File-based workflows in broadcasting are often considered to be at the leading edge of technological change.
In fact, file-based workflows are everywhere, even in the consumer domain. Almost everyone has imported MP3 files into iTunes, Winamp or other players. Until quite recently, most MP3 files had no embedded tags or metadata, no artist name, no genre, and nothing to classify the content. This, of course, made it extremely difficult to search for content by any criterion other than the title.
This illustrates some of the strengths and weaknesses of a file-based approach in both the consumer and professional domains.
On the plus side, the file-based approach has shown improved efficiency through the lack of a need for tape-based transport or storage, far easier process automation, and the ability to easily handle multiple formats.
However, it is all too easy for a file-based infrastructure to end up as a series of digital islands that have poor interoperability and little metadata flow, where it is difficult to add new functions to the production chain or upgrade the system, and where it is hard to get real visibility on content and gain standard views of asset usage, traffic and so on.
The fact is that it works, and almost every broadcaster has now adopted file-based workflows at some level. However, it is probably fair to say that few of them are reaping the full benefits of having done so.
Since we're now at the stage where tape has been largely eliminated from the system, the emphasis is moving from a technology-led approach to one where business requirements are beginning to dominate. This invariably means adopting media asset management (MAM) as a way of life. MAM is the key to unlocking the economic advantages of the file-based approach, but there are hurdles to overcome along the way.
Increasing the value of media assets
As we have already mentioned, the addition of metadata to an MP3 file makes it more valuable to the consumer. Here, as in the professional domain, metadata is vital, both to make an asset searchable and thus to give it value. Though metadata is important, it's far from easy to implement.
Metadata are usually located in different systems and are quite often lost because of the lack of integration and interoperability between those systems. And what's in the mind of the individuals along the processing chain, usually the most valuable metadata of all, is lost as well if there is no interface to enter relevant metadata. This is an essential part of the role of MAM; it should manage files, be the glue between multiple systems and gather relevant and structured sets of metadata, thereby providing the means of enriching the content and adding to its value.
Correlating human workflows and the back office
In the modern broadcast environment, every action in the creative and production process has a financial implication of one kind or another. It has actually always been that way, but compartmentalization and the lack of suitable systems to gather and process data has made it impossible to manage. This is one of the potential benefits of MAM.
A system should be able to handle both the human aspects of workflow (who has to enter which metadata, who is responsible for verifying a given file, who should edit which package) and at the same time correlate this with the back-office systems that are managing metadata (both technical and descriptive), and essence aspects (file format conversions, subtitles, multiple file wrappers, etc.). Increasingly, these back-office systems are also recording and managing the financial aspects by logging the use of resources and providing continuous measurement of key human and financial performance indicators.
Such a back office needs to be capable of adapting to many possible workflows. For instance, if a user decides to send a clip to an edit workstation, the system needs to make the essence files available (in the right format) and also deliver relevant metadata such as the title, author, duration, plus any available time-coded logs or shot lists associated with that asset that will be useful for the editor. And of course, when the edit is finished, the new content item must go back into storage with all of its elements recorded and indexed for future use.
Giving better visibility
So, a modern MAM system must track everything: user actions, back-office processes, metadata flow and so on. It should also provide a dashboard to monitor and display such important elements as bandwidth and the availability of both the core system and the different components connected to it (such as video servers, edit suites, ingest points and networks). All this information is extremely valuable. When properly gathered and processed, it allows an organization to make relevant technical and business decisions to improve operations at every level.
Protecting and enhancing future investments
There is an increasing trend for MAM systems to be based on the principles of a service-oriented architecture (SOA), even though it is still true that everyone has a distinct vision of SOA and that SOA compliance is something of a variable. Still, SOA is a well-proven concept that provides the agility needed to integrate new components in the workflow, upgrade existing systems, and improve overall integration and interoperability. Perhaps most importantly, SOA provides a means of maintaining systems over time, and it allows the deployment and integration of several specialized MAM systems to carry out different functions in a coordinated manner. Here, as in other domains and sectors of IT, SOA is becoming a standard approach and is a way of improving the future-proofing of a given application or system.
Barriers to MAM adoption
While there are advantages, there remain some significant technical obstacles and complexities in the path of the full adoption of file-based workflows, from both the perspectives of the user and the manufacturer.
Standards and formats
Standards and formats are a nightmare for many, if not most, solution vendors. There are so many variant norms and formats, and they are evolving so fast that it is almost impossible to produce a piece of software or equipment that understands and handles every format in every iteration correctly.
Also, one manufacturer's interpretation of a standard may differ slightly from another manufacturer's. The difference may not be great, but it can be crucial and may result in content processed in one system being completely incompatible with another system even though they both comply with the same standard. This is perhaps not surprising when the definition of the standard can be many hundreds of pages long. There is too much room for variable interpretation. An example is the widely adopted MXF OP1a. Everyone has their own understanding of MXF wrappers. If you expect to connect a video server, some NLEs and a production system sourced from different suppliers just by using MXF, it's unlikely to work out of the box.
A MAM system must be capable of taking such limitations into consideration and adapt to emerging technologies, integrate system upgrades, handle rewrapping, do on-the-fly conversions and so on. It's not easy, but it's very important in an uncertain world.
Fast-paced workflows
There is an important second area of complexity: fast-paced workflows. These are typically needed in news and sports. The basic issue is that you do not want to have to wait for a given ingest to finish before you are able to start working with the content. (This was, of course, the case with tape-based workflows.) The material needs to be immediately available for all users, even while the recording is still going on, especially in the case of breaking news, and obviously, at any time in a live sports environment.
This is complex, and even more so when you have to integrate multiple systems that all have their own flavors of essence and wrapper. For instance, you might want an Apple Final Cut Pro workstation to start editing content that is still being ingested into another system and recorded onto shared storage, and at the same time, access metadata that are still being generated on the fly.
Such integrated fast-paced workflows in a multivendor environment, which are capable of editing while recording is still going on, are probably the most complex things to manage in a tapeless production environment. They require a high level of expertise and integration in the following areas:
- To make content available for everyone, it obviously needs to be stored centrally. Sufficient capacity and bandwidth are basic requirements. Other interoperability features such as edit-while-record require precise management of file-locking and access control across system boundaries.
- When several systems need access to the same files, it may be that the file needs to be written in or capable of being converted on the fly to multiple formats and wrappers. For instance, within one infrastructure, you require a given content item to be available in H264, and in DVCPro HD with a QuickTime wrapper for Final Cut Pro, and with a separate wrapper for Avid.
- It may be necessary for multiple separate processes to read the content of a file while recording is still taking place to generate both metadata and useful essence files such as proxy versions. The back office of the MAM must be capable of managing this automatically and taking into consideration all the different aspects of the systems and workflows in a transparent manner.
Conclusion
Building and implementing such versatile and agile MAM systems is not easy, and we have not even mentioned the human aspects such as training.
Even though the hype in the industry tends to be mostly about new technologies such as 3-D, the successful, full integration of file-based workflows is arguably the most important challenge facing the industry today, and it is capable of delivering the greatest business benefits in the long run.
The effective monetization of content coupled with the truly cost-effective use of resources are both crucial to the economic future of the industry, and in both cases, MAM is the key.
Raoul Cospen is director of marketing at Dalet Digital Media Systems.
