The Digital Dilemma 2: Perspectives From Independent Filmmakers, Documentarians and Nonprofit Audiovisual Archives
The Digital Dilemma 2: Perspectives From Independent Filmmakers, Documentarians and Nonprofit Audiovisual Archives
The Digital Dilemma 2: Perspectives From Independent Filmmakers, Documentarians and Nonprofit Audiovisual Archives
Oscar, Academy Award, and the Oscar statuette are registered trademarks, and the Oscar statuette the copyrighted property, of the Academy of Motion Picture Arts and Sciences. The accuracy, completeness, and adequacy of the content herein are not guaranteed, and the Academy of Motion Picture Arts and Sciences expressly disclaims all warranties, including warranties of merchantability, tness for a particular purpose and non-infringement. Any legal information contained herein is not legal advice, and is not a substitute for advice of an attorney. All rights reserved under international copyright conventions. No part of this document may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system without permission in writing from the publisher. Published by the Academy of Motion Picture Arts and Sciences Inquiries should be addressed to: Science and Technology Council Academy of Motion Picture Arts and Sciences 1313 Vine Street, Hollywood, CA 90028 (310) 247-3000 http://www.oscars.org Printed in the United States of America Library of Congress Cataloging-in-Publication Data The Digital Dilemma 2 Perspectives from Independent Filmmakers, Documentarians and Nonprot Audiovisual Archives 1. Digital preservation Case Studies. 2. Film Archives Technological Innovations 3. Independent Filmmakers 4. Documentary Films 5. Audiovisual I. Academy of Motion Picture Arts and Sciences Science and Technology Council II. Andrew Maltz III. Milton R. Shefter ISBN 978-0-9840150-2-3 LCCN 2011941780
CONTENTS
PREFACE 1 EXECUTIVE SUMMARY 3 1 INDEPENDENT FILMMAKERS 7
Introduction A Brief History of Independent Filmmaking Independent Filmmaking and Digital Materials Born Digital Materials Active Management of Digital Materials Data Migration Survey Methodology, Results and Analysis Summary
DOCUMENTARIANS 19
Introduction A Brief History of Documentary Filmmaking Survey Methodology, Results and Analysis Summary
ONLINE APPENDIX 115 END NOTES 117 BIBLIOGRAPHY 121 ACKNOWLEDGMENTS 125
PREFACE
The digital revolution in lmmaking began some 25 years ago with the introduction of digital sound and visual effects. The changes have kept coming at the industry in waves, and the most recent areas to be transformed by digital technology cinematography, mastering and exhibition brought to light an unanticipated problem: how to preserve and maintain access to digital motion picture materials. The Digital Dilemma, published in 2007 by the Academy of Motion Picture Arts and Sciences, articulated the growing concern about the longevity of digital motion picture materials and other valuable digital data. That report explored the issues facing those responsible for preserving digital data in the medical, military and geoscience arenas, and found that they all shared the same problem: there was no guaranteed long-term access to their digital data. This report was produced through a partnership between the Library of Congresss National Digital Information Infrastructure and Preservation Program (NDIIPP) and the Academy. Although the Library had completed major studies on lm preservation in 1993 and television and video preservation in 1997, the impact of digital technologies on the independent lmmaking and nonprot audiovisual archive communities did not become evident until well into the current century. This project, a follow-up to The Digital Dilemma, examines long-term preservation of digital motion picture materials from their perspective. The Library and the Academy felt a separate study was necessary
Preface
PREFACE
because independent lmmakers both narrative lmmakers and documentarians create, and nonprot archives collect and store, a sizeable portion of the nations audiovisual cultural heritage. These chronically under-resourced communities should not be allowed to fall through the cracks. From 2008 through 2011, broad surveys were conducted, many representatives were interviewed, and archival case studies were undertaken to examine the current preservation practices and digital preservation concerns of independent lmmakers and nonprot audiovisual archives. These communities are decentralized and loosely coupled, and thus very different from the highly structured and commerce-driven Hollywood studios. We therefore developed an approach that we believe has resulted in the most comprehensive study done to date on the challenges facing these communities, and is representative of their current status relative to digital preservation. Relevant historical and technical background information was included in this report to provide sufcient context for otherwise uninitiated readers. This report also offers suggestions from those interviewed and surveyed, as well as from members of the team that produced this report, that may help independent lmmakers and nonprot audiovisual archives navigate their way toward a safer environment for their digital works. Some of these suggestions appeared in the earlier film and television preservation studies referenced by this report, but have not yet been implemented. The fact remains that digital data cannot survive unattended, and with the passage of time, answering the call to action becomes increasingly urgent.
Milt Shefter, Lead, Digital Motion Picture Archive Project Andy Maltz, Director, Academy Science and Technology Council
EXECUTIVE SUMMARY
The Digital Dilemma, published in 2007 by the Academy of Motion Picture Arts and Sciences, focused on issues of digital motion picture data longevity in the major Hollywood studios and included comparative investigations of scientic, government and other major enterprises and industries. Among the reports conclusions was that although digital technologies provide tremendous benets, they do not guarantee long-term access to digital data; compared to traditional lmmaking using motion picture lm stock, digital technologies make it easier to create motion pictures, but the resulting digital data is much harder to preserve. Long-term preservation maintaining access to content for 100 years or longer is a key requirement for studio archives. Meeting this requirement necessitates professionally managed digital storage systems and processes at substantial, perpetual operational and capital expense, and oftentimes major enterprise reorganization. This reality will exist as long as technology obsolescence remains an integral part of the digital storage technology business model. Independent (indie) lmmakers operating outside of the major Hollywood studios supply 75 percent of feature lm titles screened in U.S. cinemas, despite facing substantial obstacles in doing so. As digital moviemaking technologies have lowered the barrier to entry for making lms, competition among indie lmmakers seeking theatrical distribution has increased. Without the benet of studio backing, these lmmakers must navigate the distribution waters on their own.
Executive Summary
EXECUTIVE SUMMARY
New digital distribution platforms may make it easier for indie lmmakers to connect their lms with target audiences and possible revenue streams, but these platforms have not yet proven themselves. Most of the lmmakers surveyed for this report have given little thought to what happens to their work once it is completed. Most pay for some type of storage for the master version of the completed work, but few store their lm masters in proper environmental conditions or manage their digital masters using appropriate preservation practices. Many depend on distributors (traditional theatrical distributors, packaged media, pay TV) or new streaming platform providers to take responsibility for preservation. In general, independent lms that beat the odds and secure some form of distribution do so after a much longer time period than movies produced by the major studios. This time period quite likely exceeds the shelf life of any digital work; that is, by the time distribution is secured, the digital data may become inaccessible. Most of the lmmakers surveyed and interviewed for this report were not aware of the perishable nature of digital content, or how short its unmanaged lifespan is compared to the 95-plus years that U.S. copyright laws allow lmmakers to benet from their work. Documentarians, dened in this report as independent lmmakers specializing in nonction topics, have access to funding sources that are not generally available to the broader group. These funding sources include grants as well as work-for-hire contracts. Unlike narrative lmmakers, documentarians achieve distribution primarily through broadcast and pay television; only a relatively small number achieve wide theatrical distribution. Many documentarians license archival footage for their work, and those surveyed noted the shift to acquisition of historical footage from lm to videotape beginning in the 1970s, and then to digital formats in the early 2000s. Surveyed and interviewed documentarians did not seem concerned about or aware of the possibility or likelihood of digitally acquired historical footage being lost. To the contrary, they believed that the Internet and todays digital technologies offered unprecedented access to historical footage. Most surveyed indie lmmakers, including documentarians, expressed two primary concerns: getting their work seen by an audience and moving on to the next project. They were therefore focused on securing distribution, with an eye to some measure of revenue generation. Today indie filmmakers face greater challenges in getting their work accepted to film festivals, which historically have been their primary path to theatrical distribution. They have consequently pursued nontheatrical distribution platforms such as direct-to-video and the newer Internetbased video-on-demand services, which can provide an easier path to a paying, if smaller, audience. Unless an independent lm is picked up by a major studios distribution arm, its path to an audiovisual archive is uncertain. If a lmmakers digital work doesnt make it to such a preservation environment, its lifespan will be limited as will its revenue-generating potential and its ability to enjoy the full measure of U.S. copyright protection.
Executive Summary
The archives estimate that their collective digital holdings will grow from approximately 183 terabytes in 2009 to more than 2.7 petabytes by 2014, a 15-fold increase that will result in indi vidual collection sizes in excess of 100 terabytes. Although archives can take advantage of digital technologies to provide greater access to their holdings, they are generally ill-equipped and inadequately resourced to properly store and manage such relatively large collections of digital materials for the long term. Well-established, time-tested analog preservation practices do not apply to digital holdings; digital materials are fundamentally different from motion picture film and other analog materials. Suitable long-term preservation and access mechanisms for digital motion picture materials have not yet been developed. At nonprot audiovisual archives, the decision to start digitizing analog materials, as well as the digitizing itself, quite often precedes the establishment of a digital preservation program. The digital les are typically created to satisfy an immediate need for end-user access or to preserve deteriorating materials already in a collection so the programs overall design and implementation are often deferred. The broader digital library community, which deals with smaller digital les in smaller numbers relative to audiovisual archives, has made progress in addressing digital preservation issues. While the motion picture industry has increased collaboration around these issues, independent filmmakers and nonprot audiovisual archives suffer from a dearth of nancial resources and active collaborative forums. This report describes proposals that may improve the outlook for these groups: Facilitating collaboration among representative organizations from these communities on issues of funding, technology and practice Organizing cooperatives to share technical infrastructure and knowledge Offering more educational opportunities at industry conferences, lm festivals and lm schools and greater exposure to the technical standards activities of major Hollywood studios and motion picture industry organizations
EXECUTIVE SUMMARY
5
The nal destination for many independent lms as well as much historical footage is one or more of the hundreds of nonprot audiovisual archives in the U.S. (and hundreds more worldwide) that actively collect materials in support of their particular missions. Many years can pass between the creation of such content and its entry into an archive. Archives surveyed for this report stated that increasing amounts of digital materials are entering their facilities through two mechanisms: analog holdings being digitally reformatted, and collections being created in digital form.
Executive Summary
EXECUTIVE SUMMARY
The digital dilemma is far from solved. Unless preservation becomes a requirement in planning, budgeting and marketing strategies, it will remain unsolved for independent lmmakers, documentarians and nonprot audiovisual archives alike. These communities, and the nations artistic and cultural heritage, would greatly benet from a comprehensive, coordinated digital preservation plan for the future.
INDEPENDENT FILMMAKERS 1
Introduction A Brief History of Independent Filmmaking Independent Filmmaking and Digital Materials Born Digital Materials Active Management of Digital Materials Data Migration Survey Methodology, Results and Analysis Summary
More than three times as many independent lms as studio pictures are released each year.
INDEPENDENT FILMMAKERS
Introduction
The metrics of feature lm distribution indicate that more than three times as many independent (indie) lms as studio pictures are released each year. Of the 706 feature lms released in domestic theaters in 2010, the six major Hollywood studios and their subsidiaries released 174 and other entities produced and screened 532.1 In 2009, members of the Independent Film & Television Alliance, a trade association for independent producers and distributors, produced more than 400 independent features.2 As a group, independent lmmakers typically distinguish themselves from their studio counterparts by having both total creative control of their lms and limited nancial resources with which to make them. As one interviewed lmmaker astutely noted, the term should be dependent lmmakers because they are dependent on friends, relatives and other contacts to help with the production, nancing and distribution of their lms. That said, a lms indie provenance does not preclude it from achieving wide critical or box ofce success. Since 1980, more than half of the lms that won the Academy Award for Best Picture have been independent productions, including such recent winners as Crash, No Country for Old Men, Slumdog Millionaire, The Hurt Locker and The Kings Speech. 3 Films developed by studios specialty lm divisions, or acquired by studios for art house or other targeted distribution, are generally 9
INDEPENDENT FILMMAKERS
not considered independent films. And while indie films may seem to comprise a fairly new facet of the motion picture industry, these productions have a long and illustrious history going back to very origins of cinema.
In 1908, the Motion Picture Patents Company (also known as the Edison Trust) was formed, and with it came an awakening of certain lmmakers to the inherent problems of one corporation being able to control their art form (and their business).6 These lmmakers believed in preserving the artistic integrity of lmmaking as well as their own authority to distribute and exhibit their films and took matters into their own hands. Escaping the Trusts patent-based monopoly on lm equipment and raw lm stock was a major factor contributing to both the development of Hollywood as the center of the motion picture industry and the American independent film movement. The rst acknowledged group of filmmakers to launch an independent lm studio in the United States was composed of Mary Pickford, Charles Chaplin, Douglas Fairbanks and D.W. Grifth, who together formed United Artists in 1919.7 Their aim was to gain more control over their work, and the power of self-determination was to become a major current in independent lmmaking from that point forward.
i
The history of independent lmmaking outside the U.S. is beyond the scope of this report.
10
With the introduction of portable and relatively inexpensive cameras during World War II, anyone (theoretically) could write, direct and produce a lm without studio support. Like the wide availability of todays digital cameras and post-production tools, this had a democratizing effect practical and economic barriers to entry were reduced. Several acclaimed lmmakers, such as Maya Deren (Meshes in the Afternoon), Kenneth Anger (Fireworks) and Raymond Abrashkin (Little Fugitive) came out of the productive 1940s. In the 1950s and 60s, artist-run organizations such as the New American Cinema Group and its offshoot, the Film-Makers Cooperative, began to support and encourage lms as works of art.8 Improved, inexpensive technology also gave rise to the exploitation lm genre, represented most notably by the films of producerdirector Roger Corman. Corman mentored some of the next generations top directors on their earliest lms Peter Bogdanovich (Targets), James Cameron (The Terminator), Jonathan Demme (Caged Heat) and Ron Howard (Grand Theft Auto), to name a few and his lowbudget, quick-turnaround output matched and sometimes exceeded that of the major studios.ii 9 As the 1960s led to The New Hollywood, directors such as Francis Ford Coppola (Dementia 13, Youre a Big Boy Now), Dennis Hopper (Easy Rider, The Last Movie) and Martin Scorsese (Boxcar Bertha) helped change the studio system from a completely in-house production line to an acquisition-balanced approach that placed financing, marketing and distributing activities alongside production a business model that is maintained to this day. On the other side of the country, filmmakers including John Cassavetes (Shadows, Husbands) and Robert M. Young (Nothing But a Man) led the independent movement in New York, continuing through the 1970s and 80s with Melvin Van Peebles (Sweet Sweetbacks Baadasssss Song), Charles Burnett (Killer of Sheep), John Sayles (The Return of the Secaucus Seven), Spike Lee (Shes Gotta Have It) and others. In 1979, the Independent Filmmaker Project (IFP) formed as an advocacy group for indie filmmakers, followed by the Sundance Institute in 1981 and Film Independent (formerly IFP/ West) in 2005. These nonprots and other similar organizations continue to nurture independent filmmaking today.10 The indie movement has been a consistent source of signicant cultural content in American lms. Roger Corman, accepting a 2009 Honorary Award from the Academy of Motion Picture Arts and Sciences, stated: the finest films being done today are done by the original, innovative filmmakers, who have the courage to take a chance and to gamble. So I say to you: keep gambling... 11
The lms listed here are these lmmakers early independent lms.
11
INDEPENDENT FILMMAKERS
INDEPENDENT FILMMAKERS
Almost all motion pictures produced today regardless of the capture medium reach a point of digital existence.
card. The elimination of raw lm stock and laboratory processing expenses reduces production costs even further. But the same digital-versus-lm trade-offs exist for low-budget independent lms as they do for studio pictures: the digital data that represents the recorded images and sound needs to be handled very differently from motion picture lm and analog audio tape. This topic is covered in detail in the Archives section and reference materials, but some discussion of digital motion picture storage and handling practices here will help provide context for non-technical readers to understand the survey and interview sections that follow.
12
Motion picture film deterioration is easily managed with proper environmental controls, simple inspection procedures and passive detectors that warn of possible chemical degradation. By contrast, digital data of any type is subject to invisible failure mechanisms at many levels the actual recording media, the data reading and writing system in the digital storage device, the data interface that connects the storage device to a computer, the computer network that connects individual machines, and the many levels of software that control the overall system. Although digital systems can be made to be highly reliable, they must be constantly monitored to detect failures or conditions that may lead to failures, and preventive or corrective actions must be taken to avoid data loss. Regular backup of digital data (copying to another storage medium and transporting to a remote physical location) and verication of primary and backup data are examples of active data management processes.iii
Data Migration
All digital technologies hardware and software have a nite useful life. Hard disk drives eventually fail, all computers are replaced by newer models (often with different electrical connections to peripheral devices and new storage media), and computer software never stays at version 1.0. It is true that new generations of technology-based products bring new capabilities and sometimes even cost savings, but one of the consequences of these ongoing advancements is technology obsolescence, that is, the continual need to replace old hardware and software with new and sometimes very different hardware and software. The same is true for the digital data containers called le formats that hold the 0s and 1s in logical arrangement. File formats evolve over time, and maintaining compatibility among different generations of le formats is not always a priority for technology suppliers. An extreme example of this dynamic is eight-inch oppy disks holding 1980s-vintage documents in the then-popular WordStar word processing format such documents are very difcult to access today. One common strategy for dealing with technology obsolescence is to regularly copy digital data to new le formats and storage media, which is called migration. In practice, migration requires regular capital investment for upgrading hardware and software, and ongoing operational expenditures to reliably perform the data transfers from the old formats and storage media to the new, which in turn requires dedicated staff with specialized training. This strategy works for some applications and has been used for many years as part of the lm preservation process, e.g., creating copies on new lm stock, but for technical and operational reasons, its implementation becomes exceedingly difcult and expensive for the large amounts of digital data generated in motion picture production.
13
INDEPENDENT FILMMAKERS
motion picture data is an indirect access medium there are several technological layers between the digital 0s and 1s that represent the images and the fully realized visible picture.
INDEPENDENT FILMMAKERS
were collected via an online survey form and interviews with approximately 150 indie lmmakers, industry leaders, marketing consultants, lm festival staff and others who are involved with various aspects of independent lm production and distribution. The survey questions were developed and then made available through Filmmaker Magazine, an IFP publication whose editorial content specically targets the independent filmmaker community. The survey was expected to reach a sizeable pool of indie lmmakers not easily accessed through traditional oneon-one or roundtable interviewing methods, and was expected to produce representative views, but not necessarily statistically accurate or precise data. Subjects agreed to be interviewed without attribution, and their wide-ranging viewpoints are important contributions to developing an understanding of the independent lmmaking community and how its digital dilemma might compare to the one facing the major studios. In the Filmmaker Magazine story that accompanied the survey, indie lmmaker Tom Quinn (who wrote, directed and produced his debut feature The New Year Parade about the famous Philadelphia Mummers) foreshadowed much of what was about to be discovered by noting: archiving is not a topic that comes up often when filmmakers trade war stories. 12 The following sections summarize the survey and interview responses, and provide selected survey data as well as representative views on the storage and handling of materials, the costs of and responsibility for preservation, and the development of alternative distribution mechanisms. The survey results are reprinted in their entirety in the Online Appendix. The open-ended questions elicited both consensus views and conicting opinions, and representative statements were selected for inclusion where appropriate. Survey respondents were asked to identify themselves by job function. They were allowed to select more than one job function, since many independent filmmakers wear many hats.
14
As previously mentioned, environmental storage conditions are very important for extending the life previouslyand digital videotape as well as lm. Slightly more than half of respondents maintained As of analog mentioned, environmental storage conditions are very important for extending the at least some and digital videotape as well as lm. Slightly more than half of respondents maintained life of analog level of environmental controls for both their production and master-level content. It is importantlevel of environmental controls for a shelf in aproduction and master-level content. at least some to note that storing hard drives on both their cool, dry room, as some respondents reported, is notto note that storing hard Hard drivesshelf in aoccasionally powered on respondents It is important a preservation strategy. drives on a must be cool, dry room, as some to maintain internal lubrication, and data integrity must be periodically validated. reported, is not a preservation strategy. Hard drives must be occasionally powered on to maintain internal lubrication, and data integrity must be periodically validated.
15
INDEPENDENT FILMMAKERS
content on computer hard drives, and more than half (52 percent) used motion picture lm.
INDEPENDENT FILMMAKERS
Many respondents noted the substantial and ongoing expense of migration: hardware and software have half-lives of only 5 to 10 years at most. We will have to upgrade our digital copies every 5 to 10 years. Unless the cost of those upgrades is nominal (and you know it NEVER is), it will cost a lot more than preservation of lm elements.
iv
16
I think its out of respect for the effort that went into the project and the medium that I preserve the lm at my own expense. Others suggested sharing the costs and responsibility for archiving, which could be approached as a cooperative venture: It would be great if a collective were started to house digital transfers of lms in one central server location. We need archives that make it easy for independent image makers to donate their work. And those archives need to have the wherewithal nances, storage space and staff to preserve the work and store it for the very long term. Im really terried that once I die, all the work Ive created will vanish with me. Also suggested was a free service: There should be a place a library or other institution where we can have our works archived. It should be free and sponsored by a major cultural institution. The survey answers and comments reect a substantially expressed view that the burden of preservation should be borne by other individuals and/or institutions, even though most of those surveyed who are preserving their work are also paying for it.
17
INDEPENDENT FILMMAKERS
When asked who actually pays for preservation, almost two-thirds of the respondents answered I do. It is worth noting that those who identied themselves as lmmakers consistently gave this answer. Besides an ongoing nancial incentive from potential distribution deals, some lmmakers had other motivations to cover the preservation costs:
Summary
INDEPENDENT FILMMAKERS
One respondent was concerned enough to write: Im beginning to understand the long-term implications of indifference. The conict between cost and convenience was also evident: There is no argument that there are great advantages to digital acquisition and distribution. But it may be that we are accepting short-term advantages with long-term loss. I am very concerned that the next generation will not have the rich archive of historical and classical lms that we have today. Though Id love to shoot on lm, its too expensive for my budgets. When I consider digital preservation, I think its important to remember that nothing is forever, including hard drives and DVDs. Everything needs to be backed up over and over if its held long enough. Thats why we make lms after all, isnt it? To tell stories that will be preserved forever? I am all for protecting my lms and any other lm, for that matter, as long as possible. A survey respondent who stated he is a new father put it in this perspective: I do think about how and when my children will see my work, and in turn, their children. I do consciously try to make statements about my vision of the world and I certainly want to share that with as many people as possible, so of course that means people both now and those unborn.
Summary
The survey responses represent disparate viewpoints, which is to be expected from such a diverse group of creative individuals. While some indie lmmakers recognized the need for and value of preservation for cultural or nancial reasons, getting their created content in front of an audience was their rst and foremost goal.
18
DOCUMENTARIANS 2
Introduction A Brief History of Documentary Filmmaking Survey Methodology, Results and Analysis Summary
19
Documentaries are a distinct type of independent lm. The original concept of a documentary lm was, as its name suggests, to document reality.
20
DOCUMENTARIANS
Introduction
Documentaries are a distinct type of independent film, and this is recognized by honorary and professional organizations: the Academy of Motion Picture Arts and Sciences has a Documentary Branch, whose members have distinguished themselves in short and/or feature-length theatrical documentaries, and the Academy of Television Arts & Sciences includes documentarians in its Nonction Programming Peer Group. As documentary lmmakers generally rely on access to historical footage for their storytelling, a careful look at this segment is required to sufciently represent the independent lmmaker perspective.
21
Documentaries generally have greater access to funding sources such as nonprot foundations.
audience who had never seen photography come to life before. Soon people with cameras would travel the world and visit really exotic places, places people had only heard about but had never seen. Thomas Edison and the Lumire brothers, who did some of this photography, and others lmed these faraway places and the lives of the people who lived there, and the audiences were enchanted. One hundred and ten years later, these kinds of movies are still informative and fascinating. 13 The original concept of a documentary lm was, as its name suggests, to document reality. Also called actuality films, some of the earliest examples captured aspects of daily life.14 This term was rst used in a New York Sun review of Robert Flahertys lm Moana in 1926. Reviewer John Grierson, who later became a documentarian, wrote about these lms cinematic potential. He described the creative treatment of actuality, which predicted reenactments as part of the documentary repertoire. In referring to nonction lms, he included travelogues and instructional lms.15 The invention of the 16mm lm format in 1923 was conceived as a medium for producing home movies. Independent and government documentary lmmakers later embraced it as an affordable and practical medium for their work. Through the years of the Great Depression and World War II, 16mm and 35mm newsreels produced independently and by the studios accompanied the presentation of feature-length narrative lms on cinema screens and in newsreel theaters. Today, television rather than theaters is the primary market for documentaries a result of the growth of broadcast television since the 1950s. In the early days of the medium, local and network news programs were recorded on 16mm lm, and many local television stations also produced documentaries on lm.16 Stations maintained libraries of these recordings, because management realized that lmed content was an asset that could be used in the production of future programming. The impact of documentaries is enormous consider eye-opening studies of social institutions, such as Frederick Wisemans Titicut Follies and piercing, deeply human portraits, such as the Maysles brothers Salesman. Historian Stephen Ambrose observed that more Americans get their history from Ken Burns than from any other source, and indeed, Burnss documentaries on the Civil War, jazz, baseball and Prohibition, have drawn television audiences in the tens of millions.17 It is worth noting that in his nine-part series on baseball, Burns made extensive use of 70- to 80-year-old 16mm black-and-white archival
DOCUMENTARIANS
22
A short list of contemporary documentarians who also made their mark in television includes Robert Drew, Henry Hampton, D.A. Pennebaker, Charles Guggenheim and Barbara Kopple. Today independent documentaries rarely get broad theatrical distribution, but there are notable exceptions, such as Davis Guggenheims An Inconvenient Truth and Michael Moores Roger & Me, Fahrenheit 911 and Sicko. v The rules and rigors of making and marketing documentaries are generally the same as for other independent lms, with certain key distinctions. Documentaries generally have greater access to funding sources such as nonprot foundations, which may underwrite production costs. And broadcast and pay television outlets have become major commissioners and purchasers of documentaries as well as avenues to much larger audiences than could be achieved through theatrical release. Those interviewed for this report said that of the documentaries screened at most independent lm festivals, a large majority have television distribution deals in place prior to the festivals, and those that do not have such deals by that time probably never will. Nonetheless, short and feature-length documentaries have opportunities to raise their proles though festival and theatrical screenings, especially if they go on to receive Academy Awards or other recognition. While documentarians do commissioned work, or work for hire, they can and do create documentaries on spec and then try to sell them. This is no different from a director of narrative lms, who can either accept work as a director-for-hire or raise funds for a personal passion project. Except in those work for hire situations, independent narrative lmmakers and documentarians appear to face common challenges in nancing their projects, getting their work before an audience and securing distribution.
23
DOCUMENTARIANS
newsreel footage, which he augmented with contemporary interviews with historians, surviving athletes and journalists from the period.
DOCUMENTARIANS
24
The environmental conditions under which analog and digital source materials are stored are known to have a major impact on the lifespan of recorded media, but among the surveyed groups, use of temperature and humidity-controlled storage for master-level content remains relatively low: 18 percent of AMPAS respondents,12 percent of ATAS respondents and 8.5 percent of IDA respondents store their source materials in climate-controlled environments. Answers of no climate control and dont know were 38.7 percent for AMPAS, 37 percent for ATAS, and 50.7 percent for IDA. Overall, documentarians responses on this topic were not substantially different from those of the larger independent lmmaking community.
25
DOCUMENTARIANS
DOCUMENTARIANS
26
The few respondents that commented on this question were aware that digital copies are not a guarantee of long-term access: I am not sure any format is really safe. I know that it is not as secure as when stored as lm. I feel depressed and hopeless because there seems to be no intelligent agreed upon solutions for long-term archiving.
DOCUMENTARIANS
However, when presented with a multiple-choice question about their feelings about data migration for these purposes, none of the respondents selected Its a bad idea. Substantial numbers selected I am relieved (22, 40 and 43 percent for AMPAS, ATAS and IDA, respectively), and smaller numbers selected I have some concerns (11, 38 and 36 percent). The range of responses suggests a general lack of awareness regarding the risks associated with migration the risks of data corruption and other process errors, as well as the risks of doing nothing at all.
Summary
DOCUMENTARIANS
period will always be available. In the interviews, however, documentarians did not seem to recognize that when todays current events become tomorrows history, an audio or visual record of these events may not exist, because todays digital systems do not offer guaranteed long-term access. In other words, 25 years from now there may not be much usable archival footage pertaining to the world as we experience it today.
Summary
While documentarians are a distinct type of independent lmmaker, their experiences and practical challenges are similar to those of narrative lmmakers because of the widespread adoption of digital production and post-production tools. In both surveys and interviews, documentarians cited several benets of digital technologies, including greater freedom to shoot covertly, more editorial exibility and better compatibility with digital distribution platforms. Some recognized that digital lms are not forever, but they generally did not take steps to ensure long-term access to their completed work. Moving on to the next project seemed to be a higher priority.
28
29
While digital technologies have made it easier for independent lmmakers to create their movies, they have also fractionalized distribution channels.
30
31
their lm would capture a theatrical distributors interest. Today, indie lmmakers have a variety of digital distribution options to consider that may result in their work reaching an audience and generating revenue. Given that archival activities typically occur for independent lms after distribution, the changing nature of lm festivals and the rise of self-marketing and self-distribution were deemed important research topics for this report. Motion picture marketing and distribution are two different disciplines requiring different skill sets, but both are needed to connect a movie and its audience. Marketing is the strategy and tactics for creating anticipation among members of a movies target audience; distribution is the mechanism that gets the movie in front of paying customers. As the surveys and interviews for this report showed, unless todays indie lmmakers perform both functions, they will likely have to pay for outside expertise in these areas. With or without expert help, most interview subjects still considered lm festivals as the starting point for independent lm marketing and distribution. Film festivals are organized, multiday events, usually in a single geographic locale with several screening venues. Festival programming ranges from celebrity-focused Hollywood premieres to showcases for specic genres, formats, subjects or countries of origin; the offerings vary by program length (feature versus short form) and sometimes include retrospectives, tributes to individual artists, lmmaker panels and other events. For the better part of the last 80 years, festivals have been the launch platform for many talented lmmakers seeking independent or studio-backed distribution for their work. But as festivals explore broader and broader programs, the number of available screening slots often decreases within each category and genre. The challenge of raising an indie lms prole becomes more difcult when the lm must compete with other attractions that are key to promoting the festival itself.
Today there are hundreds of lm festivals around the world that offer independent lm screening opportunities. The major North American festivals include the following: The Toronto International Film Festival, which started in 1976, is the most widely attended lm festival worldwide. The Seattle International Film Festival runs for more than three weeks, and in 2011 it screened 450 feature lms and approximately 150 short lms.21 The Los Angeles Film Festival showcases more than 70 new works in American and international cinema, along with premieres, preview screenings and lm education programs. New York Citys Tribeca Film Festival, founded in 2001 by Robert De Niro, Jane Rosenthal and Craig Hatkoff, has a mission to assist lmmakers with reaching the broadest possible audience, as well as promoting New York City as a major lmmaking center.
The Telluride Film Festival in Colorado and the South by Southwest Film Festival in Austin, Texas, are also considered signicant festivals for independent lm. Documentaries are also well represented at lm festivals, and in their book A New History of Documentary Film, authors Jack Ellis and Betsy McLane credit festivals as the platform where most documentaries that have been produced outside the mainstream are discovered for cinema, television and alternate distribution. Notable documentary-focused lm festivals include Torontos Hot Docs and the Full Frame Documentary Film Festival in Durham, North Carolina.22
33
to Park City in 1981, adjacent to a major ski resort. Redford founded the Sundance Institute that same year, and its Feature Film Program and Documentary Film Program now offer several labs annually to independent directors, screenwriters, producers and composers.19 The Sundance Institute Documentary Fund provides grant support to documentarians exploring contemporary issues, and the Institute also maintains the Sundance Collection at UCLA to conserve and archive independent lms.20
The 2010 New York Film Festival received about 1,700 submissions but screened only 138, including short (three-minute) avant-garde lms (8 percent) The 2011 Los Angeles Film Festival received 4,521 submissions and screened 153 (3 percent) The 2011 South by Southwest Film Festival received 4,900 submissions and screened 293 (6 percent) The 2010 San Francisco International Film Festival received 3,200 submissions and screened 181 (6 percent) The 2011 Chicago Film Festival received 3,640 submissions and screened 194 (5 percent) The 2011 Telluride Film Festival received 1,423 submissions and screened 84 (6 percent) 23
As more indie lms are made and festivals explore programs that extend beyond new indie productions, indie lmmakers are nding it more and more difcult to get their movies seen at these events. This dynamic was articulated by one of the surveyed lmmakers: Today, with the new electronic cameras and editing systems, its easier to make an indie feature, but harder to get (it) seen. Despite the low acceptance rates, festivals are still considered a major force in getting indie filmmakers exposure. If screenings are successful, they will generate a following for the lmmaker and enough buzz around the lm that could lead to a pick-up by a distributor. Most people recognize that it is very difcult to make a living as an independent lmmaker, and that those challenges underlie indie lmmakers rush to secure distribution. The revenue from each lm they make must pay off investors, post-production facilities, equipment rental houses, and cast and crew members who deferred compensation; it must also become the funding stake for their next project.
34
These new distribution mechanisms are raising questions about independent feature lms traditional route to theaters, as a marketing professional responding to the survey noted: The industry is changing, and the old theatrical exhibition model may not work for the indies. We may see day-and-date (simultaneous) release in theaters and online. According to marketing consultant Cheryl Boone Isaacs, Once the marketing strategy is set, a digital consultant can be brought in to interact with digital aggregators to enable a lmmaker, production company or other rights holder to understand their cash outow in advance, evaluate costs for expenses and marketing services, gain access to collection stats, and review their collected funds balance from retail platforms, such as iTunes, Netix, Hulu or cable VOD operators like Comcast, Time Warner, etc.26 Aggregators charge a commission for this service, but even with commissions, marketers claim that these new do-it-yourself (DIY) distribution options can lead to potentially greater income for independent lmmakers. Filmmakers attempting to realize the potential benets of these new distribution opportunities have to add yet another skill to their multidisciplinary role that of marketer. Filmmakers and independent marketing consultants interviewed for this report stress self-marketing as a new need, because compared to traditional theatrical distribution, the barriers to entry for electronic distribution are relatively low, and quite often no sales agent or other middleman is required to effect the sale. However, most marketers interviewed for this report still recommend using a sales agent for foreign distribution, as the DIY approach is difcult to execute given the territorial fractionalization of the international marketplace. Dispensing with the middleman requires lmmakers to take a detached view of their work. This is not always easy, because it takes passion and dedication to get a lm from concept to completion. Marketers interviewed for this report say their rst question to a lmmaker always is: Who is the audience for this movie? All too often, they say, lmmakers have no clear answer to this question, and therefore no apparent market for their lms. During the research phase of this report, many indie lmmakers were seen at various lm festivals pitching their projects from computer laptops and tablet screens on the trunk of a car or inside a van. And the predominant question asked by viewers was: Who is it for? Who will want to spend money and two hours of their time to see this? The interviewed marketers also stated their belief that indie lmmakers should make features that have long lives to take advantage of the Long Tail theory of retail sales. First covered in 2004 by Chris Anderson in Wired magazine, the Long Tail theory proposes that large amounts of digital content for which there is low demand can collectively generate signicant revenue, for very low distribution costs, in a digital delivery environment such as the Internet. Those interviewed for this report believe that the theory applies to all entertainment content, including independent lms and documentaries. The belief in the existence of future markets provides both an incentive and obligation to future-proof for future revenue. Whether the Long Tail theory proves true or not, given that copyright protection lasts for 95 years (or longer),27 there seems to be sufcient commercial
35
losses from unauthorized duplication of electronic versions of a movie, and the impossibility of putting an end to the copying once it starts.25
justication for indie lmmakers to be concerned with preservation for long-term access to their created digital content.
36
37
As audiovisual content becomes increasingly digital, all archives are now facing, or soon will face, the challenge of digital preservation.
38
39
proactive collecting activities. The proactive effort usually brings in independent lms long after their commercial lives are over. These archives are, in many cases, the repositories of last resort. A cross section of archives was surveyed to develop an understanding of how archives acquire independent lms and how digital storage technologies are impacting their operations. The archives were categorized by types of content they collected, collection focus, organizational model (stand-alone or part of a larger institution such as a library or museum) and whether the archives were actually handling and storing digital materials. A smaller sample of archives was selected for direct interviews, and four archives, somewhat typical and representative of the whole, were chosen for in-depth case studies to more fully investigate how independent, public and nonprot archives are dealing with their digital dilemma.
40
The National Film Preservation Act, rst passed in 1988... codied the importance of motion pictures as an artform and a record of our times.
The ve largest lm archives in the United States the Library of Congress, UCLA Film & Television Archive, George Eastman House, the Museum of Modern Art and the Academy Film Archive handle nitrate lm on a broad scale. This capability primarily distinguishes these institutions from all other U.S. lm archives.37 In the 1960s, during its rst years of existence, the American Film Institutes media campaign articulated the need for lm preservation. The campaign noted that the major repositories collecting and preserving Hollywood feature lms were not motivated by the lms potential commercial value, but for their artistic or cultural value. It is interesting to note that beginning in mid-1980s, in anticipation of future revenue from the nascent home video and pay television markets, the major Hollywood studios began investing considerable resources to extend the life of their audiovisual assets by building new storage facilities, upgrading existing ones, and restoring and copying these assets to new recording media. The National Film Preservation Act, rst passed in 1988 and updated several times since, codied the importance of motion pictures as an art form and a record of our times. 38 The Act and its subsequent revisions established the National Film Preservation Board, the National Film Registry and the National Film Preservation Foundation (a public/private partnership), as well as articulated a national plan to preserve the countrys lm heritage.39 Audiovisual archives, especially the nonprot public archives, came into being because archivists, curators and historians accepted that moving image and recorded sound materials were as legitimate a part of societys cultural heritage as books, documents, photographs and other media. As a result, the universe of audiovisual media preservation extends well beyond Hollywoods borders. In the last few decades, it has become apparent that many of these archives have custody of lms that have been ascribed permanent value, i.e., enduring historical and cultural signicance, even if these lms
41
It is important to note that new productions of documentaries about Americas participation in World War II and other historic events would be difficult, if not impossible, to create without continued access to holdings of the National Archives, the Library of Congress, UCLA Film & Television Archive (which holds the Hearst newsreel collection), the University of South Carolina (which holds the Fox Movietone collection) and other audiovisual archives around the world. However, despite the extraordinary efforts of these institutions, silent-era newsreels survive only fractionally as remnants of a once innovative and widely accessed news source for the general public.
did not generate much revenue in theatrical release. The various collections range in size from a handful of rare lms, such as home movies acquired along with personal papers of important historical or cultural gures, to collections consisting of thousands of reels from large, complex, local and national television news libraries. In a large institution, preserving lm collections may represent only one of many important programs and activities, such as book rebinding, paper conservation and digitizing photographs. Most specialized lm collections closely support an institutions mission. For example: The Country Music Hall of Fame and Museum holds lms about leading country music performers and culture. The Hoover Institution holds lms about the history of communism and Eastern Europe. The Human Studies Film Archives of the National Anthropological Archives (Smithsonian) holds ethnographic and anthropological lms about the worlds cultures and peoples. The U.S. Holocaust Memorial Museum holds lms about the history of Jewish life in Europe during the Holocaust. Northeast Historic Film holds lms about the history of and life in the New England region. Anthology Film Archives holds lms with a specialized focus on avant-garde, experimental and independent lmmakers. Motion pictures have been used to not only entertain, but to document and communicate almost every aspect of American life since the turn of the last century. In the United States, the widely dispersed and diverse use of motion picture lm and now, digital recording technology, has made motion picture preservation a highly decentralized activity spread across hundreds of institutions and organizations.
42
Digital preservation from the audiovisual archivists perspective must rst begin with the basic concepts of archiving, of which preservation is just one part. Archival work requires actions such as appraisal, arrangement and description (cataloging), preservation, management and providing access. For analog materials, preservation is an umbrella term that includes conservation (storage in archival conditions), preservation (cleaning and reformatting) and restoration (proactive work to return the item to as close to its original state as possible). The critical conservation step means essentially do no harm. If analog materials are stored in a cold, dry environment in appropriate containers, their life expectancy will be extended with minimal human intervention.40 Also known as the store and ignore approach, this relatively passive strategy is not possible with digital media. Recent denitions of digital preservation by professional associations stress le management and related actions. For example, the Association for Library Collections and Technical Services denition reads: Digital preservation combines policies, strategies and actions to ensure access to reformatted and born digital content regardless of the challenges of media failure and technological change. The goal of digital preservation is the accurate rendering of authenticated content over time. 41 The United Kingdoms Joint Information Systems Committee denition reads: Digital preservation is the series of actions and interventions required to ensure continued and reliable access to authentic digital objects for as long as they are deemed to be of value. This encompasses not just technical activities, but also all of the strategic and organisational considerations that relate to the survival and management of digital material. Digital objects will cease to be accessible without active management and intervention. The biggest risk to the accessibility of digital objects is the continual development of computing hardware and software. Many digital les or formats are dependent upon a particular computing environment for accurate presentation of their content. Any change to the rendering environment could result in change to the rendered representation of a resource (or result in not being able to render the resource at all). 42 If anything, the term digital preservation is a misnomer. Some use the term digital archiving, which is different from what traditional information technology (IT) staff consider archiving (regular backup to digital data tape with or without verication steps). Others prefer data curation, which emphasizes the active management of a digital le throughout its life cycle.43 Whichever term is used, preserving digital media is an active process that comprises all stages of traditional archival work and important additional actions, some of which are very complex. Some of these additional actions must be taken even before the digital audiovisual data is created selecting le formats and storage media, for example, and considering data-handling workows that facilitate downstream preservation. In traditional analog archiving, archivists typically face substantial backlogs of incoming materials, but they can appraise and catalog them well after they are received, because analog objects are generally
43
able to remain in an input queue for long periods without decaying. By contrast, digital collections must be appraised and cataloged at the time of their creation, because there is insufcient time, resources and technical information available to process them for guaranteed long-term access once the digital collection reaches the archive. The fundamental difference is that the unmanaged life expectancy of digital materials is much, much shorter than that of their analog counterparts. Several other actions are required to preserve digital media: copying to new media and le formats (migration), maintaining redundancy and verifying data integrity, as well as scheduling and managing the actions themselves. A further consideration is that assessing and cataloging digital collections require specialized hardware and software tools that may vary from collection to collection, and these tools require specialized technical skills to operate. For purposes of this report, the term digital preservation is used to refer to both reformatting (using digital techniques to preserve analog or digital originals) and preserving the digital les themselves.
44
There are many types of les (also called le formats), many variations on individual le formats and many kinds of codecs. The survey respondents identied 26 different moving image le formats with 15 different codecs, and 15 different audio le formats with 6 different codecs. Many of these formats are supported today by commodity operating systems running on popular computer platforms, but some require specialized software and/or hardware. Accessing these les requires at least basic computer literacy, and in some cases more technical skills, especially when it comes to long-term access. Metadata is organized by the specic function it serves: technical, descriptive, administrative or preservation. Among the several metadata schema that are used in archival applications, PREMIS (Preservation Metadata Implementation Strategies) and PBCore (Public Broadcasting Core) are two of the most common. Essence data is quite often wrapped together with the metadata; MXF (Material eXchange Format) is an example of such a wrapper, which may also be referred to as a le format. Technical metadata is most easily generated at the time of content creation, because the hardware and software that create a digital le know the les technical details. They are therefore able to incorporate this information, even though they may not be designed specically to do so. Other types of metadata are usually generated by archive personnel, but as previous studies
45
any way, that is to say, excessive cleanup of the original material or introducing digital artifacts should not be allowed.
have shown, the earlier the metadata is created in the production process, the more likely it is to be created at all.44
46
Respondent Proles
Twenty-one archives with audiovisual holdings completed a detailed survey on their digital preservation activities. These archives represent a cross section of institution size, mission and collection focus: Academic Film Archive of North America Anthology Film Archive Archives of Appalachia, East Tennessee State University* Berkeley Art Museum and Pacic Film Archive Film & Media Archive, Washington University in St. Louis* Franklin Furnace* Getty Research Institute Harvard Film Archive Hoover Institution, Stanford University New York Public Library, Dorot Jewish Division New York Public Library, The Rodgers and Hammerstein Archives of Recorded Sound Northeast Historic Film Stanford University Libraries and Academic Information Resources University of Texas at Austin, Harry Ransom Center University of Virginia Library Walter J. Brown Media Archives and Peabody Awards Collection, University of Georgia* Wisconsin Center for Film and Theater Research Witness Anonymous (3 archives) Archives listed with an asterisk (*) also took part in the in-depth case studies for representative views from a diverse group of audiovisual archives, and these case studies are included in the Appendix. Of the surveyed archives: 10 identied themselves as university-based archives 5 as independent nonprot organizations 2 as state archives 2 as public libraries 1 as a museum 1 as an independent moving image archive The surveys and in-depth case studies requested detailed information from the archives on:
47
The nature of their collections What content types they are digitally reformatting for preservation How they are digitally reformatting content (le formats and technical details, in-house vs. outsourcing) What content they are receiving as born digital les Their digital infrastructure Policies for the preservation of digital materials Funding strategies The following sections summarize the nonprot audiovisual archives survey and interview responses regarding the most critical issues they seem to be facing today. Direct quotes were edited for readability without altering the respondents intent. For more details, the complete case study reports are provided in the Appendix, and the survey data is provided in the Online Appendix.
48
As described earlier, digital content is handled according to its intended function: digital les created to provide user access to an archives holdings do not need to be handled with the same diligence and care as les created for preservation purposes, because they are simply working copies that can be re-created if they are lost or damaged. Surveyed archives reported that the desire to provide digital access to their holdings usually got their rst toe in the digital pool. They began to build digital capabilities in response to this need, but they generally did not build sufcient infrastructure at that time to meet the requirements for digital preservation. Digitizing for access The following table illustrates how the surveyed archives ranked their motivations for digitizing for access, with 5 being most important and 1 being least important. Importance of content and user request received the highest rankings.
The archives were also asked to rank the types of moving image and audio content most often digitized for access purposes. Their top 10 responses are summarized in the following tables.
49
Digitizing for preservation In selecting the content to be digitally preserved, the surveyed archives have been applying a mix of criteria that include uniqueness or importance of content, format obsolescence, and condition of the original object. Uniqueness is a key metric in the selection process. If an object was on an obsolete or deteriorating format, but was not unique, an archive would consider digitizing the item for access but not necessarily for preservation. If an item was unique, and the format was obsolete or deteriorating, it was a prime candidate for preservation. The following table illustrates how the surveyed archives ranked their motivations for digitizing for preservation, with 5 being most important and 1 being the least important.
As with access, the archives were also asked to rank the types of moving image and sound content most often digitized for preservation:
50
The surveyed archives have been preserving unique moving image items such as eld recordings, lectures, oral histories, interviews, performance documentation and home movies, as well as orphan works. Orphan works are those works for which the copyright holder cannot be located, or the creator lacks the means to preserve what limited copies exist (works such as an experimental lm or video).45 The uniqueness criterion explains why broadcast television content was not among the top 10 content types digitized for preservation, although it ranked high in digitizing for access. Duplicate copies of older, previously-aired television programs are quite often held at several different archives as well as at the originating network or television station, so such content is generally not a priority for preservation. As with moving image content, the majority of audio content that was digitized for preservation was selected because of its uniqueness. An interesting audio preservation dynamic was noted by one archive with collections of independent lmmakers production elements. If there wasnt
51
...analog recordings made more than 100 years ago are more likely to survive than digital recordings made today.
enough funding to preserve an entire lm, the archive digitized the soundtrack for preservation, because soundtracks are often recorded on magnetic stripe lm, which is prone to deterioration by a chemical mechanism called vinegar syndrome.46 Summary Digitizing original audio materials for preservation was widely acceptable to the surveyed archives, and they were in agreement on a preservation le format and technical specications (discussed later in this report). Surveyed archivists said that standards work by the Audio Engineering Society (AES) and other organizations made them comfortable with their technical decisions regarding preservation of digital audio les. It is worth noting that a standardized audio preservation le format is not sufcient to ensure long-term access. The State of Recorded Sound Preservation in the United States: A National Legacy at Risk i n the Digital Age was published by the Library of Congresss National Recording Preservation Board (NRPB) after the surveys and interviews for this report were completed. The studys ndings state that newer content, such as born digital audio, is at greater risk of loss than older recordings on such media as 78 rpm discs, and that analog recordings made more than 100 years ago are more likely to survive than digital recordings made today. Institutions, archives and libraries hold an estimated 46 million recordings, yet the study found that major areas of Americas recorded sound heritage have already deteriorated or are inaccessible. One of the NRPB studys conclusions is that the advent of digital technologies and distribution platforms has made preservation issues and access issues for sound recordings inseparable.47 The authors of the NRPB study state that digital technology alone will not ensure the preservation and survival of the nations sound history. 48 This is consistent with the findings of other digital preservation studies: policies and procedures that ll technological gaps also must be put in place to achieve long-term access to digital materials. For moving image content, the surveyed archivists agreed that lm should continue to be preserved as lm, that is, new lm copies should be made if the older ones deteriorate. One reason for this may be that under proper storage conditions, lm has a longer life expectancy than electronic media. The archivists also felt that more visual information is held in a lm frame than could be digitally captured with their current technical capabilities. Digitizing video originals for preservation is a gray area. Surveyed
52
Reformatting Content
Some of the surveyed archives have been practicing a two-pronged approach to preserving audio and moving image content, preserving both digital les and analog audiotape and videotape. Although they acknowledged that tape stock supplies are dwindling, they were comfortable with analog preservation and were reluctant to abandon the practice unless they had no other choice. If audiotape and videotape originals existed on fully obsolete media, or required unavailable or unserviceable playback equipment, these archives digitally reformatted them to provide a serviceable master record for the near term. The surveyed archives reported that they are also reformatting born digital content. The main reason they did this was immediate usability: One thing weve experienced with born digital formats from a few [digital] camera types is the need to up-convert the original so that it can be edited [for access purposes]. At present we archive both the original format and the up-converted les. Four archives mentioned that on occasion, they have rejected born digital content. They gave a variety of reasons for doing so:
Digital video les harvested from the Internet may carry viruses that could damage a collection The items had been removed from their original context, so no provenance information was available There were too many le formats to juggle Their parent institution had not yet supplied a repository with a proper digital asset management system to store, preserve and stream these materials They lacked the software to view the digital les
The surveyed archives received born digital oral histories/interviews, eld recordings and lectures, with moving image le formats outnumbering audio le formats. Interviews suggested two possible reasons for this larger number of image le formats are digital videos ease of use for acquisition and the fact that there are more moving image formats than audio formats.
archivists believed that Betacam SP and Digital Betacam videotape stock might be harder to acquire in the near future, and therefore they felt that they must begin digitizing their tape-based video holdings. The lack of a clear standard preservation le format and technical specications for digitizing video was unsettling to them, but it was not the main factor inuencing their digital video preservation decisions. As will be discussed later in this report, other factors such as storage capacity and technical infrastructure were stronger inuences.
The immediate need is typically the creation of access or preservation les; designing and implementing a digital preservation program generally happens much later. Whether they were focused on access or preservation, the surveyed archives had varying digitization capabilities: 13 of the 21 archives (62 percent) could create digital preservation les from 17 different analog formats in-house. Five archives reported digitally preserving analog video, and only one could make preservation les of lm. The creation of relatively large quantities of digital audio les is primarily due to archives widespread in-house capability to digitally preserve audio content. Several of the surveyed archives built in-house digitization labs to create digital preservation les for their audio and/or video originals. Three of these archives received grants that enabled them to buy equipment to build in-house labs, and two others mentioned receiving funds from donors. Another three said developing a digitization and digital preservation program was a necessity, because their original archival objects were on obsolete or fragile analog media. In these cases, the equipment for an in-house lab was supported by either the parent organization (two archives) or by the archive itself (one archive). One archive built upon its institutions established still-image digitizing infrastructure. Two specically mentioned having a good relationship with their institutions IT department, which would be able to help with digital storage issues. A table listing all film, video and audio formats mentioned by survey respondents and where content is digitized can be found in the Online Appendix.
Survey Methodology, Results and Analysis Lack of funding Lack of technical infrastructure Lack of trained staff Lack of institutional support File formats and technical specications The surveyed archives were asked to list the le formats and related technical specications they used for reformatting moving image and analog audio originals. They were asked to provide information regarding archival originals (original analog sources that are reformatted for preservation) and born digital les for three use cases: Preservation: Creating a digital le from the archival original at the highest image quality, and using the least amount of data compression. Preservation les are not regularly accessed, in part because their large size makes them difcult to manage. Mezzanine: Creating a digital le that is the equivalent of a copy master that is used to make other duplicates, generally for viewing purposes. A mezzanine le generally uses a data compression scheme to reduce its le size for easier handling, with an often corresponding (but acceptable) reduction in image or sound quality. Access: Creating a digital le that is provided to users for content access. An access le uses a consumer-friendly le format and is typically highly compressed. Because mezzanine les are transitional by nature, the following sections discuss ndings related to preservation and access les only. Moving image preservation le formats In total, 16 of the 21 survey respondents (76 percent) created digital moving image les. Of the archives responding to this question, all except one preserved lm as lm; the sole archive that digitally reformatted lm for preservation digitized at a pixel count of approximately 2K x 1K and stored this data in DPX les.viii Surveyed archives gave these reasons for preferring photochemical preservation: Films long life expectancy if properly stored The belief in the superiority of lms image quality over that of its digital surrogate The high cost to digitize lm and the attendant high cost to store and manage the resulting digital les Lack of a standard digital preservation le format
viii
DPX les are digital image les conforming to SMPTE 268M-2003, a standard published by the Society of Motion Picture and Television Engineers.
55
Although some archives reported digitizing analog video for preservation, there was no consensus on the particular le format and technical specications. The archives not digitizing analog video for preservation preferred creating analog tape preservation masters from their videotapes for several reasons: Familiarity with analog tape formats Comfort with storing easily accessible physical objects on a shelf Lack of a digital infrastructure Belief that analog transfers are more true to the information on the original media Table 7 lists the le formats (and technical details, where provided) used by the surveyed archives for moving image preservation masters. The lack of consensus on a standard format for moving image preservation is underscored by the use of 12 different le and tape formats with varying technical parameters.
Moving image access le formats All 16 archives that created moving image digital les also created access les. Surveyed archives with a more sophisticated technical infrastructure provided online access to these les, and those without such infrastructure created DVD copies for on-site access.
56
Audio preservation le formats In total, 15 of the 21 archives surveyed created digital audio les. These archives were clearly comfortable with digitizing audio for preservation purposes. The technical parameters of 96 kHz/24-bit were used for preservation in 10 out of the 15 digitizing archives, and all have settled on the use of the WAV format and the closely related BWF (Broadcast WAV format). Archives interviewed in the case studies noted that the International Association of Sound and Audiovisual Archives (IASA) technical recommendations inuenced their decisions on le format and technical parameters.49 These archives believed that there are established standards for audio digital preservation, whereas none yet exist for moving image digital preservation.
Audio access le formats All 15 of the surveyed archives that created audio digital les also created access les. One archive made only access les (no preservation les). The availability of online access versus
57
physical media-only access was similar to moving image access les: archives with a more sophisticated technical infrastructure provided online access, and those without provided CD-only access. The table below lists the digital audio formats used for access les, ranked by total number of archives using each format:
Data storage As stated previously, le size is one factor that determines how archives choose to store their digital preservation les. Cost and technical support requirements are two others. Larger archives, or archives that are part of a larger institution, generally have more internal nancial support than smaller, stand-alone archives. If the larger archives do not use their parent institutions IT infrastructure for data storage, they generally have the nancial resources to purchase and operate their own enterprise storage systems and data tape drives. By contrast, archives with limited resources or no parent institution must nd affordable means to store their preservation les. Surveyed archives with limited resources stated a preference for using inexpensive portable storage devices, such as external hard drives or recordable optical media, which they believed could be as easily stored and managed as the analog media in their archives. From a digital preservation perspective, it can be more convenient to store digital les on fully managed enterprise storage systems equipped with data tape drives than on physical carriers such as portable hard drives or recordable optical media. Storing les on disparate media types, such as consumer-grade portable hard drives, digital tape, and independent computer workstations or servers, requires archives to perform different digital preservation actions for each media type and system. This is extremely labor-intensive and increases the likelihood of human error. Physical carriers also degrade over time. Hard drives must be regularly powered up to distribute internal lubrication and perform error detection operations; leaving a hard drive in a powered-down state for an extended period increases its risk of malfunction and may reduce its operational lifetime. Surveyed archives were asked to provide their current and projected digital storage requirements, which are shown in Table 11. While continuing increases in the storage capacity and price/performance of digital storage technology will likely offset some of the additional hardware requirements, the projected 1,500 percent growth in archives cumulative storage requirements between 2009 and
58
The surveyed archives that rely on physical carriers clearly wanted fully managed enterprise storage systems, but they lacked the necessary funding and technical expertise. Their comments articulated their conundrum: LTOix with fast RAIDsx. Impeded by lack of funds, institutional priorities. A system of RAID servers. We lack funds and technical support to do this. Dont know what system. Funding, funding, funding and technical expertise. I would like to use LTO3 or 4. Lack of personnel, funds and an existing system impede us from implementing a new system. Would prefer more low-cost redundant disk forms (e.g., racked SUN thumpers, or MAID devices like Copanxi). Factors = cost + enterprise support issues. Organizational impediments were also described: Obviously, the current digital storage arrangement for digital time-based media les is insufcient for the burgeoning needs of a world-class university library. We hope to be depositing les in an archive maintained by [the universitys] Information Technology and Communication division. The thing holding us back from this is the slow grind of academic bureaucracy, really. Processes and workows must be determined before anything is stored properly. With respect to capacity, 11 of the surveyed archives had current storage requirements of less than 20 terabytes, and 4 others ranged between 21 and 40 terabytes each. Seven of the archives did not have enterprise storage systems and only used portable hard drives and/or data tape. On average, the surveyed archives expected their storage needs to increase 14-fold over the next ve years as a result of continued reformatting activities and incoming digital collections. Metadata As discussed earlier in this report, metadata is essential to describe, locate and preserve digital les. Technical, descriptive and administrative metadata fully describe digital les, and technical metadata
ix
LTO is the acronym for Linear Tape-Open, an open standard for digital data tape. The numerical sufx indicates the revision of the format, which is updated approximately every two years. http://www.lto-technology.com/index.html.
x RAID is the acronym for Redundant Array of Independent Disks, which is a technique for connecting together several individual hard disk drives for increased capacity, reliability and/or data transfer rates. xi Sun Microsystems, Inc. (acquired by Oracle Corporation in 2010) and Copan manufacture and sell high-performance digital storage systems. MAID is the acronym for Massive Array of Idle Disks, which is a strategy to reduce the power consumption of digital storage systems.
59
2014 will still force the archives to purchase new storage systems and implement comprehensive data management practices to be able to handle the sheer number and types of les.
is essential for proper playback of digital media files. Preservation metadata enables digital preservation by maintaining information such as le validity, access history and migration status. Archives were asked whether they tracked technical and preservation metadata, and if they did, where it was tracked and which standards were used. Twelve archives tracked metadata to varying degrees (shown in Table 12 below). The extent to which they captured and maintained metadata varied widely for most archives, it did not go beyond capturing lenames and storage locations, e.g., identifying the computer server, portable hard drive or other physical media on which the le was stored. This represents a small subset of the metadata necessary to fully manage digital assets for preservation purposes. Since the application of metadata is highly dependent on individual workows and business practices, archives were asked to list standards from which they selected technical and preservation metadata elds (they could list more than one), as well as whether the denitions of their metadata elds were specic to their organization.
It is interesting to note that the metadata source listed most often was Internal, that is, surveyed archives primarily used technical data elds they created on their own. This was followed by PBCore, AES (for audio content), SMPTE and PREMIS, which are the most widely known metadata standards and schema for audiovisual content. Archives were also asked whether they stored their technical metadata in independent databases, embedded in the digital le, or in digital le wrappers such as MXF and METS. Eleven of the 12 archives that tracked technical metadata were using separate databases. At least some of the time, the archives were also beginning to track the technical metadata in more than one location. Seven of the archives that used a separate database also embedded metadata in the le itself, and two wrapped their les in MXF and/or METS. That many of the surveyed archives embedded metadata in the le as well as stored it in a database indicates that they believed doing so aids digital preservation actions. Resource constraints typically kept the amount of embedded metadata at a minimal level, but as one archivist explained, the handling of metadata is evolving:
60
Software tools Tools to manage the creation of digital les, metadata and preservation actions are essential for any preservation program. Several archives wrote in their surveys that even though they were digitally preserving analog source materials, they did not consider these efforts a formal digital preservation program because they had neither the proper storage infrastructure nor the means to manage and preserve the digital les. The surveyed archives tracked their digital content in different ways; very few had a digital asset management system that held both the metadata and the digital media les. It is interesting to note that only one surveyed archive had tools specically created for digital asset management and some level of digital preservation (Artesia and Ex Libriss DigiTool). Interestingly, this archive also reported that it had not yet established digital preservation policies. Less than half (8) of the 18 archives that handle digital les were using some form of database or DAMS to track their digital le actions. Four of these archives used off-the-shelf database management software such as FileMaker Pro and Microsoft Access. The rest used a mix of open source digital repository software (Fedora Commons), a home-grown database, a traditional archive catalog system (MAVIS), and as mentioned earlier, a combination of two commercial DAMS products. In general, the surveyed archives have taken a low-cost database approach to managing the data that describes their digital objects. They believed that it was important to manage this data, regardless of which software was used. Process data backup Six of the surveyed archives stored their digital les on an enterprise storage system and relied on their organizations IT department to backup the les on digital tape. Another three archives stored and managed their digital les on their own systems. Those not utilizing tape backup were nding other ways to create redundancy. For example, some were duplicating les on portable hard drives. A few were creating only one copy of their moving image preservation les, a practice they justified by having insufcient resources to fully manage the characteristically large moving image les. All the surveyed archives creating digital audio preservation les saved redundant copies, a reection of the relative ease of saving multiple copies of smaller les. The archival principle of geographic dispersal, i.e., distributing copies to different buildings or other remote locations to avoid catastrophic loss, was not widely practiced. Only four of the surveyed archives indicated that they practice geographic dispersal.
61
Whether or not the metadata is recorded in database systems outside the le itself depends on available resources for a given project. All les ingested thus far into our preservation repository have extensive metadata wrapped in METS, but we are beginning to rethink this approach, as it tends to be resource-intensive and portends to not scale well.
Process xity checks and le validation When asked which digital preservation actions they performed, the surveyed archives most often reported checksum calculations (a mathematical procedure run on digital les as part of the process of data verication and error detection) and le validation. Performing these actions is not without its challenges: MD5 checksums are run on all les. The checksum data is stored within the FileMaker database along with the other technical metadata that is extracted from the les. And not to complicate things, but the checksums have come up recently as an issue that we havent quite worked through in terms of when to run. They tend to slow down our process a lot, and we need to take that into account. So our process may actually change. Eight of the 18 archives that stored digital les performed xity checks via checksums, although none stated that they scheduled the checksum calculations more than once. Three of the archives performed le format validations, but did not necessarily repeat validations on a dened schedule. It is important to note that because digital les are always subject to corruption mechanisms, le format validations and xity checks must be performed regularly for the life of the les.
62
63
Despite these challenges, the vast majority of the surveyed archives (18 of 21) were digitally reformatting analog originals for preservation or access. Older organizations with analog workflows and institutionalized practices reported that they found it difcult to gain internal support for moving from an analog environment to a comprehensive digital infrastructure. They discovered that the shift involved major operational changes and expense to retrot legacy workows, staff practices and mindset.
We have little control over the storage of our preservation materials (it may be managed by our IT people). This may be a point of contention in the future. Cost and institutional barriers will be the main factors that impede or slow down our development of a system. Identifying exact responsibilities between units and departments, as well as cost-sharing issues, will be the major problems with our plans going forward. One archivist suggested creating reports and documents explaining digital preservation from the archivists perspective to help both groups come to a common understanding of digital preservation, as well as to help archivists justify their assertions to IT staff.
Funding
Surveyed archives reported the following on their funding sources:
Archives generally reported very tight operating budgets, although audiovisual archives with a nancially strong parent institution could benet from that strength. A 2006 report by the Metropolitan New York Library Council (METRO), a member organization of almost 300 public, higher education and special libraries in the New York City metropolitan area, found that 46 percent of its members received all or partial digitization funding from their parent institution.50 In 2008, Primary Research Group, a for-prot company that combines research and business trends forecasting, published a report on collection digitization trends in international libraries and museums. The report found that U.S. libraries received 55 percent of their digitization budgets from their parent institution: Spending on digitization has graduated out of the grant-dependent category and is now a mainstay in library budgets for many institutions.51 While an archives digitization budget is often funded by its parent institution, it is dened here as funds allocated to digital preservation activities from the archives operating budget. This is separate from funding that covers digital projects for the institution as a whole. For example, a parent institution might pay for the creation of an in-house digitization lab, but the labs staff and their activities are supported by the archives operating budget. For more detailed information on selected archives funding strategies, see the case studies in the Appendix.
64
Archivists Recommendations
All the surveyed archivists expressed some knowledge of what needed to be done to preserve access to their digital les, and understood the basics of storage, redundancy and migration. Many expressed concern that resource, organizational and technical issues would hinder their ability to design and implement comprehensive digital preservation programs. When asked what they thought should or could be done to assist them in developing adequate digital preservation programs for their audiovisual materials, the archivists responded with the following recommendations: Digital moving image preservation format standards: The archival community has not reached a consensus on a standard digital preservation le format and associated codec for moving images. In pursuit of such standards, archives and standards-setting bodies should not simply aim for a single format and its related technical details, but consider a range of formats that address archives varied technical infrastructures. Recognizing that it might be difcult for archives with limited storage and management capabilities to reformat according to one highest-performance standard, the community should take a multi-tiered approach, described in clear language, so each archive could select the formats and codecs that it could best support. It is worth noting that as of this writing, there are multiple efforts underway that are expected t o lead to a set of digital moving image preservation format standards: the Federal Agencies Digitization Guidelines Initiative Audio-Visual Working Group, the Image Interchange Framework project and the Interoperable Master Format project. Establish in-house or cooperative digital reformatting facilities: Having sufcient equipment and staff for digitizing in standard le formats helps archives with large collections do a larger volume of digital reformatting for preservation and access. Archives that have established inhouse transfer capabilities report lower le-handling costs, which allows them to digitize more content than they would if they outsourced to more expensive digitization services. Those archives without the funding or collection size to justify in-house digitization facilities proposed developing cooperative fee-for-service arrangements with archives that are so equipped. Several of the surveyed archives also suggested building a digitizing co-op, where several archives would pool their resources to build a shared facility. Foster relationships with IT departments: Archives at institutions with enterprise storage capability should be encouraged to build relationships with the departments that manage their institutions digital storage systems. As IT policies and practices for general business operations do not meet an audiovisual archives needs, open dialogue among all parties rather than an us versus them dynamic will lead to more productive dialogue and collaboration. Through this type of collaboration, archivists could learn more about technology and data management, and IT staff could learn about audiovisual archival practices. Develop affordable, fee-based digital preservation relationships with neighboring institutions: Archives without enterprise storage capability must nd other means to safely store their digital les. Given the typical size of archives budgets and the relatively high cost of commercial data
65
Archivists Recommendations
Archivists Recommendations
storage services, the surveyed archives suggested forming partnerships with nearby universities and libraries that have appropriate digital storage infrastructures to provide a minimum level of managed data storage at a price they could afford. Practice geographic dispersal: Archives should store redundant sets of digital les off-site. If they cannot afford off-site storage, they should explore partnerships with other archives in which each participating archive stored copies of anothers les. Varying administrative, security and liability policies present signicant challenges in this scenario, but however it is achieved, archives should practice geographic dispersal to lower the risk of catastrophic loss. Archive-oriented metadata and le management tools: Archives recognize that along with managing their digital media les, they also need to create and manage various types of metadata to preserve those les. They would like to see software tools designed specically for audiovisual archives rather than production-oriented DAMS or systems built from generic database management software. They also want their digitizing software to produce usable metadata in accepted schema such as PBCore and PREMIS. Further education for audiovisual archivists and managers involved in digital preservation: Many of the surveyed archivists understood the two basic digital preservation concepts: storage and backup. But most did not have the training or knowledge needed to ensure preservation of their digital objects for the long term. Archivists want a deeper understanding of the principles of digital preservation, especially with regard to audiovisual collections; they believe this will help them in fundraising, working with their institutions managers, creating partnerships and building digital preservation programs. Suggested approaches include workshops that go beyond the panel discussions that are typically incorporated into larger, more general archival conferences, with a focus on digital preservation concepts, digital preservation standards (to the extent that they exist), best practices and the application of ideal goals and principles to the reality of their individual situations. The Bibliography of this report contains resources that may be helpful to those who are interested in developing digital preservation education programs for audiovisual archives.
66
67
Collaborations and best practices are insufcient by themselves to resolve the digital dilemma. The underlying technologies must take archival lifetimes into account.
68
69
Progress Report
Progress Report
With this more complete picture of the challenges involved in preserving digital motion picture materials, it is worth reviewing The Digital Dilemmas ndings on industry needs as well as its call for near-term and long-term steps toward sustainable preservation of digital motion picture materials. Through such a comparison, progress can be assessed, and notable distinctions that exist for documentarians and nonprot audiovisual archives can be identied. The corresponding pages in The Digital Dilemma are referenced for convenience.
70
Interim Options
Collaborations: The Library of Congresss National Digital Information Infrastructure and Preservation Program (NDIIPP) is an excellent example of how a diverse group of organizations can be brought together and have their energies focused on a common set of problems. The Library recently published a major report on NDIIPP, which details the programs accomplishments and proposes next steps. One of its more significant new initiatives is the National Digital Stewardship Alliance, which seeks to develop a framework for a national digital collection as well as to strengthen and enable public-private partnerships, among other activities. From the motion picture industrys perspective, NDIIPP provided a context for its own collaboration namely the Academys Digital Motion Picture Archive Framework Project which resulted in productive work on file formats, metadata, open source software and data storage research with contributions from the major studios, manufacturers and the research community. However, the diffuse nature of the independent lmmaking and nonprot audiovisual archive communities makes it more difcult to create a productive, collaborative environment to address their needs.54 Standards Development: The nonprot audiovisual archives surveyed for this report expressed a clear desire for digital moving image preservation standards. It is encouraging to report that as of this writing, there is signicant energy within SMPTE and the U.S. Governments Federal Agencies Digitization Guidelines Initiative being directed toward image le formats and related technical details. Metadata standardization has still not been achieved, but pre-standards development work is underway at several organizations, including the Academy.
Interim Options
Independent lmmakers face the irreversible loss of their work during the lengthy period between commercial exploitation and acceptance by an archive (neither of which is a guaranteed event). Archives that have not yet developed an approach to preserving digital materials will eventually face a crisis as their digital holdings grow, whether through acquiring new collections or reformatting their aging analog holdings. At some point, they will be unable to access some percentage of their digital les. The digital preservation needs of the motion picture industry as a whole have not changed. In this regard the needs of independent lmmakers and nonprot audiovisual archives seem more urgent, given the changing dynamics of theatrical and non-theatrical digital distribution, the duration of copyright protection, the lack of a dened path to an archive and severe resource limitations. It also seems unreasonable to burden nonprot audiovisual archives with a dilemma they did not create, and unrealistic to expect them to organize themselves to tackle the dilemma without help. The studies by the Library of Congress mentioned earlier in this report and the nal report from the Blue Ribbon Task Force on Sustainable Digital Preservation and Access (see Funding section that follows) all contain calls to action for improved funding mechanisms and collaborative efforts. In effect these studies suggest sharing the burden of devising practical solutions that will enable all community members to maintain access to important cultural, historical and artistic works. Until the underlying operational and technology obsolescence problems are solved, however, a renewed and revised call to action in three key areas seems to be justied:
71
Interim Options
Collaboration: Identify key stakeholders and representative organizations that can make it their top priority to collaborate on the most pressing unsolved problems facing nonprofit audiovisual archives. As an organized group, they might take on the following initiatives: Cooperatives and resource sharing: There is a wide range of operational capability, technical infrastructure and financial resources across the archival community and in related elds, and many examples of successful collaborations in which the whole is greater than the sum of the parts. The Library of Congresss NDIIPP effort, the many projects undertaken by program partners, and the Sundance Institute-UCLA Film & Television Archive collaboration that benets independent lmmakers are good examples. Those interviewed for this report most often mentioned shared digital storage, knowledge transfer and staff training as candidates for cooperative efforts. Research and prototyping projects: Many ideas have been put forth by lmmakers and archivists alike, such as revenue-sharing models to support digital preservation, lower-cost approaches to creating lm archival masters, implementing born archival xii production strategies, and reducing the long-term cost of digital storage through the use of open source software tools or even peer-to-peer networks. Ideas such as these are worth investigating as research and proof-of-concept projects that may lead to viable digital preservation technologies and practices. Funding: A recurring theme among all surveyed groups was the lack of funding for digital preservation. The nal report by the Blue Ribbon Task Force for Sustainable Preservation and Access discusses the funding gap at length and offers concrete suggestions for several archival contexts. Representative organizations in each of the communities covered by this report should review the Task Forces recommendations, and where appropriate, coordinate their efforts so that their constituents will be able to take advantage of applicable strategies and partnerships. Documentarian and author Betsy McLane suggests that the only way to assure preservation is to have funding organizations insist that applicants include preservation in their budgets.55 This concept is being implemented in the scientic research community by the National Science Foundation (NSF), which now requires that all grant proposals for NSF projects include a data preservation plan if scientic data is generated as part of the project.56 Education: As long as independent lmmakers and archivists stay on the technology treadmill, they will face the continuous need to refresh and update their knowledge of technologies and practices. The same is true for motion picture producers and executives. Moving image archive programs should offer continuing education classes covering the preservation of digital materials, and archives should require their staff to take them. Industry conferences focused on the content supply chain are good forums for presenting these issues to producers and executives. Likewise, lm festivals are excellent venues at which to provide independent lmmakers and distributors with the knowledge to help keep digital content alive until the content reaches a suitable archive. Finally, college curricula for aspiring lmmakers and archivists should be updated to cover digital preservation issues so that succeeding generations will be prepared to meet the challenges of long-term preservation of digital materials.
xii
Born archival is the concept of creating digital les that are immediately able to be archived, i.e., they are formatted for archiving and all necessary metadata is created at the time of le creation.
72
There is no escaping the fact that digital technologies enable independent lmmakers to explore and extend the art form in ways that are simply not possible with motion picture lm. The price to be paid for these new capabilities, however, is either the loss of content to digital decay, or accepting the responsibility of working with technology providers to articulate and satisfy industry requirements for the long-term preservation of digital data, achieve satisfactory backwards compatibility and implement standards. Collaborations and best practices are insufcient by themselves to resolve the digital dilemma. The underlying technologies must take archival lifetimes into account.
73
Time is and will be the greatest enemy of future access to digital data.
74
CLOSING 6
The Digital Dilemma ended with a call to action for cross-industry efforts to increase cooperation, develop standards and ultimately achieve guaranteed long-term access to digital motion picture materials. From the Academys perspective, there was and is no bias toward one storage medium or system over another, as long as the replacement for the lm archive system meets or exceeds the lm archives performance characteristics. This report focused on those individuals and organizations with substantially fewer nancial and staff resources than the major corporations and institutions covered in The Digital Dilemma. Independent lmmakers and documentarians expressed two primary goals: getting their work seen by an audience and moving on to the next project. Having been made aware of the risks to their unmanaged digital materials, they now have a new concern: ensuring that their digitally created work survives long enough for them to accomplish both. For nonprot audiovisual archives facing the digital dilemma, very little will change until they get adequate funding, stafng and technological support. Key members of the interviewed and surveyed groups generally acknowledged the importance of raising the visibility of these issues at workshops, lm schools, lm festivals, industry conferences and applicable standards development organizations. The underlying problems facing the industry are the constant need to monitor the health of digital data and the mandatory, periodic replacement of digital storage systems and data le formats because of their nite operational lifetimes. 75
Closing
A primary sales pitch by digital storage system vendors is that storage always gets cheaper, and while that has proven true with respect to storage media, price efciency is signicantly offset by rising energy and labor costs as well as more sophisticated technological support systems and data management policies that such systems require. Increases in data storage density show no signs of abating, but history suggests there is little reason to believe that the already huge (and growing) amount of important digital data will somehow manage itself. While some claim that following simple data backup procedures can preserve one or even a handful of digital motion pictures, these procedures do not scale for larger collections held at the hundreds of audiovisual archives worldwide. For these archives, the wide variety of incoming data formats and media types, coupled with the archives limited nancial, technological and staff resources, make it impossible for them to do much more than shelve the material while they wait for the largest and best-funded institutions and organizations across the private and public sectors to solve the problem in a way that might yield trickle-down benets. Time is and will be the greatest enemy of future access to digital data. For independent lmmakers, a longer time-to-market as well as an expectation that the Long Tail theory will prove true for much-needed future revenues requires a digital content sustainability plan. Indie lmmakers ignore the limited lifetime of unmanaged digital data at their own peril. For nonprot audiovisual archives, continued deferment of a comprehensive digital preservation program will result in mission failure. The interim options presented in this report offer some possibilities for temporarily extending the accessibility of digital content. The authors of this report hope that there will eventually be a standardized, globally adopted solution that will address the technology obsolescence issue. Until that happens, and without immediate mitigating action, our moving image and recorded sound heritage is in danger of beginning to disappear in a few years. Facing this danger begins with answering key questions during the production of economically or culturally valuable digital work: Whose responsibility is it to preserve independently created work for future use? What would be the economic and cultural impact of losing this work?
CLOSING
But the broader questions remain: What will it take to create digital preservation standards and achieve their universal adoption? Who will assume the leadership role in solving the digital dilemma for the independent lmmaking and nonprot audiovisual archive communities?
The time for doing studies and dening problems has passed. The issues are clear. The steps to answering these critical questions are also clear, and they start with you, the reader. It is the hope of all who worked on The Digital Dilemma and The Digital Dilemma 2 that the next report will be titled The Digital Solution.
76
77
78
Q: Lets start at the beginning. What led you into the digital realm?
Peter Mavromates: When I was a young teen, there were three things I wanted when I grew up: a Trans Am, a darkroom, and a reel-to-reel audio tape recorder. We wont discuss the Trans Am. Lets call it a youthful indiscretion. I still dont have the darkroom and, with the advent of digital photography, I am somewhat conicted about this. Regarding the reel-to-reel tape recorder, I had this wonderful idea that I would edit together reels with my favorite songs to impress my friends and get rid of the deadwood on many of my record albums. Also, by bumping my records to tape, I would protect the records from the inevitable scratches that would occur from repeated playing. My best preserved albums are, of course, my least favorite albums!
Q: So, how did that lead to your use of electronic media and digital systems? would completely change. I only thought that current technology would simply get better, have more bells and whistles added to it. It never crossed my mind that new technologies would actually supplant old technologies. Now, I re up my computer and a few key strokes later, I am playing only my favorite songs, and no wear and tear. My fantasy of making best of tapes is nothing but a playlist today. Same result, but a different way to achieve it. Needless to say, I no longer need or want a reel-to-reel tape recorder. Q: Director David Fincher was an early adopter of digital image capture and you worked with him on several features. What was the basis for his decision to do this?
PM: My vision of utility was spot on. What I missed entirely was that the way we treated media
PM: As David began to shoot digitally, he would constantly be quizzed about the quality
of lm versus digital capture. He would avoid that question because that is not how he sees things. He looks at the digital capture paradigm honed on his productions and says, This is simply a different way to work. As for my own tastes, I nd magic in all forms of image capture whether they be chemical or electronic. I have videos shot on a Palm Treo of my kids riding a carousel. The blooming quality of the highlights injects nostalgia into the core of every pixel that wouldnt be the same if captured on 35mm lm or High Denition videotape. No doubt, in my later years, these inferior images will pierce my heart with more emotion than any 70mm print.
By the time we began production on Zodiac, David had already shot ve commercials using the Viper FilmStream camera. He liked its size. He found the method of capturing on hard drives superior in a number of ways to capturing on videotape. Data capture offered random access and no image data compression. It offered quick start and stop. There was less reloading. On the set he could view the actual les in real time, so there was no doubt what was captured. After a take, he could get his actors back to starting positions and begin his next take in less than 10 seconds, and he could slate digitally, that is, key metadata was burned into the rst six frames of every take and he was into the meat of the action more quickly. This was a way of working that allowed for a tremendous building of momentum that would result in greater focus on the part of actors and crew, and paid off with being able to pack in an extra 45 to 60 minutes of shooting per day relative to shooting lm.
80
we get done the more insufcient it is. No matter how much we improve processing and storage, somehow it always seems as if it is only half as much as we need. There was always too much to load and not enough time to render.
PM: The modern curse is loading and rendering. With all the power we now possess, the more
As dailies owed in from the set, our pipeline basically broke down into two basic functions: archiving and processing edit media. Footage came from the set on D.MAGs. These were robust hard drive arrays built by the company S.two Corp. They held about 34 minutes of footage. We would mount the D.MAG, archivexiii its contents on two LTO3 data tapes, and then begin rendering edit media. Our computers basically worked 24/7 for the entirety of the shoot. When we started shooting Zodiac, there were only 25 D.MAGs on the planet, all owned by either S.two or The Camera House, our camera rental vendor. Our rst week of shooting was out of town in parts north of San Francisco while our edit room and data lab stayed in Los Angeles. This meant that before D.MAGs were shipped, they had to be cloned to backup D.MAGs for safety from loss in transit (something not possible in a lm paradigm). We were constantly calculating shipping times, archiving speeds, and render ratios because, in the out-of-town cloning scenario, 25 D.MAGs were on the edge of being insufcient. It was like running the rst mile of a marathon uphill. As we squeaked by the out-of-town shoot and the crew headed back to Los Angeles, we knew that we would manage to enjoy some downhill days. Yet, even on the downhill days, loading and rendering was a 24-hour affair. As we settled into a routine and reected on our pipeline, what came into view rst was what was not there. There were no scratches. If David locked off a camera, there was no weave. There was no dirt (positive or negative), no light leaks, no pressure marks, no water stains, no perf damage, and no roll outs. And, because David would delete takes on the set, there were no takes that the director never wanted to see again. By shooting data, our footage was already pre-clipped and broken down into takes. Also, what we did have was two negatives, geographically separated, and every single VFX (visual effects) plate at our ngertips should we want instant advice on a VFX shot. There were denitely hiccups here and there. Render errors and lost directories caused the occasional heart attack, but in the end, I can only recall one half-frame error that remains unexplained and one half of one scan line that was unrecoverable but easily reparable. Considering that we shot the equivalent of more than 1.5 million feet of lm, I nd that failure rate completely within reason. Q: What le format do you use on your LTO tapesxiv?
PM: We captured DPXxv les on the set. So, DPX les are backed up onto the LTO tapes.
xiii xiv
Archive in this context means creation of backup protection copies of camera original data. LTO is the acronym for Linear Tape-Open, a digital data tape recording format.
xv DPX les are digital image les conforming to SMPTE 268M-2003, a standard published by the Society of Motion Picture and Television Engineers.
81
Q: Weve heard that one of the negative issues is loading the raw captured data into the downstream processing computers and then rendering it out with full frame information in post. What was your experience in this regard with Zodiac?
Q: What kind of hardware was used and how often was it refreshed?
PM: Both Zodiac and The Curious Case of Benjamin Button were captured on the S.two data
recorders. The backup LTOs were done in editorial on an Advanced Digital Information Corporations Scalar, a tape library evaluation and integration system. Q: Who operates all of this equipment?
PM: On both of those movies (Zodiac and Benjamin Button), Wayne Tidwell was the data
capture engineer on the set. On Benjamin Button, the entire back up of dailies was handled by an apprentice during production. Q: Is that an expensive labor cost?
PM: First, we drink the beer, and then call 911. Then, wed call The Camera House for a recommended replacement. Actually, we would probably call The Camera House rst! Q: How often are the LTOs audited? back on the system in their entirety.
PM: Checksum vericationxvi is done on every tape. Then we spot check tapes by loading them
Q: You used an electronic dailies distribution system. What was it, and how did that contribute to the efciencies of your post workow?
PM: During shooting, dailies were distributed through a web-based digital dailies system called
PIX. A limited number of authorized viewers were issued log-in passwords to view dailies during a limited, predetermined window of time. Aside from being an efcient and secure method to distribute dailies, PIX offered the ability to frame accurately tag notes to dailies and edits posted on the system. In this way, David was able to communicate with the edit room on a level far more sophisticated than the script notes that came from the set. From day one, he was providing notes on best and worst takes with reasons why he liked or disliked certain takes. He digitally drew on frames and issued instructions for matte paintings and retouches. The result was that by the time production ended, the editors had a huge amount of highly specic feedback directly from the director. In post-production, this level of feedback spread to visual effects, sound design, and scoring.
Checksum verication is a process for verifying digital data integrity.
xvi
82
Q: Moving on to your next major project with David Fincher, Benjamin Button, were you able to use what you had developed and learned on Zodiac? Did you improve on it?
PM: So, what could be made better than what we did on Zodiac? To be sure, Zodiac was
far from nirvana. It was a hard uphill slog that required constant vigilance from Joe, Andreas, and the edit team. A large part of this vigilance was due to the newness of the paradigm. There was an unspoken fear about whether the data was really there and whether it would really come back. It was a fear, I must add, that was more prevalent in the older crew members than the younger ones. Older crew members lived with memories of lost data and accidentally kicked-out computer plugs in other aspects of their lives. The younger crew, raised on bits and LCDs, thought there was nothing more natural than pushing data around. Yet, beyond wanting faster processing and more storage there were a few things to improve on.
One area to improve was to develop the ability to scrub metadata. On Zodiac, as drives moved from set to post, we inherited typos in the metadata entered on the set. Scene numbers or camera rolls would be incorrectly entered on the set, and then would move through post as broken data. You can imagine that, with a director picking up momentum on the set and literally allowing about 10 seconds between takes, the pressure on the data capture person was immense. Just as on lm, where slates have incorrect information, our data digital slates also had errors. We wanted a way to x those errors before they entered our pipeline. Another related problem was that we wanted less or no processing time to create our edit media.xvii S.two rose to the occasion and designed a deck they called the i.DOCK, for ingest and recording data. It did two things for us. It offered the opportunity for an assistant editor to review and repair the metadata before ingest, and it provided real-time edit media creation upon ingest. It basically worked like a tape deck. This provided the single biggest improvement in our pipeline compared to the one on Zodiac. The irony was that, in a way, this was a step backward. Ultimately, the Zodiac edit media creation is conceptually correct, that is, render the edit media, but it was simply taking too long, an issue that we knew would disappear over time as processing speeds continued to improve. As we ne-tuned the dailies ow on Benjamin Button, we tried to improve other parts of our digital lives. PIX became more entrenched in our workow and was used more deeply in casting, location scouting, visual effects, publicity, and even publishing a book which contains over 170 pages of imagery from the movie itself as well as from the making of the movie. Throughout postxvii
Edit media is a low-resolution copy of the master material recorded on-set for use by the editor on a digital editing system.
83
Zodiac was edited on Final Cut Pro by Angus Wall, our editor and owner of Rock Paper Scissors (RPS). Angus later pulled in Kirk Baxter, another Final Cut devotee. Also from RPS came the brain team of Joe Wolcott and Andreas Wacker. In general terms, Joe concentrated largely on hardware while Andreas handled software issues. Put another way, Joe made sure cameras and audio recorders talked to hard drives which talked to edit systems think timecode, timecode, timecode. Andreas made sure that the immense volume of data could be tracked and organized, safely archived, and restored for our digital neg cut. With guidance from these two, our edit team delivered VFX plates directly from the edit room, integrated VFX nals, and conformed reels sent to the DI house. Among the advantages here are no negative shipping, no charges for lm scanning, and instantaneous delivery of VFX plates.
production, PIX, an Internet-based project information exchange, became a quick and simple way to share edits of scenes with a wide range of people such as the sound designer, composer, publicity department, VFX companies, and title houses. Gone were DVDs and videotapes. Q: One of the issues with complicated projects such as Benjamin Button was the concern about coordinating visual effects from different vendors. How did your team handle that? PM: In the later stages of VFX, David, his producer Cen Chafn and I started to visit the local VFX facilities twice a week, on Tuesdays and Fridays. By working digitally, we were able to review a large volume of shots in a short period of time. At its peak, the review sessions covered more than 150 shots at four companies within a four-hour period. At these sessions, notes were given which were often executed and uploaded to PIX that afternoon. David would write additional comments on PIX and would see responses to those comments the following day. So, instead of seeing something on Tuesday and then coming back Friday for the next iteration, by using PIX, Fridays screenings were sometimes five or six iterations beyond Tuesdays work. This is an example of the level of involvement by the director and the momentum built in post-production. In a two-year period, PIX handled over 300,000 assets covering about 873GB used by more than 300 users on three continents. We had photo and publicity approvals coming from as far away as Australia and France, and VFX shots coming from as far away as San Francisco and Mexico City. All of this was largely managed by the three of us from a building in Hollywood. Q: The digital intermediate, or DI, is the modern mastering process for a digital feature. It is also expensive and time-consuming. How did you handle that part of the post workow? PM: With the advent of the digital intermediate, nishing a modern motion picture has already made the transition to a digital life cycle. We did not have to reinvent mastering a movie. Think of doing a DI without the scanning. That was us. I ran into a coloristxviii I knew who bragged that he was on a project that was developing their negative and then scanning 100% of the shot negative at 2K. I smiled wickedly and said, We are scanning our negative on the set! We showed up at the DI facility with conformed les ready to work. We had no erroneous key codes, no dust busting, no negative cutting, and no moving 800 boxes of negative. Beyond that, the DI went as normally as is now the case on all projects. Most of my frustrations in the DI do not revolve around resolution as one might expect from the is digital as good as lm debate in which people argue about whether a lm frame is 3K or 4K. In digital, as in lm, if there is a resolution issue, the rst place to look is the lens. Beyond that, I would trade resolution for color depth. That is where the pitfalls are. And this is true in the DI whether acquisition is lm or digital. For all the selling of competing digital cameras with exposure latitudes of 10 and 12 and 14 stops, in digital, as in lm, there is nothing like a properly exposed negative! Just ask any colorist. Q: Is there a lm-out at the end from the nished digital master les?
xviii
A colorist adjusts image colors in post-production for scene matching and artistic reasons.
84
Q: If this is an interim solution what do you do over the long term with the digital les? PM: If youre asking is the lm-out an interim solution no. It is a delivery requirement and although the lm-out is certainly an archival element, we consider the DI les the mother lode. For both now and later, the solution is data migration. Q: Moving on to the third project, Iron Man 2, this was a sequel and you werent involved in the rst feature. So, were the systems already in place and could you introduce what you and David had learned from your past collaborations? PM: I joined Marvel Studios one week before the beginning of principal photography of Iron Man 2. Already in place was a pipeline in which all negative shot would be scanned at EFilmxix using their CinemaScan process, in which lm is scanned at 3K and captured at 2Kxx. The les would be used to create edit media, and could be used for the DI unless it was decided to re-scan at a larger le size. After we started shooting, I was called by an executive who asked, If the negative were lost or destroyed, could we still nish the movie? This was an insurance question. The answer was, Yes, So, now there exists a pipeline Iron Man is not the rst movie using this process in which a lm-based production creates a second negative of all of their material. One immediate advantage of this pipeline is that the publicity department can create trailers without ever touching the negative. Q: We hear so much about lost digital data and various other problems what safety net did you use? PM: When audio went digital, I realized that my fear of losing the audio, before completion of production, evaporated. I knew that a good production mixer would maintain a backup. Also, digital les were fed to the edit room each day. Once that happened I knew that there were two copies of equal quality and geographically separated. At the end of the shoot, I would purchase the mixers hard drive, clone it, and then send it to the sound editing facility. Now there were three digital copies, not counting the digital les in the edit systems. At that point, it was difcult to lose the audio. It is now likewise with digital picture. As explained earlier, when hard drives were delivered from the set, two LTO backups of each drive were made and geographically separated. This is the rst safety net. Then, at the end of shooting, the editors assembly was used as a guide to begin reloading digital les from the LTO backup tapes (in our case, DPX les). Once the media required for the assembly was loaded, an LTO backup was made and moved out of the edit room. Although not a 100% safety net, it was probably at least 75%, a partial second safety net. Next, when the lm was locked, and the DI les conformed, the conformed reels were backed up to LTO, a third safety net. The conformed les were then sent to the DI house and loaded on to their system, a fourth safety
xix xx
EFilm Digital Laboratories is a post-production facility in Hollywood, California. 3K and 2K are abbreviations for digital image pixel counts of approximately 3072 x 2120 and 2048 x 1080, respectively.
85
PM: The lm-out is done only of the nal edited, color-corrected DI master les.
net. Once that happened, there were three existing conformed negatives: one on LTOs made in the edit room, one on the edit room edit system, and one on the DI companys server. So, as in sound, at this point, it becomes pretty hard to lose the movie. Q: The rendered DI, the digital master that is the result of all your hard work, must obviously be protected. How do you insure that?
From the above elements are made: Hundreds, if not thousands, of 35mm prints Hundreds, if not thousands, of Digital Cinema Package les, although these are erased at the end of a theatrical run Many thousands of DVDs and Blu-ray discs Dozens, if not hundreds, of TV broadcast masters in Hi Def and Standard Def The point is that in our modern media-addicted world, it becomes nearly impossible to wipe a movies existence off the face of the earth. Q: In this new world, what is the physical archival media required to protect this asset, the nal movie? This le most closely represents the intent of the lmmaker. Furthermore, the digital representation of this le most closely expresses Davids intent in his digital movies.
PM: From our perspective on Zodiac and Benjamin Button, it is the nal rendered DI le.
Although David is widely regarded as a master at moving the camera, he is equally adept at not moving it. One specic example of this is in Zodiac in the incident at Lake Berryessa in which the Zodiac killer stabs a couple on a quiet, sunny afternoon, killing Cecilia Shepard and seriously wounding Bryan Hartnell. In a digital screening, this scene is horrifying in its omniscience. Although subtle, I feel that in a lm screening, the lm weave lets the audience off the hook. They get to feel a process between the event and themselves, therefore experiencing a slight amount of protection from the horror. In the digital screening, it is like Alex having his head strapped down in A Clockwork Orange and being forced to watch violent lms. The rendered DI is the Holy Grail. It is the asset most closely tied to this lmmakers intent. When the DI was completed on Benjamin Button, the DI house ran two sets of LTO tapes as archival backups. In a DI facility, these backups cost between $12K and $15K. In addition, in our own edit room, we made two additional sets. Our
86
Q: Can you go through the budgeting process and give us an idea of the various costs of the various master elements? $15K for two migrations within the next ve years and then about $8K every ve years after that. Assuming a YCM costs $90K and would last the proverbial 100 years, that same $90K would buy about 50 years of migration. That is based on the assumption that the migration price would stay the same as opposed to dropping, which I believe is the likely scenario, and assumes that the ve-year migration span would not expand to longer stable periods, which, I believe, it likely would. Also, over time, loading DI les for migration becomes simpler as more facilities and manufacturers shift from an acetate-based distribution paradigm to a digitally based one, at least thats what I think.
PM: Consider that the current market rate for data migration for a standard set of LTOs is about
On the YCM side of things, the rst thing that happens is that the image suffers a loss in quality, rst, in a loss of resolution and second, with the introduction of lm-based artifacts not present in a digitally acquired project weave, dirt, chemical stains, and registration anomalies. The last thing that happens is a recombine which is a sophisticated, expensive process. Over time, say the proverbial 100 years, as the world learns to spin data on its pinky, the expertise and technology required for recombining separations becomes more scarce, and, therefore, more expensive. Again, thats my considered opinion. Q: What is the role of the post-production houses that have invested considerable money in the new technologies? has its CinemaScan process and facilities like Plaster City and nextLABxxi cater to the growi n g digital acquisition market. And there is even the in-house digital lab as done with Zodiac and Benjamin Button.Although our experiences on Zodiac and Benjamin Button were successful, many producers are not comfortable with that responsibility. As such, a market has arisen with companies and people expert in digital workow solutions. It is certainly intimidating. In the course of the projects discussed in this article, I have dealt with the Viper, F23, SI2K, RED, Phantom, DALSA, Canon 5D, XDCAM, F900, and Weisscam. There are still the F35 and the ARRI digital cameras for me to experience and, no doubt, a number of new ones in the pipeline. It is no wonder that producers want to work with facilities that bring expertise to the table. Q: Finally, to put the three features in perspective, please lay out the pipelines in simple, broad strokes.
xxi
PM: The very idea of a lab in our post processes today has been turned on its head. EFilm
nextLAB Mobile is a portable production workow system offered by FotoKem Industries, Inc.
87
sets cost about $400 per set. The studio also ran a set of digital seps, that is, 35mm YCM archival separations shot from the rendered nal DI les. These cost between $70K and $90K. The argument is that the YCM is a passive backup that requires little attention other than professional storage, while the LTOs require active data migration. In discussions I have had with colleagues, data migration is looked upon with the same enthusiasm one has for cleaning hair out of a drain. In actuality, I believe it is not that bad, and it is something we have been doing already in other parts of the industry like television production.
the details: Zodiac was shot with the Viper except for ve shots done on lm and used for slo-motion VFX. Benjamin Button was shot primarily on the Viper, with specic scenes shot on F23, and a healthy amount shot on Super 35mm 4-perf and 3-perf. Iron Man 2 was shot primarily on Super 35mm 4-perf, but also used 16mm, VistaVision, F900, RED, SI2K, Canon 5D, XDCAM, and Phantom. One of the charms of the DI suite is that it is where all the different formats can be stitched together. It can be a real playground for those inclined to experiment.
PM: I have tended to deal with the primary medium of capture. For clarication, here are
88
Four archives generously participated in case studies for this report: Archives of Appalachia, East Tennessee State University Film & Media Archive, Washington University in St. Louis Franklin Furnace Walter J. Brown Media Archives and Peabody Awards Collection, University of Georgia Archivists with primary responsibility for their collections were interviewed and asked questions that delved behind the statistics they had already provided in their surveys. They all digitally reformat content for preservation, and are at different stages of developing digital preservation programs. The interviews were initially conducted in mid-2009 with a follow-up review in mid-2011. All archives have some in-house digital transfer capability. Each one can create digital preservation les for some audio formats, and two can create preservation les for common analog video formats. Three of the archives are in larger institutions with enterprise storage capability managed by an IT department, and are in discussions to build a cooperative arrangement allowing the archive to use the IT storage.
Archives of Appalachia
http://www.etsu.edu/cass/Archives/
1. Organizational Overview
Organizational Type: Unit within a research center at a university. The Archives of Appalachia is a unit of the Center for Appalachian Studies and Services (CASS) at East Tennessee State University. The Archives collects content relating to the people of Appalachia, and is the repository for the East Tennessee State University Archives. The other CASS unit is the Reece Museum, which holds physical objects related to the history of Appalachia. An administrative ofce coordinates the study and research programs with campus academic departments. The Archives was founded in 1978, and began digitizing analog media in 2002. It is the only campus department that preserves audiovisual items. While it has a large paper and photograph collection, it does not have a trained conservator on staff, so it does minimal paper conservation work. Instead, it has chosen to develop expertise in audiovisual preservation given the content strengths in its sound and moving image collections. Collections Content Types Not surprisingly for an archive that focuses on the history of a people, the vast majority of content held in the Archives are oral histories, interviews and eld recordings. 89
A2. Archives
Audio content includes: oral histories/interviews (40%), eld recordings (35%), radio entertainment broadcasts (5%), performance documentation (10%), radio documentaries/essays (6%), nonction source material (raw news) (2%) and literary readings (2%). Moving image content includes: oral histories/interviews (20%), eld recordings (20%), outtakes (documentaries/nonction) (20%), home movies (10%), documentaries (1923 - )(10%), performance documentation (10%), television broadcasting (includes aired news programming) (6%), news source material (newsreels, newslm and raw news footage, or B roll) (1%), industrials/educational (1%), literary readings (0.5%), animation (0.5%), experimental lm (0.5%) and childrens programming (0.5%). Collections Media and Formats The majority of the audiovisual collection is sound recordings. The Archives holds approximately 15 million manuscripts, a large photographic collection containing nearly a quarter of a million images, a music and folklore collection of over 25,000 sound recordings, and 6,000 lms or video recordings. Born Digital Content The Archives began receiving born digital audio and video eld recordings in 2000. Files were provided on DAT, CDs, MiniDV videotapes, and DVDs. They have also received digital photographs in JPEG format, and digital manuscripts and transcripts. Databases containing scientic data were rejected by the Archives, because it did not have the software applications required to manage and run the databases.
90
The Archives follows the guidelines under Section 108 of the Copyright Act. It does not feel restricted by copyright in preserving its collections, and preserves copyrighted works as long as it does not distribute the preserved works without permission. Staff tracks rights research and information in an accession database that includes any restrictions on a collection, with supporting documentation on rights in a case le. Digitization as a Preservation Action The Archives has established an in-house transfer facility where it can make preservation digital les for audio content, and transfer video formats to Betacam SP videotape for preservation. Given the Archives limited budget and the large amount of audio content in its holdings, its independence has enabled it to digitally preserve 14,000 audio items. The Archives keeps old playback equipment for parts ( audio tape recorders, variable speed turntables, cassette decks, videotape recorders for U-matic, PAL, Betacam SP, helical scan open reel, and VHS formats). It also has digital conversion equipment for 8mm and 16mm lm. The Archives facility has the capability to transfer reel-to-reel tapes, audiocassettes, and disc recordings (LPs, 45s, 78s, and acetate eld recording discs). For extremely rare and damaged recordings, the Archives makes two audio preservation masters: one digital, and the other analog on reel-to-reel tape. Staff feels that analog-to-analog transfer is true preservation. They state more condence in analog, as they know how long tapes will last, and what must be done to preserve them. They know how to inspect the physical media. In contrast, digital les can disappear without warning. That is, the physical carrier can fail, and technology can change so quickly that digital le formats can become obsolete, rendering the digital les essentially unplayable. Once the Archives remaining tape stock supplies are depleted, it will discontinue parallel analog-to-analog transfers because the tape stock is no longer manufactured. For digital audio preservation, the Archives creates WAV les at 24-bit/96kHz. Archives staff selected this format and resolution after researching audio le format standards and practices described in professional documents, and reviewing advice from other archivists. They also spoke with the Universitys Broadcasting and Academic Technology Support departments to be sure that the Archives selected le formats were compatible with those used or supported by campus entities with which the Archives might work in the future. Film and 2 videotape holdings must be sent to off-site facilities for preservation. Film is preserved photochemically as lm; video formats other than 2 are copied to Betacam SP and digitally transferred to gold DVDs in-house. The Archives does not yet have the digital storage capacity or infrastructure to create uncompressed digital video les. The Archives has not migrated born digital les to next-generation formats.
91
Digitization for Access The Archives considers its primary service to be providing access to content, which necessitates creating digital access copies. It is using iTunesU (http://www.etsu.edu/itunesu/) to provide access for audio recordings, and a Facebook page to provide access to photographs as well as to communicate with users beyond the campus. However, it is selective about what has been released on the Internet because of copyright concerns. Staff perceives an attitude shift from a world where archives are open 9:30 - 5 for on-site research, to 24/7 digital access. At the time of its interview in July 2009, the Archives iTunesU site had been live for ve weeks. In week four, its site had 480 downloads. The following week there were 1,066 downloads. Archives staff tracks usage statistics to demonstrate the value of the collections and services to the University administration in support of its argument for increased nancial support. The Archives creates access-only copies in two scenarios. It creates access copies as it processes new collections. It also creates access copies on demand when a user requests an item that has not already been digitized as part of a larger project. Only access copies are made of lm and video originals. If a lm has been photochemically preserved, the Archive will create a DVD access copy. The Archives in-house transfer facility can transfer 16mm lm, 8mm, U-matic, and VHS as MPEG-2 digital les on DVD and/or AVI les. The Archives recently acquired Adobe Master Suite for creating moving image les, and will work with Academic Technology Support staff to learn how to use it. Transfers of 35mm lm, Super 8, 1 videotape, and 2 videotape must be outsourced. Stafng and Workow At the time of the case study interview, processing, digitization, and preservation of the Archives audiovisual content was performed by a small staff, student workers, and volunteer retired engineers. Besides the Archives Director, there are two full-time archivists who process collections among other duties, and two full-time support staff positions. One support staff member manages the reading room and provides secretarial and nancial management support. The other is a media specialist who spends all of his time digitizing and preserving sound, video and lm recordings, and scanning photographs. A volunteer retired engineer helps the Archives to keep the vintage equipment in working order. Ten students each work 5 - 10 hours per week to assist in processing collections and digitizing photographs. When a collection is processed, staff separates and arranges the audiovisual items. The collection is inventoried at the item level and assigned accession numbers, and the inventory list is included in the collection nding aid. Once the items are described, digital access les are created. If the original objects are sound recordings, preservation les are also simultaneously created. Born digital items are processed in the same way: they are assigned accession numbers, and listed in the inventory. Archives staff makes copies of the born digital les, and stores them on external hard drives, CDs, and DVDs.
92
93
Both the Reece Museum (the other collecting unit in the Center) and the Archives use PastPerfect for cataloging collections and physical objects. The collection nding aid is encoded in EAD using Archon (an open source nding aid-EAD creation tool) and placed on the Archives website, where it is keyword searchable. Neither of these two systems notes whether an object has been digitized, although the Archives will begin including a note on what items have been digitized in the collection nding aid. The Archives maintains an internal spreadsheet to track digitized items.
boxed and stored in the media storage room. The location is tracked both in a manual paper system and in ARCHON or the Archives accession database. The total preservation le storage to date is not large 400 GB that is duplicated on other media but there are over 14,000 preservation and access files stored on CDs and DVDs that must be managed. Even if the Academic Technology department stores the preservation les on its servers, it will only do traditional IT-level tape backups of the data. There will not be checksums, le validations and verications, or technical or preservation metadata created. Therefore the Archives must develop means to manage its own digital preservation actions.
5. Funding Strategies
Funding for the Archives digital preservation work comes largely from the units own budget. The Archives budget is created from University/State funding (55%), a Tennessee Higher Education Center of Excellence grant/State funding (40%), 3% from foundations, and 2% from revenue generated by educational duplication requests, and licensing footage and stills to documentary productions and other publications. The Archives Director allocates a portion of that budget to preservation activities. In the past, federal and foundation support paid for large content-based reformatting projects: We have been able to digitize a good portion of our large collections of eld recordings or oral history recordings through the help of three large grants from NEH and NHPRC and several smaller grants from the Grammy Foundation and the National Film Preservation Foundation. However, they have not applied for grants in the past few years due to other demands on the Directors time. Staff believes that any grant funding for building a digital preservation program would require that the Archives guarantee sustainability of the program, which reinforces the need for a cooperative digital preservation program plan.
94
1. Organizational Overview
Organizational Type: Unit within a special collections department at a university library. The Film & Media Archive is one of ve collecting units in the Department of Special Collections at the Washington University Library. It specializes in collecting and preserving documentary lm and other media which chronicle Americas great political and social movements with a particular emphasis on the African-American experience. Other collecting units within Special Collections are: Manuscripts, Rare Books, the Modern Graphic History Library and the University Archives. The University Archives has lm, video, and audio in its holdings, and looks to the Film & Media Archive for guidance on caring for those media types. The Archive opened to the public in 2002, and its initial acquisition was the Henry Hampton Collection (Henry Hampton was a 1961 graduate of Washington University). It began outsourcing digitization of analog media in 2004. The University Library has shown strong support for its new Special Collections unit, and has funded cold storage, staff, a research area, and an in-house digitization facility. The Archive is currently working with other Special Collections units that digitize holdings to develop a digital preservation plan that will be integrated into the Librarys general digital preservation program. Collections Content Types The Archives cornerstone collection is the Henry Hampton Collection, consisting of footage shot and collected for Hamptons production company, Blackside, Inc., notably the acclaimed series Eyes on the Prize. The collection includes outtakes, archival footage and interviews, as well as scripts and paper research. The collections prestige attracted donations, including production elements from documentary lmmaker William Miles. Because the Archive focuses on documentary production, much of its collections are raw elements created or acquired to include in such lms. Moving image content includes: news source material (newsreels, newslm and raw news footage) (35%), oral histories/interviews (35%), outtakes (documentaries/nonction) (15%), television broadcasting (including aired news programming) (5%), documentaries (1923 -) (5%) and industrials/ educational (5%). Audio content includes: oral histories/interviews (85%), nonction source material (raw news) (5%), radio outtakes (5%) and performance documentation (music and theatre) (5%). Collections Media and Formats The Archive has over 75,000 items that include paper materials (scripts, photographs, research, etc.) and audiovisual materials. Moving image formats include 16mm lm (prints as well as
95
production elements such as negatives and workprint), some 35mm lm, and many video formats (2, 1, open reel, U-matic, Betacam SP, VHS, 8mm, MiniDV, DVCam, D1, and D3). Audio formats include open reel audio tape, cassettes, CDs, DAT, and LPs. Born Digital Content The Archive has received a small amount of born digital content. These are primarily oral histories and interviews from Blackside captured on DV and DVCam digital videotape formats, as well as data les on a variety of disk formats.
96
Digitization as a Preservation Action As a new archive that was founded during the digital era, the Archive has not experienced the analog-to-digital preservation paradigm shift that many other archives are facing. The Archive knew from the beginning that it would use digitization techniques for preservation, although it prefers photochemical preservation for lm originals. In the Archives rst few years, it copied small amounts of content onto Betacam SP videotape, and then changed to DigiBeta videotape. They now only preserve to digital les. From the time of its founding in 2002 until 2008, all digitization work was outsourced. In 2009, the Archive installed an Apple Mac Pro-based workstation equipped with an AJA Kona LHe video card. The workstation can digitize various videotape formats such as U-matic, Betacam SP, VHS, and S-VHS, as well as audiocassettes and audio tape. Because these are the primary audiovisual formats in the collection (excepting lm), the Archive is able to do most of its digital preservation work in-house. Film and other formats not supported by its facility are outsourced. As of July 2009, the Archive has digitized for preservation 10 lms, 10 videos and 300 sound recordings. Three types of digital les are created: preservation, mezzanine, and access. Preservation les are created as uncompressed 10-bit 4:2:2 in the QuickTime container. The Archive tested several mezzanine le formats, comparing le sizes and image quality, ultimately choosing the DVCPRO50 codec as its mezzanine format. Broadcast quality audio is captured as 24-bit/96kHz WAV les. Other reference audio is captured as 16-bit/48kHz WAV. The Archive does not create mezzanine les for audiocassette recordings of meetings or phone interviews; for these content types, it creates only preservation and access les. Born digital content presents new challenges, as the Archive does not have equipment to read all digital formats. In 2010, it purchased a Sony J-30SDI compact player to enable the playback of Digital Betacam. More recently, it acquired a Tascam digital audio tape deck to handle its growing collection of DATs. Digitization for Access As a general policy, the Archive does not digitize originals purely for access. If it has multiple copies of content (for example, an interview shot on 16mm lm with a video copy), it will create an access le from the video copy in-house, as outsourcing lm is too expensive. In taking this step to access, the Archive does not consider this footage as preserved. Video access les are multiplexed MPEG-2 program streams, which can be burned to DVD when needed. Audio access les are MP3 at 128 kbps (Stereo) and 64 kbps (Mono). Stafng and Workow The Archive has six full-time staff: Film & Media Archivist (the unit head), Cataloging and Preservation Archivist, Digital Archivist, and three Archive Assistants.
97
The Archive licenses the content it owns, thus creating a very small revenue stream to support its operations. The Archive will waive its fees if a user needs to pay for preservation costs in order to access footage, because the Archive receives an access copy.
Staff hold monthly meetings to discuss all aspects of the Archives functions. Generally, they discuss progress, problems, and future needs. Digital projects tend not to be one-offs but part of larger projects that can take weeks or months to complete. If a user requests an item, that takes priority. All items pulled for digitization are tracked through MAVIS, which is the cataloging database used by the Archive. MAVIS has the capability to track workow as well as descriptive and technical data. The Film Archive is the only department that uses MAVIS. The Digital Archivist creates the digital les. A le naming convention has been developed for digital assets: the root unique identier for le names is based on the unique identier assigned by MAVIS. The les are then moved from the workstation to servers maintained by the Library IT department. Preservation, mezzanine and access les are moved to the Special Collections department servers, which the IT department manages but the Archive can directly access. The Archive uses mezzanine les for transcoding to various access les based on user requests, so it must have control of and unfettered access to those les. The Cataloger and the Digital Archivist create content records in MAVIS. The Digital Archivist adds information stating that the analog originals were digitized, the le names, technical metadata, and the les locations. Staff can run reports from MAVIS to track what has been digitized.
98
The three le types (preservation, mezzanine and access) are stored on a Xiotech Emprise 7000 Production Mirror. Files are backed up incrementally each business day and more completely on a biweekly, monthly, semiannual and annual basis. Preservation les are stored on slower units that spin up on demand. Mezzanine and access les are maintained at higher availability. As described, the three le types are stored on different servers according to their intended use. The IT department makes LTO4 tape backups from all servers, but the Archive has no input on how often that takes place, nor are its les separated from other les stored on the entire storage area network (SAN). The les are mirrored on other servers. Backup tapes are stored in the Library, the same building as the IT departments servers. In the system currently in place, the les are mirrored on other servers. There are daily partial backups of the SAN and weekly backups of the entire SAN, and the tapes are stored for the long term. The Archive has been actively digitizing content in-house since January 2009. As of July 2011, the Archives preservation les total 8.4 TB. Mezzanine and access les take up 4.5 TB. The Archive anticipates its preservation storage will increase at 3 - 5 TB per year. It does not currently pay for the storage, which is provided by the Library. However, if the Archive receives a grant for a reformatting project in the future, funding for servers or other storage will be incorporated into the project budget.
99
Technical Infrastructure
Persistent communication and quarterly meetings between the Archive and IT have helped the IT department understand the Archives digital preservation requirements. In addition, the Archive has become aware of the challenges that digital preservation presents to the IT department. Most importantly, the Archive has learned that it must express clear and concrete needs in order to receive increased support. Time must be allocated for IT to order necessary equipment and integrate that equipment into the current infrastructure. Archive staff now believes that they are all working toward the same goal: to create a digital preservation program with clear procedures and workows. There is still a limit on the Archives staff s reformatting workload: with only one Digital Archivist, there is only so much work that can be done in-house.
5. Funding Strategies
The Washington University Library has been generous in building the Film & Media Archive, and purchasing equipment for an in-house digitizing facility that allows an economical means to preserve and provide access to the collections. The Archives operations largely fall within the Librarys annual budget (overhead, staff, supplies, etc.). Digitization costs also fall into this category with the exception of outsourcing. As such, 85% of annual digital reformatting costs are covered in-house, 10% covered by users (paying to digitize for access) and 5% from other sources (including any earned income from licensing). Occasionally, projects are funded by Special Collections, grants, or donors. Because these are sporadic and do not happen every year, they have not been included in the percentages. A very small percentage of the budget is funded by revenue from licensing footage and photos. Providing online access to important and high-visibility materials is a fundraising tool as well as part of the Librarys mission. The Dean of Libraries can promote the Archive to the University as an example of good work and request more funding for Special Collections. As the Archivist notes: As more and more library resources become available online (books, journals, etc.) what distinguishes us from any other library are our special collectionsthose unique materials that do not exist anywhere else. We have beneted directly from the value now placed on our collections, which has allowed us to do so much. Special Collections is one of the largest draws for donor money and other nancial contributions to the Library. Donors to the Film Archive are familiar with Eyes on the Prize, and it can be used as a fundraising tool. Similarly, the collections are often used as a recruitment tool for potential faculty. Prospective faculty candidates are often given a tour of the Film & Media Archive if their research focuses on the civil rights movement, 20th century American history, lm studies, or other areas of our collections. Even with the history of strong support from the University, the Archive anticipates that internal funds for outsourcing digital preservation work will be limited. External funds will be raised for those activities. Fortunately, support should remain consistent for storage and backup needs. The Archive is actively involved in fundraising beyond grants. It created an Advisory Board, which raised enough funds to motivate the University to set up an endowment. The endowment has a line item for digital preservation, which the Archive can draw upon to support its digitization activities.
100
http://www.franklinfurnace.org/
1. Organizational Overview
Organizational Type: Independent nonprot arts organization. Franklin Furnace was founded in 1976 as a bookstore for artists books in New York City, and quickly evolved into a performance and installation art space. Performances, readings, and installations were documented on videotape, audiotape, and still photography, resulting in an important archive of variable media (artworks that can change with each instantiation). Since 1997 it has no longer operated as a performance venue, but continues to sponsor artists work and archives documentation of the sponsored art. Since 2006 the organization has been actively preserving its unique archive as part of its Unwritten History Project to make all of its archival event records accessible online. Franklin Furnace began digitizing analog content in 2005. Collections Content Types Franklin Furnaces archive consists almost entirely of performance and installation documentation. Collections Media and Formats The majority of the collection is still images (slides, photographic prints, programs, yers, etc.). There are almost 800 video recordings in formats ranging from open reel video, U-matic, VHS, Betamax, Hi8, MiniDV, CD, and DVD. Their audio collection consists of approximately 63 items: 54 audio cassette tapes, two 7 vinyl discs, one reel-to-reel tape, and ve CDs. Born Digital Content Franklin Furnace began receiving video born digital content from artists whose work it sponsored in 2000. No born digital audio has been contributed. Digital les are delivered on DVD or off-loaded from external hard drives. Artists have contributed QuickTime, MPEG2, and .vobxxii les.
vob is an abbreviation for Video Object File, which is the data le format for DVD-Video discs.
101
Franklin Furnace
at Franklin Furnace. These tapes will become part of the permanent collection eventually, but the organization will not take steps toward preserving them until the primary collection is processed. Given the small size of the collection, Franklin Furnace intends to preserve all sound and video archival originals. This will be a slow and ongoing process, as preservation activities are entirely funded by grants. Its preservation priorities are loosely chronological, beginning with the oldest tapes and further rening based on tape obsolescence and condition. Franklin Furnace purposely does not rate preservation priorities on the external perceived importance of the artist, believing instead that all artists work has equal cultural value. Grant funding in 2006 - 2008 provided the opportunity to preserve 20 videos on open reel from the organizations rst ten years (1976 - 1986). Previously, the open reels were transferred to VHS; MiniDV copies were made from the VHS tapes as an interim preservation strategy. With the grant funds, Franklin Furnace is creating Betacam SP videotapes and 10-bit uncompressed digital les from the original open reel videotapes. It is important to note that Franklin Furnace does not view preservation as object-based (the specific video or audio recording), but as event-based. This is the heart of the Variable Media concept, which proposes that an artists performance or installation be preserved rather than just a video. Thus its preservation priorities tend to be focused on preserving all items related to an event (stills, video, yers and invitations) as a means to preserve that specic occurrence of the performance. A performance could have variations based on location or other variables. Preservation and Copyright Franklin Furnaces copyright situation is complex. Copyright does not affect its preservation policy; it only affects whether the preserved material is published on the website. To provide online access to documentation, Franklin Furnace must secure permission from the videographer or photographer who shot the documentation, as well as from the artist whose work is represented in the documentation. Different permission policies were practiced over the years. In the past, Franklin Furnace was most concerned with obtaining permission from videographers to stream their work on their website. Franklin Furnace is now revisiting its policy for artists whose works are represented in its legacy collection (works performed prior to 2000). Franklin Furnace does not seek permission from an artist prior to preserving documentation of their work, but will ask permission before publishing the work on their website. Only two or three artists have denied permission to publish stills or video on the website. 2000 - (9 born digital) 3% of collection Franklin Furnace has signed permissions from artists and videographers to provide access to the recent born digital documentation. 1990s - 2000s videos (272 VHS) 80% of collection Approximately 80 to 90% of these videos were created by one videographer who shot everything on Hi8. He gave Franklin Furnace VHS copies and kept the originals. Franklin Furnace has release forms for all his work. Approximately 50 tapes shot by other videographers require release forms, and permissions have not been obtained from artists whose work was documented during this period. Beginning in approximately 2000, the rights situation reversed: Franklin Furnace acquired rights permissions from artists they have presented, but not from the videographers.
1970s videos (20 open reel tapes) 6% of collection In the mid-1970s, the New York State Council on the Arts was concerned that avant-garde art activity was not being documented. It created a Media Bureau, which videotaped various art activities and gave the tapes to the arts organization presenting them. The presenting organizations own the rights to the videos themselves. The video equipment was stored at The Kitchen, an alternative performance space in New York City. Organizations made appointments with The Kitchen to document their events. New York State paid the videographers. All of Franklin Furnaces open reel videos were created through this program. Since 2009, Franklin Furnace has included explicit language in the contracts with their sponsored artists and their videographers and/or photographers to give permission to provide online access to their documentation. Franklin Furnace would prefer documenting sponsored artists performances as it did in the past rather than receive work from artists, but this is primarily because of quality control concerns rather than rights. However, there is no current funding to produce the documentation. Digitization as a Preservation Action Video digitization for preservation work is outsourced to vendors. Franklin Furnace performs most audio digitization in-house. Franklin Furnace can transfer VHS to MiniDV in-house, but considers that a stop-gap measure until funding is received to do full preservation of the tapes. Sound recordings are encoded as WAV les at 16-bit/44.1kHz. MP3 access les are created as well. 34 of its 63 items have been digitized thus far. Partial audio digitization was done in 2006. Franklin Furnace follows a two-tiered track in preserving video: it creates an analog Betacam SP preservation master as well as a 10-bit uncompressed digital le. Factors that inuenced this decision were partly philosophical, and partly economic. Philosophically, staff believes that it is still essential to convert archival originals to analog formats as well as to digital. They believe that more information can be extracted from analog copies later than from digital les. They nd that the quality of the image is different in digital formats; it looks too sharp, especially for low-end quality originals as are found in their collection. As one staff member said, Low-end tapes shouldnt look so good. We could always process the les to make it look like the original, but thats processing which isnt preservation. Staff said that by the time analog tape stock is completely discontinued, digitizing techniques should be improved so that the image is truly equivalent to the original. Franklin Furnace also had to consider the costs in its ability to support and use a particular format. It originally planned to create uncompressed digital les, a DigiBeta videotape, a Betacam SP
103
1980s videos (38 U-matics) 11% of collection Franklin Furnace considers this subset of the collection to be its main copyright problem. During this time, individual artists paid for their own videographers, so either the videographers or the artists own the rights to the documentation. In a few instances, Franklin Furnace did hire videographers. To provide access to the tapes not produced by Franklin Furnace, the organization would need to nd the videographers and obtain release forms. However, it will nonetheless preserve these tapes while working to acquire releases.
videotape, and a DVD. Its vendor quoted $100 to make a DigiBeta copy, $30 for a Betacam SP copy, $150 to create a one-hour uncompressed digital video le. Franklin Furnace chose not to make DigiBeta copies because it would not be able to use that format without purchasing or renting a DigiBeta deck. Franklin Furnace discussed its preservation format options with art museums that have similar content in their collections. It learned that museums are creating either Betacam SP videotape or 10-bit uncompressed digital les, so the organization believes its preservation actions are in line with museum best practices for variable media works. Franklin Furnace has not transcoded (reformatted) born digital les for preservation. However, it has transcoded RealPlayer les to QuickTime for access. Digitization for Access Currently, researchers primarily access audiovisual content on-site at Franklin Furnaces Brooklyn ofce, where they can view access les (QuickTime, DVDs, and MP3) or original tapes if those tapes have not been transcoded. Future primary access will be provided through Franklin Furnaces website. Just over 20 video clips currently stream online on the websites Video Directory section. Streaming services are provided by Streaming Culture, a project of the City University of New York to help nonprot cultural organizations stream rich media. Franklin Furnace provides artists VHS tapes and other media to Streaming Culture, and the service creates QuickTime les. The access les are stored on Streaming Cultures servers; no preservation le is made. Franklin Furnace digitizes for access only (no preservation les are made) in two situations: 1) when there is not enough funding to create preservation masters, but access to the content is critical for a project, or 2) the work is recent so does not require preservation, and Franklin Furnace wants to stream the content on its website. Stafng and Workow Three staff members are involved in preservation activities. The Executive Director decides which group of tapes is to be preserved. The Senior Archivist selects and pulls the individual tapes and creates catalog records for each tape before sending them to the vendor for encoding and transfer to Betacam SP videotape. Uncompressed and QuickTime digital les are returned on a portable hard drive, and .vob les are returned on a set of two DVDs. When the tapes and les are returned to Franklin Furnace, the archivist copies the les to another hard drive. He sends the archival original tapes, Betacam SP, and one hard drive to off-site storage. The duplicate hard drive and DVDs are stored in Franklin Furnaces ofce. Franklin Furnace creates audio preservation les in-house and copies them to a server in its ofce. Copies on CDs are stored off-site The Cataloger adds the hard drive number for a videos uncompressed le to its catalog record in
104
105
Franklin Furnaces FileMaker Pro database. The hard drive shipping box also has a printed list of le names stored on the hard drive taped to the outside. Digital audio information has not yet been added to the audio database records, but Franklin Furnace will follow the same procedure as with the video database. It will embed a small MP3 le directly into the FileMaker database record for access and include the path and le name for the preservation copy into the catalog record. There are only 34 artists who have audio work that is digitized. Franklin Furnace nds that the les are easy to nd due to their limited quantity, so a database reference is not needed at this time.
5. Funding Strategies
Funding for Franklin Furnaces preservation activities is completely supported by grants on a project basis. A grant from the National Endowment for the Humanities in 2006 - 2008 supported digitizing and making accessible records of performances, installations, exhibits, and other events produced by the organization during its rst ten years. The organization anticipates that grant-funded preservation will continue, and is considering strategies to help its grant proposals succeed. It considers its website to be a key access and marketing tool. Providing online access to content is an incentive to digitize, as the public face will convince funders that Franklin Furnaces content is unique and important. Franklin Furnace is considering producing and distributing a DVD of performance art highlights, but it is doubtful that the resulting revenue would be allocated to digital preservation. Cooperative arrangements with artists whose works are represented in Franklin Furnaces collections are also being explored. Franklin Furnace could digitize an artists work in exchange for making the work accessible on Franklin Furnaces website. The artists would receive a digital copy and their original tape would be returned. The artists would not be paid, as the digitization would be performed as an in-kind service. It is hoped that funders would be interested in supporting this arrangement, because it would help preserve and provide access to work that would otherwise not be available.
106
A2. Archives: Walter J. Brown Media Archives and Peabody Awards Collection
1. Organizational Overview
Organizational Type: Special collections department at a university library. The Walter J. Brown Media Archives and Peabody Awards Collection is one of three separate special collections departments at the University of Georgia Libraries. It specializes in broadcasting collections (both television and radio), and audiovisual records documenting Georgia and the South. The other special collections are the Hargrett Rare Books and Manuscript Library, which also manages the University of Georgia Archives and the University of Georgia Records Management Program; and the Richard B. Russell Library for Political Research and Studies. The Russell Library collects audiovisual content as well, and creates audio and video oral histories of Georgia politicians. The Archives was founded in 1995, and began digitizing analog media in 2006. While other campus units collect audiovisual content (for example, the Journalism School and the Athletics Department), only it and the Russell Library have staff trained and dedicated to audiovisual preservation. Audiovisual preservation is well integrated into the University Libraries goals and priorities, and to preserve and make accessible the Archives collections is one of its major mandates. Collections Content Types The Media Archives has three large broadcasting collections: the Peabody Awards Collection (all radio and television entries to the Awards since they began in 1940), the WSB Newslm Collection (5 million feet of newslm from this Atlanta station dating from 1949 - 1981), and the WALB Newslm Collection (newslm from this Albany, Georgia, station dating 1961 - 1978). Additional collections have been added over time, especially in the area of Georgia and regional history, but the collection focus is broadcasting. Moving image content includes: television broadcasting (aired programs) (70%), news source material (outtakes, newslm, etc.) (20%), oral histories/interviews (4%), educational/industrial (2%), field recordings (1%), documentaries (1%), home movies (1%), literary readings (0.5%) and music videos (0.5%). Audio content includes: radio news broadcasts (20%), radio entertainment broadcasts (20%), radio documentaries/audio essays (20%), oral histories/interviews (10%), eld recordings (10%), performance documentation (10%) and childrens programming (10%). Collections Media and Formats The Archives has over 300,000 titles on analog formats including radio transcription discs,
107
A2. Archives: Walter J. Brown Media Archives and Peabody Awards Collection
audio reels, 16mm lm, 2 video, 1 video, U-matic, open reel video, VHS, Betacam SP and DigiBeta videotape. Born Digital Content Born digital content is received as part of the Peabody Awards entries, but the Archives receives only access copies submitted on DVDs and CDs. Since the 2009 Peabody Awards entry process the Archives has been accepting digital les along with analog video formats. Entrants can submit digital les as uncompressed MOV les at the highest resolution the entrant can deliver on Blu-ray disc, USB ash drive, or external hard drives. If the entrants submit les on hard drives, the hard drive must be formatted for Apple Macintosh computers. For master audio content the Archive is requesting BWF les (submitted on CD or DVD) or MP3 les on CD.
108
A2. Archives: Walter J. Brown Media Archives and Peabody Awards Collection
Digitization as a Preservation Action The Archives has established an in-house reformatting facility where it can make preservation-quality les from audio and video formats, and access les from 16mm lm. In general, lm is preserved photochemically (this work is outsourced), but some lm originals were preserved digitally. For the Civil Rights Digital Library, 10 hours of 16mm lm were digitized to high resolution AVI les as specied by the Institute of Museum and Library Services (IMLS), the funder for that project. The AVI les were transcoded into three separate streaming formats for access. An additional 90 hours of lm have been digitized in-house to QuickTime ProRes 422. As of July 2011, the Archives has created preservation digital les for 100 hours of lm (10 to AVI, 90 to ProRes 422), 4,000 videotapes, and 350 sound recordings. The Archives keeps old equipment for both use and parts. Three le formats are used as digital preservation les for moving image content: JPEG 2000 wrapped in MXF, ProRes 422 (HQ), and 10-bit QuickTime in MOV les. The facility uses two SAMMA Solo systems to encode U-matic tapes and create JPEG 2000 (50 mbps) preservation les. For a grant project, it created a second analog preservation master on Betacam SP videotape. Once its Betacam SP stock is depleted, it will no longer make analog preservation masters and will only create digital les. The Archives also use Final Cut Pro on four Apple workstations to create QuickTime (10-bit, 230 mbps) and ProRes 422 (73 mbps) les. ProRes 422 les are preferred over 10-bit QuickTime les because of le storage limitations, although the Archives will create 10-bit les on request (for example, when working with producers who are licensing footage or have permission to copy a Peabody program). Analog audio formats (transcription discs, audio reels, DAT) are encoded as BWAV les at 24-bit/96kHz. Audio preservation le formats are well established; the Archives referred to the IASA-TC04 standard and best practices it found in researching the topic. The Archives hired an audio technician in late 2010 to do audio transfer and preservation work. The Archives selected its preservation le formats based on formats the vendors hardware supported, asking colleagues, reading the Association of Moving Image Archivists ListServ (AMIA-L) and researching what formats had the widest support across institutions, vendors, and broadcasters. The SAMMA Solo system creates JPEG 2000 les, and after research the Archives decided to use that format as its primary preservation le format. QuickTime ProRes 422 les are created for digital lm transfers. The preservation les are maintained on a dedicated Apple Xserve with a redundant RAID digital storage system by the Libraries IT department, although IT does not normally support Apple hardware or software. JPEG 2000 les are only stored on LTO3 tape. Two copies are made, with one copy stored in the Archives and the other off-site. No JPEG 2000 les are stored on servers. Born digital content received as Peabody Awards entries is currently received on CDs and DVDs, but this changed somewhat with the 2009 entries. Audio les on the CDs are copied to MP4 for streaming from the Peabody Awards database locally to the campus community. Approximately
109
permission to stream the content. Archives staff notes that there has been a positive response to their requests.
A2. Archives: Walter J. Brown Media Archives and Peabody Awards Collection
500 programs have been transferred. If the Archives receives a dye-based optical disc (e.g., CD-R or DVD-R) it is transferred immediately to a digital le for eventual migration to LTO tape. Entrants submit two copies of DVDs or a videotape; one copy is placed in the Archives, and the other is used as an access copy. Digitization for Access The Archives digitizes 16mm lm for access and uses photochemical processes for preservation. Film content digitized for access includes home movies and newslm. Otherwise, all audio and video that are digitized in-house have access les created simultaneously with the preservation les. As mentioned above, radio content received on CDs is transcoded to MPEG4 les, and attached to its cataloging record for local campus access. Stafng and Workow The Archives has a staff of 5.5: Director, Media Archivist, Peabody Archivist (who also catalogs Peabody items), Moving Image Digital Conversion Manager, Audio Technician and one half-time Cataloger who works on non-Peabody content. It also employs student assistants. Content to be digitized is selected by the Director, or upon user request. A shared spreadsheet used to track the digitization process is stored on Google Docs. Staff uses Google Docs rather than sharing documents on a networked server because Archives documents are created on an Apple computing platform, and the IT department does not support Apple documents. Archives staff inspects and cleans the archival original prior to encoding. U-matics are digitized to JPEG 2000 les using the SAMMA Solo systems; 16mm lm and all other formats are encoded to QuickTime ProRes 422 using the Apple workstations. Derivative les are created as needed: 400 kbps 360x240 Flash, and a 1 mbps 720x480 QuickTime streaming le. The streaming formats F4V and MOV use the H.264 codec. A newslm reel can contain several unrelated outtakes or clips that are spliced together into a 2,000 foot reel. When a newslm reel is digitized, the digital outtakes are broken into individual les with their own le names based on the unique identier for the archival original clip. Derivatives are given to the Cataloger to attach to the catalog record. The catalog record notes the original reel from which the clip originated, so it would be possible to locate the original source reel or even digitally re-create the reel if necessary. There is not enough staff to catalog all content that is digitized, so many of the non-Peabody items have minimal metadata. The Archives has a two-tiered approach to digital storage. While their approach is described more fully in the Technical Infrastructure section below, it is noted here because the Archives backs up preservation les to LTO3 tape itself as part of the workow. The two SAMMA Solos are an early model with an LTO3 tape drive internal to the machines, so the Archives can copy the JPEG 2000 les to LTO3 tape within its own department. All other les are stored on dedicated servers to be eventually migrated to LTO. The Archives will soon be attaching an LTO5 drive to the Xserve and eliminating the LTO3 drives installed in the SAMMA Solos.
110
A2. Archives: Walter J. Brown Media Archives and Peabody Awards Collection
Physical Carriers JPEG 2000 preservation les are stored on two copies of LTO3 tape; all other preservation les are stored on servers maintained and backed up by the Libraries IT department. Audio preservation les are also stored on gold CDs as well as transcoded and copied to the server. Peabody Awards television entries received on DVDs are left as DVDs unless they are dye-based (e.g., DVD-R); les are not copied to servers. Digital Asset Management The Archives currently does not use an integrated digital asset management system, though this is coming in the near future as the number of digital les increases. The Archives has recently purchased an Apple Final Cut Server for the Xserve storage system, which assists in managing its digital assets. It manages digital production workow through Excel spreadsheets. In addition, access les are attached to cataloging records for streaming. The Peabody cataloging record with a connected instantiation database lists the digital le formats, le names, and where they are located (names of servers and LTO tape number). Technical data created by the SAMMA Solos during the encoding process is stored separately in an XML le. Technical and Digital Preservation Metadata Technical and digital preservation metadata is limited to le names and storage locations, and technical metadata created by the SAMMA system during encoding. The XML le created by SAMMA (see the Technical Infrastructure section below) adds technical data about the JPEG 2000 and derivatives creation; this data is saved in the XML le and not exported to a database at this time. As the Archives further develops a digital preservation program, it intends to use PREMIS for digital preservation metadata. Technical Infrastructure As mentioned previously, the Archives operates in a hybrid technical environment. It is responsible for copying large preservation les (JPEG 2000 and ProRes 422) to LTO3 tape with a migration to LTO5 beginning in 2012 within its department, while the Libraries IT department stores all other les on servers it manages and backs up. IT creates the server directory structure for the Archives. The Archives asks for more directories and space as collections are digitized, as it can monitor its storage usage itself. The Archives pays for the storage; as of July 2009, 60 TB is allocated for its use. The storage servers are not shared with any other Librarys department, although the streaming server is shared with another department . The servers are duplicated by the IT department for backup in a mirrored RAID system. The Archives Director, in consultation with the Moving Image Digital Conversion Manager, decides when and which les are migrated to LTO tape when server space needs to be freed up.
111
A2. Archives: Walter J. Brown Media Archives and Peabody Awards Collection
ProRes 422 les are stored on the server, although at times they are copied to LTO. This decision depends on the collection, and how the Archives intends to use the les after capture. Files are only copied to LTO when a group or a directory of les total 400 GB of storage (the storage capacity of the LTO3 tape they use). In these circumstances, the preservation les are transferred to the SAMMA Solo workstations, then written to LTO. (SAMMA LTO drives are networked to the Archives dedicated servers.) The Archives LTO3 tape drives are part of their SAMMA Solos. The University was a beta tester for SAMMA, and the early models produced by the company installed LTO drives into their systems. Front Porch Digital, the company that has since purchased the SAMMA product line, no longer includes LTO drives. The Archives has also encountered some software issues related to LTO use with the SAMMA system, and will soon replace the LTO drives with an LTO5 attached to the Xserve. This will enable them to move the SAMMA les to the Xserve and back up les to LTO5. The SAMMA Solos produce an MXF le that contains the JPEG 2000 le. The MXF le is read and validated through the SAMMA systems. Digital Rapids software then produces three derivatives, two QuickTime and one Flash. If there is an error in the MXF le, there will be an error in all derivatives, and therefore validation of the MXF les is critical. Technical data on the analog transfer is written to XML les.
112
A2. Archives: Walter J. Brown Media Archives and Peabody Awards Collection
5. Funding Strategies
The Archives digital preservation efforts to date have been funded 75% by grants; 20% is allocated from the departments budget, and 5% comes from revenue earned from licensing footage. The Archives does not anticipate that these percentages will shift, but it needs more staff and infrastructure support. Planning for a new special collections building has taken up time that could have been devoted to fundraising, and once the building is completed the Archives Director will focus on grant writing. At the moment, the Archives is only receiving small grants for preservation work that is outsourced. The Archives Director believes that in order to raise funds to preserve content, an archive needs to know what it has in its collections. Funders are captured by content. The Archives Director also considers it an irresponsible practice to digitize items merely to discover its content, although sometimes that is the only option. Physical materials must rst be processed, noting media format and creating basic content descriptions using written information on the item and from other documentation. A funding priority is for more processing and cataloging staff; all digital preservation actions will follow once content is identied. Hiring additional staff for digitizing, especially when it is time to refresh the data, will also become a priority. As the Archives Director notes, Broadcasters have gone down this [digital] path so there is no turning back. Digital le creation requires another level of commitment for archives that involves leadership, IT assistance, digital storage, money, metadata creation, le management, le migration, and the people to keep it going. The ease of access is fantastic but it comes with a fairly steep price tag for small institutions.
113
LTO tapes can still be read on new tape decks. The Archives Director believes that external studies and research papers help support the archivists concerns to IT staff. The two departments need to work together: one has technology, the other has archival knowledge. If a common understanding cannot be reached, Its the collections that pay, not us.
114
ONLINE APPENDIX
The complete set of survey data used in the creation of this report is available at: www.oscars.org/tdd2
115
116
END NOTES
1 Theatrical Market Statistics. Motion Picture Association of America. 2010: 13. 6 July 2011 <http://www.mpaa.org/Resources/93bbeb16-0e4d-4b7e-b085-3f41c459f9ac.pdf > 2
IFTA FAQs. Independent Film & Television Alliance. N.d: 1. Los Angeles, CA. 5 July 2011 <http://www.iftaonline.org/sites/default/les/IFTA_FAQs.pdf > Academy Awards Database. Academy of Motion Picture Arts and Sciences. Oscars.org. 2011. 17 June 2011 <http://awardsdatabase.oscars.org>
4 Timeline for Inventing Entertainment: The Motion Pictures and Sound Recordings of the Edison Companies, 1847-1879. American Memory. Library of Congress. Washington, DC: 1999. 5 July 2011 <http://memory.loc.gov/ammem/edhtml/edtime.html>
William Kennedy Laurie Dickson. Edison Kinetoscopic Record of a Sneeze. 1894. Prints & Photographs Reading Room. Library of Congress. Washington DC. Prints & Photographs Online Catalog. 24 Aug. 2011. <http://www.loc.gov/pictures/item/95505651/>
5 6
Eileen Bowser. History of the American Cinema: The Transformation of Cinema 1907-1915. New York: Macmillan Inc., 1990. 29-33.
7 Dave Kehr. Four Stars Bright Idea Still Shines 90 Years On. New York Times 27 March 2008. 6 July 2011 <http://www.nytimes.com/2008/03/27/movies/27unit.html>
Jonas Mekas. The New American Cinema Group Inc. History. The Film-Makers Coop.com. Film-Makers Cooperative. N.d. 6 July 2011 <http://www.lm-makerscoop.com/history.html>
8 9 Richard Harrington. The Essential Roger Corman. Washington Post 14 September 2007. 6 July 2011 <http://www.highbeam.com/doc/1P2-7596700.html >
Academy Awards Acceptance Speeches. Oscars.org. Academy of Motion Picture Arts and Sciences. 14 November 2009. 13 May 2011
12 Jason Guerrasio. Continuing Dilemma. Filmmaker Magazine Winter 2010. 14 July 2011 <http://www.lmmakermagazine.com/issues/winter2010/reports-continuing-dilemma.php> 13
Caleb Deschanel. Astronaut as Filmmaker. Science and Technology Council public program. Samuel Goldwyn Theater, Beverly Hills 14 July 2009. Robert Sklar. Film: An International History of the Medium. New York: Harry N. Abrams, Inc., 1993. 27. Jack C. Ellis and Betsy A. McLane. A New History of Documentary Film. New York: The Continuum International Publishing Group Inc., 2005. 3-4. Leo Enticknap. Moving Image Technology from Zoetrope to Digital. London and New York: Wallower Press, 2005. 39.
14
15
16
17 About the Filmmakers: Ken Burns. PBS.org. Public Broadcasting Service. 2011. 6 July 2011 <http://www.pbs.org/kenburns/lmmakers/>
117
End Notes
END NOTES
18
Marijke de Valck, Film Festivals from European Geopolitics to Global Cinephilia, Amsterdam University Press, 2007. 47-48.
The Sundance Collection at UCLA. UCLA Film & Television Archive. 2011. 18 July 2011 <http://old.cinema.ucla.edu/collections/Proles/SundanceAtUCLA.html>
20 21
37th Seattle International Film Festival Wraps with 2011. Golden Space Needle Awards. SIFF.net. Seattle International Film Festival. 12 June 2011. 7 July 2011 <http://www.siff.net/> Ellis. 341-343. Anon. Personal Interviews. 2010-2011.
22
23
24 Anthony Kaufman. Why VOD Is Turning into a Protable Avenue for Indie Filmmakers. Filmmaker Magazine. Fall 2009. International Film Project. 15 Apr. 2011 <http://www.lmmakermagazine.com/issues/fall2009/industry-beat.php> 25
Richard Verrier. When Pirates Pillage Indie Films. Los Angeles Times. 28 Sept. 2011: B1. Cheryl Boone Isaacs, Marketing Consultant. Personal Interview. 29 Aug. 2011.
26
27 U.S. Copyright Ofce. How Long Does Copyright Protection Last?. Library of Congress. 10 Mar. 2010. 8 July 2011 <http://www.copyright.gov/help/faq/faq-duration.html> 28
About. Outfest.org. Outfest Legacy Project for LGBT Film Preservation. 8 July 2011 <http://www.outfest.org/legacysite/about.html>
29
National Film Preservation Board. Public Moving Image Archives and Research Centers. Library of Congress 2011. 31 Aug. 2011 <http://www.loc.gov/lm/arch.html>
30 31
Motion Picture and Television Reading Room. Motion Picture Broadcasting and Recorded Sound Division. Home page. 27 July 2011. 26 Aug. 2011 <http://www.loc.gov/rr/mopic/> Charles Buckey Grimm. A paper print pre-history, Film History, Vol. 11, No. 2, migr Filmmakers and Filmmaking (1999): 204-216. Indiana University Press. 26 Oct. 2010 <http://www.jstor.org/stable/3815323> Motion Picture and Television Reading Room. Motion Picture Broadcasting and Recorded Sound Division.
33
34
35 Richard A. Sanderson. A Historical Study of the Development of American Motion Picture Content and Techniques Prior to 1904. New York: Arno Press Inc.: 1977, 1-33. 36
United States, Cong. Senate. An Act to Establish a National Archives of the United States Government, and for Other Purposes, 73rd Cong., 2nd sess. H.R. 8910. The National Archives. archives.gov. 3 Jan. 1934: 13. 27 Aug. 2011 <http://arcweb.archives.gov/> Preservation Without Access Is Pointless. The Committee for Film Preservation and Public Access. 12 Feb. 1993: 2-3. 28 July 2011 <http://www.loc.gov/lm/pdfs/fcmtelmprespubaccess.pdf> Redening Film Preservation: A National Plan. National Film Preservation Board. loc.gov. Library of Congress. Aug. 1994. 8 July 2011 <http://www.loc.gov/lm/plan.html>
37
38
H.R. 4867: National Film Preservation Act of 1988, 100th Congress: 1987-1988. GovTrack.us. Civic Impulse, LLC. 6 Sept. 2011 <http://www.govtrack.us/congress/bill.xpd?bill=h100-4867&tab=summary>
39
National Film Preservation Foundation. The Film Preservation Guide: The Basics for Archives, Libraries, and Museums. San Francisco: 2004. 4-42.
40 41
Denitions of Digital Preservation. ALCTS Preservation and Reformatting Section Committee Working Group on Dening Digital Preservation. 24 June 2007. 28 Aug. 2009 <http://www.ala.org/ala/mgrps/divs/alcts/resources/preserv/defdigpres0408.cfm>
42 Digital Preservation Brieng Paper. Joint Information Systems Committee (JISC). HEFCE. 20 Nov. 2006. 18 Apr. 2011 <http://www.jisc.ac.uk/publications/briengpapers/2006/pub_digipreservationbp.aspx>
118
End Notes
44
Science and Technology Council. Long-Term Management and Storage of Digital Motion Picture Materials: A Digital Motion Picture Archive Framework Project Case Study. Los Angeles: The Academy of Motion Pictures Arts and Sciences, 2010. 2-49.
45 United States Copyright Ofce. Report on Orphan Works, Library of Congress Jan. 2006: 1. Washington, DC. 7 July 2011 <http://www.copyright.gov/orphan/orphan-report.pdf> 46
47 Council on Library and Information Resources and the Library of Congress. The State of Recorded Sound Preservation in the United States: A National Legacy at Risk in the Digital Age, National Recording Preservation Board. Washington, DC. Aug. 2010: 3-39. 29 July 2011 <http://www.clir.org/pubs/reports/pub148/pub148.pdf>
Library Releases Final Study Showing Dire State of Sound Recording Preservation and Access. loc.gov. Library of Congress. 29 Sep. 2010. 29 July 2011 <http://www.loc.gov/today/pr/2010/10-194.html>
48 49
International Association of Sound and Audiovisual Archives. Technical Committee. IASA-TC 03, Vers. 3. The Safeguarding of the Audio Heritage: Ethics, Principles and Preservation Strategy. Dec. 2005. 28 Aug. 2009 <http://www.iasa-web.org/sites/default/les/downloads/publications/TC03_English.pdf>
Metropolitan New York Library Council. 2006 Digitization Survey Final Report. Oct. 2006: 3. New York, NY. 4 July 2011 <http://www.metro.org/attachments/contentmanagers/79/2006_digsurveyreport.pdf>
50 51
International Survey of Library and Museum Digitization Projects. New York: Primary Research Group Inc. c2008: 17. 29 July 2011 <http://www.primaryresearch.com/>
The American Archive. cpb.org. Corporation for Public Broadcasting. 29 July 2011 <http://www.cpb.org/features/americanarchive/>
52
FY 2011 Operating Budget. cpb.org. Corporation for Public Broadcasting. 29 July 2011 <http://www.cpb.org/aboutcpb/nancials/budget/cpb_fy2011OperatingBudget.pdf >
53 54
Library of Congress. Preserving Our Digital Heritage: The National Digital Information Infrastructure and Preservation Program 2010 Report. Library of Congress Digital Preservation Publications. Jan. 2011: 5. 13 May 2011 <http://www.digitalpreservation.gov/library/resources/pubs/docs/NDIIPP2010Report_Post.pdf> Betsy McLane. Personal Interview, 30 June 2011.
55
56 Data Management for NSF Engineering Directorate Proposals and Awards. nsf.gov. National Science Foundation. 6 Sept. 2011 <http://www.nsf.gov/eng/general/ENG_DMP_Policy.pdf >
119
END NOTES
DCC Charter and Statement of Principles. Digital Curation Centre. 2010. 8 July 2011 <http://www.dcc.ac.uk/about-us/dcc-charter>
43
120
BIBLIOGRAPHY
Academy Film Archive of North America. 2010. <http://www.afana.org/> Academy of Motion Picture Arts and Sciences. Science and Technology Council. The Digital Dilemma: Strategic Issues in Archiving and Accessing Digital Motion Picture Materials. Beverly Hills, CA: AMPAS, 2007. Academy of Motion Picture Arts and Sciences. Science and Technology Council. Digital Motion Picture Archive Framework Project. 2011 AMPAS. <http://www.oscars.org/science-technology/council/projects/dmpafp.html> Academy of Television Arts and Sciences. 1995-2010. ATAS. <http://www.emmys.tv/> American Film Institute. History of AFI. 2011 AFI. <http://www.a.com/about/history.aspx> American Library Association. Denitions of Digital Preservation. ALA Annual Conference. Washington, DC. 24 June 2007. <http://www.ala.org/alamgrps/divs/alcts/resources/preserv/defdigpres0408.cfm> Anderson, Chris. The Long Tail. Wired 12. 10 Oct. 2004. CondeNet Inc. 17 Aug. 2007 <http://www.wired.com/wired/archive/12.10/tail.html> Anthology Film Archives. N.d. <http://anthologylmarchives.org/> Archives of Appalachia at East Tennessee State University. N.d. <http://www.etsu.edu/cass/archives/about/contactus.aspx> Association of Canadian Archivists. What Is an Archivist? Mar. 2004. ACA. <http://www.archivists.ca/sites/default/les/Attachments/Outreach_attachments/What-an-Archivist.pdf> Audio Engineering Society. 2011. AES. <http://www.aes.org/> Audio Visual Working Group. 6 Dec. 2010. Federal Agencies Digitization Guidelines Initiative <http://www.digitizationguidelines.gov/audio-visual/> Baseball: A Film by Ken Burns. Dir. Ken Burns. Public Broadcasting Service Home Video. 28 Sept. 2004. Columbus International Film and Video Festival. 2011. Film Council of Greater Columbus, Inc. <http://www.chrisawards.org/> Comcast. 2011. <http://www.comcast.com/default.cspx> Country Music Hall of Fame and Museum. N.d. Country Music Foundation, Inc. <http://countrymusichalloffame.org/>
121
Bibliography
BIBLIOGRAPHY
EAD: Encoded Archival Description. 1 Sept. 2011. The Library of Congress. Standards. <http://www.loc.gov/ead/> Entertainment Identier Registry (EIDR). <http://mesalliance.org/blog/tag/movielabs/> Entertainment Technology Center. Interoperable Master Format. 2011. University of Southern California. <http://www.etcenter.org/imf/> Federal Agencies Digitization Guidelines Initiative. Visual Working Group. N.d. loc.gov. <http://www.digitizationguidelines.gov/audio-visual/> Fedora Commons Repository Software. N.d. <http://fedora-commons.org/> Franklin Furnace Archive, Inc. N.d. <http://www.franklinfurnace.org/about/institutional_history/organizational_overview.php> Full Frame Documentary Film Festival. Center for Documentary Studies 2010-2011. <http://www.fullframefest.org/> George Eastman House. International Museum of Photography and Film. The Motion Picture Collection. 2000-2011. GEH. <http://www.eastmanhouse.org/collections/motion-picture.php> Getty Research Institute. N.d. J. Paul Getty Trust. <http://www.getty.edu/research/> Grammy Foundation. 2011. The Recording Academy. <http://www.grammy.org/grammy-foundation> Harry Ransom Center. The University of Texas at Austin. N.d. <http://www.hrc.utexas.edu/> Harvard Film Archive. 2011. Fine Arts Library of the Harvard College Library. <http://hcl.harvard.edu/hfa/> Hoover Institution. 2011. Board of Trustees of Leland Stanford Junior University. <http://www.hoover.org/> Hot Docs. Canadian International Documentary Festival. <http://www.hotdocs.ca/> Hulu. 2011. <http://www.hulu.com/> Independent Film & Television Alliance. IFTA FAQs. N.d.:1-5. Los Angeles, CA. <http://www.iftaonline.org/sites/default/les/IFTA_FAQs.pdf> International Documentary Association. 2011 IDA. <http://www.documentary.org/> International Standard Audiovisual Number (ISAN). 2003-2010. <http://www.isan.org> Interoperable Master Format Project. Entertainment Technology Center at USC. 2011. <http://www.etcenter.org/imf/> Image Interchange Framework. 2011. Academy of Motion Picture Arts and Sciences. <http://www.oscars.org/science-technology/council/projects/iif.html> Joomla! Streaming Culture. N.d. <http://www.playbackarts.com/beta/> Library of Congress. N.d. USA.gov. <http://www.loc.gov/index.html> Library of Congress. Sustainability of Digital Formats Planning for Library of Congress Collections. Material Exchange Format (MXF). N.d. USA.gov. <http://www.digitalpreservation.gov/formats/fdd/fdd000013.shtml> Los Angeles Film Festival. Our History. Film Independent. 2011. <http://www.lmindependent.org/about/our-history/>
122
Bibliography
National Anthropological Archives and Human Studies Film Archives. National Museum of Natural History. Smithsonian Institution. N.d. <http://www.nmnh.si.edu/naa/index.htm> National Archives. N.d. The U.S. National Archives and Records Administration. <http://www.archives.gov/> National Science Foundation. 15 Apr. 2011. NSF. <http://www.nsf.gov/> Netix. 2011. Netix, Inc. <https://www.netix.com/> New York Public Library. Dorot Jewish Division. 2011. The New York Public Library. <http://www.nypl.org/locations/schwarzman/jewish-division> New York Public Library. The Rodgers and Hammerstein Archives of Recorded Sound. 2011. The New York Public Library. <http://www.nypl.org/locations/lpa/rodgers-and-hammerstein-archives-recorded-sound> New York State Council on the Arts. 2005. <http://www.nysca.org/> Northeast Historic Film. N.d. <http://www.oldlm.org/> Open Text Digital Media Group. (Artesia) 1999-2011. Open Text. <http://digitalmedia.opentext.com/> Preservation Metadata Implementation Strategies. 2011. Library of Congress. Standards. <http://www.loc.gov/standards/premis/> Public Broadcasting Metadata Dictionary Project. N.d. Corporation for Public Broadcasting. <http://pbcore.org/> Reece Museum. A Unit of the Center for Appalachian Studies and Services. N.d. East Tennessee State University. <http://www.etsu.edu/cass/reece/> Society of American Archivists. So You Want to Be an Archivist: An Overview of the Archives Profession. 2011. SAA. <http://www2.archivists.org/profession/> Society of Motion Picture and Television Engineers. SMPTE Standard for File Format for Digital Moving-Picture Exchange (DPX), V 2.0, 268M-2003, 5 May 2003. Print. South by Southwest Film Festival. SXSW Inc. 2012. <http://sxsw.com/lm> South Carolinas Flagship University. University Libraries. Moving Image Research Collection. N.d. <http://library.sc.edu/mirc/> Stanford University Libraries and Academic Information Resources. 2011. Stanford University. <http://www-sul.stanford.edu/> Sundance Film Festival. Sundance Institute 2000-2011. <http://www.sundance.org/> Telluride Film Festival. National Film Preserve 2011. <http://www.telluridelmfestival.org> Time Warner Cable. 2004-2011. Time Warner Cable Inc. <http://www.timewarnercable.com/> Toronto International Film Festival. TIFF History. N.d. Tiff.net. <http://www.tiff.net/abouttiff/organization/tiffhistory> Tribeca Film Festival. About Us. Tribeca Enterprises LLC 2011. <http://www.tribecalm.com/about/>
123
BIBLIOGRAPHY
Bibliography
BIBLIOGRAPHY
United States Copyright Ofce. Copyright Law of the United States of America and Related Laws Contained in Title 17 of the United States Code. Circular 92. 31 Aug. 2011 <http://www.copyright.gov/title17/92chap1.html> United States Holocaust Memorial Museum. N.d. <http://www.ushmm.org/> University of California, Berkeley Art Museum and Pacic Film Archive. 2011. UC Regents. <http://www.bampfa.berkeley.edu/> UCLA Film & Television Archive. 2011. UCLA. <http://www.cinema.ucla.edu/> University of Virginia Library. N.d. University of Virginia. <http://www.lib.virginia.edu/> Walter J. Brown Media Archives and Peabody Awards Collection. 2010. University of Georgia Libraries. <http://www.libs.uga.edu/media/> Washington University Film & Media Archive. 4 Aug. 2011. Washington University. <http://library.wustl.edu/units/spec/lmandmedia/> Wisconsin Center for Film and Theater Research. 2011. Wisconsin Historical Society. <http://www.wisconsinhistory.org/wcftr/> Witness. N.d. <http://www.witness.org/index.php?Itemid=51&id=68&option=com_content&task=view> Yee, Martha M. Moving Image Cataloging: How to Create and How to Use a Moving Image Catalog. Westport, CT: Libraries Unlimited, 2007. Zhang, Allison, and Don Gourley. Creating Digital Collections: A Practical Guide. Oxford: Chandos Publishing, 2009.
124
ACKNOWLEDGMENTS
This report would not have been possible without the support of the members of the Science and Technology Council, the consultants who performed the research and interviews, the many experts from the independent lmmaking, documentary and nonprofit audiovisual archive communities who generously contributed their time and knowledge, and the support and funding provided by the Library of Congress. This report, in its various draft forms, was reviewed by several of the individuals listed below to assist the Council with delivering a nal version that meets Academy standards for overall quality and accuracy. The reviewers comments and identities will remain condential to protect the peer review process. The reviewers have not seen the final version of this report, nor do they necessarily support the reports conclusions and recommendations. The Academy of Motion Picture Arts and Sciences is wholly responsible for the contents of the nal report. Science and Technology Council Bill Kroyer, Co-chair George Joblove, Co-chair Andy Maltz, Director Project Lead Milt Shefter
125
Acknowledgments
ACKNOWLEDGMENTS
Section Leads Independent Filmmakers: Documentarians: Nonprot Archives: Bob Fisher, Bob Hoffman, Gregory Lukow Ed Carter, James Moll Eddie Richmond, Linda Tadic (lead consultant)
The Council would also like to thank all of the individuals and organizations that participated in the creation of this report:
Ruta Abolins Geoff Alexander Steven Anastasi John Bailey Beth Bird Dan Birman Gilda Brasch Peter Broderick Brandon Burke Ken Burns Karen Cariani Ed Carter Paul Chapman Liz Coffey Elizabeth Cohen Matthew Cowan Shari Dalconzo Bruce Davis Ira Deutchman Geralyn Dreyfous Maxine Fleckner Ducey Harrison Engle David Farneth Ray Feeney Carl Fleischhauer Hannah Frost Sid Ganis Brian Gerber Mark Gill Randy Gladden Nancy Goldman Shawn Guthrie Bob Harvey Laurin Herr Charlie Herzfeld Ted Hope Brad Hunt Sara Hutchison Cheryl Boone Isaacs Michael Katchen Laura Kim Carol Krumbach Ben Kufrin Ellen Kuras Kendra Kurosawa Grace Lile Evan Lim Mark Lipsky Michael Lumpkin Scott Macaulay Peter Mavromates Laura Lee McKay Michael Morelli Patrick Murphy William Murphy Jarod Neece John Nein Brian Newman Bob Niemack Rob Nilsson Michael Pogorzelski Steven Reich Norma Myers Riddle Leigh Rockey Andy Romanoff David Rowntree Bill Russell Marjan Sania Laurie Ann Schag Eddie Schmidt Sandra Schulberg Lucia Schultz Karan Sheldon Leon Silverman Michael Terry Susie Tofte Rick Utley Norma Vega Moises Velez Sara Velez Tom Vice Pi Ware David S. Weiss Kenneth Weissman Martha Wilson Steve Wilson Beverly Wood Andy Young
126
Acknowledgments
Academic Film Archive of North America Anthology Film Archive Archives of Appalachia, East Tennessee State University Berkeley Art Museum and Pacic Film Archive Film & Media Archive, Washington University in St. Louis Franklin Furnace Getty Research Institute Harvard Film Archive Hoover Institution, Stanford University New York Public Library, Dorot Jewish Division New York Public Library, The Rodgers and Hammerstein Archives of Recorded Sound Northeast Historic Film Stanford University Libraries and Academic Information Resources University of Texas at Austin, Harry Ransom Center University of Virginia Library Walter J. Brown Media Archives and Peabody Awards Collection, University of Georgia Wisconsin Center for Film and Theater Research Witness
127
ACKNOWLEDGMENTS