Motion Picture Conservation at the Library of Congress

by David Francis, Chief National Audio-Visual Conservation Center

From 1893, when Thomas Edison first used 35mm perforated film in his Kinetoscopes, until 1951, when acetate, or safety, film replaced it, the film base chosen by the industry was made of cellulose nitrate, an extremely flammable substance that, like all polymers, undergoes continuous chemical decomposition.

This problem more than any other has dominated film archivists' thinking since the first archives were established in the early 1930s. Most works of art spend the first part of their lives in private hands because the assessment of artistic importance often requires hindsight and national museums and art galleries are traditionally conservative in their acquisitions policies. Except for some modern sculpture, which also suffers from polymer degradation, most works of art normally survive the experience.

Film, however, with its beginnings in fairgrounds and the vaudeville theater, developed its own form and language during the 1920s and 1930s and is only now being recognized as an art form and an important historical record. To a large extent, films did not survive their time in private or, more accurately, commercial hands. Only about 10 percent of the movies produced in the United States before 1929 still exist. Hence, the film archivist is continually under pressure just to ensure the future of films content. As for the artifacts themselves, there is no way the archivist can prolong their lives.

In the past the Library's policy has been to make exact copies of all the nitrate films in its collection, leaving it to others in the future to determine whether the duplicate would win the approval of the original production team. However, Congress's establishment in 1988 of the National Film Registry and the deliberations of the National Film Preservation Board have started to change the public's attitude toward the country's film heritage. This is providing a more favorable climate for "film restoration," that is, the reconstruction of a film so that it represents as nearly as possible the intentions of its maker.

Nevertheless, the federal government, like governments the world over, has realized that it cannot singlehandedly pay for the preservation of all aspects of the national heritage, which by definition grows with the passing of time. The search for partners in this preservation enterprise has been directed toward businesses, charities and wealthy private individuals. These groups, however, are now besieged by requests for funds, and more and more require an identifiable commercial benefit or need to see a tangible result from their generosity, such as media attention, before allocating resources.

In the past, the preservation of an important film has not given donors the same satisfaction as saving a major painting for the nation. How then can film preservation be made more appealing?

It would undoubtedly help if archivists looked at their own image. The film archive must be moved from the world of commerce to the realm of the fine arts. The terms "vaults" and "laboratory," "printing" and "processing" are associated with an industry. A name like our own Motion Picture Conservation Center, staffed by conservation specialists and film historians, seems appropriate for an organization responsible for safeguarding part of the national heritage.

There must also be a shift in emphasis from preservation-- the action of duplication or making the best copy from existing material--to restoration--the determination of the original form prior to copying and the reconstruction of the work from a multitude of sources. In short, the scholar and the technician must work more closely together.

In 1991, the Library of Congress Motion Picture Conservation Center in Dayton, Ohio worked on Frank Capra's 1939 classic "Mr. Smith Goes to Washington," one of the 25 titles on 1989's National Film Registry List. Surprisingly, there were only two nitrate elements still available for such an important film: the original negative, which was full of "replacement footage," and a heavily used release print made from the European negative that the Library borrowed from Britain's National Film Archive. In addition, we had at our disposal two safety fine-grain(1) film elements from the producers, Columbia Pictures, and a rather flat safety negative from which we recently produced the print shown by David Packard, an important supporter of LC's film preservation efforts, in Palo Alto in his tribute to Frank Capra.

The nitrate negative presented an unexpected problem. A large part of the sequence at the Lincoln Memorial had been blown up from 16mm to 35mm. This was likely inserted in the negative in the mid-1950s. As a result, all prints made from the negative during the past 35 years have significant image degradation in this section of the film.

Our first task was to make certain that we knew which version of the film we were trying to restore. Capra had a reputation for changing his films, even after previews and sometimes even after the film had been in limited release. This meant that the original script and even the published script were not necessarily reliable source material. We therefore had to turn to contemporary reviews, articles about the making of the film and biographies of the members of the creative team. In the case of "Mr. Smith Goes to Washington," there appeared to have been no changes after release. The National Film Archive print therefore became a valuable reference source.

Hollywood tended to dispatch the European duplicate negative as soon as the film was completed so that it could be released in major cities while the positive (it was hoped) critical reaction received in the United States was still firmly in the public's mind. Since the National Film Archive print was made from the European duplicate negative for London release, it was probably an accurate record of the film's form when it left Columbia's Hollywood Laboratory.

The next step was the preparation of a restoration script. It had to indicate accurately all of the sections in the original negative that needed to be replaced and from whence the replacements had to come. In the end the script prepared by Ken Weissman, Dayton Laboratory supervisor, totaled nearly 30 pages.

The restoration process is costly and time-consuming. From the original negative we have to make two fine grains. One will be held as security in case the first fine grain is damaged during reconstruction and we are unable to print another fine grain from the nitrate negative. The other will form the basis of the restored master. At the same time the nitrate optical sound track is rerecorded on magnetic tape and a safety optical negative produced. This operation now has to be done by commercial laboratories; I hope that someday the Library will exploit all the expertise it has acquired in the field of sound restoration and tackle this within the National Audio-Visual Conservation Center.

We then had to find the appropriate material in other copies to replace damaged sections and missing or incomplete shots in the original negative. The resulting sections then had to be converted into the fine grain format so they could be cut into the restoration master. The sound track and picture elements were then be matched, producing a fine grain with numerous splices and many untimed inserts(2). This copy is the restoration master because it is the closest to the original.

However, we still have to produce a "performance copy," as well as satisfy the copyright owner's need for access to the restored version. We can either produce a one-light negative or a partially timed one(3) if the differences between the original material and the insert sequences are too prominent. If the copyright owner also wants material, then we would make two such negatives at this stage, one in the form of a composite for the copyright owner, the other with separate sound track for the Library. From the latter we will make a timed answer print. This will then be screened in a normal theatrical environment, because the purpose of the performance print is to help re-create the experience of cinema and therefore, it should look its best on the screen of, say, a 500-seat theater. Further timing adjustments will be made and it is hoped the second answer print(4) will meet archival "show print" standards and be available for screening.

The above is a straightforward account of what we have to do in restoring a film but does not take into account the difficult ethical problems. For instance, it is now possible to remove all the scratch and hiss from variable area and variable density sound tracks using electronic enhancement techniques. Such changes will undoubtedly make the track more acceptable to today's ears. I believe the restoration master optical sound track should only be equalized to reflect material from different sources (if sections have been edited into the sound track from elsewhere) and not electronically enhanced. On the other hand, there is no reason why a separate rerecording should not be made using all available sound track enhancement techniques for public reissue use.

In the same way, except for archival audiences, it may be necessary to trim slightly incomplete shots or even cut out, say, two frames of damaged image or insert material in a show copy so that these technical imperfections do not interfere with modern audiences' enjoyment of the film. The cinema is basically an entertainment medium.

In short, we have not lost the opportunity to present the film almost as it was seen originally, nor do we forget that today's audience has a different perspective from that of the moviegoers who saw "Mr. Smith Goes to Washington" in 1939.

Of course, the Library's preservation specialists and their counter-parts elsewhere in America cannot restore all 100,000 films in their collections to the standard described above. The basic process of making the best copies of nitrate originals, that is, preservation, will still be essential. Then at least archivists will be able to undertake full restoration in the future. Indeed, this will be easier when archives the world over have copied all their nitrate holdings, accurate and comprehensive copy data has been captured and made accessible, and international cooperation in the field has become the rule rather than the exception.

There will of course be those who feel this is too great an undertaking for the federal government to finance, even with the help of the film industry and the sponsor. These people suggest that we must therefore copy some material onto an inferior format--16mm film, videotape, etc.--or indeed copy only the most worthy among the 100,000 films that have fortuitously survived. I oppose both suggestions. We have recently celebrated the 100th anniversary of film and soon will celebrate the end of the first century in mankind's history whose development has been chronicled by the motion picture camera. By 2000 I feel certain that the importance of the nation's film heritage will finally be given the recognition it deserves. If we could turn back the clock and save all the great manuscripts and works of art that were destroyed or allowed to deteriorate, I am sure we would jump at the opportunity. Today, we are in a position to save all extant motion picture film from the nitrate era and avert the reproaches of our descendants for a failure to allocate what in today's terms are minimal resources to the safeguarding of our heritage.

To those who would promote the idea of holding part of this heritage on an inferior format, I would simply remind them that each new technological advance in the fields of image recording and dissemination requires a master of the highest quality. Never will this need be more apparent than in the revolution that is just around the corner, namely digital television. The new system will be able to capture and disseminate more of the information contained in the motion picture frame than ever before. The Japanese market makers need to be able to demonstrate the advantages of higher quality images. They will undoubtedly seek out film archives with inducements. We must be ready to meet the challenge.

Not only are we undergoing a quality revolution but also a quantity revolution. Today every moving image, whether fact or fiction, is potentially valuable. The archives are the great untapped storehouses. But we must not sell ourselves short. We must cooperate with the copyright owner, the sponsor and the disseminator and use government funds for the "everyday images" that do not have a benefactor. We must also shout about our achievements, make certain that the archive's name is above the credits and have a say in the way in which the products of our labors are displayed, presented and performed.

1. Safety Fine Grain: A type of positive copy of a film specifically made for duplication purposes.
2. Untimed Insert: Inserts into a film are "timed" to make them match the material into which they are inserted; an untimed insert does not match.
3. One-light Negative is made using an average overall exposure. Timing changes are then made when the positive prints are produced from it. This helps achieve integration between original material and insert sequences.
4. Answer Print: The first print from a newly made negative. Such prints are used to help achieve the highest "show print" standards.