface — not just a vendor-specific optimization.

And that’s where the Media eXchange Layer (MXL) comes in.

MXL, developed under the leadership of the EBU, isn’t introducing the idea of shared memory — but it is amplifying the conversation. What makes MXL noteworthy is that it proposes an open, interoperable model for shared memory access. By moving beyond proprietary implementations, MXL lays the groundwork for a more composable, multi-vendor ecosystem.

In short, it’s not just about making media accessible faster. It’s about making it more accessible to more people and systems, more easily.

What’s Next

In the final post of this series, I’ll share how we’re building on this momentum — extending the shared memory model beyond current limitations in scope, infrastructure, and format.

We’ll introduce a new approach that’s not just compatible with MXL — but completes the mise en place promise by making it global, compressed-stream aware, and orchestrated by design.

Because the kitchen doesn’t just need to exist.

It needs to be ready — every drawer unlocked, every ingredient instantly within reach.

About the Author:

Mike Cronk

Mike Cronk is Vice President of Strategy at TVU Networks, where he helps scale the company’s award-winning cloud and IP-based live video services. With over two decades of experience at AWS, Grass Valley, and as founding Chairman of AIMS, he has been a driving force in media workflow innovation and cloud transformation.

• Previous page