[table of contents | journal index]

Film / Digital / Film

Michael Friend

Introduction

With close to 90% of the silent cinema, and 50% of all the film produced before 1950 now irretrievably lost, and virtually all of the images captured on film susceptible to similar loss within the next century or two, there is a sense of involuntary irony that attaches to the term "preservation." What we call "preservation" in the film archives is, in comparison to many other areas of museum science, a misnomer. There is no such thing as preservation. The motion picture is one of the most ephemeral objects yet created in the cultural sphere, with a life span shorter than almost any previous medium. There is in fact no adequate system for the long-term retention of the motion picture. The base and emulsion of the motion picture are subject to deterioration, and therefore necessitate duplication in order to extend the life of a film, yet the method of reproduction is subject to physical laws which guarantee loss of some part of the image with each generation of printing, and this loss is cumulative and irrevocable. And whether we regard film as elements of a cultural heritage or as business assets, we are nevertheless starkly confronted with a situation that involves the administration of cultural and economic resources in a complex relationship to a technology which does not produce stable, permanent products.

From the beginning of film production in the last century, technicians have been concerned with improving the processes of film production and editing, processing, duplication and projection. In the classical era of the cinema, the general practice was to make prints off the original negative, subjecting the best element to the harshest treatment in handling, exposure to contaminants and mechanical stress. Modern practice tends to reserve the original negative for the production of secondary facilities used to create prints or videotransfers. One of the most important innovations has been the introduction of the printing negative. In every area of preservation, more attention is being paid to the conservation of the original elements of a subject as the key to producing superior images. Rising expectation concerning image quality has been a forty year trend brought about by technical and cultural factors. Perhaps the first step in the process was the advocacy of the cinema as an art form, which led to a more profound consideration of its innate qualities. The discovery of value in old film libraries through television rebroadcast of the films of Hollywood's "Golden Age" helped to put an end to the malign neglect of so called "media assets." Up until relatively recently, archives and studios would frequently destroy serviceable nitrate original negatives after duplication. Although this practice has largely been stopped and the process of depositing nitrate materials with archives is now a more or less standard practice, many important original negatives have been needlessly eliminated. Subsequent preservation work has shown this policy to have been unfortunately short-sighted, as we learn that in this world in search of the perfect image there is just no substitute for the original negative, and preservation technology continually improves, making it possible to improve the quality of materials derived from an original negative. (A similar development occurred in the area of newsfilm, around 1975, as many television stations transferred parts of their newsfilm to videotape and then deaccessioned the original film). In the 1960s and 1970s, some of the studio libraries and archives copied 35mm nitrate titles onto 16mm film for reasons of economy on the assumption that 16mm was adequate for all future needs.

In the mid-1970s, the introduction of the home videocassette recorder and pre-recorded cassettes (free of terrestrial broadcast defects, usually telecined on modern equipment) provided a new market in home entertainment, and a market where technical quality was a competitive element. This era saw the growth of cable television (which presents a better image than most terrestrial broadcast), the deployment of film libraries to produce "instant" cable channels, and general improvement in television production and dissemination. Expansion of the international television market (global distribution of films on video for broadcast and home media) further increased competition in the domain of image quality, and like the development of cable as an outlet for film libraries, helped to subsidize film preservation in the studio context.

With the advent of videodisc (with its self-proclaimed interest in quality, better transfers and better film elements to work from), technology took another step closer to optimizing the NTSC image. Retrospective cinema has made a transition from filler (the old all-night movie, often film-chained from mediocre 16mm television library prints, panned and scanned, edited, faded, cut up by scores of commercials) to programming of choice (premium cable channels, letterboxed videodisc "restored" versions) as viewers become more sophisticated and attentive to programming.

The demand for quality in the home video market was in part matched by an increasing concern for film quality in theater presentation (better film stocks, digital sound systems) that came with the resurgence of the cinema as a competitor to the electronic media. The blockbuster action films of the last two decades depend on event-style presentation (wide-screen stereophonic spectacle) that demands maximal performance from the medium of cinema.

All of these developments have raised the level of quality required of film resources, and have also produced a revenue stream which has capitalized preservation (from better storage to superior duplication and finer videotransfers) in the studios and among their archival partners. This trend is likely to continue as forms of digital and high definition television come onto the market. These trends, along with the desire of film curators to present films in a form most closely approximating their original release, is driving both studio libraries and archives to enhance the quality of their preservation work. One of the most important revelations for the archivist in all of this is that there is no final state of preservation. Improvements in all parts of the chain of duplication technology (printers and processes such as step contact and wet gate printing, better film stocks, better lenses for optical copying, and now, various digital tools) have made it possible to extract ever more information from the original negative, and hence to make improvements in the presentation of older films in their original format as well as in ancillary media. Improvements in duplicating technology as well as the fragility of the medium of film make it crucial to retain the original elements as long as possible as the ultimate source of the best image quality.

Archivists have been deeply influenced by the experience of trying to preserve the first fifty years of film. It is deeply unfortunate that the quality of most of what has come down to us from this era is extremely compromised, not just because we have lost so much, but also because our efforts to recover this tradition have necessarily been conducted in the face of lowered expectations. We must take a hard look at the second half-century, and strive to raise the standards of preservation to the point where we truly achieve the effect of the original release.

1. Preservation in the Absolute Sense

If by film preservation we mean the perpetual retention of all of the information in the original negative, we must confront the fact that there is currently no true preservation. All of preservation falls away from the perfect standard; the efforts of archivists to restore films, while absolutely essential, are bound to fail in an absolute sense. Picture information which has been lost cannot be restored. When a film is scratched or torn or faded or riddled with embedded dirt, there is no way to restore all the pictorial information that has been lost. Restoration efforts aim at recovering the maximum amount of valid data, of authentic information. But every decision, every process in the act of photographic preservation alters, however incrementally, the original, and introduces both intentional and unintentional changes in the image. The preservationists seek to reduce as much as possible the unintentional changes, and make correct decisions in the case of changes that require some degree of intentionality.

back to table of contents

2. Three Sources of Problems in Preservation

Despite increasingly sophisticated processes for storage and handling of film, these three main technical factors, the limits of photographic reproduction as an analog process, the instability of the support media for motion picture images and sound, and the cumulative effects of storage and handling (residual chemistry, subsequent contamination, abrasions) remain persistent and universal problems for film preservation. In part, the film archivist can be defined as someone who inherits the cumulative problems of this history, and whose role is to apply such remedies as may be possible. And usually, the archivist comes into the picture about mid-way through the life of a film negative, after the original element has been subjected to sub-optimal treatment. And at this point, the impossible task of trying to return the picture to its original state begins.

Film itself is an unstable substance. By the year 1913, film professionals were aware of the problem of nitrate film decomposition. In 1955, approximately five years after the introduction of triacetate "safety" film, the "vinegar syndrome" (acetate decomposition) was discovered. In less than a decade after the introduction of Eastman Color negative, problems in color shifting were documented, and the deterioration of magnetic soundtracks was noted within two decades of the widespread acceptance of this medium. Thus, all of the media on which moving images are stored will eventually degrade, necessitating the imperfect act of photographic (analog) duplication.

Preservation begins with proper storage and handling. The better a film is stored, the longer it will be a viable element, and more of its information will be recoverable. The less damage and contamination sustained in handling, the less special processing and compensation will be necessary to make the image look acceptable again. Thus, proper storage and handling are the most cost-effective and successful aspects of preservation.

The medium of film is limited by the nature of photographic reproduction, which is analog, mechanical and chemical in nature. These attributes are not in and of themselves bad. The photochemical image is one of the densest forms of information storage and obviously the most superb medium yet for capture and display of images, and particularly for the art we call the motion picture. But the analog nature of photographic reproduction, which involves a series of relays (the film stock, lenses, mechanics and chemistry of each iteration of shooting, developing and printing film) which imposes a theoretical limit on the quality of reproduction possible within a photographic system.

Technical enhancements to the duplicating process and better methods of film handling have been developed over the years to minimize tears and other gross physical damage. Laboratories became aware of the need to work in a very clean environment and to keep the film free of dirt and contamination. Lab preparation and printing processes (cleaning and scratch removal, "coatings", wet gate printing) were developed to minimize the transmission of non-photographic defects (abrasions, ground in dirt, etc.) from one generation to the next. These methods and processes, which were not all successful, have continued to evolve to the present. While these processes offer incremental improvements in technique, none of them fundamentally alter the basic limitations of photographic reproduction.

3. Preservation Practices: When More Than Photographic Duplication Is Required

Traditionally, when we speak of "restoration" in a preservation context, we mean the activity which has as its objective the recreation of the "integral" version of the film, the activity of bringing together all of the images present in what is considered to be the original state of the film, at the highest fidelity to the original; replacing damaged, technically inferior or inconsistent sections with material from a superior source in another element. These changes in the original version occur not only through physical deterioration or mishandling, but through loss of elements, censorship or other forms of editorial revision, and they usually require restoration work at the macro-level: replacement of missing or damaged footage from a superior source (when available), comparison of different versions, etc. Film restoration is primarily a curatorial process that involves knowledge of the history of a film and its context, connoisseurship and subjective decision-making, as well as research and technical expertise.

Reconstruction of a film in the electronic sense operates at a micro-level, by using technical devices to adjust characteristics within the individual frame at the level of (for example) the grain structure of the image, the restoration of lost color, or reconfiguration of a frame that has been subjected to gross damage such as a tear or flicker. Although some of these technical operations have photographic correlatives, they are optimally performed in a digital (electronic) mode rather than an analog (photographic) mode.

Enhancement of Picture Quality: Photographic solutions to problems such as negative dirt, color fading, deep abrasions and other forms of overt physical damage to the emulsion usually consist of processes which mask or render transparent these aspects of the image in the transfer process. Digital solutions approach the problem by essentially painting the problem out of the image and if necessary, replacing lost picture information from data resources in other parts of the image or from surrounding frames. Electronic processes such as scratch removal and negative dirt extraction ("dust-busting"), grain and noise reduction are approximations, processes designed to substract and replace "noise" or non-signal information to an image which has been altered physically over the course of time. Many of these processes are additive in so far as they attempt to replace information that has been lost as a result of damage or degradation. This is the case where a program isolates a spot that it identifies as negative dirt, and replaces that spot with an average of the color of the surrounding pixels. Although this processing yields an image which is subjectively superior, the process exceeds the bounds of strict duplication. And because of the nature of the photographic image, the computer assessment of true information requires human verification and intervention. It is entirely possible for such programs to remove valid pictorial information if they are not properly deployed. As preservation practice comes to include use of digital reconstruction tools, it may be necessary to specify the application of such processing only to output images, and to retain the original transfers in an unenhanced state in order to be able to render the most accurate digital transcription of the original element as may be needed in future work.

Computer Processing vs Human Subjectivity in Preservation: Some processes, such as the "restoration" of color (as opposed to "colorization," a video process by which new colors are subjectively used to change the color from that of the faded original) depends upon a mathematical assessment of the differential between the current state of an image and its original color content. This type of reconstruction consists primarily of extrapolation, of taking a faded image back to its original state in a digital environment, based on an understanding of the rates of color loss in each layer of the film, and then rendering the image again on film. It is intended to be transparent, and capable of rendering without subjective distortions the original state of the film's color information. In essence, this type of restoration works with data resources which still exist within the image to reestablish the prior state of the image.

The set-up and oversight of the digital restoration process always requires informed human intervention, but digital tools can be generally divided between those that are subjectively applied (that is, where a specialist manually "repaints" or otherwise manipulates images directly) and those that are primarily applied by the computer programming (that is, most "real-time" systems that run a set software through the entire resource of moving images to be restored). The more individual decisions about the look of the images, the more human intervention will be required.

Archivists need to explore the philosophical as well as practical significance of human intervention with respect to issues such as the moral rights of the artist, and the historical authenticity of a restored film. Although standards have yet to be developed (it is likely that a range of practices geared to different applications and markets), the ubiquitous presence of digital options for enhancement of the moving image prefigures the universal deployment of such tools, and thus, historical and curatorial issues will necessarily have to be confronted.

back to table of contents

4. Against Technological Romanticism

The current capabilities of electronic imaging have in the last couple of years attained a stage where digital technology can now solve some of the problems that cannot be solved in the domain of film. Particularly in the area of tears, digital reconstruction can accomplish work not possible in the photographic realm. Certain printing defects such as incorrect framing are relatively trivial to correct digitally (so long as all the picture information remains to be extracted and repositioned). Intermittent problems involving flicker, fading or discoloration can also be corrected using various digital techniques, especially when there is ample picture information available in surrounding frames to extrapolate the original condition of the imperfect frames. Embedded negative dirt and some types of abrasion can be relatively effectively removed, although this work begins to approach the limits of acceptability because these techniques are based on excising damage artifacts from the image and then synthesizing the image area to be reconstructed from surrounding picture information. When the damage occurs in areas of complete dark or light, or areas with a consistent and replicable pattern, such synthetic reconstructions can be made virtually invisible. However, physical damage in areas of complex and non-redundant image data cannot be satisfactorily restored. And problems such as the accretion of contrast, characteristic of multi-generational photographic duplication, cannot be reconstructed by digital (or photographic) means. Despite the many gains that have been made, and are currently being made with new technologies for image restoration, not only is there no practical method of recovering data lost from a given piece of film, but there is no theoretical basis for recovering information which has passed out of the physical medium due to changes in the state of the image. Thus, the best protection for film continues to be conservation of the original negative and protection masters under the best possible conditions. A great deal of research into factors which produce image and support degradation are refining our concepts of long-term conservation and storage technologies. Digital restoration should not be considered as an alternative to the traditional methods of film preservation. But digital techniques in conjunction with high quality laboratory work allow us today to enhance the quality of the preservation work we are doing in the practical sense as we inherit materials which have already sustained damage, contamination, fading and deterioration(NO1).

There is a tendency to recoil in the face of these facts and hope for some transcendental solution to preservation. Often, that solution is characterized as an electronic form of preservation. And sometimes, that solution is alleged to be just around the corner. But despite the rapid development of scanning, processing, storage and output devices, there is no "electronic preservation" medium.

5. Problems with Electronic Technology as Archival Media

For aesthetic, historical, and technical reasons, film is the current standard for the long-term retention of the moving image. None of the past or contemporary electronic formats, from magnetic tape to optical disc media, can be considered "archival." For all of its shortcomings, film remains the best aggregate choice for long term retention of the moving image. This is not to suggest that a better medium won't be developed, but as the need for digital storage media intensifies in all the public, private, industrial and research applications of electronics, those media will eventually emerge.

The first and most obvious reason is the primacy of the original negative or master as the source of all valid picture information, the ultimate and most authentic image source. It is necessary to retain all of the data in the original film negative for the future value of moving image resources, and this is currently beyond the capabilities of practical contemporary systems. Although today, it may seem that the limits of visible resolution have been achieved, it is important to realize that requirements of new display systems may differ substantially from current electronics, and having a higher resolution source may produce superior imagery in those as yet undeveloped display systems. At EVERY historical moment in the past (as cited briefly above), assumptions that less than full preservation would be adequate for present and future needs has proved disastrously wrong(NO2). A substantial part of today's studio and archive preservation budgets are devoted to salvaging, insofar as any redress at all is possible, the losses of previous generations of "preservation."

Current technical capabilities in the areas of telecine (videotransfer devices that illuminate, read, encode and transfer the data from the film), data storage (reliable, long-term mass storage capable of rapidly, compactly and economically storing data on the order of 7.5 terabytes per feature film), processing (digital techniques for application of image-enhancing programs such as those discussed above to massive amounts of data in reasonable time frames), and output (transfer of the data back to film via laser recorder, electron beam recorder or other "writing" device, again, at reasonable speed) do not permit the acquisition, accurate transmission and reconstruction of all of the data in a 35mm original negative. In fact, researchers are still struggling with the basics of the film/digital/film interface, and it is this work that will ultimately demonstrate the practical viability of this system.

Although there are many factors which affect the actual life span of a given film, substantial research and massive practical experience give the archivist at least relative certitude about the performance and longevity of the several components of motion picture film. However, the lack of an equivalent, adequate system for the long-term retention of the electronic image precludes committing to any of the available electronic storage media as an alternative to motion picture film.

Both videotape and optical media such as videodisc have demonstrated short-term failure in various and unexpected ways, through practical experience. None of the tape or disc formats has been subject to rigorous testing, although various tests are now being performed. The longevity of metal-evaporated tape, for example, may prove to be substantial, and the theoretical possibilities of optical media seem even more promising.

However, electronic media are evolving so rapidly that current storage devices will soon be obsolete, if only because their data density and data transfer characteristics will be inadequate for the next generations of electronic media. Since the very beginning of the age of information technologies, the pace of innovation has been intense, and there is no reason to believe that this pace will diminish in the near term. Archivists have witnessed the succession and diversification of video formats, equipment, display systems and storage media (types of tape and tape stock, optical storage media, magnetic drives, etc.), and the equally rapid obsolescence of these systems. Theoretically, all of these systems will be interoperable in the digital domain, but when, how and at what cost remain open questions as the "digital domain" continues to evolve without the slightest regard for archival stability(NO3).

Also, techniques employed to affect the migration of the image from one system to another (from film to HD, for example) and the methods used to optimize an image within a system, are subject to constant development. For example, telecine transfers over the last 10 years have improved continually, making transfers obsolete in as little as two or three years. Practical experience has produced an understanding of the artifacts that are produced in the stages of image migration, and practices which eliminate those artifacts and refine the signal continue to come on-line, improving the quality of the image.

Film and digital imaging are not perfectly equivalent systems, and not yet truly interoperable. There are differences between these media in the basic form of data storage (analog vs digital), differences in the way data is stored and encoded, in the display systems, and in the concepts of gamma, chroma, and other characteristics of the image inherent in the two systems. The migration of a film image into the digital domain and back out to film again is not yet a transparent or satisfactory process. Until reliable techniques and standards have been developed to effect this full transfer, digital storage cannot be considered a true preservation medium for motion picture film.

Finally, the economics of electronic imaging which would be adequate for motion picture preservation are forbidding. The term cost of color separations and other requisite preservation elements conserved in cold storage over the long run is far lower than the cost of even rough parity in the digital domain. The mechanical simplicity and universality of 35mm film, the installed base of relatively inexpensive film printing and projection facilities, and the comparatively unproblematic compatibility of motion picture technology with all existing systems of distribution and presentation, all contribute to the long-term superiority of film over current electronic storage media.

Therefore, regardless of the evolution of electronic imaging, we cannot say today that electronic imaging alone can solve the overall problem of preservation.

back to table of contents

6. Some Technical and Aesthetic Problems

Given the problematic nature of the film medium, it is clear that archivists need to find a superior alternative for long-term retention of moving image resources. But there are sound technical reasons why preservation continues to be film-based, despite the limitations of the photographic medium. Many of us participate in the broad discussion which seeks to define an ideal archival medium for moving images, and one day there may be a superior medium that is both process- and image-transparent, robust, economical and universal.

But preservation is not just a technical activity. Although the primary element of preservation work is the film, archivists are also concerned with the context, the culture of cinema. Film preservationists must be concerned not just with the absolute quality of the image, but with its authenticity, with the presentation of the film image in conditions which approximate the ideal viewing ambient of the historical motion picture. Therefore, film preservation is not just the retention of an image, or an acceptable or approximate image, but requires the production of a historical image (the original image, as a correlate of the original viewing experience and medium). While new media such as video are extremely important tools for access and increasingly for preservation, the motion picture archivist must be able to show film, in a theater, to an audience. But ultimately, this is not enough. The creation of a reliable trans-filmic digital storage medium, capable of producing historically and technically accurate printing elements will be important in any scenario that envisages the scarcity of motion picture film.

7. The Problems of Context, Taste and Authentic Reference

Curators need to know how a film should look; that is, they need to understand thoroughly how the film appeared when it was originally presented. One of the difficulties of color restoration is finding the proper color without an accurate or authoritative reference. Color film processing allows for extensive latitude, and a wide range of effects can be produced from a given negative. These effects are always the product of choices, and such choices need to be informed by historical and aesthetic considerations. The differences between imbibition Technicolor and Eastman Color add to the problems of recreating the original look of a film. The same problem obtains with black and white prints, to a lesser extent.

How will we be able to authentically "preserve" films when the characteristics of the available exhibition medium (types of printing processes and film stocks) is substantially different from the film processes of the original release and there is no modern equivalent?

The viewing context of the motion picture changes with each adjustment and change in technology, and with these changes in deployment and characteristics of the medium, there is an eventual change in taste that responds to the loss of older characteristics and their replacement with new characteristics. The audience for motion pictures, including the specialized groups who generally attend archival screenings, is very different from audiences of the various historical periods of the cinema. Audiences have been conditioned by the style and look of modern film and video, and are accustomed to the visual and aural effects produced by the latest stages of film technology. Today's audiences expect sharper pictures, more controlled color effects and cleaner sound than was possible in the past. This presents the curator with the problem of deciding how much "improvement" is appropriate for a given film. For example, when Technicolor three-strip negatives or separations are transferred to Eastmancolor interpositive or internegative stock and then printed, the effect can be considerably sharper and somewhat less grainy than an original Technicolor release print. Similarly, sound processing allows the modern archivist to remove pops and clicks, hiss and rumble from early 1930s sound tracks. But since the general sound design of the film took these system limitations into account, the resulting "improved soundtracks" will often seem unbalanced and less acceptable than the original tracks with their primitive recording and reproduction artifacts. If we examine the movie audiences of the various eras of motion picture production technology, we could write a history of compensatory paradigms, a history of the parts of the apparatus and its effects that are repressed or screened out "automatically" in the experience of seeing and hearing a film. These would include the obvious (taking musical accompaniment to the silent film as a given rather than synchronized sound that is modelled on the "real" world; accepting monochrome and tinting as valid devices for stimulating emotional participation rather than a failed attempt at pictorial realism; subconsciously screening out hiss and rumble in the soundtracks of the early sound era, accepting image imperfections such as grain, negative dirt, and color inaccuracies, ignoring the image field distortions characteristic of early Cinemascope, accepting the conventions of synchronized monophonic sound, and then the conventions of multi-channel recording, etc.).

It is relatively easy for the archivist to reject certain false arguments about how a film should be refashioned using modern technology (i.e., colorization's famous notion that "They would have made the film in color if they could have afforded it..."), but there is no clear and unambiguous standard for how much technology should be introduced to "improve" a film. In the early days of preservation, when relatively little could be done, there was less chance that the techniques of preservation were likely to fundamentally change the aesthetics of a film. Today, "apparent" restorations which alter original information in the film and make the film look and sound better than it did historically are quite possible, and archivists and curators need to re-evaluate the goals of preservation practice to establish standards for restoration based on aesthetic and historical considerations.

8. The Future of Film Preservation

Today, preservationists have better tools than ever before for the reproduction of the photographic image. But the historical era of the motion picture may be ending, and when that era ends, so will the economies of scale which have allowed the preservation "industry" to subsist at the margins of mainstream film printing and processing. Film has been a predominant medium for 100 years, and this longevity tends to make us complacent. In fact, motion picture film is one of the older industrial technologies, and one which faces competition from a variety of rapidly developing media. Film is a complex of processes each of which seems to derive its economic strength from one or another application in the moving image world. Often, a loss of one part of this complex affects other parts. If we look at recent history of the media, there are a number of trends that we can expect to overtake the motion picture early in the next century. These include:

A. Loss of Exhibition Infrastructure: With the advent of electronic projection, there will almost certainly be a rapid disappearance of venues, equipment and trained operators for motion picture film projection. As the film product disappears, this element of the infrastructure will collapse, followed by the more gradual disappearance of the archival retrospective film exhibition that was an active force in the cultural sphere from the 1960s to the 1980s.

B. Loss of Work Volume and Economies of Scale: As the motion picture becomes a "taking" medium and is replaced by electronic projection, the economies of raw stock production and processing will be lost, and film will become a specialty item, a museum product rather than a product of mass distribution.

C. Loss of Processing Infrastructure: As the volume of work for film printing declines, laboratories will have to downscale in both facilities and personnel. Economic realities will make it difficult for more than a few laboratories to survive, leading to a further concentration in the field. Few laboratories will have the equipment and technical expertise to do quality film work, and few will be available to perform the old processes at high quality. This is already true in the case of black and white film. As soundtrack development takes us beyond the era of optical sound, not only will the optical sound reproduction capabilities disappear in many places, but black and white processing which is largely today the domain of sound tracks, will be further marginalized.

D. Loss of Media: It is easy for archivists and historians of technology to list media that are no longer reproducible in their original forms. The Vitaphone disc recording and playback system and the Two-color Technicolor process are examples of the scores of technological innovations which had distinct aesthetic and technical characteristics, but which no longer exist as motion picture technologies. Cellulose nitrate and the traditional high-silver content emulsions have disappeared, as has the Technicolor imbibition dye transfer printing process. These are among the many original forms of the motion picture that can only be reproduced in the form of later forms motion picture media which have different technical characteristics. The rise of the ENG camera eliminated the market for 16mm newsfilm, and videotape distribution made 16mm television syndication prints obsolete. These may be the two factors most responsible for the radically diminished availability of 16mm film prints for educational and repertory markets. Kodachrome, which was a medium of preference for many independent film-makers, has all but disappeared, and is only available in 16mm format. The loss of 35mm Eastman Color film as a print medium is not impossible to imagine.

In the light of these facts, it is important to perfect the interface between the film system and digital imaging technologies. Initially, archivists must come to terms with the new class of tools which allows the extension of traditional methods of preservation in the digital domain. In this first stage of accommodation with the digital world, techniques and standards should be developed which allow a completely transparent exchange between the film and video domains, so that film may be scanned, processed and output without being inflected by any type of modification or artifact introduced by the process. As time goes on, electronic media of all types will displace more and more of the traditional domain of film. Archivists need to appropriate these new tools and to develop practices which conform with the requirements of the curatorial and historical imperatives of the film archive movement. This is especially important since electronic systems of storage, transmission and display are increasingly supplanting film projection, and a refinement of the transfer processes between film and digital imaging will allow more authentic images to be presented in non-film formats. In the end, archivists will almost certainly find themselves reliant on electronic means to preserve images originally captured on celluloid. Major technology suppliers are becoming aware of the need for an archival medium to support long-term retention of large scale digital data resources, and the ultimate archival medium may turn out to be very different from film. But we will only be able to take advantage of such a solution if we have mastered the film/digital/film interface.

back to table of contents

9. Toward an Archival Medium

We cannot expect the development of the ideal archival medium to be developed by film archives, or even by the film industry. But the intense development of digital technology is likely to provide the pieces of the puzzle that are required for an archival medium. FIAF, through its Commissions, should begin to work toward the specification of such a medium. Some of the primary characteristics would include the following.

A. Transparency: It is of primary importance to develop the technical basis for a film-digital-film transfer that is truly transparent. By this we mean a system capable of capturing all of the data in a given element (negative, fine-grain, print of any gauge and process), storing that information in a digital format, and rendering the information subsequently on film in such a way that no artifacts are introduced, and so that the resulting film copy is indistinguishable from the original. Only by exploiting the transparent duplication possible with digital storage and processing can true preservation (in the sense of duplicating all of the information in an original image with complete fidelity) be accomplished. Eventually, perfect duplication will be the standard, and no one will accept the limited forms of duplication common in today's preservation environment. However, the actual techniques of such duplication have yet to be specified or refined to the degree required by archives.

B. Flexible High Density Storage: Since the capacities of electronic data storage and manipulation are theoretically superior to the photomechanical moving image, it is logical that an archival medium should be developed in an electronic form. This storage will have to accommodate information equivalent to 35mm original negative. Because such a medium will also be required to absorb data from sources other than the average 35mm negative, including the new extremely high-resolution negatives (i.e., Kodak 5245), 65mm or IMAX, Showscan, etc., as well as originating media using larger frame size, various aspect ratios, frame rates higher than 24 frames per second, and eventually, 3-dimensional imaging, the archival medium will have to have the flexibility to encode different amounts and types of information per frame.

The ideal archival medium must be capable of accepting all current forms of film and electronic imagery via telecine, scanner, a/d conversion, etc., and also be capable of transparent migration to present and future formats. A minimal requirement for the preservation archivist is that the material stored in an archival medium be convertible back to its original form. The major challenge will be the reconstitution of 35mm film from high density video storage via devices such as electron beam or laser recorders. The reconstitution of iconic data from the archival medium will have to be a reliable, economical and hardware-independent process, although such operations will not necessarily be constrained to real-time conversions rates.

C. Universality: It is likely that various international broadcast standards and video formats will persist into the indefinite future. Motion picture film will also continue as an originating format, perhaps with more variations than currently deployed. However, a universal storage code for iconic data could transcend these divisions. Since there are many frame ratios and different data densities depending on originating format, the storage code for the archival medium will have to be a universal standard independent of specific equipment, applications or systems. Standards for encoding iconic material will have to be implemented to support this diversity. Data in this universal form would not be hardware dependent, thus alleviating the current problems of equipment obsolescence. Conversion equipment and techniques would be required to migrate imagery from storage in the form of universal digital code to present and future display formats and standards, and provision would have to be made to have such equipment perennially available. The code might also be used to store still data, a factor which would increase the user's base of the code and therefore help to insure its viability as a recoverable data format.

D. Durability, Robustness and Economy: The physical medium for preservation will have to be capable of efficient, very high density storage, and to retain data under normal storage conditions for three to five hundred years. Although it can be predicted that such a medium will never be inexpensive, the cost factors associated with this universal archival medium should be such that it allows the retention of all of the motion pictures that archivists may choose to protect. As a maximal base figure, the cost of protecting a film in this universal archival medium should approximate the current cost of film preservation using motion picture film, and the cost of conversion from this medium back to film should be roughly equivalent to the cost of making a print or pre-print element.


Notes:

1. Some of the latest thinking on these issues is provided in The Effects and Prevention of "Vinegar Syndrome" by A. Tulsi Ram, David F. Kopperl, Richard C. Sehlin, Stephanie Masaryk-Morris, James L. Vincent and Paige Miller; Journal of Imaging Science and Technology, Volume 38, Number 3, May/June 1994, pgs. 249-261, and American National Standard for Imaging Media - Processed Safety Photographic Films - Storage (Proposed ANSI Standard) by the Secretariat, National Association of Photographic Manufacturers, Inc.; February 1994, iv + 41 pgs.

2. see, for example Resolution-Independent Film Scanning: How Independent is Independent? by Peter R. Swinson; SMPTE Journal, February 1995; pgs. 82-84

3. see Ensuring the Longevity of Digital Documents by Jeff Rothenberg; Scientific American; January 1995 pgs. 42-47.

Film / Digital / Film

En su detallado artículo sobre las perspectivas de la película y de los medios digitales como sistema de conservación a largo plzo, el autor nos propone primero una reseña histórica del pasaje de la película al medio digital y a la televisión de alta definición.

Siendo la definición de la preservación absoluta la conservación perpetua de toda la información contenida en el negativo original, no hay en el sentido propio posibilidad de preservación. Para ello se precisan técnicas que la duplicación fotográfica sola no puede garantizar...

Sin embargo, advirtiéndonos sobre el peligro que corremos al incurrir en lo que el autor llama el romanticismo tecnológico, nos invita a analizar de manera sistemática las ventajas y desventajas de los medios electrónicos como sistema de archivaje y las relaciones que existen entre film e imagen digitalizada.

Los archivistas deberán poder disponer de alternativas para la conservación a largo plzo de imágenes, sin perder de vista la forma en que ésta se presentaba originalmente.

Evaluando detalladamente los diversos aspectos de ambos sistemas, el autor concluye a favor de la película como medio de conservación a largo plazo y del medio digital como técnica de acceso del futuro.