The Brooklyn Rail

For years, most writing on film restoration has been dominated by a debate over how best to future-proof our entire moving image history: on digital tape or analog film? High-octane rhetoric makes the question urgent. “It’s a Silent Fire,” claims a Hollywood Reporter headline from March of 2024, referring to the potential for decay in Hollywood’s digital repositories, which have grown in size and scope as fewer and fewer films are stored, restored, and preserved on more reliable analog film. “An archival tragedy,” reads a further description, “reminiscent of the celluloid era.”

That such a tragedy did take place in the “celluloid era” (more rightly, the prewar studio era) is incontestable: the field of film restoration as we now know it was founded in honor of scores of films lost to shoddy preservation practice and unstable nitrate film. Whether we’re on the verge of repeating such a disaster, this time in the cloud or on hard drives, remains to be seen. Analog advocates continue to push for on-film restorations, in part in anticipation of some potential digital implosion. But in an interview for IndieWire, published just a few months after the Hollywood Reporter article, the former chief of the Library of Congress’s National Audio-Visual Conservation Center, Gregory Lukow, claimed such anxiety was a gross overreaction. Digital decay, he observed, “was a lede fifteen years ago.” The piece ends, sunnily, with an affirmation of analog and digital methods’ equal pride of place in film restoration and filmgoing culture. 

But that isn’t quite right, either. Much coverage of the digital versus analog film restoration debate misses the fact that the battle is already over—or, at the very least, close enough to call. In the United States, new film stock is produced exclusively by Eastman Kodak, which paused all manufacturing in November 2024 to meet concentrated demand. It’s not yet clear what effect this will have on film prices, which have increased steadily over the past decade. (In January 2025, Kodak Alaris, the UK spin-off of Eastman Kodak, announced price hikes on most of its analog products.) Though directors like Martin Scorsese, Christopher Nolan and, more recently, Brady Corbet have pushed to keep the medium alive in new projects, celluloid’s availability to those looking to restore old works has become increasingly limited. As of January 2025, there are only three film labs capable of analog—or photochemical—restoration operating in the United States. Elsewhere, the picture is equally bleak: a few years back, Germany’s federal archive rerouted funding from its analog lab toward digital efforts, citing an inability to sustain both.

The shift toward all-digital methods in film restoration, not to mention exhibition and distribution, announces film’s rapidly encroaching obsolescence. Even as individual archives, niche cinematheques, and committed analogophiles push for the perseverance of some celluloid culture, the writing is on the wall—or, in this case, the recording is on the digital tape. Film is, or will soon be, gone. This reality lends an elegiac pall to arguments for outputting to analog formats even after digital restoration. (“My thing is,” says Anthology Film Archives restorationist John Klacsmann, “Why not do it while it’s still possible?”) It has also changed the terms of this field-defining conversation. The question is no longer whether digital or analog technology is more appropriate to film restoration. The question is: what is film restoration, or what will it be, after film? 

*

Since 2007, when the Academy of Motion Picture Arts and Sciences published its oft-cited “Digital Dilemma” paper, concerns of the type outlined in the Hollywood Reporter piece have dogged the film restoration community. Offering a wide-ranging assessment of digital archiving methods’ suitability for long-term preservation, the Academy paper stressed the shortcomings of Linear Tape-Open (LTO), the most popular digital archiving format then and now. Archivists were alerted to the necessity of maintaining regular digital migration to avoid losing media to a lag between hardware advancements, human error, or both.  

While some were quick to dismiss the Academy’s alarm, many others responded with a renewed insistence on analog technology. Outputting digital restorations to film seemed a surefire way to prevent overreliance on a changeable digital interface. Stored properly, polyester film (as opposed to nitrate or acetate) can last for a hundred years, or longer. Even less stable formats have a decent shelf-life: we know because we’re still finding films from a hundred years ago. Plus, the barrier to accessing film stock seems, compared to the risk of hardware clash, relatively low. The image that analog purists return to is at once oversimple and eminently persuasive: some future film lover, armed with a flashlight, illuminating a strip of celluloid. Alone in the dark with a fragment of cinema. No other tools necessary. 

For a time, arguments in this vein seemed to have won out. Then, recently, as film stock prices continued to rise and pandemic delays disrupted in-person workflows, things changed. “It’s only in the past ten years or so that everything has gone digital,” says Klacsmann. MoMA’s Dave Kehr, film curator and co-programmer of the annual To Save and Project restoration festival, concurs: “Methods of distribution and exhibition are all favoring digital now. At this point, we will [output a restoration to] 35 mm if we have the grant money for it, but that’s becoming more and more rare.” With this shift has come renewed inquiry into the stakes of film restoration efforts themselves. “If we’re not preparing films to last as far into the future as possible,” Klacsmann asks, “then what are we doing?” 

This is precisely the sort of question thrown into sharp relief by celluloid’s decline. How much is film restoration a fundamentally archival project? To what degree will it, or can it, continue to be? For experts, answers vary. Some, like Peter Bagrov, senior curator of the George Eastman Museum’s Moving Image Department, see film preservation and film restoration as interrelated but still distinguishable efforts. “Film preservation,” says Bagrov, “is a broad term that encompasses film restoration as well as many other things.” Jay Weissberg, director of the Pordenone Silent Film Festival, agrees that there’s some overlap: “I’m quite happy with the terminology of preservation and restoration being slightly fluid.” But the difference, if there is one, is quite literally material. “Preservation, to me, means keeping the material that exists,” says Weissberg. Whereas restoration would involve “looking harder at that material,” potentially cleaning it of mold or scratches, scanning it digitally, or combining it with other artifacts to produce a reconstruction. 

For Klacsmann, however, there is more direct traffic between the relatively passive process of preservation and the active efforts of restoration: “Preservation is about duplication for longevity, while restoration involves active tools to bring the film back to its original state.” But the latter includes, he adds, “making archival film elements that can last hundreds of years.” 

Someday soon, film restoration stands to lose this aspect of longevity. There will still be old film material to grapple with—if not always, then for a long time to come. (Weissberg notes that there’s quite a backlog: “Every archive has stacks and stacks of film cans that they’ve barely been able to open.”) It’s the ability to produce a new archival element that has been and will be most severely limited by the rising cost of celluloid and the shuttering of photochemical labs. To the extent that film restoration is ever defined by efforts to return films to their original state through a preservation of their original medium, the scarcity of celluloid amounts to an almost ontological threat. 

But there are those who don’t see it that way. For them, film restoration means something a bit different: it is as much about access—broadly defined—as it is about archiving. “To me,” says Larry Blake, a longtime sound supervisor and one of the original critics of “Digital Dilemma”-induced anxiety, restoration and preservation are essentially the same:

Once you’ve [digitally] restored a film, you’ve got a pile of ones and zeroes, on hard drives or servers. There’s no meaningful distinction between restoring and preserving. The goal is to make something such that in ten, fifty, a hundred years, people can see and hear a film exactly the way the director did, period. Film in any form is not going to do that, much less a film print. 

Film prints fall short here for a couple of reasons. The first is that they are, by today’s standards, far less capable than digital when it comes to presenting a reliably excellent range of sound and image. “You scratch a film print, it’s scratched forever. You cannot show it in a theater. If I preserve a DCP [digital cinema package, the standard exhibition format], you can show it in two hundred thousand theaters around the world.” Exhibition is the other concern when it comes to analog restoration: “There will be no more film projection in a hundred years,” Blake predicts, “or if there is, it’ll only be at the Academy.” 

This proclamation will come as a surprise to some. For now, film projection is enjoying a twilight burst of popularity, wrenching watchers off their couches for the always-on-film showings at West Hollywood’s New Beverly Cinema, or the annual “Big & Loud” 70 mm series at New York’s Paris Theater. In a culture that has largely left the cinema house behind, on-film screenings have a unique appeal. There is a singularity to the experience of seeing this print in this space at this time—a sense of immediacy that, even if the print is (and it almost always is) a copy of a copy of a copy, is not present in a DCP. “We want to experience more of our senses,” says Bagrov of resurgent audience interest in film projection. “We tap the keyboard, and we look at the monitor. But when we read books or see films, we like to touch something different. In the past, when something went wrong at a screening—a print broke or was out of focus—audiences would yell at the projectionist. Now, when something happens, people are excited. It’s a sign of the physical that’s been lost in our life.”

Still, even Bagrov admits that audience interest in film projection is primarily limited to cinephilia’s urban, coastal hubs. Most theaters do not have the capacity to project film (or, if they do, they do not have projectionists trained to run them—which comes to the same). Though a 35 mm or 70 mm screening may be a selling point in some circles, it’s rarely enough of one for studios or distributors to justify the expense. The reality is that film projection, as rare and fine as we know it to be, leaves many audiences—or potential audiences—in the dark.

“Archivists are an inherently conservative people, literally,” agrees Kehr, back at MoMA, “but even the most dedicated analog people are starting to give into the reality that, if you want your films to be seen, you really have to lean into digital.” Kehr notes that digital practices make it much simpler to share restored material, both with audiences and with other institutions, because archives are less protective of digital files than they are of rare or easily damaged prints. When it comes to how he sees a movie, on film or digital, in theaters or at home, Kehr isn’t particular. “My ethos has always been: by whatever means necessary,” he says. “I got my basic film education by watching late night television. I’d sit up until four o’clock in the morning and watch a film intercut with used car commercials. And you know what? I still came out with a pretty good appreciation of John Ford.” 

*

Pursued too long, the question of what film restoration is becomes a Möbius strip—a fitting, if not satisfying, form for it to take. For every point (digital allows us to see more of the director’s original vision) there is a counterpoint (what if that original vision was on film?), and for each perspective there is, twisted a different way, another. That kind of cyclicity is oddly comforting: it suggests the end of the discussion is never the ending. In this way, the question of what the “end” of film means for film restoration becomes a chance to countenance all the ways in which the practice resists finality. 

Every archivist, restorationist, curator, analogophile, digital defender, silent film nut, or indie features fan I spoke to for this piece offered a different view of what film restoration was, had been, or could be. Even the clearest definitions of the field left ample room for indeterminacy. That ambiguity, like an imperfection in a projected film print, is a sign of life. Film restoration is a vocational profession: those in it are in for little but the thing itself. That’s clear not only in the work they do, but in the ardent way they think and talk about that work. The film vs. digital debate may be effectively over, but the fervor that animated the conversation hasn’t gone anywhere. If disagreement begets uncertainty, and uncertainty begets interpretation, clarification, revision—well, those are just a few ways to make an old thing feel new again, which is to say: those are just a few things film restorationists do best.  

Case in point: this year, the oldest professional organization for film restoration, the International Federation of Film Archives (FIAF), will revise their code of ethics for the first time since it was adopted in 1997. The overhaul is about many things, says Bagrov, who is FIAF’s acting president, among them the need to update language specific to photochemical preservation for a digital age. But one change stands out. “It’s a very strict and, I would say, catchy formulation,” says Bagrov. “It’s this: ‘Restoration is not creation.’”

Per Bagrov, the claim prompted a flurry of intra-field conversation. Of course it did. Defined in the negative, restoration becomes more creative than ever. With only one option off the table, the alternatives for what film restoration could be are endless. It’s this endlessness, more than anything, that should animate our understanding of the practice today. With digital methods and formats the new standard, the limits for what can be done to restore a film are less clearly defined than ever before. All that means, at base, is that questions restorationists have asked themselves for years are being asked with more urgency now. Whether film restoration is an archival effort is one such question. Whether film material is necessary to achieve the “original” look of a film is another. What kind of clarity you can access with digital technology, and to what degree that might surpass authenticity, are others still. 

Both film-specific and highly subjective, such questions are, perhaps, not answerable. At least, not finally. Rather, they must be posed again and again, each time a film is restored. “I don’t think it’s possible to do restoration without these questions,” says Amy Heller, co-director (with husband Dennis Doros) of Milestone Films, an independent distributor focused on restoring films from outside the Hollywood canon. Heller is referring here to the uncertainty she faced over whether to make a significant change to a film’s title cards—even at the behest of its cinematographer. That question was raised, she says, because often, while working on a restoration alongside members of its original crew, “they remember things. They remember it was hot that day, or it rained in the middle, or I had a terrible camera operator that day, or the best boy was sick, or the lab tech was drunk. It’s a very human process, in that way.” 

Restorationists, of course, don’t always have a film’s original creators available to weigh in. In the absence of immediate interlocutors, they’ll turn to old reviews, notes and diaries, the recollections of directors who made similar films or the audiences who saw them. In this way, memory is always the material of film restoration. This description is the closest we may come to a definition. What is film restoration, after film? What it’s always been: a human process, founded on open questions, asked again and re-asked. In this way, it’s future-proof—more present, still, than past.