Sound on Sight has been nice enough to publish my short history of the real-time film. Check it out. Or here it is in this blog:

The Past, Present, and Future of Real-Time Films

What do film directors Alfred Hitchcock, Stanley Kubrick, Agnès Varda, Robert Wise, Fred Zinnemann, Luis Buñuel, Alain Resnais, Roman Polanski, Sidney Lumet, Robert Altman, Louis Malle, Richard Linklater, Tom Tykwer, Alexander Sokurov, Paul Greengrass, Song Il-Gon, Alfonso Cuarón, and Alejandro Iñárritu have in common? More specifically, what type of film have they directed, setting them apart from fewer than 50 of their filmmaking peers? Sorry, “comedy” or “drama” isn’t right. If you’ve looked at this article’s headline, you’ve probably already guessed that the answer is that they’ve all made “real-time” films, or films that seemed to take about as long as their running time.

The real-time film has long been a sub-genre without much critical attention, but the time of the real-time film has come. Cuarón’s Gravity (2013), which was shot and edited so as to seem like a real-time film, floated away with the most 2014 Oscars, including awards for cinematography, directing, and editing. Now, its director of photography, Emmanuel Lubezki, is back with another fellow Mexican, Iñárritu, and a film, Birdman: Or The Unexpected Virtue of Ignorance (2014), that purports to tell its story in one extended shot. If the real-time virtuosity of Birdman winds up facing off against Linklater’s Boyhood for this year’s Best Picture Oscar, let us be the first to say that that will be no small irony, as though Billy Beane (played by Brad Pitt in Moneyball) had been using a certain strategy for the Oakland A’s for years, tweaked it, and shepherded the A’s to the World Series only to compete with another team who mastered his former strategy. Or, perhaps more comparably, as though Sidney Lumet had been working on a certain kind of film, perfected it, and then lost critics and audiences to someone else’s version. As we’ll see, that pretty much happened.

This article is the brief history of the real-time film. But first, why does anyone bother to make a film appear as though it happened in real time? The shortest answer is gravitas. As a device, real-time underlines the urgency and difficulty of solving a given problem in the amount of time it takes for French people to eat dinner. Imagine, if you will, halfway through the exquisite 80-minute conversation that is Linklater’s Before Sunset, a title card saying “THREE HOURS LATER.” We would have felt an immediate diminishment of the gossamer-web preciousness of Jesse and Celine’s reunion. Real-time isn’t appropriate for genres like the musical or broad comedy – these genres present life “with the bad bits cut out.” If cinema should be “pure entertainment,” as a few philistines have implied, then real-time stands as a rebuke to turning off your mind and letting the editor and director tell you where and when to look. Real-time shares this with horror: done poorly, it brings tedium, but done well, it forces an uncomfortable identification that serves to make dramatic pathos feel closer to the bone. Most films are dreamlike and often hokey; real-time films short circuit the usual distanciation by putting us in the same, often claustrophobic position as the actor-characters. They implicitly say to us, “Oh, you’re running out of time to finish this film? We’re running out of time to solve this problem!”

But real-time films share more than forced gravitas. They also tend to contain characters who put great faith in the power of language to solve problems, often to indicate the folly of such faith. In the forced realism of cinema, at first, words soothe, then they clunk, then we can only hope against hope that they are enough. Even Gravity, perhaps the least dialogue-cluttered real-time film, includes Matt (George Clooney) insisting that Ryan (Sandra Bullock) talk through her problems. By giving “plastic reality to subjective material,” as Peter Dellolio put it, the “synthesis of real-time and filmic space forces the viewer to absorb narrative information on multiple, often distastefully ironic levels.” Real-time suggests almost a forced purification, on the order of the Dogme 95 manifesto, though beyond that document’s rules. Perhaps “real” real-time films shouldn’t include any flashbacks or flashforwards or any tricks that directly call into question narrative flow…or perhaps filmmakers should be commended for establishing and then occasionally breaking form.

Real-time films share the non-freedom of freeing themselves from the implied potential elision of time in every single edit unbridged by diegetic sound (in most films). Real-time films acknowledge cinema’s properties of mummification, of freezing time into a sort of forced nobility: they both construct and deconstruct that nobility, giving us the thrill of continuity and the agony of contingency. This was what Gilles Deleuze was getting at in The Time-Image when he wrote that, in most modern films, “perceptions and actions ceased to be linked together, and spaces are now neither co-ordinated nor filled.” For cinema to be true to itself, Deleuze wrote, it should unfold images in the same manner that we know from our lives and bodies and selves…in real-time. It’s easy to respond, “But not every story should be in real-time!” True, but filmmakers don’t often use real-time even when asked to essay a 2-hour-or-less story. There are hundreds of filmic adaptations of plays that took place in one place and time (say, A Raisin in the Sun (1961)), but few that consist exclusively of 60 minutes (or more) of contiguous real-time. (To “make it more cinematic,” scenes and flashbacks are often added.) With all the expense and planning that goes into films, real-time can never be other than a difficult deliberate choice on the filmmakers’ parts. Let’s give them their due.

For some time now, critics have marveled at the power of the extended shot, as seen in films like Orson Welles’ Touch of Evil (1958) and Steve McQueen’s Hunger (2008), and in some ways, the real-time film extends some of the same qualities: prolonged tension, apparent acting without a net, a how-are-they-doing-that? feeling, the camera becoming the proverbial “bomb under the table.” (Hitchcock famously said that after viewers see the bomb under the table, they are riveted by even the most mundane dinner conversation.) Perhaps the single-shot film – and here we are only permitted to speak, as we shall at more length, of Rope (1948), Sleep (1963), Empire (1964), Running Time (1997), Timecode (2000), Russian Ark (2002), The Magicians (2005), PVC-1 (2007), La Casa Muda (2010), Silent House (2011), and Birdman – is just the real-time film squared. But before we can conclude that, we have to ask where did the real-time film come from? And where is it going?


For Deleuze, the European postwar situation moved cinema from the “movement-image,” where all meaning was contained in rapidly moving plots and performances, to the “time-image,” where signs could no longer be trusted, and images had to work harder to represent reality. Deleuze cited spaces and situations “we no longer know how to describe,” like deserted yet inhabited buildings, razed cities under reconstruction, and the people bearing witness to this, rather than doing things with purpose and confidence as in classical narrative. Film scholars will not be shocked to hear that Deleuze traced the “time-image” to the immediate postwar, Italian neo-realist films of Visconti, Rossellini, and DeSica – films like Ossessione (1943), La Terra Trema (1948), Roma Citta Aperta (1945), Paisa (1946), Sciuscia (1947), and La Lidri Bicicletta (The Bicycle Thieves, 1948). None of these films claimed its events took place even during one day, never mind two hours, but they certainly showed the world an aesthetic of (seemingly) unlit street-level realism, (seemingly) unmade-up actors who were their alienated characters, and a disquieting sense of long periods of (screen) time wasted and squandered.

Hollywood was well aware of neo-realism; in 1947, the Academy Awards created the first-ever Oscar for a Foreign Language Film in order to recognize Sciuscia (Shoeshine). No doubt, some American filmmakers were mildly peeved to hear waves of critics praising new Italian films as “realistic” in a manner that implied that 20 years of Hollywood films had been anything but. At the time, one of the few filmmakers with both the position and predilection to react to “neo-realism” was Best Picture Oscar Winner (The Gentleman’s Agreement, 1947) Elia Kazan, who consciously filmed Panic in the Streets (1950) in a neo-realist style. Another one was Alfred Hitchcock, who also had a Best Picture (Rebecca, 1940) under his belt, and who also liked formal experiments; he’d made Lifeboat (1944) all in the one location of the lifeboat, the action taking place over a day, a night, and another day.

Rope (1948) is based on a play loosely based on the famous Leopold and Loeb case of two pretentious, nihilist peers murdering a third, and daring their professor to unearth their crime. It’s quite astonishing how well Rope holds up, considering it could barely have been more sui generis. It was the first film Hitchcock produced himself, not under the aegis of British or American producers. It was his first film in color and his first to star Jimmy Stewart. It was almost certainly the first American film to feature homosexual characters as two of three lead roles – not that such representation was exactly unproblematic, as well documented in both book and film of The Celluloid Closet. And oh yes, it was the first real-time film as well as the first single-shot film…sort of. Hitchcock liked promoting it as a film without cuts – or at least as far as that was possible in the 1940s, when a camera’s roll couldn’t hold more than 10 minutes of film – but Hitchcock cut even when he didn’t need to.

Rope is about a pair of 20-something “school chums” who kill a third chum just before a dinner party, then try to persuade their Nietzsche-loving professor into admiring their crime. As with the best of films, Rope’s form and style work in synchronous harmony: the nowhere-to-hide-ness of the form suits the increasingly uncomfortable reality of Phillip and Brandon. Considering Hitchcock didn’t know if anyone had ever done such a thing before, his hand is remarkably assured, the dollying camera feeling more like an interrogator, the setting sun suggesting a long dark night for the pro(-an)tagonists. (Few films on this list dare go from day to night.) If Rope is the granddaddy of real-time films, it’s a distinguished ancestor who can feel safe that he did the proper trailblazing for his descendants. Rope anticipates almost everything that would characterize the real-time film, like the theatrical syntax, like the march of time being both too short and too long, like the sinking feeling that all of the culture espoused by the leads won’t be enough to save them in the end.

Now that Sight and Sound has seen its way to anointing another Hitchcock film (Vertigo, 1958) the Best Film Ever, it’s probably not difficult to move to anoint Hitchcock as the originator of the real-time sub-genre (and its own single-shot sub-sub-genre). However, what most people don’t know is that Hitchcock followed Rope with another real-time experiment called Under Capricorn, starring Ingrid Bergman, which did not produce anywhere near as felicitous results. Under Capricorn is about reunited lovers in 19th-century Australia who learn each other’s dark secrets. The film never clicks; it’s a historical thriller without thrills. Or as Hitchcock told Truffaut, “As an experiment, Rope may be forgiven, but it was definitely a mistake when I insisted on applying the same techniques to Under Capricorn.” Not that Under Capricorn was ever going to be quite as real-time as Rope, but we shouldn’t remember Hitchcock as a master of the long take, or otherwise, who couldn’t do anything wrong. Hitchcock regretted the project from the start, involving himself only to reunite with his star of Spellbound and Notorious, as he told Truffaut: “I was literally intoxicated at the thought of the cameras and flashbulbs that would be directed at Bergman and myself at the London airport.” (The master of suspense also recounted how “Bergman got angry with me one evening because of those long shots” and kept yelling at him after he’d left the room. Bergman, apparently frustrated with Hitchcock’s version of realism, wrote to Roberto Rossellini to put herself in his hands, went to Italy to star in Rossellini’s Stromboli [1950, as memorialized in the Woody Guthrie/Billy Bragg song “Ingrid Bergman”], had a baby with the married director, was denounced on the Senate floor, and may or may not, with the similarly decamping-for-Europe Grace Kelly, have inspired the misogynist betrayal themes of Vertigo, Psycho [1960], and The Birds [1963]. All this over real-time scenes!)

In 1949, Under Capricorn shared space in newspaper film sections with a movie called The Set-Up. Director Robert Wise had something to prove since his editing of Citizen Kane had catapulted him to the directors’ ranks (before he would win two Best Picture and two Best Director Oscars in the 1960s), and The Set-Up remains the only case of feature-length real-time joined to the intuitively appropriate, claustrophobic, fear-riddled atmosphere of film noir. The old story of an old boxer’s one last chance is invigorated by the real-time tension and a surprisingly sensitive performance by the Adonis-like Robert Ryan in the lead. Wise explains on the DVD commentary that he filmed the street-clock seen in the opening and closing shots set to several intervals (a few minutes apart), so that however he had to edit it, the movie would play out in real-time. At a tough, lean 72 minutes, The Set-Up is one heavyweight with no extra paunch; on the commentary, Wise and Martin Scorsese both marvel at the way the near-music-free real-time heightens the street-grime. Though Wise uses a few clever dollying cameras, on the commentary he criticizes directors who force audiences to notice the camera. Presumably, as we’ll see, Wise wouldn’t have been thrilled with all of his real-time descendants.

A funny thing happened on the way to more real-time films…called television. 1949 marked the television’s first appearance in a Sears-Roebuck catalog; by 1950, ABC, NBC and CBS were producing variety shows, like those hosted by Milton Berle and Ed Sullivan, as well as original programming performed live like Man Against Crime (starring Ralph Bellamy), The Trap, The Clock, Sure As Fate, and Lights Out (starring Leslie Nielsen). Videotape was still a couple of years off, which is both why we can’t see them today and why shows had to be performed live. Though these fiction shows were of course interrupted by commercials, they nevertheless made the “real-time” aesthetic seem part of television, at the exact moment when Hollywood wanted most to distinguish itself from television.

In early 1952, along came a film that split the difference: real-time, as real as neo-realism, yet with the sort of outdoor panoramas and pulse-pounding action that neither TV nor neo-realists had done. High Noon is about a middle-aged sheriff, Will Kane, marrying and retiring while a bandit he long ago incarcerated is returning with a gang to enact revenge on Kane’s town. Howard Suber wrote, “The running time of the story almost exactly parallels the running time of the film itself…High Noon is one of the few American films that follow the classic Aristotelian principle of the ‘classic unities’ of time, place and action.” Actually, Phillip Drummond carefully explains that the film isn’t quite real-time – some of its pictured clocks are off, including the one behind Will and Amy during their wedding, showing 10:35, in an 80-minute film that certainly gets past the stroke of noon – but that’s mere quibbling, because director Fred Zinnemann achieves the time-TIME-TIME! effect all the same. High Noon both inaugurated and perfected cinema’s version of the well-worn radio-play device of saying something like: “you have an hour until something terrible happens” and then underlining that tension by forcing the audience to live through that hour. (One reason 007 films never seemed all that realistic is that they would show us two minutes left on the bomb, but the rest of the film’s editing made us doubt how many of those 120 seconds would be elided or stretched. The first Die Hard [1988] did slightly better – John McClane names Gary Cooper over John Wayne to attempt to invoke real-time tension.) Drummond’s book on High Noon walks readers through the film’s remarkably realized artistry, including Gary Cooper’s career-defining performance…and the drama, not unlike that confronting the more daring TV producers, behind screenwriter Carl Foreman’s blacklisting because of the film’s script, which supposedly advocated accommodating North Korea during the heart of the McCarthy era. Oddly, to contemporary audiences, Will Kane’s lone vigilantism seems right-wing.

The Academy Awards seemed to understand the nuance, and prepared to hand High Noon the Best Picture Oscar (Variety wrote “High Noon was a cinch to win”) when, to the crowd’s astonishment, the envelope revealed High Noon had lost to The Greatest Show on Earth. (According to Oscar lore, NBC was so taken aback that the cameraman failed to find the winning film’s producer until he was onstage. No tears for director Fred Zinnemann: he won Best Director the following year for the Best Picture-winning epic From Here to Eternity.) Putting TV aside, if Hollywood produced a total of about 5000 westerns (Suber’s estimate), we might say that a strain of them got sparer and more stripped-down leading up to High Noon…perhaps even culminating in the stark, all-too-real church scene at the heart of High Noon. From there, westerns – and most other films – could only add more details. How could anyone improve on High Noon? Instead Hollywood produced a few knock-offs, like 3:10 to Yuma, whose real-time begins in its second half. The wave of realism (and what historians call the “social problem” film) crescendoed even in Italy, and Hollywood would combat the television aesthetic with big-scale productions that certainly encompassed more than 2 hours of their characters’ lives.

Here’s the second part.…and here it is on the blog:


Throughout the 1950s, Hollywood’s relationship with television was fraught: TV was a hated rival but also a source of cheap talent and material, as in the case of the small-scale Marty (1955), which won the Best Picture Oscar. These contradictions were well represented by the apparently “televisual” 12 Angry Men (1957), which began life as a teleplay concerning a jury with a lone holdout who must, and eventually does, convince his fellow jurors of the defendant’s innocence. Its writer, Reginald Rose, persuaded one of Hollywood’s biggest stars, Henry Fonda, to become a first-time producer of the film version. Fonda and Rose took basement-low salaries in favor of future points, and hired a TV director, Sidney Lumet, for next to nothing because Lumet wanted a first feature credit. Technically, there’s an opening bit on the courtroom steps that keeps this from being a true real-time film, but considering the subject matter, the remaining 90 minutes play out in a remarkably flashback-free, tautly paced real-time. Lumet had learned well from live TV; he prepared 387 camera set-ups for the film, half of which he used in the film’s final half-hour, to tighten the screws. Upon its release, 12 Angry Men seemed a case of low budget equating to low returns – the $350,000 film didn’t even earn back that much, and Fonda would never produce again. Yet critics liked it, and 12 Angry Men managed Oscar nominations for Picture and Script – both of which it lost to the epic Bridge on the River Kwai. Esteem for the film has grown since its nominations, and it now seems remarkable that it’s the only film in the Imdb top 15 to have been made before 1972. (Too bad, Citizen Kane, Casablanca, and The Wizard of Oz!)

Lumet wrote in Making Movies, “In the late fifties, walking down the Champs Élysées, I saw in neon a sign over a theater: Douze Hommes en Colère – un Film de Sidney Lumet…fortunately for my psyche and my career, I’ve never believed it was un Film de Sidney Lumet.” Lumet knew his place in the pecking order, knew that TV was his bread and butter at the time (he directed 26 TV movies and episodes after 12 Angry Men)…and he also knew he had higher ambitions, but those would not include compromising artistic material. If a Tennessee Williams play like Orpheus Descending needed to cover days and locations, then Lumet could direct it that way, as when he turned that play into the excellent The Fugitive Kind (1960) starring Marlon Brando. If a story was best served by presenting it in real-time, then Lumet would do that, even if the movie starred, say, Katharine Hepburn. Long Day’s Journey into Night (1962) is about an older wife and husband, cripplingly dependent on drugs and alcohol (respectively), and their interactions with their two adult sons. True to its title, at 174 minutes, it’s the longest movie that could ever be considered a real-time film, and that’s no doubt a large part of why critics don’t remember it well when drawing up their all-time top 100 lists. Lumet altered Eugene O’Neill’s great play for the sake of real-time; though the opening scene is written to take place in the house at breakfast, Lumet sets it outside around a table at an indeterminate time (no one says “breakfast”), to suggest characters falling into old patterns as they return to the house…and to suggest more claustrophobia. Lumet also snips away at any dialogue that suggests that this is a “reunion,” furthering the implication that the sons may have to live this torturous day again and again and again. With towering performances by Hepburn and Ralph Richardson as the parents, Long Day’s Journey into Night is a harrowing masterpiece, even if it’s not quite real-time.

While Lumet was working on Long Day’s, French filmmakers were putting in their own long days taking to heart the signs on the Champs Élysées – they did think of their films as personal expressions, they had a lot to say, and they were determined to exploit the freedom of new light-weight cameras. One of their best films is Cleo From 5 to 7 (1962), as true a real-time film as there is, and probably the most feminine version of the form. The young, attractive Cleo has a problem: her doctor feels she may have cancer, and she wanders around Paris for exactly 90 minutes waiting for the results. As Steven Ungar’s book shows, director Agnès Varda didn’t consider herself part of what was already anointed the nouvelle vague, though the feeling wasn’t mutual. The film seizes on the New Wave’s focus on alienated youth, and even its foregrounding of style: titles tell us what the clocks say, precisely calibrated to the film’s running time. A more appropriate, yet less catchy, title would have been Cleo From 5 to 6:30. The femininity of the narrative is only partly because it’s Cleo’s story and she’s in almost every shot…it also has to do with Cleo’s perspicacity, with the absence of familiar melodramatic situations. Middle-class alienation has rarely been given such supple, unforced form. This is just Cleo reflecting on life – granted, wondering if she’ll be diagnosed with cancer – and that’s enough. (Okay, there are minor narrative-driven sequences, and a few men worth talking to.) Here, the real-time is almost a rebuke to other versions of real-time, though entirely appropriate on its own terms.

Last Year at Marienbad (1961), winner of the Golden Lion at the Venice Film Festival and condemned to forever typify the word “artsy,” may be considered a real-time film…or a real-time nightmare. The plot, if you can call it that, concerns a man trying to convince a woman that they met before, and fell in love before, at the same palatial chateau where they now attend some sort of party. Marienbad occasions an exploration of memory and identity even while avant-garde devices dare viewers to interpret the film in various ways. Director Alain Resnais had already fluidly compressed and stretched time in Night and Fog (1955) and Hiroshima Mon Amour (1959), but this time he allied himself with writer Alain Robbe-Grillet (who co-directed, according to many), known for the nouveau roman, a kind of book where the plot is implied rather than stated. Sure enough, Marienbad is elliptical to the point of frustration (for many), its opening chime heard at the finale, its partygoers apparently trapped in a perpetual present. Plentiful flashbacks almost take this film out of real-time consideration, except that the flashbacks are by definition unreliable. Robbe-Grillet said that the “entire story of Marienbad happens neither in two years nor in three days, but exactly in one hour and a half.” One might say the same of any 90-minute film, but that comment resonates differently after you’ve watched an elaborate courting ritual that travels down every imaginable corridor…of memory.

In a sense, Resnais and Robbe-Grillet were playing in, and taking to extremes, the surrealist landscape that Luis Buñuel had so assiduously established in films like Un Chien Andalou (1929) and L’Age d’Or (1930). Buñuel, then in Mexico, took Marienbad, aspects of his own career, and the real-time subgenre to their natural conclusion when he made The Exterminating Angel. In Buñuel’s words, it’s “the story of a group of friends who have dinner together after seeing a play, but when they go into the living room after dinner, they find that for some inexplicable reason they can’t leave.” Perhaps it’s not quite real-time, as Marsha Kinder points out: “we spectators are constantly confronted with continuity errors—repetitions, inconsistencies, contradictions—which can be missed if we focus too exclusively on narrative drive.” Yet this feels within the bounds of establishing the rules so that you can break them…something we desperately wish the characters would do, if they weren’t over-committed to mysterious, inexplicable, bourgeois codes, not unlike…finish this sentence how you like. It’s hard not to think of the film in terms of Buñuel’s brilliant, but interrupted, career: after hostility from Europe and the U.S., he was “in exile” (sort of) in Mexico from 1946 to 1962, during which time producers would (mostly) pay him to make mere melodramas, until the wicked satire Viridiana (1961) became an international smash, allowing him to return to a France that, with the New Wave in full bloom, was obviously ready for him. Kinder calls The Exterminating Angel a “pivot” to Europe as well as Buñuel’s way of proving what he could do in Mexico when granted full artistic control. It may be the best film made in Mexico in the 20th century.

Real-time was just one of many aesthetics that the New Wave played around with; Jean-Luc Godard popularized jump cuts, which are pretty much antithetical to real-time, and yet Godard became famous because of Breathless (1960), a film that, twenty minutes into a lively narrative, stops itself to spend the next twenty minutes in an apartment with lovers in casual real-time conversation. Working in America and typically with much less money, Andy Warhol was well aware of what the nouvelle vague had done…and hadn’t done. Warhol knew that the en vogue French filmmakers purported to combine a stripped-down, almost documentary style with a sort of stylish insouciance, and in many ways Warhol did them one better, with even less regard for what might be called conventional narrative. Kiss, Haircut, Blow Job, Eat, and Henry Geldzahler, all made in 1963 and 1964, were short, silent, black-and-white films that purported to present quotidian events in real-time, unadorned by anything that might be considered distraction, including editing, camerawork, or acting (some question that last one). The films are proudly strange, mostly joyful only for the pleasure one feels at seeing something that doesn’t fit into any previous categories. Perhaps they expanded some notions of cinema.

However, this is an article about real-time feature films, and the real reason to mention Andy Warhol is his two most famous non-shorts, Sleep (1963) and Empire (1964). In the summer of 1963, Warhol told a friend that he wanted to film a person sleeping for eight hours; Warhol’s problem was that his only camera was a silent 16mm Bolex that could only film for four minutes at a time. Warhol made the film anyway, and perhaps ran out of patience (or money) after about four hours of filming; Warhol asked that Sleep be projected at two-thirds its normal speed in exhibitions, making it seem that his friend John Giorno was sleeping for a bit less than six hours. Sleep is its own subversive meditation, forcing us to question why we tend to prioritize being awake. After saving up a bit of money and borrowing a 16mm Auricon that could film for 33 minutes at a time, Warhol executed an experiment that even today retains its stature as one of cinema’s most audacious: on July 25, 1964, Warhol filmed the Empire State Building beginning at 8:06pm and finishing at 2:42am that morning. Again, Warhol asked that his film be projected at 16 frames per second, and so the landmark film is a 12-hour film of a landmark. When asked about the point of Empire, Warhol said it was to “see time go by.” Indeed, Empire may stand as Deleuze’s time-image incarnate, a constant, all-too-real unfolding of a barely inhabited mass construction. There’s something oddly reassuring about the very existence of a 12-hour film that we know, that we can accurately describe, without even seeing it. The longest amounts of time can be somehow captured, like sarcophaguses.

The reassuring placidness of Empire stands in stark contrast to 1964’s two other famous real-time experiments. (One of them ends with a fateful encounter with the Empire State Building – coincidence?) Around the time of the 1962 Cuban Missile Crisis, Sidney Lumet decided to adopt Eugene Burdick and Harvey Wheeler’s 1962 novel Fail-Safe, and even persuaded his once-burned-twice-shy star of 12 Angry Men, Henry Fonda, to come along for a “featured” role as the President. Fail-Safe (1964) is about a technical malfunction that sends nuclear-missile-equipped U.S. Air Force jets to the U.S.S.R., forcing leaders of both countries to work to avoid a worldwide apocalypse. Ari Schulman recently wrote in Slate, “the crisis’s real cause is the logic of the nuclear system at every level—its institutions, structures, procedures, and rationales. This isn’t a movie about why we should fear machines or the people who control them. It’s about how managerial systems can bring about just the things they’re designed to avert.” Normally, parallel action feels like kind of a fun way to be in two places at the same time; here, the real-time aesthetic makes each location change remind us that a terrible reckoning is that much closer. Lumet makes an unusual choice early on, filling up the screen with titles that say “5:30AM: NEW YORK CITY” and “5:30AM: WASHINGTON D.C.” One could say, then, that this film isn’t real real-time – it can’t be 5:30 in New York and 5:30 five minutes later in D.C. But Lumet never brings up any other titles or clocks, leaving us with the discomfiting feeling of a rather consequential timer having begun…and unlike in High Noon, we can’t know when time is up. The real-time aesthetic only adds to the horror-laced drama, and critics and audiences hailed it as the best real-time film ever…right?

No, they didn’t, and that’s largely because of Dr. Strangelove: Or How I Learned to Stop Worrying and Love the Bomb (1964), its own almost-real-time version of many of the same plot events, played as scathing satire. (Thomas Allen Nelson wrote, “The film has a running time of 94 minutes that closely approximates the fictional time it covers.”) Kubrick had purchased the film rights to Peter George’s 1958 novel Red Alert before the Fail-Safe novel was even written, and Kubrick threatened to sue to keep the obviously similar film from being made. That failed, but Kubrick managed to keep Fail-Safe out of theaters until Dr. Strangelove had finished its run, and that was good enough for government work: arriving second, Lumet’s melodrama seemed wooden compared to Kubrick’s high-spirited black comedy. (“You can’t fight in here! This is the war room!” is one of Strangelove’s classic lines, and one that could almost be applied to the Lumet-Kubrick standoff.) Dr. Strangelove is by any standards an all-time classic, with an archly cynical tone that no movie has ever really replicated. Perhaps that’s because no other darkly humorous fable has ever been told in real-time. In any event, Hollywood and the public loved Dr. Strangelove, and lavished it with awards and nominations.

Perhaps the timing was circumstantial, but after three somewhat underappreciated real-time masterworks, Lumet then gave up on the form for a while. In fact, he went the other way in terms of borrowing from the French New Wave, bringing flash cuts (under-one-second shots intruding on the main action, like unwanted memories) to his next film, The Pawnbroker (1965), and in turn to Hollywood, according to Matt Zoller Seitz.

By 1965, then, the real-time film had refreshed, invigorated and emboldened the western, the whodunit, the film noir, the legal thriller, the family melodrama, the war drama, the black comedy, the existential exploration, the avant-garde movie, and perhaps the horror film (depending how one reads Buñuel). After all that, why work to make a real-time film when it had already covered those bases? To put it another way, what was left? Well, for one, science fiction. Fantastic Voyage (1966) is about a group of scientists who must shrink to microscopic size, enter a scientist’s body and remove a deadly blood clot…within an hour, or they’ll grow back to normal size. Fantastic Voyage is every inch a studio production, with none of the artistic idiosyncracies I’ve been discussing, yet with the possible exception of a few opening scenes, it’s real-time all the way, even using the device to the point of potential incoherence – the film ends during the moment that the scientists grow back to full size, without actually telling the audience if their life-saving mission was successful. That’s not the film’s only flaw, yet the clunky dialogue tends to be forgiven because of the sheer imaginative beauty seen while navigating the human bloodstream. Fantastic Voyage absolutely remains a fabulous spectacle, and the real-time style makes it essential to see at least once.

Then, the real-time film vanished like men under a shrink-ray. After Fantastic Voyage, there would not be another artistic attempt at real-time for 10 years. One is left with the impression that the first-time-handheld-camera early 1960s really was a time of rare experimentation, and that by comparison, the great films of the late 1960s and early 1970s…wanted to paint on wider canvases.