Folkstreams

Back to Homepage

Folkstreams Guide to Best Practices in Film Digitization

Prepared by Heather Barnes, August 2007 and updated by Tom Davenport and Steve Knoblock in August 2010.

Originally funded by the Institute for Museum and Library Services (IMLS) and updated with funds from the American Folklore Society.

Table of Contents

1.0 Background
1.1 Introduction
1.2 History of the Folkstreams project
1.3 Scope of the project

2.0 Discovering and acquiring films
2.1 Selection criteria
2.2 Acquiring films
2.3 Copyrights and permissions
2.4 Contextual information

3.0 Film and video digitization
3.1 General film digitization issues
3.2 Folkstreams digitization workflow
3.3 Working with Colorlab
3.4 Web stream creation

4.0 Website Infrastructure
4.1 Site development
4.2 Programming

5.0 Metadata

6.0 Preservation
6.1 General preservation issues
6.2 Summary of the Folkstreams archiving process
6.3 Finding aid and catalog records

7.0 Advertising and Outreach

8.0 Evaluation
8.1 Evaluation criteria
8.2 User feedback
8.3 Website statistics

9.0 Appendices and links
9.1 Folkstreams films currently streaming on website
9.2 List of database fields
9.3 Website feedback survey
9.4 Contributors and staff

10.0 Works cited
10.1 Full bibliography

1.0 Background

1.1 Introduction

The Folkstreams project has two goals. One is to build a national preserve of hard-to-find documentary films about American folk or roots cultures. The other is to give the films renewed life by digitizing them and making them available on the project's website, www.folkstreams.net The films selected for the project by the Folkstreams committee were created by independent filmmakers during a time of prolific documentary filmmaking - one that began in the 1960s, and was made possible by the development of portable cameras and the capacity for synchronizing audio and video. These films focus on the culture, stories, and crafts of otherwise-unnoticed Americans from many different regions and communities.

The filmmakers were driven more by sheer engagement with the people and their traditions than by commercial hopes. Their films have unusual subjects, odd lengths, and talkers who do not speak "broadcast English." Although many have won prizes at film festivals, or have been screened in college classes or broadcast on PBS, they have never found outlets in venues such as theaters, video shops or commercial television. But they have enduring value. They emerged from the same intellectual movement that gave rise to American studies, regional and ethnic studies, the "new history," "performance theory," and investigation of tenacious cultural styles in phenomena like song, dance, storytelling, visual designs, and ceremonies. They also respond to the intense political and social ferment of the period.

The filmmakers and the researchers they collaborated with explored performances situated in a community's customary work, worship, and play. Beneath their colorful surfaces often lie serious issues of physical, psychic, and social survival under duress. For understanding what they saw the filmmakers relied more heavily on observant and knowledgeable community members than on outside "experts." They conveyed understanding through action and symbol as often as by "talking heads."

Many of the films are linked to significant published research. Folkstreams draws on this research to accompany and illuminate the films. The films themselves add powerful dimensions to folklore scholarship. They offer direct observation of unfamiliar worlds. Many of these are now receding into the historical past, but the creators of Folkstreams hope that the example of these films may stimulate contemporary filmmaking on subjects and people still ignored by mainstream corporate media.

1.2 History of the Folkstreams project

The idea for Folkstreams grew out of attempts by documentary filmmakers to gain greater exposure for their films. As independent filmmakers, they did not have access to the standard advertising and distribution systems of mass market film distribution. Neither movie theaters nor commercial television networks would show them. Video stores, when they spread across the country, wanted the Hollywood blockbuster hit. Public television sometimes broadcast the films, particularly in early years, but its programmers were uneasy with several characteristics of these documentaries. The films often ran in odd lengths that did not fit into the time slots crystallized for television. They lacked the stars to draw an immediate audience. The language of the subjects was a barrier. They spoke dialects colored by race and region and class or even languages like Cajun French. Audiences might lack the background to understand the social worlds that the films showed.

The documentaries to which public-television programmers instead gravitated typically had national historical subjects presented through scripted narration woven with archival photographs, newsreel footage, and talking heads of scholars. If independent filmmakers could not work through existing media institutions, they also found that they had no good way to advertise and sell to the general public. They targeted libraries and schools, but had no effective way to acquaint them with their films or to make a living by selling them at prices that would promote purchases.

Unable to make a living solely from his documentaries, Tom Davenport developed From the Brothers Grimm, a successful series of dramatized adaptations of fairy tales translated into American settings. This led him and his wife and partner Mimi Davenport in 1999 to construct a website for their feature-length film, Willa: An American Snow White. They quickly saw that the Internet had the potential also to connect documentary filmmakers with niche audiences. A website streaming major films on American vernacular culture could introduce audiences worldwide to important works they might otherwise never see. Enhancing the public's awareness of hard-to-find films could benefit viewers and stimulate video and stock footage sales for the filmmakers and their distributors. The films themselves, and the prospect of a viable career might also encourage a new generation of filmmakers to take up documentary work.

Tom Davenport also saw that the website could also include contextual information along with the films. Documentary filmmakers have often been or collaborated with musicians, folklorists, anthropologists, Americanists, or other scholars in choosing subjects and approaches and in filming and editing. They have learned that the information supplied in a classroom often made their films better understood and appreciated. So filmmakers began to collaborate with researchers to create materials to accompany their films. Folklorist Daniel Patterson, working with historian/folklorist Allen Tullos and filmmaker Tom Davenport, produced some of the earliest, a 16-page booklet of "background, transcription, and commentary" to accompany their film Born for Hard Luck in 1972. Their similar, 44-page booklet for Being a Joines (1980), gained strength from the collaboration of Joyce Joines Newman, a daughter of the central figures in the film, who participated in the production of both the film and the study materials. She combined an insider's understanding with the perspective of a professional folklorist. Two subsequent Davenport films, A Singing Stream and The Ballad of Frankie Silver, had their booklets published as special issues of The North Carolina Folklore Journal. The interlinking of documentary films with print scholarship has come to be extensive. A dozen of the films currently streaming on, or soon to be added to, the Folkstreams website have parallels in books written by persons who collaborated in the making of the films.

In summary, as Tom and Mimi Davenport have written:

The idea of creating Folkstreams.net grew out of our love of filmmaking, a respect for the traditional culture of ordinary Americans, and a desire to get our work to the general public. Heretofore, much good independent film work was like the tree falling in the wilderness with no one to hear. With the Internet and video streaming, we will be able to make a 'national park' from this wilderness where everyone can come and freely hear and see what we, and others, have labored on for so long and with such enjoyment. The idea of a 'cultural preserve' as a kind of national park of intellectual property is an important one for our times.

1.3 Scope of the project

The technical objectives of the Folkstreams project are 1) to provide broad access to a unique selection of folklore documentary films through digitization and online access, and 2) to ensure long-term preservation of archival-quality copies of the films. In 2004, the Institute for Museum and Library Services (IMLS) made a grant to the University of North Carolina at Chapel Hill's School of Information and Library Science (SILS) to digitize, archive, and create web streams of 35 films that document "traditional" culture in the United States.

Working with folklorist Daniel Patterson and other scholars, filmmakers, archivists, and computer specialists whom he recruited for a Folkstreams advisory committee, Tom Davenport created a proposal which has been the basis of successful grant applications to both of the National Endowments, and to the Institute for Museum and Library Services as well as state Arts and Humanities organizations. This initial effort resulted in a database and prototype of the Folkstreams website; the project officially began streaming films on the Internet in 2002.

The project completed its work on the first 35 films in November 2005. Collaborators include UNC Chapel Hill's School of Information and Library Science, Ibiblio.org, Folkstreams, Inc., a 501(c)3, non-profit organization, and the Southern Folklife Collection.The grant also supported creation of a special series of short films, entitled Video Aids to Film Preservation (VAFP), that demonstrate the craft of film restoration and preservation.

The films acquired by the project are first digitized and then streamed via a website, www.folkstreams.net. The website features contextual information about each film such as transcriptions, study and teaching guides, suggested readings, and links to related websites. Preservation copies of the films are placed at the University of North Carolina at Chapel Hill's Southern Folklife Collection (SFC); access/viewing copies (on DVD) are available to visitors to the SFC. In addition to grants for developing the website, the project has garnered support from many documentary filmmakers. They have supplied copies of their prints and valuable contextual information for the films.

2.0 Discovering and acquiring films

2.1 Selection criteria

Folkstreams relies on a selection committee, made up of folklore scholars and other experts, to review and recommend films for the project. Committee members have substantial experience in identifying preservation-worthy documentary films. Dr. Daniel Patterson, a renowned folklore scholar, serves as a primary advisor to the project. Daniel Patterson and Beverly Patterson, who also serves on the Folkstreams Committee, were for some years co-editors of the film review section of the Journal of American Folklore. In addition, the committee finds new films by soliciting recommendations from viewers and other filmmakers. Project director Tom Davenport notes two major criteria for selecting a film: 1) the film is generally regarded as a high-quality film, and 2) the committee's folklore scholars consider them to be valuable or significant films in the folklore genre. The existence of contextual materials enhances a film's value to the project; filmmakers' willingness to participate in the project is another important factor.

The following are comments by Dr. Dan Patterson on how the project defines and selects folklore films:

What we look for in choosing films for the Folkstreams site are the films that usually focus on oral traditions, especially those with historical depth, and also the subcultures that generate and sustain and find value in these traditions. We especially like the films that give strong performances of these traditions (whether music, narrative, craft, ceremony, work, worship, etc.) and explore their meanings and uses. We like ones in which outstanding and knowledgeable community performers and leaders ("community scholars," some call them) demonstrate and explain the traditions.

As in most fields in the humanities, definitions in folklore never win a unanimous vote. This is in part because the fields of study continuously evolve and drag along all their older history, interests, and disputes behind them.

In their definition of the term "folklore," some people in the field stress the central role of oral transmission and/or imitation of observed actions. This was one of the first characteristics of folklore that people became aware of. This orientation led them to focus on the aesthetic forms, performance styles, origins, and history of change of songs, instrumental music, tales, rituals, arts, handicrafts, games, beliefs, and customs. Other people stress, in their definition of folklore, the kind of cultures in which oral transmission normally plays a big role. In ancient times this included everyone except priests and scribes, who could read and write. With the spread of literacy it was narrowed to pre-literate tribes, or illiterates within a mainly literate society, or people who lived their lives without much reliance on writing - that is, cultural enclaves like the rural village, the rural peasant or urban working class, and the regional, ethnic, religious, occupational, gender, or age-group subcultures in a larger society.

Increasingly, folklorists have come to realize that these are interrelated and complex issues. Even in a highly literate society, everyone is a member of groups in which oral communication flourishes - gossiping fellow workers and neighbors, for example - and many of us in daily life move frequently from roles that demand highly technical literacy or are a part of a very rapidly evolving pop culture spread across the entire country to other roles that are more oral, more local, more slow to change, perhaps even hostile to the influence of the dominant society around them, which they may see as an exploiter. Some of the oral culture is rather trivial, such as jokes told around the office coffee pot; some is absolutely essential for physical or psychic survival (like the wood-chopping songs of men in the Texas prison) or importantly symbolic of one's identity and relationships (like music within some small protestant church congregations or a saint's-day ceremony in an Italian American community).

I like to think about all of these ways of looking at folklore as involving many variables, each variable actually encompassing many degrees from "little" to "much." The questions therefore become, "In what ways is this folklore and not folklore? To what degree in each of the ways?" For an example that may clarify this, see the piece I wrote on Shaker song for Davenport's film The Shakers. In that you'll find a list of variables one can use to distinguish the mainly oral tradition in early Shaker songs from the mainly literate tradition in Shaker music after 1870.

Folkstreams does not currently maintain a formal, documented methodology for adding to its catalog. An important element of building digital libraries is developing a formal collection policy. The National Information Standards Organization (NISO) argues in its Framework of Guidance for Building Good Digital Collections that a "good digital collection is created according to an explicit collection development policy that has been agreed upon and documented before digitization begins." For Folkstreams and similar cultural heritage collections, a formal collection policy would incorporate a written component that is flexible enough to incorporate new or surprising "finds" in its film discovery process, but adequately structured so as to provide users and other stakeholders with a clear understanding of the project's intentions.

2.2 Acquiring films

Since its inception, Folkstreams has added almost one hundred films to its digital collection. Its founders initially encountered some challenges in engaging filmmakers; some feared losing potential revenue from sales of their films, or were wary about providing web access to their works. Many, recognizing the value of exposing their films to new audiences, were eager to participate. Persuasive evidence for the benefits of sharing films on the Internet came when producers of the French film Amélie agreed to purchased rights to footage from Tom Davenport's film Born for Hard Luck.

As the project has grown, lending their films to Folkstreams has provided several benefits to filmmakers. Some filmmakers have experienced increased sales of films, or more intangible benefits to their careers, primarily an increased audience for their films and a boost to their reputations in the field. Most important, these films can be seen. As Tom Davenport notes, the sum is greater than the parts; as it develops a critical presence, the Folkstreams site adds value to its catalog of films.

Acquiring films involves both planning and serendipity. In some cases, it has taken the project several months to procure a film, even after establishing an agreement with the filmmaker. It may be difficult for the filmmaker to locate the film components, or, for unknown reasons, the process of getting the physical copies is delayed. Shawn Nicholls, who has worked for the Folkstreams project in both researching and acquiring the films, notes that locating films for Folkstreams is similar to doing detective work:

Several qualities are required to obtain films: persistence, ability to establish rapport, and knowing something about the film we are soliciting. There have been instances in which a filmmaker was reluctant to send his or her film to Colorlab for fear that it would get lost. I think that is an example of why it is important to develop the relationship, establish a level of trust, and show that we are willing to work out a solution.

Initially, we try to learn a bit about the film, projects that the filmmaker has worked on or is working on, and his or her production company. Most of this information comes from the Internet: biographies, reviews, articles, production company/personal sites, or film festival websites.

In some cases we've gone after films and it's been really hard to get any information on the filmmaker. In a couple of instances, I wasn't able to find any information at all, so that can make it difficult.

I'll take a title, for example, and search on the Internet; nine times out of ten, I can find the filmmaker or the production company. It's almost like being a detective, in a sense. In some cases, I can work a couple of days just trying to get information about one filmmaker, but I've got one or two films I'm thinking of right now about which I just can't find any information. I've got the title and the name of the filmmaker, but finding more information presents a challenge. It is important to keep working at it, and not to give up.

2.3 Copyrights and permissions

Filmmakers who provide copies of their films to the Folkstreams project retain full copyright to their works. A standard contract is used to formalize agreements between the filmmaker and the project; the contract can be modified based on the specifics of the agreement. (Click here to view the standard Folkstreams agreement in PDF form.) The National Information Standards Organization (NISO) emphasizes in its Framework of Guidance for Building Good Digital Collections that a digital collection "Å“respects intellectual property rights. Collection managers should maintain a consistent record of rightsholders and permissions granted for all applicable materials." [1]. Gail Hodge, in Best Practices for Digital Archiving, also notes, "One of the most difficult access issues for digital archiving involves rights management. What rights does the archive have? What rights do various user groups have? What rights have the owners retained? How will the access mechanism interact with the archive's metadata to ensure that these rights are managed properly?" [2] Folkstreams endeavors through its copyright policy to preserve filmmakers rights to their intellectual property, while providing benefits for the filmmaker through sharing the films online. In addition to the formal contract outlining the details of the arrangement between the filmmaker and the project, Folkstreams maintains a database record of each film's copyright holder, and the website contains a clear statement of copyrights on its "rights" page:

Folkstreams.net is a video streaming site. With the exception of the downloadable short clip, all films or long excerpts from films are for streaming only. No copying permitted without permission. For permission, apply to the filmmaker/distributor under "Order video/stock footage" on each film's main page or "Film Facts" page.

Copying permitted ONLY for the short downloadable clip. This is to facilitate sharing the clip via e-mail to bring attention to the film. These short clips are licensed under the Creative Commons (www.creativecommons.org) agreement which requires:

* attribution (you must acknowledge the copyright holder)
* noncommercial use only
* no derivative work (the clip cannot be changed).

You can apply to the filmmaker for exceptions to these rules.

The project has had minimal difficulties retaining content once it has been acquired, despite hesitation by a few filmmakers to allow complete versions of their films to be placed online. The success of this aspect of the project would seem to support the open, transparent arrangements in which it engages with the filmmakers. Sales of films are generally handled by distribution companies or by the filmmaker; links to the distributor of each film, if available, are included in the film's website on folkstreams.net. Filmmakers are viewed as equal participants in the project. Folkstreams emphasizes continued, full participation by the filmmakers once their films are added to the site. A filmmaker may update biographical and contextual information as he or she wishes; the website provides tools for filmmakers to control various aspects of their film's presence via the "MyFolkstreams" administration tool.

2.4 Contextual information

The Folkstreams website features background information about each film, such as a narrative summary, film facts, and distribution information. These contextual resources help build interest around the films and provide important background about characters, scenes, or topics that may not be easily accessible or understandable by those not familiar with the communities featured in the films. For example, the page for the film Born For Hard Luck is accompanied by the following contextual information:

Born for Hard Luck: Peg Leg Sam Jackson
Film by Tom Davenport
Produced by Tom Davenport with Allen Tullos and Daniel Patterson
Cinematographer: Tom Davenport
Sound: Allen Tullos with Kip Lornell
Editing: Louise B. Steig
Copyright: 1976, Tom Davenport
29 minutes, Black and White
Original format: 16mm, 1976
Distributor Contact: Davenport Films

More Film Facts

A portrait of Arthur "Peg Leg Sam" Jackson --black harmonica player, singer, and comedian who made his living "busking" on the street and performing in patent-medicine shows touring southern towns. Footage includes excerpts from one of his last medicine shows, videotaped at a county fair in 1972, and material filmed near his home in South Carolina in 1975. The performance includes harmonica solos, songs, a parody of a chanted sermon, folktales and reminiscences, and three buck dances.

A clip from this film appeared in Jean-Piere Jeunet's popular French feature Amélie. Born for Hard Luck is available on DVD from Davenport Films.

If you are interested in other films about medicine shows, we recommend Free Show Tonight, by academy award winner Paul Wagner.

If you liked Born for Hard Luck you may also like:

* Sonny Terry: Whoopin the Blues
* Land Where the Blues Began
* Free Show Tonight

In addition to film facts, for selected films viewers can browse links in categories such as Background, Making the Film, Using the Film, and Order the Film. Background contains a bibliography and additional topics related to the film. Making the Film addresses the process of creating the work. Users can click on On video or stock footage to find information about obtaining a copy of the film. Tom Davenport notes that the pop-up transcripts provided with the films are much like director's commentaries provided on a typical DVD.

Description and metadata such as transcripts, film information, and links to other resources all help to provide appropriate context for Folkstreams films so that they may be effectively used and understood. Folkstreams treats contextual information-gathering as an ongoing, "organic" process that is intended to engage all stakeholders, including viewers, the communities from which the films arise, filmmakers, and project staff. Dr. Dan Patterson comments on the process of building contextual information for the Folkstreams films:

We try to foresee what in the important content of the film will be unfamiliar or puzzling to the average person or even unnoticed by the outsider. And we try to get several short but informative essays written about these by someone who is from the culture and can interpret it to outsiders or who has studied the culture and its lore or has at least studied folklore and is interested in exploring the traditions in the film.

In some cases we are fortunate in that the writer has actually participated in the making of the film, as a folklorist working as a consultant to the filmmaker. I worked, for example, with Tom Davenport on a number of his films and (with other participants) helped prepare background materials for them. I was writing my book The Shaker Spiritual when Tom made The Shakers, and my book A Tree Accurst: Bobby McMillon and Stories of Frankie Silver when he was making The Ballad of Frankie Silver. It was Tom's film project, in fact, that stimulated me to write this book. Folklorist Bruce Jackson worked with the Seegers on Afro-American Worksongs in a Texas Prison and was writing a book on the same prison songs, later published under the title Wake Up Dead Man; he wrote the background materials for this film on Folkstreams. And so on.

A sizable number of films on the website are paralleled by books and articles written by folklorists involved in the films, and these materials are routinely drawn upon for the background materials. Probably the most delicate problem posed by the writing of the background materials is that the perspective of the writer may differ from that of the community shown in the film. The writer needs to find a way to write respectfully and tactfully about the material and at the same time honestly. The background materials offer not just praise but also analysis and insight.

The diversity of the films is so great that there is no set list of things to include in the packet of background materials. We do, however, try to have as many as possible of the following:

1. One or more essays on traditions presented in the film and the culture with which the film deals.

2. An account of the making of the film-usually written by or in collaboration with the filmmaker.

3. A biographical note on the filmmaker, linked where available to the filmmaker's Internet homepage.

4. Notes on persons appearing in the films as major performers or consultants, including biographical facts, information about books and articles written about them, other films about them, recordings they have made, honors received.

5. Suggested readings and links. These include books, articles, and websites bearing on the film and the traditions in it. See for example Almeda Riddle: Now Let's Talk about Singing, where we link to a site in Arkansas where a library streams the John Quincy Wolfe Collection of field recording and the viewer can listen to many performances by Almeda Riddle and other Ozark ballad singers, recorded fifty years ago. Another such link is set into Final Marks: The Art of the Stone Cutter, where the user can click and go to an archival site that streams a huge collection of photographs of 18th-century New England gravestones, including markers by the stone carving family whose shop in Providence, R.I., has had an unbroken history down to carvers who currently practice there.

6. A study guide for middle- and high-school teachers; this needs to be prepared by someone who has had teaching experience at these levels.

7. A transcript of the sound track: in some cases, this will be crucial for the future use of the film; some dialects now comprehensible will grow very hard for people to understand as years pass and the current speakers die. The transcriptions also make the content of the films accessible to the hearing impaired.

3.0 Film and video digitization

3.1 General film digitization issues

Best practices in film digitization are still emerging, and will most likely continue to evolve as technologies change and hard disk storage becomes more affordable. Creating digital objects from analog originals is a complex undertaking, the success of which depends upon many factors, including but not limited to the fragility of originals, availability of playback equipment, and intended purpose or audience for the digitized versions.[3] Much of the current research in the field of digital preservation focuses on a range of issues related to developing archival-quality repositories, or "trusted digital repositories" (TDRs) for digitized or born-digital materials. Some of the difficulties inherent in working with the resultant digital materials include the availability of multiple (often proprietary) file formats and the uncertainties inherent in rapidly changing technology. [4]

Though a complex and potentially costly process, digitization of analog materials, particularly of moving images, provides many benefits. According to the University of Maryland's best practice guidelines for digitization projects, these benefits can contribute to both preservation and access of the films [5]:

Enormous file sizes, time and labor intensive processes, and the instability of the original object all contribute to the difficulties inherent in digital reformatting of audio and moving images. There are, however, marked advantages to the digital reformatting of audio and moving image material that can compensate for the complexity of the process. These include fragile analog originals receiving less wear and tear due to repeated use, increased remote access to the content, improved intellectual access through appropriate metadata creation and increased flexibility for future use.

The Folkstreams collection consists of films in analog formats, primarily 16mm film, though some were transferred to video; contemporary films in the collection were shot on digital video. Significant expenses can be incurred when digitizing a film from analog original components. It is generally easier and cheaper to transfer from a video release print. If an existing release print of a film selected for acquisition is available, Colorlab, the company engaged by Folkstreams to process the films, uses it to create a Digibeta copy.

Much of the literature on digitization of cultural heritage materials emphasizes digitization for both preservation and access. [6,7] In the case of Folkstreams, digitization (after the initial transfer to Digibeta) is primarily intended to provide a platform for online access to the films. In other words, the online versions (MPEG-4 and Real files) of the films are created in order to be shared. The creation of compressed digital files from the film components is not intended to provide long-term, preservation-quality digital objects. It is generally assumed that, in the digitization process, there is a trade-off between compression and video quality. File sizes can be kept to a manageable level with higher compression; however, quality tends to degrade exponentially at higher levels of compression. For online video streaming purposes, the quality of transfer from video is acceptable.

Projects such as the Digital Video Preservation Reformatting Project, the objective of which was to find an optimal preservation format for thousands of Dance Heritage Coalition tapes (mostly VHS), have recommended that film and video be digitized at the highest possible quality.[8] With its primary objective of providing access to the films, however, the Folkstreams project was not as concerned about the preservation of the access files themselves; those files are intended to serve the needs of current users and to provide optimal viewing experience at a reasonable speed and file size. The original film components, as well as the Digibeta tapes, serve as the preservation masters for this project. Digibeta, a Sony proprietary format, is a generally accepted preservation-quality storage format, although it is a tape-based format and thus may be subject to some of the long-term preservation concerns inherent in tape-based media. (A fuller discussion of the limitations of this format may be found in the Preservation section of this document.)

3.2 Folkstreams digitization workflow

The following is an overview of the process used by Folkstreams to convert original film materials to archival and digital copies. For the film and video conversion process, Folkstreams has engaged the services of Colorlab, a respected film processing company located in both Maryland and New York (http://www.colorlab.com/).

First, Folkstreams consults with the filmmaker to determine how the film was made (on 16mm film or video tape) what elements (both 16mm film elements and video tape transfers) are available. Depending on these elements and their condition, the work flows varies.

For 16mm films that already have professional video transfers, project staff prefer to make our Digibeta copy from the existing transfer to tape. However, with some films, the video transfers are old and defective. In these cases, staff use either a good positive release print with optical sound, or an intermediate single strand 16mm film element, like an interpositive or internegative, to make a new "best light" transfer to video. In the case of an interpositive or internegative transfer, technicians use the 16mm magnetic mix for the sound track. Transfers from these low contrast elements like the interpositive (IP) or Internegative (IN) and magnetic track produce the highest quality transfers. Steps 1 and 2 below are for transfers from 16mm film.

Step 1: If a new transfer of a 16mm film is required, the the acquired film is checked for damage and cleaned.

Step 2: The film is converted to Digibeta at Colorlab using a Rank-Cintel Flying Spot machine. (As stated above, if the filmmaker already has a good transfer of the film or if the filmmaker made the film on video, staff use that video tape to make the Digibeta). A second Digibeta clone/copy is made for the filmmaker, and a mini-DV copy is produced for Folkstreams to use in creating Youtube preview clips.

Step 3: One Digibeta is sent to the filmmaker, the mini-DV tape is sent to Folkstreams in Delaplane, VA, and the other Digibeta tape is shipped to Southern Folklife Collection (SFC) at UNC-Chapel Hill. If the Digibeta transfer was made from a 16mm print, Colorlab puts it on a core, into an archival box, and ships it to SFC with the Digibeta tape.

Step 4: At SFC, the sound and image archivist creates an Encoded Archival Description document for the film, and a MARC21 catalog entry for UNC Library's online catalog and for OCLC.

Step 5: Metadata are entered into the Folkstreams database, and are mapped to the Dublin Core standard.

Step 6: The 16mm is placed in the SFC archives. For titles that were made on video originally, the filmmaker sends us his/her best tape master -- usually a one inch, beta, betasp, or digital beta master. Folkstreams uses that tape master to make the Digibeta copy of Folkstreams as indicated in step 2. Staff return the old video master to the filmmaker because (unlike a 16mm film print which has a long life) these masters are usually too old to retain for the archive.

Step 7: Folkstreams (at the Davenport Films studio in Delaplane) captures the mini-DV tape into Final Cut Pro. Folkstreams makes a QuickTime-DV clone of the entire film which goes into a folder on a hard drive. Using Final Cut Pro they also cut a one to two minute trailer which is compressed into H.264 and uploaded to the Folkstreamer youtube channel. A QT-DV clone of the trailer is placed in the film’s folder on the hard drive. These Quick Time DV clones of the entire film and its trailer are of high enough quality that, as video on the Internet evolves, they can be used to make future streams of higher quality. These elements (a duplicate set) are also stored on a drive in the Folkstreams archive in the Southern Folklife Collection in Chapel Hill.

Colorlab notes what element was used to make the transfer (either 16mm film or the format of the master videotape) on both the Digibeta tape sent to SFC and on the mini-DV copy sent to Folkstreams. These notes are used to create an archive record on the Folkstreams administration database, that includes the date of transfer, the master used for the transfer and any other notes about the condition of the master.

3.4 Web stream creation

The Folkstreams project initially chose to create both MPEG and Real streaming files in order to provide optimum accessibility to a wide range of users. Real Media format was especially useful for computers with slower Internet connections, such as a 28.8kps modem. But Real is a proprietary system and users must download the player. MPEG 4 was open source and produced a better picture, but users without fast Internet connections were unable to play the MPEG 4 streams. Consequently in 2008 Folkstreams began to convert all the films to Flash.

The following is a brief summary of that procedure which has continued through 2010.

* Colorlab transfers the 16mm to Digital Betacam (Digibeta) tapes and make a mini-dv copy which is sent to Folkstreams in Delaplane, VA.

*Folkstreams uses those tapes to create high quality Quick Time DV file. These QT DV files are big, usually about 6 Gigs per half hour.

* The QT DV file is loaded onto a USB stick drive and mailed to Steve Knoblock in Arlington, VA.

*Steve uses the program Flash MX Pro by Moyea to make the files for Flash. Here is his account:

Quicktime DV format video is converted to Flash format using Flash MX Pro (a commercial Windows application by Moyea designed for converting video to Flash video). The standard specifications developed for Folkstreams use are:

video
480 x 320
H.264
450Kbps

audio
44100 / 96 stereo

The 450Kbps rate was chosen in deference to lower bandwidth users (800Kbps offered better quality but some users at 750Kbps reported streaming issues).

If more than one film is ready, conversion is done as a batch process, otherwise each film is processed individually.

A watermark is placed on the video, "Folkstreams Streaming use only." and appears in the lower right corner of the frame during the first minute of the video.

Before presenting the video on the site, the flash movie is evaluated visually for completeness, to be sure the video does not stall, and for any other defects. The flash video file is then uploaded by FTP to the folkstreams server. Video is organized into a single folder with a sub-folder for each film based on its title.

Before the film is made available to the public, screen grabs are taken from the film based on artistic considerations and to best represent the film. The screen grabs are taken from the original Quicktime move using Quicktime player (QT Pro). QT Pro saves screen grabs in PICT format. A free image processing application, Irfanview
is used to convert them to a size and format suitable for our website (JPEG approximately 480 x 320). One image is typically selected to be the “cover image? for the film in our catalog.

4.0 Website Infrastructure

4.1 Site development

Folkstreams began as a collection of static HTML pages maintained with a WYSIWYG web page editor. A link to the streaming version of a film was provided on the film's detail page. The first dynamic elements to be added to Folkstreams were pop-up pages for film transcripts. A database was developed to hold contextual notes related to each film's transcript. The overlib popup note library was used to display the notes. Note data were retrieved, and an overlib compatible array was populated with note content. Note contents were presented via javascript generated by a PHP script.

Folkstreams began as a project that aimed to increase access to films. Archival issues such as storage and hard drive space, record keeping, and file formats became increasingly important as the site developed. In addition to serving dynamic content and search capabilities, Folkstreams.net now incorporates administration tools that allow staff to add contextual information, images, and related links for each film. An RSS feed has been added to the main page, and visitors can visit the folkstreams blog or "digg" the site (submit it to digg.com).

Use of Youtube.com allows Folkstreams visitors to view previews of films, leave comments about the films, and subscribe to the Folkstreams channel on the Youtube.com site. This use of Youtube.com's functionality helps Folkstreams provide opportunities for viewer feedback without the added costs and difficulties of managing a hosted forum on its own servers. As of July 2007, the channel had approximately 130 subscribers and over 225,000 film views.

4.2 Programming

Folkstreams is a multifaceted project: it functions as a catalog or collection of films, a digital library and archives, and a distribution platform. Each film in the collection has multiple realizations: the original film components, digital preservation masters (Digibeta), and digital access files. In order to ensure universal access to the films on Folkstreams.net, provide for their long-term preservation, and to share high quality digital video online, multiple streams or video realizations of each film were a part of Folkstreams' mandate.

Unlike many popular video distribution and sharing websites that serve Flash video, a somewhat more "universal" format for viewing online video, Folkstreams.net provides streaming video content. Flash does have the advantage that it can play without trouble in the most popular browsers. Initially, however, the developers of the Folkstreams project hoped to avoid being locked into a single video streaming format. Proprietary video or video streaming formats were avoided as much as possible. The first streaming video server was a Real Server donated to UNC. As an alternative, short clips were offered for some films in QuickTime format. Currently, Real Video and MPEG4 video formats and streams are provided for online film viewing. Viewing films successfully is dependent on the user's having the correct media player properly configured.

The Folkstreams database is built with open source software, primarily MySQL and PHP. The use of open source technology for archival repositories is supported by the UNESCO publication, Towards an Open Source Archival Repository and Preservation System.[14] The UNESCO project provides recommendations for implementing open source digital archives. It argues that "the digital community as a whole recognise that any digital solution should be based in open standards and automated system because all digital solutions must address the issue of technical change." The use of open source solutions for cultural heritage programs is seen as an important tool for their long-term preservation and access.

Section 9.2 provides a list of the fields in the Folkstreams database. It should be emphasized that the database was developed for the Folkstreams project as a unique project with distinctive requirements; other film digitization projects may benefit from closer integration with recognized metadata standards.

Recent work on the Folkstreams site has focused on rationalizing the code, structure and organization of the site. In addition, the project would like to expand Folkstreams.net's capabilities to respond to an increasing audience, and to support a growing collection of films.

5.0 Metadata

Metadata, descriptive information about a given resource, serves the project purpose of increasing the accessibility and visibility of online resources. There are many metadata schemes, established and emerging, that are applicable for describing digital objects, including Dublin Core, MPEG7, PBCore, METS, and MODS. [15]

Popular video sharing sites do not generally have as their primary objective the long-term preservation of their videos. Only minimal metadata, such as description, creator, and running time, are typically associated with the videos. Folkstreams, however, maintains metadata for its films at the collection and item levels in an organic database not based specifically on any one metadata schema.

Film archives often hold several distinct realizations of a film work. For example, the Southern Folklife Collection holds both a 16mm print and a Digibeta tape of Tom Davenport’s A Singing Stream: A Black Family Chronicle. Metadata are generated for each of the physical forms of a film. Staff at the Southern Folklife Collection process and create descriptive metadata for the collection, including a finding aid and MARC bibliographic records that have been made available through the University of North Carolina at Chapel Hill library catalog, RLIN, and OCLC Worldcat.

Making records for the films accessible online enables discovery of Folkstreams.net content as well as of the original films, copies of which are preserved in the archives. Providing records in this way increases the likelihood that potential viewers will locate the films. Mapping of metadata from the Folkstreams.net film database into valid Dublin Core is necessary for the collection to be incorporated into shared repositories, such as the Moving Image Collection, and to be accessible via the Open Archives Initiative protocol.

Best practices in metadata vary according to the type of resource being described. Moving images are typically described in public access catalogs using MARC and the descriptive practices outlined in the Archiving Moving Images Manual (AMIM).[16] However, the development of metadata for sharing networked resources has led to several emerging standards that may be applicable to moving images. According to the NINCH guide, digital video metadata standards are still emerging:[17]

As of 2002, there is no shared good practice for what constitutes minimum metadata for digital audio and video in the cultural heritage sector. So, whether creating A/V metadata for management or for discovery, it is recommended to use or adapt a standards-based metadata standard like METS or SMIL, and to use the standard to capture as much information as the institution can afford. Note that some technical metadata can be automatically generated; for instance, some cameras create and can export metadata about capture date, format, exposure, etc. Metadata about content is more expensive to create, since it requires human intervention. At a minimum, useful retrieval will require a basic metadata set for each file, based on a standard like the Dublin Core.

The decision about which metadata standard to use is related in part to its potential use. In other words, it is important to consider the investment of resources in terms of database management and technical expertise, with relation to a clear understanding of who may benefit from richer description. The Folkstreams project has incorporated discovery tools, such as the website's advanced search function, that provide searching on a range of database fields deemed useful to visitors to the website. The fields include:

Release Date
Date Range
Color
Format
Subject
Country
Region
State
Language
Featuring
Filmmaker

The AMIM2 guide to describing moving image archives, and other descriptive and metadata standards, provide for highly detailed description of film materials.[18] Although closer integration with these standards is possible, Folkstreams provides unique fields that serve a dual function of describing moving images in general (release date, format), and providing a context for the unique materials (i.e. folklore films) in the collection.

Although metadata standards are evolving, institutions have recognized core functions for metadata with regard to digital libraries. The University of Maryland's descriptive metadata standard for its digital libraries encompasses four areas of description.[19] It notes that "The goal of successful metadata creation is to identify the main user group of a resource and define what types of information will help that group discover and make use of that resource." First, their guide recommends descriptive metadata, which "encompasses a range of information from basic elements such as title and subject to more advanced elements such as geographic or temporal coverage and relationships." Second, it calls for administrative metadata, which deals with rights management and access issues. Third, technical metadata includes "information about the digital object itself, such as file size, file type or format, bit depth, compression ratio, etc." Fourth, structural metadata is recommended which "describes the relationships between the multiple digital files that make up that single digital object" and, lastly, preservation metadata "details important information about the digital file, including any changes in the fie over time and management history."

Folkstreams maintains metadata in most of the above categories; however, structural and preservation metadata are limited. The archival institution (SFC) holding the Folkstreams film components maintains in-house information about preservation treatments. Other than recording digitization dates and information germane to the original transfer process, the Folkstreams database does not include specific information about how the digital files are preserved after that point.

The most rigorous metadata practices go substantially beyond the Folkstreams project's own approach, which placed an emphasis on creating a unique, localized database that would serve the needs of administrators and users of the Folkstreams collection, while remaining open to integration within online databases and catalogs. This is not without precedent; the AMIA Compendium, which provides data on a wide range of moving image archives, found that most had developed local descriptive practices and used them in conjunction with adopted standards.[20] Best practice would most likely aim to achieve a balance between local requirements and universal standards. The NINCH Guide indicates that best practices in metadata creation should include the following factors:[21]

* An appropriate metadata standard (such as EAD, Dublin Core, TEI Headers, METS)
* Name authority files (for person and place names)
* Subject thesauri and classifications
* Data type controls on fields (e.g. text, number)
* Independent double checking of entries

In order to share the Folkstreams collection, the data must be translated into standard formats. Selected fields within the Folkstreams.net film database are mapped to Dublin Core. Metadata about the Folkstreams films is first entered into the project database. From this, a set of records describing the films available on the Folkstreams site has been output as an Open Archives Initiative (OAI) Static Repository. The repository of item-level metadata has been made available through the IMLS Digital Collections and Content static repository gateway at http://imlsdcc.grainger.uiuc.edu/.

Jane Hunter and Liz Armstrong have explored the use of various metadata standards for video description, and note the imperfect fit of the Dublin Core schema:[22]

Dublin Core was designed specifically for generating metadata to facilitate the resource discovery of textual documents. Although a number of workshops have been held to discuss the applicability of Dublin Core to non-textual documents such as images, sound and moving images, they have primarily focused on extensions to the 15 core elements through the use of subelements and schemes specific to audiovisual data, to describe bibliographic-type information rather than the actual content.

The disadvantage of this approach is that the semantic refinement of Dublin Core through the use of qualifiers eventually leads to a loss of semantic interoperability.

The alternative is a "hybrid" approach in which RDF (or some other framework) is used to combine both simple unqualified Dublin Core and MPEG-7 descriptors within a single description container.

MPEG-7 is a rich, emerging metadata standard for multimedia description. Attempts were made to map the Folkstreams database's native fields to MPEG-7 for the purpose of sharing records with the Moving Image Collection (MIC), as the project was advised to do during the initial IMLS grant proposal phase. However, adequate documentation for the MPEG-7 standard could not be located in time to support the database development process. (Schema found on the Web months earlier, but not saved locally at the time, were no longer available online.) Further investigation indicated that MPEG-7, while supporting moving image collections, is not ideally suited for the kind of description the Folkstreams project intends to provide at this time. In addition, the Moving Image Collection, the Library of Congress-sponsored union catalog for moving images, is able to ingest the database records in many other formats.

Click to view the record from the IMLS Digital Collections and Content static repository gateway.

6.0 Preservation

6.1 General preservation issues

One of the primary objectives of the Folkstreams project is to preserve, to the extent possible, a complete, uncompressed copy of each film. Folkstreams incorporates both digital and analog preservation; many of the films selected for Folkstreams were created originally on 16mm film stock or on videotape. Working in partnership with a film and video processing company, Colorlab, Folkstreams obtains transfers or copies of these original materials for the archives. This practice differs somewhat from traditional archival practice in that archives are assumed to hold the original components of a film work. In the case of Folkstreams, there may be original components that remain in the possession of the filmmaker, depending on the arrangement worked out with each filmmaker. However, the archives does hold a complete, archival-quality print in formats suitable for long term preservation.

Each Folkstreams film, regardless of origin, is transferred to Digital Betacam (Digibeta), a proprietary, tape-based format produced by Sony. The Digibeta format provides studio-quality, minimal-compression storage. Other components (16mm transfers, mini-DV, 8mm), if they exist, may be archived as well. Copies of each film are stored in the archives of the University of North Carolina at Chapel Hill's Southern Folklife Collection (SFC). A complete discussion of environmental requirements for proper film storage is beyond the scope of this guide; however, in general, film decay can be minimized by storing film in low temperatures and minimal relative humidity. The Association of Moving Image Archivists notes several environmental factors that also speed decay, such as "air pollutants, including gases and particulate matter, the composition of storage containers, materials kept in the vault that may be harmful, acclimation, and time out of storage." [23] The SFC archives maintain constant temperature and humidity levels, which provides a more suitable environment for film and video storage. The environmental settings in the SFC are typically 65 degrees Fahrenheit and 45% relative humidity. Ideally, the archives would like to place the films in colder storage than it currently provides, perhaps by collaborating with another institution - for long-term preservation. (Note: UNC is in the currently process of hiring a full-time moving image archivist, which will further enhance the archives' capabilities in this area.)

According to the AHDS Moving Images and Sound Archiving Study, "moving images probably present the most complex preservation problem that most institutions will have to face over the next decade or so." [24] With regard to digital files, the guide recommends the motion JPEG 2000 format for digital preservation of moving image data, but emphasizes that "there is no single best way to preserve any digital resource and in the long run a mix of emulation and migration approaches may prove to be the most effective strategy." According to the Library of Congress, digital preservation is hindered by the format issue. The LC argues that "no digital format that is inextricably bound to a particular physical carrier is suitable as format for long-term preservation; nor is an implementation of a digital format that constrains use to a particular device or prevents the establishment of backup procedures and disaster recovery operations expected of a trusted repository."[25] Much of the current literature on digital preservation recommends use of open source or non-proprietary storage and delivery formats; for example, the UNESCO publication Towards an Open Source Archival Repository and Preservation System considers the importance of open source systems for cultural preservation efforts. It recommends the following core standards for a digital preservation system:[26]

* Create and store the content on a digital file in a format which does not apply any form of manipulation which causes data loss or loss of authenticity.

* Use a format which is widely implemented and supported, and preferably, though not necessarily, open or non-proprietary.

* Use a format that has a potentially long life (digitally speaking).

* Use a format that is most likely to have available migration pathways to the next format.

* Store enough metadata to be able to facilitate identification, access and preservation processes.

* Use a reliable storage format on at least two types of carrier.

* Make multiple copies, and check and verify them regularly.

* Plan to replace carriers and software as the market demands, and plan to migrate the content to the next type of reliable carrier.

Digibeta is generally considered an acceptable format for preservation masters, according to the The NINCH Guide to Good Practice in the Digital Representation and Management of Cultural Heritage Materials. However, it is not without drawbacks. A review of Digibeta as an archival format indicates that "The product life-cycle of DigiBeta equipment is under extreme marketing pressure -- being replaced by non-linear editing systems." [27]. Jim Lindner of Media Matters, LLC, also outlines several drawbacks of relying on the Digibeta format:[28]

Digital Betacam is a proprietary format which is supported by only one vendor, as such there is no competition - everything about Digital Betacam is extremely expensive and there is no competition nor alternatives should the vendor discontinue the format as it has done with many others.

It is a compressed format which means that a great deal of the original information is lost (if the original was not born digital). "Clones" of these tapes may not be a bit for bit copy.

The "data" on these tapes is difficult to support in a file oriented world as there is no file structure with these tapes.

The cost of these tapes is extremely high as compared to alternatives - in my latest calculations it cost 7 TIMES more to store an hour of video content on a Digital Betacam tape - as data - as opposed to storing it on a Data Tape - as Data.

The tapes are very large as compared to Data Tapes that hold substantially more content in substantially less space.

There is no straight forward way to manage the condition of the content when it is stored on shelves and not in a system where condition can be checked and maintained with essentially no human involvement.

The format is clearly approaching end of life - as the broadcast industry has moved to HD and Betacam SX and IMX and other highly compressed born digital formats have become popular for news gathering in the SDI - the market space for Digital Betacam keeps getting smaller, and the format is unlikely to be supported forever - nor is it likely that a replacement in SD for it will be introduced - which is a similar situation to what happened to BetacamSP.

The optimal format for storage and preservation of (digitized) audiovisual files may soon be motion JPEG 2000. According to Ian Gilmour, Media Consultant for Australia's National Film and Sound Archives, MPEG-4 is not designed "to support lossless encoding of entire picture sequences, with any significant degree of data reduction. JPEG 2000 is a technically better solution for this requirement." [29]

The NINCH Guide affirms that, rather than rely on untested digital formats, "most projects are keeping options open by storing preservation copies on tape until such time as server-based digital formats are proven for long-term preservation purposes." [30] Steve Weiss, UNC's Sound and Image Librarian, notes that the preservation process currently planned for the Folkstreams collection includes monitoring the materials and a long-term digitization plan. Monitoring the recordings in the early years is not considered as significant an issue as format obsolescence. Within the next five to ten years, all of the tapes in the collection will be migrated to lossless digital files and will be stored on the library's network attached server (NAS) system. From that point on, the digital files will be maintained and preserved. Video tapes will eventually be migrated to digital files both for preservation and access. For film, there are plans to migrate from the original film to fresh film for preservation purposes, and from film to digital (optical disc or electronic file) for access purposes.

6.2 Summary of the Folkstreams archiving process

For each film acquired by the project, a Digibeta tape is placed at SFC along with the film's original elements. If no additional elements exist, SFC receives only the Digibeta copy. Once at the SFC, materials are assigned call numbers. The Sound and Image Librarian notifies Folkstreams staff that the film has arrived. The film is then digitized at Ibiblio.org's facilities in order to be placed on the website, and the Digibeta is returned to the SFC. After digitization, a DVD access copy is provided to the SFC.

6.3 Finding aid and catalog records

When the Southern Folklife Collection acquires a substantial number of new tapes, the Sound and Image Librarian updates the collection's finding aid, which has been formatted in encoded archival description, or EAD. There is also a MARC record for each item in the Folkstreams collection, and a link within each finding aid to www.folkstreams.net.

6.4 Video Aids to Film Preservation

A core service of the Folkstreams project is a set of video clips produced in order to instruct film archivists in specific preservation techniques related to 16mm film. Each of the video clips demonstrates basic handling and restoration procedures for motion picture film. Clips are organized by title and by chapter headings in the Guide to Film Preservation published by the National Foundation for Film Preservation. This segment of the project is intended to share an important but endangered body of knowledge that is slowly disappearing as digital filmmaking replaces analog filmmaking. Visual instruction in preservation techniques also enhances transfer of film preservation knowledge; viewers are able to see technicians working with materials and discussing their techniques in real time.

7.0 Advertising and Outreach

Various outreach and public relations tools have been utilized by the project. Formal advertising, informal or "grassroots" methods, word of mouth, and positive reporting by media have comprised the bulk of the project's efforts to share news of the program.

Tom Davenport comments that the outreach process is largely informal. The project has set up a press page on the website (http://www.folkstreams.net/pages/press.html) that enables the media to learn more about the project and special events. Another facet of outreach includes researching and targeting small community groups or electronic forums that may have an interest in a specific film. For example, a film called The Sea Bright Skiff has the potential to appeal to group such as the Wooden Boat Builders' Association. Folkstreams contacted the group to announce the film's availability. In addition, the "guerilla marketing" (as Tom Davenport calls it) has been effective for a film about a tattoo artist entitled Stoney Knows How. Local tattoo parlors in the town in which the artist grew up were contacted; many of the artists had known Stoney, the main character, and the resulting conversations lead to a sharp increase in viewers of that film.

Recently the Oxford American, issue #57 (Summer 2007) included a DVD, entitled The Oxford American Southern DVD (2007). A Folkstreams film, Born for Hard Luck, appears in the collection. In return for granting permission for the film to be placed on the DVD, the magazine provided ad space for Folkstreams.

Folkstreams also maintains a blog at http://www.folkstreams.net/blog/. The blog highlights events, issues, and reflections on the project; in addition, there are links to other folklore websites, blogs, and online resources. The project has also tested a filmmaker blog on the folkstreams website. The goal of this blog is to encourage filmmakers to provide expanded contextual information for their films. Powered by WordPress, the blog allows filmmakers to post content about the film's production, characters, editing process, and anything else the filmmaker wants to share. Filmmaker Ali Colleen Neff has created an extensive site for her film at http://www.folkstreams.net/listenright/. Currently a test project, the Folkstreams project hopes eventually to expand this tool to include more filmmakers.

Recently Folkstreams has experimented with placing short previews of its films on YouTube.com. Viewers can subscribe to the Folkstreams channel, allowing them to be notified when new films are posted; viewers can also leave comments about the films. These tools provide an immediate mechanism for viewer feedback that isn't currently available on Folkstreams.net itself; the creators of the project felt that keeping the site free of comments and ratings would help foster a congenial, egalitarian environment for filmmakers to post their films.

One of the potential areas for enhanced collaboration between Folkstreams and filmmakers, via the website, is in contextual data for films. As filmmakers produce documentary films, they create and/or gather descriptive information about the footage they shoot, the scenes they edit, and the participants or consultants who appear in or otherwise contribute to the film. This expanded metadata, although it currently follows no specific standard, may at some point provide useful contextual information for viewers and researchers. At this time, it is uncertain as to how the information gathered in the logging and editing process might be later incorporated into a film's website on folkstreams.net; however, there may be potential for metadata collaboration that benefits researchers and viewers of the film, and provides useful contextual information about the filmmaker's perspective while making the film.

On Saturday, May 27th, 2007, National Public Radio presented a story about Folkstreams (Folkstreams Documents America, An Hour At a Time, http://www.npr.org/templates/story/story.php?storyId=10426054). This piece by Weekend Edition reporter Lynn Neary resulted in a substantial increase in the number of visits to Folkstreams.net. On the day the story aired, over 9,000 visits to the site were logged, up from 2-3,000 visits per day during the preceding weeks.

8.0 Evaluation

8.1 Evaluation criteria

The document Analyzing the Flow on Folkstreams.net: A Report for the Institute of Museum and Library Services, produced in January 2006, examines Folkstreams with regard to 1) preservation outcomes, 2) film accessibility outcomes, 3) multimedia film restoration website outcomes, 4) model project outcomes. These criteria are defined as follows:

Preservation outcomes: This refers to physical placement of film and video elements in the Southern Folklife Collection archives.

Film accessibility outcomes: The program a) tracks increases in the number of viewers of individual films, and b) compiles viewer feedback via an opt-in survey placed on the website. (See a list of survey questions.)

Multimedia film restoration website outcomes: The indicators of this recognition include the number of citations of the site in the archival scholarly and professional literatures, number of links to the site from archival resources sites, and feedback from archivists and others who use the site.

Folkstreams as a model project: The program is evaluated by the number and quality of citations, including links to the project by other special-interest film digitization and preservation websites, and inquiries on practice and protocol from those involved with other similar projects.

A significant body of literature exists on evaluative protocols for digital libraries. Multiple approaches to evaluating the impact and actual use of a digital collection are possible, including usability studies, log analysis, and formal data collection via questionnaires or observation sessions. The eVALUEd project, based in the UK, has published an evaluation toolkit for e-libraries, and suggests that evaluation programs serve the following purposes: 1) strategic planning, 2) managing the library, 3) exploring how services are being used, 4) improving those services, and 5) justifying the continuation of the library to potential funders.[31] Folkstreams could possibly develop a formal plan for testing website components, such as search functions and interface, coupled with a structured plan for regularly reviewing transaction logs and use patterns.[32] The document Evaluating Digital Libraries outlines several suggested core areas for evaluation: service, usability, information retrieval, bibliometrics, and transaction log analysis.

Accessibility and compatibility considerations (in terms of the myriad of browsers available) are also important, and the World Wide Web Consortium has developed specific accessibility guidelines. The digital libraries project eVALUEd recommends focusing evaluations on the following areas: access, user support, promotion, perceptions, and user satisfaction. For the purposes of a small-scale digital library project such as Folkstreams, it seems most useful to focus on evaluation of services, website usability, and actual use patterns (log analysis). Usability of the website interface is an area that offers much potential for formal evaluation. The website currently utilizes a relatively simple user interface with a database-driven, highly structured search function. It's currently possible to search by several methods (filmmaker, subject, title, etc). A more systematic evaluation of the Folkstreams website in relation to established usability standards would likely be an useful and informative process.

8.2 User feedback

The International Organization for Standardization (ISO) defines usability as "the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction in a specified context of use." [33] Although no formal usability studies have been undertaken, Folkstreams placed a survey on the website in order to elicit general feedback from users. In addition, viewers can leave comments on the Folkstreams channel of Youtube.com.

User feedback, based on appraisal of both survey results and informal correspondence, has largely been positive. Project director Tom Davenport comments on user responses:

I receive commentary and general questions from our contact form via email. Many of these are inquiries about DVD purchases. Some are comments from relatives of people in the movies. Some are looking for films or information about films not on the site. . The responses are usually full of praise.

We decided that it would be better to let users post comments on the films on Youtube.com/folkstreamer than to post directly on Folkstreams.net. We were concerned that someone might feel overly criticized or left out. We did not want to dissuade participation from the filmmakers; we are dependent on their voluntary cooperation. On YouTube.com, public comments are a part of that site and culture.

The user feedback survey on folkstreams.net was not designed according to any formal standards for questionnaire development or survey methodology. Well-designed user studies and interface testing, whether in the form of focus groups, interviews, or surveys, would provide useful information to developers, both before launching the website and after its inception, in order to better determine website usability and effectiveness.

8.3 Website statistics

Folkstreams relies on two user tracking systems. The first system attempts to track individual views of films. It is not possible to determine whether a human or a robot visited the page and requested a video; to determine the total number of individual page visits or video requests, an application called Urchin is provided by the University of North Carolina.

From the website server end, the number of times someone "launches" a video file can be recorded in a database. When a visitor clicks on a streaming link to launch a film, a program is activated that stores a record of that action in the database. Aggregated statistics can be plotted to provide an indication of overall site traffic. There is an effort made to weed out activations by web robots, but since it is not always possible to determine whether the request is from a human or a robot, there is some uncertainty in the overall statistics.

Improvements to the statistics system are necessary to keep up with increasing traffic; for example, the current system does not delete any of the click records, and summaries are created live. This will at some point fail to be scalable. Ideally, the project will incorporate a system that periodically clears out all click records and summarizes them. This will keep the table size small; in addition, the project does not require detailed data about each user's path within the website. Changes are also recommended that will help ensure requests by robots are not counted.

There is more functionality available in user tracking applications than is currently incorporated into the Folkstreams website. A thorough study of folkstreams.net website users could utilize some of the tools outlined in Knowing the user's every move: user activity tracking for website usability evaluation and implicit interaction.[34] For example, user interface studies can track the path of users through a website, enabling administrators to analyze patterns of use, potential trouble spots or confusing navigation, and general usability. Indicators such as how long users remain on the site, which pages they visit and in what order, and how they are faring with the site's search tools, would provide useful feedback in conjunction with a formal usability study. Several papers discuss the potential to track mouse movements across a web page.[35,36]

According to Evaluating Digital Libraries: A User Friendly Guide, transaction log analysis can be used to evaluate overall traffic to a website:[3]

Transaction log analysis is a way to track unobtrusively how users are using a digital library. As an evaluator, you may wish to analyze transaction log information as part of an overall evaluation aimed at obtaining a deeper understanding of how users are navigating through your digital library, which resources they access, and any search problems they encounter. Log analysis alone usually requires too much inference, but it provides important information that may be explained with the additional data obtained from interviews, surveys, and observations.

Transaction analysis can help give administrators an indication of factors such as frequency of feature use, system response times, user actions to recover from errors, session lengths, and location of users, among others.

At this stage in the project, a cost-benefit analysis of such in depth user evaluation should be undertaken before investing in complex and/or costly applications; however, a more formal awareness of Folkstreams users and their behavior may be helpful in terms of planning site enhancements and in evaluating the success of the program. A usability evaluation would be particularly helpful in light of the diversity of users, from students to senior citizens, and with an eye to expanding current and potential audiences.

9.0 Appendices and links

9.1 Folkstreams films currently streaming on website

9.2 List of database fields

Database Field

Filmmaker Questionnaire

filminfo Film ID

Title

Alternative Title

Series

Producer (s)

Photographer(s)

Editor(s)

Sound person(s)

Date created/released

Copyright holder / statement

Permission to stream whole film

Source Title

Source Identification #

Type of identification

Duration

Color ID

Original media ID

Principal Character ID

Key Words

Short Description

Long Description

Review blurb

Review source

Acknowledgements

Funding

Awards

Country ID

Region ID

State/Province ID

Language

DistributerID

StockFootageID

Master Loc

Outtakes Loc

Submit date

Submitter

Submitteremail

Filmmaker FilmmakerID

FilmmakerFirstName

FilmmakerLastNAme

Filmmaker Lookup Table FilmID

Film_filmmaker FilmmakerID

Distributor DistributorID

Distributor name

Distributor address

Distributor email

Distrubutor URL

Distributor telephone

Copyright holder

Stockfootage Stock ID

Order stock footage contact

Order stock footage address

Order stock footage URL

Order stock footage e-mail

Order stock footage telephone

Principal Character Principal Character ID

Principal Character First Name

Principal Character Last Name

Princ. Character Lookup FilmId

Principal Character ID

Film Subject Lookup FilmID

SubjectID

Administration

admin Film ID

Verified (y/n)

Verifier

Posted to Web

Date last checked

Digitization Questionnaire

streams StreamID

FilmID

Full Film?

Stream Title

Stream Description

Stream inpoint

Stream outpoint

Streaming Type

Digitization (Streaming) Date

Digitization Technician

Digitized Filename

Digitized Location

Source Format

Source Location

Archive Format

Archive Location

Contextual Materials Q

context Contextual Material ID (autogen)

Film ID

Subject Category ID

Weight

Title

Creator(s)

Source

Description

Keywords

Date

Copyright Holder

Acknowledgements

Text

Images Image ID

Cover

Filmid

Contextual material ID

weight

Title

Creator

Source

Caption/Description

Copyright Holder

Filename

File Location

Review Reviewer

Reviewer Location

ReviewerEmail

Film ID

Review

9.3 Website feedback survey

9.4 Contributors and staff

10.0 Works cited

1 NISO Framework Advisory Group. (2004). A framework of guidance for building good digital collections. 2nd edition (2nd edition ed.). Bethesda, MD: National Information Standards Organization. Available at http://www.niso.org/framework/framework2.html

2 Hodge, G. M. Best practices for digital archiving: An information life cycle approach. D-Lib Magazine, 6(1). Available at http://www.dlib.org/dlib/january00/01hodge.html

3 Carignan, Y., Evander, J., Gueguen, G., Hanlon, A., Murray, K., Roper, J., et al. (2007). Best practice guidelines for digital collections at university of Maryland libraries. Available at http://www.lib.umd.edu/dcr/publications/best_practice.pdf

4 Humanities Advanced Technology and Information Institute (HATII), University of Glasgow, and the National Initiative for a Networked Cultural Heritage (NINCH). (2002). The NINCH guide to good practice in the digital representation and management of cultural heritage materials (Version 1.0 of the First Edition ed.)National Initiative for a Networked Cultural Heritage. Available at http://www.nyu.edu/its/humanities/ninchguide/

5 Carignan, Y., Evander, J., Gueguen, G., Hanlon, A., Murray, K., Roper, J., et al. (2007). Best practice guidelines for digital collections at university of Maryland libraries. Available at http://www.lib.umd.edu/dcr/publications/best_practice.pdf

6 Humanities Advanced Technology and Information Institute (HATII), University of Glasgow, and the National Initiative for a Networked Cultural Heritage (NINCH). (2002). The NINCH guide to good practice in the digital representation and management of cultural heritage materials (Version 1.0 of the First Edition ed.)National Initiative for a Networked Cultural Heritage. Available at http://www.nyu.edu/its/humanities/ninchguide/

7 Media Matters, LLC (2004). Digital video preservation reformatting project: A report. Media Matters, LLC for the Dance Heritage Coalition. Available at http://www.danceheritage.org/preservation/DigitalVideoPreservation1.pdf

8 Media Matters, LLC (2004). Digital video preservation reformatting project: A report. Media Matters, LLC for the Dance Heritage Coalition. Available at http://www.danceheritage.org/preservation/DigitalVideoPreservation1.pdf

9 Library of Congress. (2007) Preferences in summary for moving image content. Available at www.digitalpreservation.gov/formats/content/video_preferences.shtml

10 Humanities Advanced Technology and Information Institute (HATII), University of Glasgow, and the National Initiative for a Networked Cultural Heritage (NINCH). (2002). The NINCH guide to good practice in the digital representation and management of cultural heritage materials (Version 1.0 of the First Edition ed.)National Initiative for a Networked Cultural Heritage. Available at http://www.nyu.edu/its/humanities/ninchguide/

11 Humanities Advanced Technology and Information Institute (HATII), University of Glasgow, and the National Initiative for a Networked Cultural Heritage (NINCH). (2002). The NINCH guide to good practice in the digital representation and management of cultural heritage materials (Version 1.0 of the First Edition ed.)National Initiative for a Networked Cultural Heritage. Available at http://www.nyu.edu/its/humanities/ninchguide/

12 Humanities Advanced Technology and Information Institute (HATII), University of Glasgow, and the National Initiative for a Networked Cultural Heritage (NINCH). (2002). The NINCH guide to good practice in the digital representation and management of cultural heritage materials (Version 1.0 of the First Edition ed.)National Initiative for a Networked Cultural Heritage. Available at http://www.nyu.edu/its/humanities/ninchguide/

13 Bouthillier, L. Streaming vs. downloading video: Understanding the differences. Available at http://www.streamingmedia.com/article.asp?id=8456&page=2&c=11

14 Bradley, K., Lei, J., & Blackall, C. (2007). Towards an open source archival repository and preservation system: Recommendations on the implementation of an open source digital archival and preservation system and on related software development. NY: United Nations Educational, Scientific, and Cultural Organization (UNESCO). Available at http://portal.unesco.org/ci/en/files/24700/11824297751towards_open_source_repository.doc/

15Moving Image Collections (MIC). Cataloging and metadata portal. Available at http://mic.imtc.gatech.edu/catalogers_portal/cat_relatedmetadata.htm

16 Archival moving image materials: A cataloging manual. (2000). (2nd Edition ed.). Washington, DC: Library of Congress.

17 Humanities Advanced Technology and Information Institute (HATII), University of Glasgow, and the National Initiative for a Networked Cultural Heritage (NINCH). (2002). The NINCH guide to good practice in the digital representation and management of cultural heritage materials (Version 1.0 of the First Edition ed.)National Initiative for a Networked Cultural Heritage. Available at http://www.nyu.edu/its/humanities/ninchguide/

18 Archival moving image materials: A cataloging manual. (2000). (2nd Edition ed.). Washington, DC: Library of Congress.

19 Carignan, Y., Evander, J., Gueguen, G., Hanlon, A., Murray, K., Roper, J., et al. (2007). Best practice guidelines for digital collections at university of Maryland libraries. Available at http://www.lib.umd.edu/dcr/publications/best_practice.pdf

20 Martin, A. L., & Association of Moving Image Archivists. AMIA Compendium of Cataloging Practice. (2001). Beverly Hills, CA; Chicago, IL: Association of Moving Image Archivists; Society of American Archivists.

21 Humanities Advanced Technology and Information Institute (HATII), University of Glasgow, and the National Initiative for a Networked Cultural Heritage (NINCH). (2002). The NINCH guide to good practice in the digital representation and management of cultural heritage materials (Version 1.0 of the First Edition ed.)National Initiative for a Networked Cultural Heritage. Available at http://www.nyu.edu/its/humanities/ninchguide/

22 Hunter, Jane, and Armstrong, Liz. (1999). A Comparison of Schemas for Video Metadata Representation. Proceedings of the The Eighth International World Wide Web Conference, Toronto, Canada. Available at http://www8.org/w8-papers/3c-hypermedia-video/comparison/comparison.html.

23AMIA website, http://www.amianet.org/publication/resources/guidelines/storage/intro.html

24 AHDS, Arts and Humanities Data Service. (2006). Digital moving images and sound archiving study. Available at http://www.jisc.ac.uk/whatwedo/programmes/programme_preservation/project_movingimagesound.aspx

25 Library of Congress. (2007). Sustainability of digital formats. Available at http://www.digitalpreservation.gov/formats/sustain/sustain.shtml

26 Bradley, K., Lei, J., & Blackall, C. (2007). Towards an open source archival repository and preservation system: Recommendations on the implementation of an open source digital archival and preservation system and on related software development. NY: United Nations Educational, Scientific, and Cultural Organization (UNESCO). Available at http://portal.unesco.org/ci/en/files/24700/11824297751towards_open_source_repository.doc/

27 Mayfield, B. (2006). 8mm, super 8mm, and 16mm film to digital video tutorial. Available at http://www.film-to-video.com/tutorial_page04.html

28 Lindner, Jim. (2007, June 8) Re: Stretching DigiBeta tapes? Message posted to the Association of Moving Image Archivists AMIA-L electronic mailing list, archived at http://palimpsest.stanford.edu/byform/mailing-lists/amia-l/

29 Media Matters, L. (2007). Research report on JPEG 2000 for video archiving. Available at http://www.media-matters.net/docs/WhitePapers/IansWhitePaper.pdf

30 Humanities Advanced Technology and Information Institute (HATII), University of Glasgow, and the National Initiative for a Networked Cultural Heritage (NINCH). (2002). The NINCH guide to good practice in the digital representation and management of cultural heritage materials (Version 1.0 of the First Edition ed.)National Initiative for a Networked Cultural Heritage. Available at http://www.nyu.edu/its/humanities/ninchguide/

31 eVALUEd. (2006). An evaluation toolkit for e-library developments. Available at http://www.evalued.uce.ac.uk/tutorial/1.htmReeves, T., Apedoe, X., & Hee Woo, Y. (2003). Evaluating digital libraries: A user-friendly guide. Digital Library of Information Science and Technology. Available at http://dlist.sir.arizona.edu/398/01/DLUserGuideOct20.doc

32 Reeves, T., Apedoe, X., & Hee Woo, Y. (2003). Evaluating digital libraries: A user-friendly guide. Digital Library of Information Science and Technology. Available at http://dlist.sir.arizona.edu/398/01/DLUserGuideOct20.doc

33 International Organization for Standardization. (1998). ISO 9241-11:1998. http://www.iso.org/

34 Atterer, R., Wnuk, M., & Schmidt, A. (2006). Knowing the user's every move: User activity tracking for website usability evaluation and implicit interaction . Paper presented at the Proceedings of the 15th International Conference on World Wide Web, Edinburgh, Scotland. 203-212. Available at http://doi.acm.org/10.1145/1135777.1135811

35 Arroyo, E., Selker, T., & Wei, W. (2006). Usability tool for analysis of web designs using mouse tracks. Montreal, Quebec, Canada: ACM Press. Available at http://doi.acm.org/10.1145/1125451.1125557

36 Mueller, F., & Lockerd, A. (2001). Cheese: Tracking mouse movement activity on websites, a tool for user modeling. Seattle, Washington: ACM Press. Available at http://floydmueller.com/achievements/cheese.pdf

37 Reeves, T., Apedoe, X., & Hee Woo, Y. (2003). Evaluating digital libraries: A user-friendly guide. Digital Library of Information Science and Technology. Available at http://dlist.sir.arizona.edu/398/01/DLUserGuideOct20.doc

 

10.1 Full bibliography

Archival moving image materials: A cataloging manual. (2000). (2nd Edition ed.). Washington, DC: Library of Congress.

AHDS, Arts and Humanities Data Service. (2006). Digital moving images and sound archiving study. Available at http://www.jisc.ac.uk/whatwedo/programmes/programme_preservation/project_movingimagesound.aspx

Arroyo, E., Selker, T., & Wei, W. (2006). Usability tool for analysis of web designs using mouse tracks. Montreal, Quebec, Canada: ACM Press. Available at http://doi.acm.org/10.1145/1125451.1125557

Atterer, R., Wnuk, M., & Schmidt, A. (2006). Knowing the user's every move: User activity tracking for website usability evaluation and implicit interaction . Paper presented at the Proceedings of the 15th International Conference on World Wide Web, Edinburgh, Scotland. 203-212. Available at http://doi.acm.org/10.1145/1135777.1135811

Bouthillier, L. Streaming vs. downloading video: Understanding the differences. Available at http://www.streamingmedia.com/article.asp?id=8456&page=2&c=11

Bradley, K., Lei, J., & Blackall, C. (2007). Towards an open source archival repository and preservation system: Recommendations on the implementation of an open source digital archival and preservation system and on related software development. NY: United Nations Educational, Scientific, and Cultural Organization (UNESCO). Available at http://portal.unesco.org/ci/en/files/24700/11824297751towards_open_source_repository.doc/

Carignan, Y., Evander, J., Gueguen, G., Hanlon, A., Murray, K., Roper, J., et al. (2007). Best practice guidelines for digital collections at university of Maryland libraries. Available at http://www.lib.umd.edu/dcr/publications/best_practice.pdf

E-MELD. (2006). Digitization of video files. Available at http://emeld.org/school/classroom/video/index.html

eVALUEd. (2006). An evaluation toolkit for e-library developments. Available at http://www.evalued.uce.ac.uk/tutorial/1.htm

Frey, F. (2000). Guides to quality in visual resource imaging 5: File formats for digital masters. Available at http://www.rlg.org/legacy/visguides/visguide5.html

Gill, T., Gilliland, A. J. & Woodley, M. S. (2005). Introduction to metadata: Pathways to digital information (version 2.1). Available at http://www.getty.edu/research/conducting_research/standards/intrometadata/

Hodge, G. M. Best practices for digital archiving: An information life cycle approach. D-Lib Magazine, 6(1). Available at http://www.dlib.org/dlib/january00/01hodge.html

Humanities Advanced Technology and Information Institute (HATII), University of Glasgow, and the National Initiative for a Networked Cultural Heritage (NINCH). (2002). The NINCH guide to good practice in the digital representation and management of cultural heritage materials (Version 1.0 of the First Edition ed.)National Initiative for a Networked Cultural Heritage. Available at http://www.nyu.edu/its/humanities/ninchguide/

Hunter, Jane, and Armstrong, Liz. (1999). A Comparison of Schemas for Video Metadata Representation. Proceedings of the The Eighth International World Wide Web Conference, Toronto, Candada. Available at http://www8.org/w8-papers/3c-hypermedia-video/comparison/comparison.html.

International Organization for Standardization. (1998). ISO 9241-11:1998. http://www.iso.org/

Library of Congress. (2007). Preferences in summary for moving image content. Available at www.digitalpreservation.gov/formats/content/video_preferences.shtml

Library of Congress. (2007). Sustainability of digital formats. Available at http://www.digitalpreservation.gov/formats/sustain/sustain.shtml

Library of Congress, & National Digital Information Infrastructure and Preservation Program. (2002). Digital video archives: Managing through metadata. Washington, DC: Council on Library and Information Resources and the Library of Congress. Available at http://www.clir.org/PUBS/reports/pub106/video.html.

Lindner, Jim. (2007, June 8) Re: Stretching DigiBeta tapes? Message posted to the Association of Moving Image Archivists AMIA-L electronic mailing list, archived at http://palimpsest.stanford.edu/byform/mailing-lists/amia-l/

Martin, A. L., & Association of Moving Image Archivists. AMIA Compendium of Cataloging Practice. (2001). Beverly Hills, CA; Chicago, IL: Association of Moving Image Archivists; Society of American Archivists.

Mayfield, B. (2006). 8mm, super 8mm, and 16mm film to digital video tutorial. Available at http://www.film-to-video.com/tutorial_page04.html

Media Matters, LLC (2004). Digital video preservation reformatting project: A report. Media Matters, LLC for the Dance Heritage Coalition. Available at http://www.danceheritage.org/preservation/DigitalVideoPreservation1.pdf

Media Matters, L. (2007). Research report on JPEG 2000 for video archiving. Available at http://www.media-matters.net/docs/WhitePapers/IansWhitePaper.pdf

Moving Image Collections (MIC). Cataloging and metadata portal. Available at http://mic.imtc.gatech.edu/catalogers_portal/cat_relatedmetadata.htm

Mueller, F., & Lockerd, A. (2001). Cheese: Tracking mouse movement activity on websites, a tool for user modeling. Seattle, Washington: ACM Press. Available at http://floydmueller.com/achievements/cheese.pdf

NISO Framework Advisory Group. (2004). A framework of guidance for building good digital collections. 2nd edition (2nd edition ed.). Bethesda, MD: National Information Standards Organization. Available at http://www.niso.org/framework/framework2.html

PADI (Preserving Access to Digital Information). (2007). Digital preservation Strategies. Available at http://www.nla.gov.au/padi/topics/18.html

Reeves, T., Apedoe, X., & Hee Woo, Y. (2003). Evaluating digital libraries: A user-friendly guide. Digital Library of Information Science and Technology. Available at http://dlist.sir.arizona.edu/398/01/DLUserGuideOct20.doc

Technical Advisory Service for Images. (2007). Digital preservation and Storage. Available at http://www.tasi.ac.uk/advice/delivering/digital.html

Back to Homepage