Content Management and Edit Data Exchange
Oh Boy! Pictures, pictures, everywhere!
The field of content management is exploding. Everyone is seeking to employ the growing power of computers and networking to link their production work and distribution into a seamless environment.Broadcasters, production companies, software providers and equipment venders are all caught in a crossfire of demands fueled by the rapid emergence of these technologies and the expectations they promote.
Much progress has been made by various organizations and venders. Unfortunately these are mostly proprietary implementations. For the prospective customer this too often amounts to a patchwork of incompatible systems.
Order from Chaos - Interoperability Standards
The EBU/SMPTE Task Force was formed in response to the explosive growth in digital technologies. The members recognized the need to help head off a proliferation of incompatible standards.The investigation covered the gamut of user requirements from the nuts and bolts of image compression and transport protocols through metadata and enterprise object model design.The Task Force's final report, published in 1998, represents a monumental research effort and the best roadmap of the challenges and solutions of digital production. The report is now serving as a set of recommended actions to the EBU and SMPTE and the technology committees and work items are being reorganized to follow that outline.
The author (Harris) was a member of the Task Force since 1996, contributing perspective on edit data exchange issues. He is currently a member of W25 Wrappers and Metadata and W22 Systems. A discussion of all the topics covered in the report is beyond the scope of this paper but it forms an excellent basis from which to view the emerging art of content management.
The Essence of the Matter.
At the root of a content management system is the "content" itself - the video, audio and other sources that make up the programming that is finally presented to the viewer. But what, exactly, do we mean by "content"? Does it mean 'Clark Gable', "Frankly, my dear, I don't give a damn", Gone with the Wind, or the film it was printed on? What does "material", or "programming" actually mean, and to who?To further confuse the matter this "content" is not limited to pictures and sound, but also includes various data types such as closed captioning, HTML (the internet presentation format), and graphics. In addition, diverse industries often use the same word for different things. "Media", for instance, is used by Avid users to refer to video and audio but is used by the computer industry to refer to disk drives and other recording devices.To help overcome these confusions the Task Force has coined the word "essence". Silly as it may sound at first, the distinction goes a long way toward clarifying the discussion. "Essence" refers to the data that tells the presentation device what to present to the viewer. This is typically video or audio in some data form, but also includes graphics, text and other data which reaches the viewer in some sensory manner.This said, the more context sensitive concepts "content" and "material" can be dealt with at a higher level and the engineering issues involving essence can be more clearly stated.
The first interoperability challenge is to standardize methods for machines and software to access, move, and organize the essence itself. There are four primary engineering challenges to this.
First, there is the format of the essence data itself. This is often thought of as the 'file format' but more generally must refer to essence data organization in other usage contexts such as 'streaming' or in system memory. These include standards such as 601 and MPEG and proprietary formats like AVI and Quicktime. These are all incompatible and must be converted to be used amongst applications.
There will never be just one essence format because each has advantages in different environments and more will emerge. For instance 601 will remain an important uncompressed production format while MPEG and others are excellent for lower bandwidth distribution. Several high definition formats appear to be inevitable. The challenge is to make them all interoperable.
The file system, the method used by a computer's operating system to access the files, poses another obstacle. Windows, Macintosh, and Unix all use different file systems. This is a challenge for the world of computing in general and is particularly difficult in very high performance systems required to handle picture and sound.
Again, it is not possible to agree on a single file system because they are all widely used and each has its own advantages. Interoperability must be achieved at a higher level.
Networking and Transport
Interprocess communication and transport of the essence presents challenges at several layers. Electrical interfaces, transport and networking protocols must all interoperate. The high data loads presented by essence delivery demand optimized design. Facilities will need to integrate new systems into existing networks. Interoperability must be achieved amongst many components to provide this essential backbone.
Edit Data Exchange
In a distributed environment edit data will be central to every decision made by machines and people. New requirements such as source referencing, heritage tracking, and workflow management are added to the more familiar edit data exchange needs. Edit data will play a critical role in data integrity and library management services.
No one edit data format will serve all purposes. Character based formats will endure for their unique universality, for reverse compatibility to conventional systems, and for interfacing with external systems, such as internet. Binary forms of edit data will be needed for very high speed and compact purposes. Object formats will provide high level organizational capabilities.
Wrapping it up
The Task Force came to recommend the use of a standardized digital container, or 'wrapper', which holds the essence and associated information. Wrappers will exist in several forms such as a header in a file, or a repeating object in a stream. Each will expose itself to the system in a common manner so it can be handled appropriately.
Based on this recommendation, SMPTE formed the W25 Wrappers and Metadata technology committee to continue the work of creating standards. Harris is a member of this committee and active on several particular projects, including Unique Material Identifier (UMID), and Metadata Dictionary.
Unique Material Identifier - the UMID
The purpose of the UMID is to provide a standard means of identifying unique instances of essence throughout the system. This is a critical function because it provides the system with a method of distinguishing all copies in a globally unique way. This has utility in access, library management, and, perhaps most importantly, data integrity. The UMID will be a key component of standard methods of managing data.
I Never Meta Data I Didn't Like
Like many buzz words, 'metadata' has taken on too much meaning. In general it means 'information about the essence'. This has many facets, from technical descriptions of the data format to conceptual descriptions. To provide a basis for common understanding of components in a system is necessary to clearly define terms and data types.With the valued support of National Imagery and Mapping Agency (NIMA) the development of the Metadata Dictionary is under way. The Metadata Dictionary will be a mammoth collection of registered names and data types. Its hierarchical structure allows endless expansion and mechanisms for data formatting provide a common method of implementation.
This could become one of the most important standards in years to come, not only for the production community but for other sectors as well.
Distributed Enterprise Systems
Ultimately all these components must be integrated in an enterprise wide system. Managing workflow and access require several levels of control software. Components of the system must be able to interact with existing enterprise systems and databases. The challenge for the industry and SMPTE is to develop interoperability standards which allow venders to integrate their services across these many layers.
Unmentioned so far is timecode. Timecode is an essential component all production, the glue that holds it together. Technology's advance suggests new requirements for timecode. These include new implementation forms such as object, and additional timing factors like extended frame rates, audio sample accuracy, and date.
Harris is Chairman of the SMPTE 'Adhoc Working Group on Timecode'. The mission of the group is to develop a coherent picture of user requirements for timecode in the context of distributed digital production.
AES Audio Decision List
Responding to the need for interoperability the Audio Engineering Society (AES) has embarked on a project to develop a 'Simple File Interchange Format', an edit data exchange standard optimized for audio mixing and work stations.The proposed design is ascii based, provides sample accurate editing and includes file locators within a structure to support exchange between disk based audio systems.
Harris is Vice Chairman of AES SC06-01, the committee responsible for this, and primary author of Audio Decision List (ADL) and its supporting documents. ADL will soon be published for public comment.
OMF and AAF
Avid and Microsoft have joined forces with several vendors to form the Microsoft Multimedia Task Force. The purpose is to develop the Advanced Authoring Format (AAF) based on Avid's Open Media Framework (OMF).AAF, successor to OMF, is an extensible edit data exchange format which defines an edit data object model riding atop a file container format called Structured Storage.
The Microsoft Multimedia Task Force is seeking standardization of AAF and Stuctured Storage through SMPTE. Harris is a member of the SMPTE W25 working group reviewing these proposals. BHFT is an OMF "Champion" and an AAF "Promoter".
Avid's Open Media Management (OMM) Initiative
OMM is Avid's proposed approach to sharing edit data and essence across networks.The author was a contributing programmer on the original OMM proof of concept implementation at Avid in August, 1998. The results of that effort, displayed at IBC in Sept, showed metadata and media moving between the Avid Symphony and an Informix database running on a remote server over TCP/IP. This caused a great deal of excitement.
The OMM specification is undergoing technical changes as it develops, but the goal remains the same - seamless access to centrally stored metadata and media. It promises to be an important piece of the puzzle but until these standards issues are resolved it will be limited to dealing with proprietary essence types.
Edit Data Solutions Today
While the industry awaits the completion of these standards edit data exchange formats in the forms of OMF, Avid Log Exchange (ALE), and conventional EDLs provide a useful method of moving important information amongst applications.Brooks Harris Film and Tape, Inc offers edit data solutions today. Custom implementations of EDLMAX are available and the MaxDll Toolkit provides OEM developers with tools to deploy their own solutions. Harris is available for consulting on all these issues.
* The proper name for the project is "Joint EBU / SMPTE Task Force on User Requirements for the Exchange of Television Programme Material as Bit Streams". This is revised from the original title "EBU / SMPTE Task Force for Harmonization of Standards for the Exchange of Television Programme Material as Bit Streams". You had to ask.
The views expressed are the author's alone and do not represent official positions of any referenced standards bodies.