Descriptive Metadata for Television

An End-to-End Introduction

Mike Cox, Ellen Mulder, Linda Tadic

• Publisher: Routledge
• Year: 2006
• Topic: Languages & Linguistics
• eBook ISBN:
  9781136034978
• Subtopic: Communication Studies
• Index: Languages & Linguistics

Descriptive Metadata for Television is a comprehensive introduction for television professionals that need to understand metadata's purpose and technology. This easy-to-read book translates obscure technical to hands-on language understandable by real people.

1 What Is Metadata?

“That shows that there are three hundred and sixty-four days when you get un-birthday presents,” said Humpty Dumpty.

“Certainly,” said Alice.

“And only one for birthday presents, you know, there’s glory for you!”

“I don’t know what you mean by ‘glory,’” Alice said.

Humpty Dumpty smiled contemptuously. “Of course you don’t—till I tell you. I meant ‘there’s a nice knock-down argument for you!’”

“But ‘glory’ doesn’t mean ‘a nice knock-down argument,’” Alice objected.

“When I use a word,” Humpty Dumpty said in a rather a scornful tone, “it means just what I choose it to mean—neither more nor less.”

“The question is,” said Alice, “whether you can make words mean different things.”

“The question is,” said Humpty Dumpty, “which is to be master—that’s all.”

Alice was too much puzzled to say anything, so after a minute Humpty Dumpty began again. “They’ve a temper, some of them—particularly verbs, they’re the proudest—adjectives you can do anything with, but not verbs—however, I can manage the whole lot! Impenetrability! That’s what I say!”

“Would you tell me, please,” said Alice, “what that means?”

“Now you talk like a reasonable child,” said Humpty Dumpty, looking very much pleased. “I meant by ‘impenetrability’ that we’ve had enough of that subject, and it would be just as well if you’d mention what you meant to do next, as I suppose you don’t intend to stop here all the rest of your life.”

“That’s a great deal to make one word mean,” Alice said in a thoughtful tone.

“When I make a word do a lot of work like that,” said Humpty Dumpty, “I always pay it extra.”

“Oh!” said Alice. She was too much puzzled to make any other remark.

—Lewis Carroll

Humpty’s conversation with Alice will sound familiar to anyone who has been involved with making programs. The different disciplines involved in making programs have for many years used the same words to mean subtly different things. New technologies have brought their own words and meanings as new concepts are introduced and new ways of working become possible. A few years ago, “metadata” was a word that nobody used. Now suddenly it is everywhere, yet there is nothing new about it.

So, What Is “Metadata”?

The traditional answer is that the word “metadata” comes from the Greek “meta,” meaning “about,” so that metadata literally means “about data.” To most people, this is about as helpful as Humpty’s explanation to Alice. In the simplest terms, metadata is a particular detail of information about something else.

Program makers work with sights and sounds. In theater, these are real and involve working with real settings, real people, and real music with the audience physically present for the production. In television, the audience is remote and views or hears a reconstruction of the sights and sounds produced from some form of electronic representation of the original, which may or may not have been stored as a recording. In all these cases, the people in the audience will want some detail of the show if they are to be in the right place at the right time for the right show. At the least they will want to know the title of the show, how to find it, and what time it is to be performed—in other words, they will want some details about the show. They want some “metadata.”

What Metadata Is Not: Myths and Facts

In spite of the current hype in the industry, metadata is not magic! Neither is it a panacea to bad practice and there is no metadata cavalry galloping over the hill to rescue us from rising costs, ever tighter budgets, union demands, bad management, deadlines, or more competition. Metadata is not a threat to the quality of programs, and it is not something that can be just bolted on by buying the latest piece of equipment. Metadata is not “digital” (whatever that means), though the word is often associated with digital hardware and applications.

Most important, metadata is meaningful information in its aggregate—a single item of metadata is merely a piece of detail data and in isolation is not usually very informative. Several items of metadata grouped together are probably necessary to convey useful information. Further, information is not knowledge—only when the right pieces of information are perceived in the correct relationship will knowledge dawn. This implies increasingly complex structures as simple metadata elements are used to convey firstly information and then knowledge. This increasing complexity is reflected in the way we use metadata—not as simple data elements alone or even in groups, but in complex structures and substructures each with their own rules.

Perceptions of Metadata

The perception of metadata is one of those curious things that depends on where you stand and where you start from. One person’s important metadata is another person’s rubbish. In addition, there can be several layers of metadata. For example, a written description of a program might be considered metadata—or the description itself might have its own metadata, such as the name of the person who wrote it.

Some of the edges get very blurred: a browse or preview copy of a program might arguably be considered metadata because it is a descriptive proxy for the real thing—until the original is destroyed and you have to broadcast it! Indeed the electronic representations of sights and sounds we all are used to in television might be considered to be descriptive and therefore metadata. Proxies are frequently used in program making as a research tool and usually (perhaps erroneously) referred to as browse video or browse audio—they are not a usable copy but are instead descriptive of the broadcast-quality original.

Fortunately, the industry has come up with some basic definitions to give itself a starting point. The following definitions from the Society of Motion Picture and Television Engineers (SMPTE) were derived as a consequence of the final report of the “EBU/SMPTE Task Force for Harmonised Standards for the Exchange of Programme Material as Bitstreams” published in August 1998:

Any data or signal necessary to represent any single type of visual, aural, or other sensory experience (independent of the method of coding) is Essence.

Any one or any combination of picture (or video) essences, sound (or audio) essences and data (or auxiliary) essences is Material.

That data which convey information about Material is Metadata.

Material in combination with any associated Metadata is Content.

The definition of “essence” introduces an important concept—also a difficult one because it is challenging to conceptualize essence without instinctively attaching metadata to it in one’s own mind. As a result, there is often confusion between the program essence and the program content. At the same time, the word “material” is frequently used as a sort of slang term for almost anything to do with a program, with no further thought given about what the word might really mean.

It is important to recognize that metadata can exist before the essence about which it conveys information—for example, titles, project numbers, or shooting schedules can all fall into this category. Equally, the metadata can exist long after the essence has been destroyed, as might be the case with rushes where some of the metadata continues to have importance but the essence does not—for example, contact details or historical information.

Relationships with Current and Future Broadcasting Technologies

In traditional film- or tape-based program making, the essence as previously defined is either transmitted “live” as an electronic signal or captured onto a storage medium—either chemically on a suitable emulsion or magnetically onto magnetic particles. In these cases, a supporting plastic strip is used as a physical base for the actual storage medium to form a reel of film or tape, which is kept safe in a film tin or tape cassette along with identifying data, titles, and the like— in other words, along with basic metadata. Even a live feed will have supporting metadata about its source, title, and so on.

In the production office, project plans are drawn up, contracts issued, scripts written, and so on. All these processes produce their own metadata—but then the people involved in the work traditionally store the metadata they produced independently of each other in a variety of word processing files, spreadsheets, diaries, filofaxes, and Post-it notes.

Once the program is finished, it is passed on to the archive or library for safe keeping. Librarians will catalog and classify the content, possibly using a proxy copy, and enter the resulting informative metadata in their database so they can retrieve it in the future. However, rarely if ever is the metadata from the rest of the process passed on to them, except, perhaps, for the title, tape number, and basic technical information about recording formats. It has to be re-created, with all the associated risk of errors and lack of accuracy—not to mention the work and time involved.

As the electronic technologies of program making converge with the newer concepts and technologies made possible by the computer industry, working practices are starting to change: program material need no longer be stored on magnetic tapes but can be stored in exactly the same way as word processor or spreadsheet files—that is, literally as computer data files. Program material no longer needs to be moved from place to place by physically transporting the tape or by the use of specialized and expensive communication circuits. Many different users can work on the same program material at the same time, independently of each other.

The downside is that the old numbered tape boxes are gone. Material can be ingested into a digital, computer-based system entirely automatically, without anyone ever having seen it, and stored in some nether region of cyberspace. Files are nebulous, intangible things with no obvious way to track or find them, except by the file name—did we call it doc1.doc or clip1.avi? Is it Fredsclip2.wav or Tuesday6.bwf? 1pmmurder.mpg or pmhanging.mxf? Or did the machine give it its own number—4872bro1.abc? We know Linda was the reporter, but how does that help us now?

This is why “metadata” has become such a buzzword as new digital technologies are introduced into the workplace: it is the word the new technologies use and, while it always was important, it is becoming an increasingly crucial part of the workflow, right from its start. Applications are emerging that can automatically capture metadata such as color, texture, and sounds, or even the spoken word as text. Material increasingly has to be tightly linked to its metadata right from the beginning, at the originating camera or microphone; the relationships between the different pieces of metadata have to be preserved, and metadata has to be transported, copied, and updated as work progresses on the program. In short, the metadata has to be properly managed right from the start. In a big enterprise such as CNN or the BBC, if a piece of program gets lost inside a computer-based system, it will probably stay lost.

The Perceived Relationship with the Data Handling (Information) Technologies

Once again, this problem of managing data files is not new. Personal computers began to appear in the 1980s, and at that time little thought had been given to the problem of finding things—some of us remember the early DOS keyboard commands and the seemingly impenetrable screens and unhelpful messages they produced when all we were trying to do was to find our half-finished document:

0 File(s) 0 bytes
18 Dir(s) 9,047,680 bytes free
Directory of C:\Mike and Mirador
10/06/2004 21:36 <DIR> .
10/10/2004 22:00 <DIR> ..
07/09/2004 16:28 <DIR> Mike
30/09/2004 14:11 <DIR> Mirador
0 File(s) 0 bytes
4 Dir(s) 9,047,680 bytes free
C:\Mike and Mirador>mirador
‘mirador’ is not recognized as an internal or external
command, operable program or batch file.
F:\Mike and Mirador>cd Mirador
F:\Mike and Mirador\Mirador>
… and so on

Fortunately, the situation has improved since then, and we now work with much better tools, often graphics based, which are friendlier and easier to manage. Yet how many of us can truly say that we have never forgotten what we called a word-processing document or a spreadsheet, or lost something because of a spelling mistake?

This change has, of course, been driven by the demand from real users for tools they understand. Tools have been developed to give a pictorial view of the workings of the computer system. Because much of the demand was from people using personal computers in an office environment, office terminology was often adopted, such as “files,” “directories,” “folders,” and “cabinets.” Data was stored in binary form on cassettes or discs, which required electric motors to drive them, and the term “drive” appeared in computer language.

Finding your word-processing file has become much easier due to graphical interfaces—provided you know some simple data about the file, such as its name (or even a fragment of its name), when you stored it, which folder you stored it in, and which drive that folder is on. Most of us can manage this from memory for current work in progress or by normal office good practice in the way the filing has been structured, perhaps based on past experience. There are also simple applications to help. Search engines or book-marking systems can jog our memory as to what we named the file and some will do automatic text-based indexing. So there has come to be a perception that finding word-processing documents in a computer is the same as finding TV program files in the TV archive. In practice, however, and particularly for older files or when you are looking for someone else’s files, it has much in common with trying to find a needle in a haystack or maybe a book in a large public library. Some sort of properly structured and managed knowledge-based indexing system is needed. The importance of this is clear in many broadcast archives where collections contain several millions of hours of program content dating, in some cases, back to the 19th century (due to the inheritance of news film footage and early broadcast sound recordings). Media in these collections include wax cylinder or wire recordings, film footage from Victorian times, extensive European footage of World War I, and many samples of privately shot material. No mean haystack in which to find your needle!

The Very Real Relationship with Information Science

Information...

Table of contents

• Cover
• Halftitle
• Title
• Copyright
• Contents
• Introduction
• 1 What Is Metadata?
• 2 Types of Metadata
• 3 Metadata Schemes, Structures, and Encoding
• 4 The Impact of Technology Change on People and Metadata Processes
• 5 Identifiers and Identification
• 6 Metadata for the Consumer
• 7 Metadata in Public Collections
• Appendix 1 Sample Metadata Records
• Appendix 2 Extracts from SMPTE Documents
• Index