Project

metarex.media

is like a zip file with a timeline. It’s used like an electronic FedEx service.

Status: view timeline . IBC Accelerator 2023 Responsive Narrative Factory .

Rationale

It’s super easy to send anything by FedEx - even a baby dinosaur. You put your stuff in a box and they slap 2 labels on it.

1. The Destination - send to MetaRex HQ
2. The Manifest - a bar code that points to a database. The record in the database says Baby Dinosaur - do not drop

Magic happens and the box travels over train networks, air networks, couriers, bicycles and just gets to where you want it.

So why is it so hard to send media metadata?

That’s the problem that MetaRex solves

1. Map the mrx container into networks so that destination addresses work.
2. Define a database (register) of manifests so that you know what’s in the mrx container.

An added bonus is that the mrx register can publish cloud endpoints to manage the metadata - this can lead to no code integrations. That should be nice!

So what is it?

A project to create a body of free to use Open Source software that allows media metadata to be created, transmitted and stored over common links and file systems.

Why do I want it?

Because today, every Virtual Set and engineering heavy production is a custom build. Every QC report is unlinked to the master content and version workflows lose metadata when content is processed by third parties.

• What is it good for? Carrying binary, text, isochronous and event driven metadata. Identifying the metadata via a registry and providing links to applications and schemas that can shed light on how to process that metadata. By harmonizing the identification and transport of the metadata, an infrastructure can be built to keep the stuff that we currently lose.

• What is it not good for? We will not harmonize schemas for tripod data, lens data, QC data or any other sort of metadata carried. That’s outside the scope. MetaRex will carry any metadata through the value chain to where it’s needed. MetaRex will identify that metadata and allow applications to enlighten the metadata or remove it based on business rules.

• What is it not yet good for?: At the end of 2022 MetaRex is just a paper specification, some pictures and a prototype register. At the end of 2023 we should have code for different computer languages and a framework for common tools and apps. By the end of 2024 we should have end to end workflows with common tools and a body of free software that reduces the barrier to transporting and preserving media chain metadata.

History

Throughout 2022, at NAB , IBC and the DPP Leader’s Briefing, the MetaRex team did a lot of talking to content creators, facility owners, software vendors and many others about what were the blockers that stopped metadata being a first class citizen. The biggest issue was “The metadata gets lost”.

MetaRex was shown to work back in 1998, when the original MXF team demonstrated file workflows with metadata preservation at the BBC in a project called G-FORS (see page 219 of linked PDF). The demo was brought up to date with live networking and VFX and shown at the ARRI Broadcast Day in 2019, but in 2022 there still wasn’t a similar solution available on the market. Games companies complained that every new virtual set was a custom engineering job, just to get the metadata to the right place at the right time.

MetaRex is in full flight. We need backers to create a financial reserve from which we can give out grants to engineering teams to create the Open Source that we need to enable free tooling for all. Please make a small financial or engineering contribution if a working global media metadata ecosystem is of interest to you or your company.

The project will end in Q1 2025 with the project board deciding on the details of the project closure.