- Fred Mattocks
Technology Strategy Board
A few weeks ago, the CBC/Radio-Canada Board of Directors approved the Technology Strategy Board's (TSB) proposal to invest in an enterprise-wide Media Asset Management (MAM) infrastructure. This is a significant milestone in the Corporation's path towards its final big step into the world of digital content creation and content management. I am writing this article to let you know why we are doing this, as well as what we hope to achieve.
Let's have a look at what CBC/Radio-Canada's content production ecosystem looks like. There are many different types of technologies that we deploy in service of our content mission; in fact, there are more than ever before.
Some of these technologies are small, commodity technologies that are based in the consumer electronics market and that have the advantage of being cheap, easy to use, and flexible. A suitable example would be our current use of iPhones for Radio, TV, and digital content creation.
Other technologies are more expensive and specialised for higher-end production needs. Examples would be our use of Pro Tools for high-end audio editing, our HD studio cameras, or VizRT for graphics creation.
Finally, some are huge enterprise systems, such as our Next Generation Converged Network (NGCN), iNews, Dalet DTR, or AVID Interplay Central DTV. These systems offer the benefits of scale and interconnectivity and, as a result, provide us with capabilities that we could only imagine even five years ago. All of them are based on content coupled with data about that content (metadata) and handled as digital files, the sum total of which has allowed us to implement the concept of File-based Workflow throughout the Corporation.
The Current Context
Let's review our current situation. We have the foundation of connectivity provided by the NGCN that permits content access and transfer without geographical or technological boundaries. We have file-based production systems that allow fast, efficient content acquisition, production, and deployment. From the perspective of any single production entity, we are essentially where we want to be.
Figure 1 – CBC/Radio-Canada’s Production Ecosystem
However, when we look at this from the viewpoint of realising the full value of our content, we need to know what content we have, what formats are available, where it is, and how we can get at it across the production environment. Moreover, we also need to know what we did with it for future use. How do we find it and how do we manage it? The answer to these questions is MAM.
What does management imply in this context? Ultimately, it involves keeping track of content in ways meaningful to production and to the business so that we know what we have and where it resides. Doing so allows us to maximise the opportunities to use excellent content in a variety of ways, and to avoid spending time and money on unnecessary duplication. Knowing where we have copies will also enable better storage management and potentially reduce the peak load on our NGCN connections.
It also offers the opportunity in the file-based world to automate work that computers do better than people. An example is transcoding – a process inherent in any real-world digital production environment where content and its metadata are converted from one file format to another. Right now, that is largely a manual process, as highly skilled experts figure out how to set up the transcoding process for each type of transfer. In the future, this process will be automated and transparent to the content worker, and our experts will be free to work on problems that machines are not good at solving.
Consequently, the value proposition for MAM at a high level is that we will be able to make the best use of our content and our talent by making content accessible wherever it might be, and by ensuring that we invest our talent where it makes the most difference to the product and, ultimately, to our audience.
How is this going to happen? A lot of work has already been done. The TSB has been considering this issue for two years, and the acquisition of the system itself is only one step. A MAM works by managing metadata, and then acting on the associated content in defined workflows. We have all done this ourselves: anyone who has set up a playlist in their personal listening device has used metadata to define how content will be used.
Of course, our application is much more complex. CBC/Radio-Canada has tens of millions of content items and the Corporation produces thousands more everyday. Each item can have hundreds of attributes described by its metadata. Organising that metadata to ensure that it is consistent and error-free is a major task.
Figure 2 – The Taxonomy Approach Applied to Metadata
We have spent much of the last year creating a taxonomy for our metadata within the Corporation. That taxonomy is a map that defines all of our content and process attributes across our business. The CBC/SRC Core it is a refinement of the Eureka/Medoc taxonomy created by our archivists several years ago. It reflects the principles of the Dublin Core, the base of most modern MAM taxonomies. It is unique to CBC/Radio-Canada in that it reflects the information that we need for production and business, yet, throughout the process, we have kept the ability to interoperate with other companies and content providers or markets as a central value.
We are in the process of acquiring the system now. Once acquired, we will start the process of migrating our existing archival information into the MAM (one of the core purposes of the MAM is to make the content archives accessible and transparent to the user) and building the workflows that will define how the MAM will interact with users and content.
Rather than doing this as a big bang where we change everything all at once, we have recognised the value of a more agile approach and will carry out the implementation as a series of smaller projects, working from those that are relatively simple to those that have higher complexity. The whole process will take at least two years. For many of you, the deployment of MAM will be seen as a new search function in your existing work environment.
Part of what we will be doing in order to make all of this work smoothly is standardising work processes wherever it makes production and business sense to do so. The goal is to come up with a configuration that permits us to gain the benefits of the system by standardising the core functions, all whilst enabling customisation at the creative and production levels to accommodate the needs of different platforms, genres, and products.
MAM will touch all of us. It will provide us with capabilities that we have not had before; it will enable us to leverage the power of digital media to create new products and it will make the process of creating great content more efficient and effective. Information with more technical breadth and depth will appear in future articles in SYNC focussed on particular aspects of the project.
We are in good company. Every major media company we have talked to either has a MAM, is in the process of acquiring one, or is studying the application of one. This is the time.
-  http://en.wikipedia.org/wiki/Pro_Tools
-  http://en.wikipedia.org/wiki/Vizrt
- Feel free to consult Michel Béland's article for detailed information about the NGCN in our first issue.
-  http://www.avid.com/US/products/inews
-  http://www.dalet.com
-  http://www.avid.com/US/products/Avid-Interplay-Central
-  Feel free to consult the articles by Gregg Antworth and Pierre Lemieux on File-based Workflow in our second and fourth issues, as well as in this issue.