Kristin here—

Last week during the Society for Cognitive Studies of the Moving Image conference here in Madison, I got to talking with Prof. Birger Langkjær of the University of Copenhagen. He asked me some questions about the concept of “turning points” in film narrative as I had used it in my book, Storytelling in the New Hollywood: Understanding Classical Narrative Technique. Specifically he wondered if turning points invariably involve changes of which the protagonist or protagonists are aware. The protagonist’s goals are usually what shape the plot, so can one have a turning point without him or her knowing about it?

I couldn’t really give a definitive answer on the spot, partly because it’s a complex subject and partly because I finished the book a decade ago (for publication in 1999). It seemed worth going back and trying to categorize the turning points in films I analyzed. Describing those turning points more specifically could be useful in itself, and it might help determine to what extent a protagonist’s knowledge of what causes those turning points typically forms a crucial component of them.

Characteristics of turning points

Most screenplay manuals treat turning points as the major events or changes that mark the end of an “act” of a movie. Syd Field, perhaps the most influential of all how-to manual authors, declared that all films, not just classical ones, have three acts. In a two-hour film, the first act will be about 30 minutes long, the second 60 minutes, and the third 30 minutes. The illustration at the top shows a graphic depiction of his model, which includes a midpoint, though Field doesn’t consider that midpoint to be a turning point.

I argued against this model in Storytelling, suggesting that upon analysis, most Hollywood films in fact have four large-scale parts of roughly equal length. The “three-act structure” has become so ingrained in thinking about film narratives that my claim is somewhat controversial. What has been overlooked is that I’m not claiming that all films have four acts. Rather, my claim is that in classical films large-scale parts tend to fall within the same average length range, roughly 25 to 35 minutes. If a film is two and a half hours rather than two hours, it will tend to have five parts, if three hours long, then six, and so on. And it’s not that I think films must have this structure. From observation, I think they usually do. Apparently filmmakers figured out early on, back in the mid-1910s when features were becoming standard, that the action should optimally run for at most about half an hour without some really major change occurring.

Field originally called these changes “plot points,” and he defined them as “an incident, or event, that hooks into the story and spins it around into another direction.” Perhaps because of that shift in direction, these moments have come more commonly to be called “turning points.” But what are they? Field’s definition is pretty vague.

In Storytelling I wrote, “I am assuming that the turning points almost invariably relate to the characters’ goals. A turning point may occur when a protagonist’s goal jells and he or she articulates it .… Or a turning point may come when one goal is achieved and another replaces it.” I also assumed that a major new premise often leads to a goal change (p. 29). “Almost invariably” because I don’t assume that there are hard and fast rules. As with large-scale parts, my claims about other classical narrative guidelines aren’t prescriptive in the way that screenplay manuals usually are.

To reiterate a few other things I said about turning points. They are not always the same as the moments of highest drama. Using Jaws as my example, I suggest that the moments of decision (not to hire Quint to kill the shark and later to hire him after all), rather than the big action scenes of the shark attacks, shape the causal chain (p. 33).

Not all turning points come exactly at the end of a large-scale part of the film (an “act” in most screenplay manuals). A turning point might come shortly before the end of a part, or the turning point may come at the beginning of the new part (pp. 29-30). The final turning point that leads into the climax comes when “all the premises regarding the goals and the lines of action have been introduced” (p. 29).

Most screenplay manuals consider goals to be static. To me, “Shifting or evolving goals are in fact the norm, at least in well-executed classical films” (p. 52). This doesn’t mean that the goal changes at every turning point. Instead, the end of a large-scale part may lead to a continuation of the goal(s) but with a distinct change of tactics (p. 28).

One big advantage of talking about different types of turning points is that it allows the analyst to see how the different large-scale parts function. A well-done classical film doesn’t just have exposition and a climax with a bunch of stuff in the middle. (Field calls that long second act the “Confrontation” in the diagram above.) I believe that once the setup is over, there is a stretch of “complicating action,” which often acts as a sort of second setup, creating a whole new situation that follows from the first turning point. The third part is the “development,” which often consists of a series of delays and obstacles that essentially function to keep the complications from continuing to pile up until the whole plot becomes too convoluted. The development also serves to keep the climax from starting too soon. The third turning point is the last major premise or piece of information that needs to be set in place before the action can start moving toward its resolution in the climax.

David uses this approach to large-scale parts in his online essay, “Anatomy of the Action Picture,” which discusses Mission: Impossible III.

TYPES OF TURNING POINTS

Returning to some of the examples I used in Storytelling, let’s see what sorts of changes their turning points involve. In most of these cases, the protagonist is aware of what is happening, but there are some exceptions or nuances.

1. An accomplishment, later to be reversed

Top Hat: end of setup, Jerry and Dale fall in love, but she soon conceives the mistaken idea that he is married to her best friend. (p. 28)

Tootsie: end of setup, Michael gets a job, but the results will throw his life increasingly into chaos. (p. 60)

Parenthood: end of complicating action, the parents seem to be making progress in solving their problems. (p. 268)

2. Apparent failure, reiteration of goal

The Miracle Worker: end of development, parents remove Helen from cabin, Anne states goal again. (p. 28)

The Silence of the Lambs: turning point comes at beginning of development, Chilton makes Lecter a counter-offer, removing him from the FBI’s charge. The FBI’s tactic has failed, but soon Clarice visits Lecter to pursue her questioning. (p. 123)

Here’s a case where we don’t see Clarice learning about Chilton’s treacherous undermining of the FBI’s efforts. A few scenes later she simply shows up to visit Lecter, and we realize that she has not given up her goal of getting information from him. Clearly a turning point can occur without the main character’s knowledge, but he or she will usually learn about it shortly thereafter.

Groundhog Day: end of development. Failure to save old man. No reiteration of goal, which is implicitly that Phil will continue to improve himself. (p. 147)

2a. Failure, new goal

Amadeus: end of complicating action, Salieri declares that he is now God’s enemy and will ruin Mozart. (p. 195)

Amadeus is an example of what I call a “parallel protagonist” film. Here Salieri is aware of his own decision, but Mozart never learns that his colleague hates him so. Parallel protagonists have separate goals, but they need not be aware of each other’s goals. The same would be true in a film with multiple protagonists, to the extent that they have separate goals.

2b. Failure, lack of goal

Groundhog Day: end of complicating action, suicidal despair. (p. 144)

Amadeus: end of setup, Salieri humiliated by Mozart, conceives strong resentment but no specific goal. (p. 191)

Parenthood: end of development, the parents are all resigned to their failures. (pp. 275-76)

3. Major new premise, reiteration of strategy

Witness: end of complicating action, Carter tells Book to stay hidden. (p. 29)

The Wrong Man: end of complicating action, Manny is freed on bail; he has the chance to try and prove his innocence. (p. 39)

Terminator 2: end of development, Sarah, Terminator, and John steal chip from Cyberdine, continue in their attempt to destroy it. (p. 42)

Amadeus: end of second development, Constanza leaves Mozart, allowing Salieri the access that will permit him to fulfill his goal of killing his rival. (p. 205)

4. Protagonist/Important character makes a decision, then changes or modifies goal

Little Shop of Horrors (1986): end of complicating action, Seymour agrees to kill someone to feed blood to Audrey II. (p. 29)

The Godfather: end of development, Michael says family will go legit, asks Kay to marry him. (p. 30)

Casablanca: end of complicating action, Rick rejects Ilsa’s account of her relationship with Victor. (p. 32)

The Producers: end of setup, Leo decides to join Max in committing fraud. (p. 38)

Tootsie: end of complicating action, Michael decides to start improvising Dorothy’s lines, setting up “her” success. (p. 64)

Back to the Future: end of development. Martie decides to leave a message for Doc warning him about the Libyan attack; the result is to save Doc from being killed. (p. 94)

Desperately Seeking Susan: end of setup, Roberta apparently decides to pursue Susan, change her boring life. (p. 166)

Such decisions obviously form the basis for turning points fairly frequently, and by definition the protagonist is aware of what is happening.

5. Major Revelation, new goal or move into climax

Witness: end of setup, Book is told the killer is a cop, changes tactics, flees. (p. 28)

Witness: end of development, Book learns partner has been killed, realizes he must act. (p. 29)

The Bodyguard: end of development, revelation of Nikki as villain, attack on house, death of Nikki. (p. 31)

Terminator 2: end of setup, John discovers the Terminator must obey him, sets goals of protection without killing and of rescuing Sarah from the asylum. (p. 41)

Terminator 2: end of complicating action, Sarah realizes that the war can be prevented, conceives goal of killing Dyson. (p. 42)

Tootsie: end of development, Michael’s agent tells him he must get out of playing Dorothy without being charged with fraud. (p. 69)

The Silence of the Lambs: end of setup, Clarice finds the body in the warehouse and realizes that Lecter is willing to help her. (p. 115)

The Silence of the Lambs: end of development, Clarice and Ardelia realize “He knew her,” the clue that will lead to the solution. (p. 126)

Alien: end of development, revelation that the “company” and Ash are prepared to sacrifice the crew; will lead to decision by survivors to abandon the ship and save themselves. (p. 299)

Such revelations usually involve the protagonist knowing what is happening, but I believe there are exceptional cases where the viewer learns something that the protagonist does not. This kind of revelation often involves either villains turning out to be allies or apparent allies turning out to be villains.

Take the case of Demolition Man. At 54 minutes into the film, about nine minutes before the end of the complicating action, the audience learns that the apparently benevolent Dr. Cocteau (Nigel Hawthorne), who has developed the new pacifist society of Los Angeles, is a ruthless villain. This moment isn’t the turning point that ends the complicating action; I take that to be the escape of the “Scraps,” the rebels who oppose Cocteau, from their underground prison. Most of those nine minutes consist of the hero, John (Sylvester Stallone) meeting Cocteau and going out to dinner with him. John shows signs of strongly disliking the new society, and he refuses to kill the rebels when they escape. Still, there is no sign that he considers Cocteau a villain. Rather, John thinks of himself as a misfit and expresses a desire to leave the new Los Angeles. The turning point that ends the complication depends on his assumption that Cocteau on his side, even if John considers the leader’s bland society to be unattractive.

The development portion is full of typical delays: the scene of John visiting Lenina Huxley’s (Sandra Bullock) apartment, where she invites him to have virtual sex, and the following scene of John exploring the odd, modernistic apartment assigned to him. At 74 minutes in, or about six minutes before the end of the development, John begins to learn of Cocteau’s true nature. He confronts Cocteau in the latter’s office and nearly shoots him. The scene where John tells the members of the police department who remain loyal to him that they will invade the underground prison to capture Cocteau’s violent agent, Simon Phoenix (Wesley Snipes) marks the end of the development, with a new, specific goal determining the action of the climax.

Here is a case where for a significant portion of the narrative the protagonist remains ignorant of something that the audience has been shown. Even here, however, John displays a dislike of Cocteau and what he has done to Los Angeles society. Classical films seem disinclined to show their heroes as thoroughly deluded. John’s underlying decency and common sense make him ready to distrust a leader whom everyone else, including Lenina, admire.

6. Enough information accumulates to cause the formulation of goals

Back to the Future: end of complicating action, goals of synchronizing with lightning storm, getting Marty’s father and mother together at dance. (p. 90)

Desperately Seeking Susan: end of complicating action. Information about gangsters and growing attraction to Dez lead to convergence of Roberta’s goals. (p. 170)

Parenthood: end of setup, in this case with multiple goals formed for four separate plotlines.

7. A disaster, accidental or deliberate, changes the characters’ situations/goals

The Wrong Man: end of setup, Manny is mistakenly arrested. (p. 39)

Jurassic Park: end of complicating action, Nedry shuts down the electricity grid of the park, letting the dinosaurs loose. (p. 32)

Jaws: end of complicating action; the attack involving his son makes Brody force the mayor to accept his original, thwarted tactic of hiring Quint. (p. 34)

Alien: end of setup, facehugger attaches itself to Kane. The goal conceived shortly thereafter is to investigate it and save him. (p. 292)

Alien: end of complicating action, alien bursts from Kane’s stomach; goal becomes to save the crew and ship. (p. 295)

Desperately Seeking Susan: end of development, Roberta’s arrest leads to a low point, and she opts to turn to Dez for help. (p. 172)

Amadeus: end of first development, death of Leopold, which Salieri later exploits in his plot against Mozart. (p. 201)

The Hunt for Red October: end of complicating action, Ramius realizes he has a saboteur aboard and needs Ryan’s help; Ryan starts planning actively to help him. (p. 233)

8. Protagonist’s tactics are blocked or he/she is forced to use the wrong tactics

Jaws: end of setup, Brody’s desire to hire Quint rejected. (p. 23)

Back to the Future: end of setup, Doc sets time machine’s date; after new large-scale section begins, attack by Libyans accidentally sends Marty back to 1955. (p. 85)

Groundhog Day: end of setup, Phil trapped in repeating time, goal of becoming a network weather forecaster destroyed. He soon opts for irresponsible self-indulgence. (p. 139)

9. Characters working at cross-purposes resolve their differences

Jaws: end of development, the three main characters bond, allowing them to cooperate to kill the shark. (p. 35)

The Hunt for Red October: end of development, Ramius and Ryan make contact, with Captain Mancuso’s help, they start working together. (p. 237)

10. A supernatural premise determines a character’s behavior

Liar, Liar: end of setup, Max wishes that his father would have a day where he is unable to tell a lie. Also end of complicating action, where Max fails to cancel the wish. (p. 38)

11. The protagonist/major character succeeds in one goal, allowing him/her to pursue another

Liar, Liar: end of development, Fletcher wins his court case, freeing him to try and regain his wife and son. (p. 39)

The Hunt for Red October: end of setup, Ramius learns that the silent feature of his new submarine works; he apparently intends some sort of attack on the US but is in fact plotting an elaborate ruse to defect. (p. 225)

These examples suggest that protagonists usually do know about the major events that form turning points, or they learn about them shortly after they occur. The main exception would seem to be when there are two major parallel characters, one of whom is to some extent villainous, as in the case of Amadeus, where one of the characters is duped. This reminds me of The Producers, where Max and Leo leave the theater early, convinced that they have succeeded in their goal of creating a box-office disaster. Lingering in the theater, the narration allows us to see the transition from audience disgust to fascination. Only after the audience comes into the bar where the two partners are celebrating do they realize what has happened (“I never in a million years thought I’d ever love a show called Springtime for Hitler!”) To understand the irony and/or humor in such situations, the film spectator must at least briefly know more than the main characters do.

The examples also confirm that character goals seldom endure unchanged across the length of a film. Revelations, decisions, disasters, supernatural events, and presumably others kinds of causes frequently cause major shifts in characters’ goals or at least in the tactics they employ in pursuing them. Even in what seems like a fairly straightforward thriller like Alien, the crew’s assumptions that they must loyally protect their spaceship are completely reversed at the end of the development, and the narrative turns into an attack on corporate greed and ruthlessness. Whatever one thinks of classical Hollywood films, they are usually more complex than the three-act model allows for.

Note: My title comes from Robert Southwell’s poem of the same name.

[July 11: Jim Emerson has made some insightful comments on this entry on his Scanners blog.]

Grover Crisp and Lea Jacobs outside the University of Wisconsin–Madison Cinematheque.

DB here, again:

He holds sway over thousands of movies and the transfer of hundreds of DVDs. At home he has a high-definition set, but he gets local channels on a rabbit-ears antenna. If that isn’t a working definition of a Film Person, I don’t know what is.

Over the years our UW Film Studies area has hosted many visiting Film Persons, including archivists Chris Horak, Mike Pogorzelski, Joe Lindner, Schawn Belston, and Paolo Cherchi Usai. (1) Being a department centrally interested in film history, we’re eager to screen recent restorations and learn the ins and outs of film archivery.

This year our Cinematheque screened several Sony/ Columbia restorations, running the gamut from Anatomy of a Murder to The Burglar. We wound up our series, and our season, with a stunning print of Frank Capra’s American Madness (1932). To conclude things with a bang, Lea Jacobs, Karin Kolb, and Jeff Smith brought Grover Crisp, Sony’s Senior Vice President of Asset Management, Film Restoration, and Digital Mastering.

Kristin and I had met Grover at the Cinema Ritrovato festival in Bologna, and after hearing him introduce some restorations, we knew he had to come to Madison. It was a Bologna screening of American Madness that convinced me that we had to show this sparkling print on our campus. Fortunately for us, Grover squeezed a Madison visit into his schedule and suffered the caprices of American Airlines’ delays. We had about two days of his company.

It was a great learning experience. On Thursday 8 May he addressed our colloquium, where he showed two shorts: a 1933 Screen Snapshots installment explaining the process of motion picture production, from script to final product, and a trim 75-second short by his colleague Michael Friend tracing the evolution of filmmaking technology. On Friday Grover introduced American Madness and took questions.

Just a note about terminology: Archives engage in both film preservation and film restoration. Preservation means, as you’d think, saving films for the future. This involves maintaining the materials (prints, negatives, camera material, sound tracks) in surroundings that minimize harm and deterioration. Restoration demands more resources. It involves trying to create an authoritative version of the film as it existed in some earlier point in history. But that’s a complicated matter, as we’ll see.

Archives and assets

Grover brings a Film Person’s sensitivity to the mission of film preservation. What does that mean? For one thing, he takes the long perspective. Film archivists think about how to store and maintain films for decades, even hundreds of years. In an industry geared to short-term cycles, booms and busts and jagged demand curves, a commercial archivist like Grover, or Schawn of Fox, has to reconcile traditional museum standards with market demands, strategic release considerations, and current studio policies. In short, he has to think both about saving the film for a century and about hitting a DVD street date.

Grover explained his obligations at Sony Pictures Entertainment. He manages all film and television materials. That means preserving everything for any future use and restoring some items for video and theatrical screening. It also means that he keeps abreast of current TV and film productions. This gives him a chance to “embed archival needs” into production and postproduction processes. For example, from 1991 to the present, YCM separation masters are prepared for every film. (2) And today, when a film is finished and output digitally onto a negative for release prints, another negative is assembled from the camera negative, the most pristine source, and that goes into cold storage.

Grover supervises a staff of about twenty-five people, six of whom are overseeing film laboratory and audio work full-time. Unlike other studio archives, Sony does not hand off preservation tasks to outside companies. Being a Film Person, Grover asks that for films a new negative be created, and fresh screening prints be prepared.

What impelled a big company like Sony to invest so much in preservation and restoration? Grover explained that the rise of home video taught most studios that their libraries had some value. In particular, executives were impressed by Ted Turner’s 1986 purchase of MGM for its library, which he could recycle endlessly on TBS. They realized that the libraries were genuine assets. Old movies enhanced the firm’s value (if it were to be sold, as many were in those days) and could be exploited through new media platforms, like cable.

The arrival of DVD made all the studios return to their libraries to make high-quality versions of their top titles. Now the high-definition format of Blu-ray will re-start restoration because of the leap in quality it offers.

Sony moved early, partly because it understood new media. When Sony bought Columbia Pictures in 1989, the Japanese executives envisioned a synergy between Sony’s hardware and a movie studio’s software. “One of the first things they asked,” Grover recalls, “was, ‘What’s the condition of the library?’” Soon afterward, Sony set up the first systematic collaboration between studio archivists and museum-based archivists. The Sony Pictures Film Preservation Committee included archivists from Eastman House, the Library of Congress, UCLA, MoMA, the Academy, and other institutions. In addition, Grover modeled Sony’s restoration policies on the exemplary work of legendary UCLA archivist Bob Gitt (whose work on sound restoration we’ve already saluted in another entry).

Digital restoration has been part of the Sony program. The 1997 restoration of Capra’s Matinee Idol (1927), from a Cinémathèque Française print, was the first live-action high-definition restoration from any studio.

Sony holds between 3600 and 5000 titles. It’s impossible to be more precise at this point because in hundreds of cases the rights situation is uncertain. But of the core library, about half to two-thirds of the films are restored. Crisp’s team is at work on about 200 titles at any moment! Many more are restored than warrant DVD release, but they may be available On Demand and via TCM, which has recently acquired cablecast rights to several older Columbia titles.

Restore, yes! But what, and how?

We’re grateful for preservation, but for cinephiles, restoration holds a special thrill. A “restored” Intolerance or Napoleon or von Sternberg movie gets us salivating. This year Cannes ran nine restorations on its program.

But what is a restoration?

We expect that a restoration will provide a film with a better-quality image or soundtrack than we’ve had so far. We often expect as well that a restoration will provide a more complete version of the title. Yet there are problems with both expectations.

First, who’s to say what the quality of the original is? Grover encountered this difficulty in restoring Funny Girl. It was released in the classic Technicolor dye-transfer process, usually considered the gold standard for color cinematography. He assembled five original Technicolor prints and all of them looked different. It turns out that Technicolor prints varied quite a bit. (3) The kicker was that Grover couldn’t make the new dye-transfer 35mm print of Funny Girl match any of these reference prints. That didn’t stop it looking fabulous when it premiered in LA.

What about restoration that adds footage to an existing version of a movie? My previous blog entry mentioned an example, the wild and crazy Nerven (1919), to be issued on DVD by the Munich Film Museum. Stefan Drössler, Munich’s curator, has done a remarkable job of incorporating new footage into the longest version of Nerven to date. (The new material includes “living intertitles,” in which actors drape themselves around gigantic letters.) But Nerven remains incomplete, lacking about a third of its original material.

Moreover, it’s possible to “over-restore” a film. That is, by adding footage culled from many versions, the restorer may be creating an expanded version that nobody actually saw.

Original version? What does that mean? Paolo Cherchi Usai has reflected on this at length. (4) Is the original what the film was like on initial release? This is tricky nowadays for new titles, because as Grover pointed out, most films are “initially released” in several versions. Since Hollywood filmmaking began, different versions have been made for the domestic and the overseas markets, and often the overseas versions are longer. Sometimes a film opens locally or at a film festival and then is modified for release. Major films by Hou Hsiao-hsien and Wong Kar-wai have played Cannes in versions longer than circulated later.

Now suppose that the film was modified by producers or censors for its initial release somewhere. If you can determine the filmmaker’s wishes, should you restore the film to what she or he wanted it to be? That flouts the historical principle of getting back to what people actually saw. Consider the shape-shifting Blade Runner. Only now, twenty-five years after its release, has the U. S. theatrical version appeared on video.

Or what do we do when a filmmaker, after the release, decides that the film needs to be changed? Late in life, Henry James felt the urge to rewrite many of his novels, and the revised versions reflect his mature sense of what he wanted his oeuvre to be. Why can’t a filmmaker decide to recast an older film? Such was the case with Apocalypse Now Redux and Ashes of Time Redux. Since the later version represents the artist’s latest viewpoint, should that be the authoritative version?

Grover encountered a variant of this problem when he invited one cinematographer to help restore a film. The cinematographer’s own aesthetic and approach to his work had evolved over the years, so he started to tweak things in a way that was taking the look of the film to a level much different than the original achievement. Grover had to steer him back to respecting the film’s original look.

On the brighter side: Grover brought in master cinematographer Jack Cardiff to advise on the restoration of A Matter of Life and Death. Cardiff said that one scene needed a “more lemony” cast. Grover tried, but each time Cardiff said it wasn’t right. Finally, Grover confessed that he just couldn’t get that lemony look. Cardiff nodded. “Neither could I.”

Grover offers this nugget of wisdom: It is almost impossible to get an older film to look the way it does when it was originally released. The color will never look as it did, nor will sound sound the way it did. Film stocks have changed, printing processes have changed, technology in general has changed. Every version is an approximation, though some approximations may be better than others. Take consolation in the fact that even when the movie was in circulation, it may have already existed in multiple versions.

What do consumers want?

All archivists explain that compromises enter into the preservation process down the line. One of the revelations of Grover Crisp’s visit was the awareness of just how many of these compromises come into play, and how some of them are tied to what DVD buyers expect.

Many films from Hollywood’s studio era are soft, low-contrast, and grainy. Project a 1930s nitrate copy today, and though it will be gorgeous, it’s likely to look surprisingly unsharp. Things apparently didn’t improve when safety stock came along in the late 1940s; projectionists complained that those prints were even softer and harder to focus than nitrate ones. As recently as the 1970s, films were not as crisp as we’d like to think. I’ve examined the first two parts of The Godfather on IB Tech 35mm originals, and though they look sharp on a flatbed viewer, in projection the grains swarm across the screen like beetles. (I should add that many of today’s films also look mushy and grainy in the release prints I see at my local.)

But I think that DVD cultivated a taste for hard-edged images, perhaps in the way that music CDs cultivated a taste for brittle, vacuum-packed sound. It’s not surprising, then, that video aficionados are often startled when a DVD release of a classic movie looks far from clean. This is not always attributable to a film’s not being cleaned up to the fullest extent the technology allows. Some films are inherently grainy, gritty, or soft-looking. With the advent of Blu-ray and HD imagery for the home, those films’ look is often mistaken for problems with the transfer, or signs that the distributor didn’t care enough to spend the money to thoroughly make it new. “New” in this case means contemporary.

So there are compromises, according to Grover, that the studios are having to deal with:

What do do about a film that is really grainy? Do you remove the grain, reduce the grain, leave the grain alone? Everyone seems to have a different answer, and it often puts the distributor in a difficult situation. These are questions that will be answered over time, as the consumer gets used to seeing images in HD that truly represent the way a film looks.

Another compromise involves sound. “Now you’re getting to an ethical area,” Grover remarks. In restoring early sound movies, Sony removes only clicks, pops, and scratchy noises (while still keeping the original, faults and all, as a reference). Sound problems are compounded in films from the early 1950s. Many releases at that time were shot in a widescreen process but retained monaural sound. Yet avid DVD buyers want multitrack versions to feed their home theatres. “If it’s not 5.1, we get complaints.” So Grover’s engineers “upmix” mono, as well as two-channel stereo, to 5.1—though they strive to keep the 5.1 minimal and retain as much authenticity as possible. Not on every film, of course: often the original mono track is included on the DVD.

On the positive side, Sony held the original tracks for Tommy (1975), originally released in Quintaphonic. (Old-timers will remember that audiophiles were urged to upgrade to this, and some LPs were released in that format.) Grover was pleased that he could replicate the theatrical version’s 5-channel mix on the DVD. (More details here.)

Now for a hobby horse of mine. The 1950s-1960s standards for multichannel sound were not those of today’s theatres. Today’s filmmakers funnel important dialogue through the front central speaker. But in multi-track CinemaScope and other widescreen formats, dialogue was spread to the left and right speakers as well, and these were behind the screen. That meant that a character standing on screen left was heard from that spot as well. (Remember, these screens might be seventy feet across.) Long ago Kristin and I saw an original 70mm release print of Exodus (shot in Super Panavision) in a big roadshow house in Paris. The ping-pong effect of the conversations between Ralph Richardson and Eva Marie Saint was fascinating.

Today, however, it might be distracting. In a home theatre, where discrete tracks present sound from offscreen left or right, it would seem downright weird. Although Grover didn’t comment on this, it’s clear that the big-screen classics from the magnetic-track era are remixed for DVD to suit current theatrical and home-video standards. This makes it very difficult for researchers to study the aesthetics of sound design from that period. Even if you visit an archive, that institution almost certainly isn’t able to project the film in a multi-channel version. (5) Is it too late to ask DVD producers to replicate, on a second soundtrack, the original channel layout? This would be a big favor to the academic study of the history of sound, and some home-theatre enthusiasts might develop a taste for the old-fashioned sonic field.

Last questions

Q: Do Grover and his colleagues take notice of online chat about DVD releases?

A: Yes, because it’s important to know what many participants in those conversations want from a film’s release, and they may also know things (from a hardcore fan’s pespective) about a film that is useful to know. Grover’s staff members can learn about missing scenes and other variant prints. But sometimes the Net writers are working from incomplete information and can get things wrong. A fan fervently announced that one word was intelligible in the “original” version of a film but not on the DVD. Yet the theatrical release version muffled the word. It turns out that the word was audible in a remixed TV version, which the fan had used as reference. Likewise, one critic who found the color on the Man for All Seasons Special Edition DVD too vibrant was evidently using a VHS version as the reference point. (For what it’s worth, the color in the Man for All Seasons theatrical release I saw was extremely vivid.) And Grover and his colleagues never intervene in the online debates.

Q: Why do DVDs seem different from projected prints—often much sharper?

A: An original film frame, in camera negative, has more resolution than can be captured in a print. Most prints are a generation or two past the material that a DVD transfer uses, so they gain a certain softness. Granted, however, in making high-definition video masters sometimes the tools to sharpen and de-grain the image are overused.

In addition, it’s generally a good idea to get your home video display calibrated. Most home displays are way too bright, sometimes brighter than a theatre screen. If you darkened the room and had the display dialed down, DVDs would look more film-like.

Q: When will we get the Boetticher westerns?

A: It’s the most frequent question Grover gets asked. Soon, soon: A boxed set of restored titles is on the way.

Q: And what about all the Capra titles?
A: Sony plans to restore each of them. Columbia struck many prints of them, and so there is some negative wear. Some negatives no longer exist. On one title, Say It with Sables (1928), Sony has neither negative nor prints.

There’s a too-good-to-be-true backstory here. Capra had a ranch in Pomona, which upon his death was bequeathed to Pomona State University. After some years, people found in a locked stable his private collection of prints struck in 1939. The cache included Lost Horizon, You Can’t Take It with You, It Happened One Night, and the best-quality print of Mr. Smith Goes to Washington Grover had yet seen. There were also photographs Capra had shot of premieres and vacations. Thanks to the good offices of Frank Capra, Jr., Sony acquired the prints and used them in its restorations.

As for American Madness, Grover called it a “training film” for later Capra productions. The theme of faith in the little man, manifested in a bank run that tests a humane banker’s alliances in the community, points ahead to It’s a Wonderful Life and other films. Sony holds a complete soundtrack and a nearly complete original negative. UCLA held a complete nitrate print donated by the Los Angeles Parks and Recreation Department; evidently the film was screened in parks as public entertainment. That print wasn’t in good condition, but it allowed Grover’s team to fill out certain scenes. The nitrate print’s footage is noticeably lighter and grainier in a few places. American Madness wasn’t a digital restoration, but if it were done today Grover would probably use CGI to blend in the alien footage.

Talkies, with a vengeance

It was fine to see the film again; its brisk inventiveness held up. The gleaming and geometrical images of the opening, which acquaint us with the daily routine of opening the bank vault, might have come out of Metropolis. One helter-skelter montage sequence, complete with canted framings and chiaroscuro lighting, looks forward to Slavko Vorkapich’s delirious contributions to Mr. Smith and Meet John Doe.

The tactics of depth composition that we find in many 1930s movies reappear here, and the illumination has dashes of noir.

I especially like the dynamic pans that carry Walter Huston and Pat O’Brien in and out of the main office; I suspect that they’re cut together faster and faster as the climax gets near. And the huge set of the bank lobby, publicized at the time as the biggest set yet built on the Columbia lot, remains not only impressive but functional. It establishes a cogent geography to which Capra adheres strictly while filming it from a great variety of angles.

But this is not a treat just for the eyes. American Madness flaunts its mastery of emerging talkie technique. The rising action is accentuated by the steady increase in volume of the growing crowd in the bank lobby. The clerks’ scattershot morning chat in the vault is captured in microphone distances and auditory textures that suggest a hollow, sealed-off space. Hawks’ rapid-fire patter in Twentieth Century (1934) has an antecedent here, and at the same studio. The actors speak at a terrific clip, scarcely pausing between lines, and sometimes the dialogue overlaps. We even get competing lines, two or more speeches rattled off at once. (And did Hawks get the idea for His Girl Friday’s variants on telephone chatter from O’Brien working the receivers at the climax?)

In films like this, American movies talk American. Consequently, I’m inclined to regard as an in-joke the glimpse we get of a marquee that’s advertising another Columbia picture: Hollywood Speaks.

Thanks to Grover and Sony Pictures Entertainment for a wonderful series and an enlightening brace of talks. American Madness is available on the DVD set The Premiere Frank Capra Collection.

(1) Chris Horak has been head archivist at George Eastman House, the Munich Film Museum, Universal, and the Hollywood Museum; he’s now Director of the UCLA Film & Television Archive. Mike Pogorzelski is Director of the Film Archive for the Academy of Motion Picture Arts and Sciences; Joe Lindner is Preservation Officer there. (Both are former Badgers.) Schawn Belston is Vice President of Asset Management and Film Preservation at Twentieth Century Fox. Paolo Cherchi Usai is Director of ScreenSound Australia, the national archive, and director of the recent film Passio. Kristin interviewed Mike and Schawn for this blog entry.

(2) YCM separation masters are made by copying a color film onto three black-and-white films. Thanks to filtering, these preserve color luminance at the different values of yellow, cyan, and magenta. Since these monochrome versions are not as susceptible to fading, they can be used for archival preservation. Combined, they can recreate the original color image.

(3) This confirms my impression when I’ve seen different Technicolor prints of the same title. You sometimes get this effect with a composite print, in which the color values change from reel to reel. Anecdotally, I’ve heard that Technicolor staff in the studio days would sort reels according to their dominant color (“That’s the yellow pile over there”).

(4) See Chapter 7 of Silent Cinema: An Introduction (London: British Film Institute, 2000). This book, incidentally, is a good introduction to the sheer fun of archive work. Others communicating the same enthusiasm are Roger Smither and Catherine A. Surowiec, This Film Is Dangerous: A Celebration of Nitrate Film (Brussels: FIAF, 2002) and Dan Nissen et al., eds., Preserve Then Show (Copenhagen: Danish Film Institute, 2002).

(5) This problem renders the activities of Britain’s National Media Museum in Bradford all the more important. There you can see classic films in a great many formats and sound arrays.

American Madness.

15 June: Thanks to Kent Jones for correcting a name slip.

19 June: Breaking News: Grover has just been promoted to Senior Vice President for asset management, film restoration, and digital mastering. Nice timing; wish we could say our blog put him over the top, butVariety explains that it was good old-fashioned talent. Congratulations to Grover!

Went the Day Well? A traitor gives information to a German infiltrator, who takes notes against a memorial to World War I dead

Kristin here—

As David and I sit here revising our Film History textbook for its third edition, we’re looking forward to the annual Il Cinema Ritrovato film festival in Bologna (June 28-July 5). We blogged from there last year and plan to do so again this year.

One of the festival’s themes in recent years has been World War II. Last year the program included Alberto Cavalcanti’s Went the Day Well?, which he directed for Ealing Studios in 1942. Partly out of a sense of duty, since I cover Great Britain for the textbook, I went to the screening. It turned out to be a well-made, engaging, and surprisingly affecting film.

Alberto De-Almeida Cavalcanti

Alberto Cavalcanti is one of those names that people who take courses in film tend to be vaguely aware of, and that for only a handful of titles. He got his start as a set designer on Marcel L’Herbier’s L’Inhumaine (1923). His first film as a director, the 1926 experimental city symphony Rien que les heures, is a minor classic of the era. Coal Face (1935) is one of the better known British documentaries of the 1930s. Cavalcanti worked for Ealing for several years in the 1940s, and a modern cinema buff is likely to know his two episodes of the collectively directed Dead of Night (1945), including “The Ventriloquist’s Dummy,” the most highly regarded of the batch. Devotees of Ealing or of classic British comedy will also be familiar with his Champagne Charlie (1944). His other best-known films are Nicholas Nickleby and They Made Me a Fugitive (both 1947).

In an era of globilization, Cavalcanti’s career looks positively modern. Born in Rio De Janeiro, he made his early films in Paris. He worked in British documentaries during the 1930s before moving to Ealing, and he remained in England throughout the 1940s. The rest of Cavalcanti’s filmmaking career saw him on the move, including projects in East Germany, Israel, and Brazil. He had a brief stint in TV in the early 1970s before he retired. He died in Paris in 1982. He tended to sign his films simply “Cavalcanti,” as is the case on Went the Day Well?

British Pluck

Based on Graham Greene’s short story, “The Lieutenant Died Last,” Went the Day Well? is a “what if?” tale. A group of Germans disguised as British soldiers deceive a small village’s citizens into letting them billet there and conduct practice exercises. In reality they are there to set up observation posts in preparation for an invasion.

When the film was being planned, there were many rumors in England that a German invasion was imminent and much speculation as to whether the country could resist such a thing successfully. The film has no doubts. In fact, its brief frame story is set in 1945, and the narrator, one of the inhabitants of the fictional village Bramley End, reveals that Germany has been defeated. The narrator shows us the grave of several German soldiers who died in the attempted infiltration.

From the start, then, the spectator is aware that the apparent British soldiers who arrive in Bramley End are disguised Germans and that they will be defeated. This seems like an odd tactic, letting us know so much, and yet it makes perfect sense. The point of the early part of the film is to introduce us to the villagers who will form the collective hero of the plot. Seeing them going about helping British soldiers settle into the local hall and various households would hardly be dramatic. As it is, knowing more than they do, we are in suspense as to whether and how they will discover the truth and react.

At first they seem the conventional characters of a comedy or light drama—the sorts who would be the supporting roles in a Hitchcock film like The Lady Vanishes. They innocently cooperate in typical British fashion, giving directions and offering tea and spare bedrooms. The initial third or so of the film is pleasant in tone, despite our underlying awareness of the enemy being let into these people’s midst.

Yet the little scenes of ordinary life swiftly and subtly set up the traits in these people that will allow them to fight back later. The local shopkeeper and telephone operator may be a bit dotty and forgetful, but she also bridles when one of the Germans manhandles a boy who had been snooping around the equipment in the soldiers’ truck. In the same scene, that boy is established as spunky and curious. Went the Day Well? manages to set up several major characters that we come to care about, as well as many recognizable supporting figures. There is no one protagonist. In fact, the film’s top-billed actor, Leslie Banks (familiar to modern audiences mainly from The Most Dangerous Game, the original Man Who Knew Too Much, and Henry V), is a traitor, secretly aiding the Germans. That’s Banks in the photo surmounting this entry.

Naturally once the villagers discover who their guests really are, they come through with English pluck and resourcefulness–the women as well as the men. The film doesn’t sugar-coat the conflict, though. A number of characters one would not conventionally expect to die do meet heroic ends, and the tone rapidly turns more serious roughly halfway through the action.

Went the Day Well? is a handsome film, too. British films of the 1930s sometimes look like low-budget Hollywood imitations, but by the 1940s the studios had more resources. Cavalcanti’s design sense serves him well, and the beautifully lit scenes and crisp editing look like something MGM could have been proud of. William Walton contributes a brief but suitably grand score.

Sources

Unfortunately Cavalcanti’s films aren’t easily available on DVD in the U.S. The only current way to buy Went the Day Well? on region 1 DVD is as part of a set, the “British War Collection.” (The set was originally released as individual discs as well, and several copies of that edition of Went the Day Well? are offered on eBay.) Amazingly, the only other in-print films are Le Capitaine Fracasse (a re-discovered 1929 French film with a young Charles Boyer, co-directed by Cavalcanti), Nicholas Nickleby, and They Made Me a Fugitive.

Those with multi-region players can visit Amazon.uk, where one can find Went the Day Well?, the three films just listed, Dead of Night, and Champagne Charlie.

Went the Day Well? seems like the kind of film ripe for a Criterion edition, if we may drop a hint in their ear.

Added May 11:

David Cairns, director and scriptwriter, critic and blogger, and Cavalcanti fan has responded to this entry with one of his own, discussing other aspects of Went the Day Well? (Beware, his entry has more SPOILERS than mine.) As he rightly points out, the early portions of the film not only set up the bucolic village and its apparently conventional British country folk, but they provide an object lesson for citizens not to behave in such a naive and unquestioning fashion. Germans might be anywhere! I was mainly writing about the narrative conventions that would hold for a modern audience watching a decades-old classic. There’s no doubt, though, that for a contemporary audience there would have been that implication. The point is fairly explicitly made when one character finally finds a clue that makes her begin to suspect something is wrong, while another dismisses her ideas–with we all the while knowing that the woman who becomes suspicious is right.

I don’t think I buy another claim David makes about a subtext for the film: “the film becomes more subversive. According to Cavalcanti, a pacifist, his objective was to show that when war comes to even a place as charming as Bramley End, the people become monsters. Without the slightest change in underlying personality, peace-loving and jocular countryfolk pick up weapons and set about slaughtering their fellow humans.”

I don’t find the villagers in any way to be monsters, even in their most violent acts against the invaders. For one thing, it’s self-defense in wartime. The starkly violent scene of the Germans ambushing and mowing down a group of men from the Home Guard makes that clear from before the point when the villagers respond. I certainly can’t believe that the audiences of 1942 would have seen the villagers as descending into some primitively violent instincts. After all, this film came out only a little over a year after the end of the Blitz. In fact, the Germans had attempted to infiltrate England in preparation for the Battle of Britain, which in turn was to culminate in a land invasion. All infiltration attempts were thwarted, and poor intelligence was a major contributing factor to what became Germany’s first significant loss of the war. There would have been few people in Britain who could conceive of pacifism as a response to what they had recently lived through.

For another thing, the villagers don’t kill ruthlessly. All three women who kill someone pause afterward, clearly upset and even nauseated . The men are more blithe about it, but they’re seldom seen killing except in pitched battles, and just as often the village menfolk are the victims up until the rescuing soldiers arrive at the end. I’d argue that the characters’ underlying personalities are assumed to include the courage and determination that surface in time of need.

Thanks to David for his thought-provoking comments and for backing me up in urging that cinephiles should seek out this and other Cavalcanti films. He mentions They Made Me a Fugitive in particular.

DVD reviewer Slarek has reviewed Went the Day Well? on DVD Outsider. (Again, SPOILERS.) Among other interesting comments, he reveals that it was shot in Turville, Oxfordshire. I think Bramley End is probably supposed to lie in the south of England, where most of the German aircraft and the putative invasion would have entered the country. Still, the rolling landscapes and hedgerow-lined fields of Kent and Sussex are fairly similar to those of Oxfordshire. As Slarek proves with a photo, the village is not all that much changed today.

DB here:

On some of Bill Kinder‘s business cards, the I in PIXAR is represented by Buzz Lightyear, the blustery, not-too-swift astronaut of Toy Story. It’s a typical gesture of self-deprecation from the studio that showed that computer animation didn’t have to be just plastic surfaces and mechanical expressions. Pixar is cool, geeky, and warm all at the same time. Its films are both smart and soulful, made by movie fans for movie fans, and for everybody else. Like the best of the Hollywood tradition, Pixar movies have the common touch and still offer the most refined pleasures.

Kristin and I have already written admiringly about Pixar on this site (here, here, and here). It’s quite likely that this studio is making the most consistently excellent films in America today. So we were delighted when our colleague Lea Jacobs arranged for Bill to come to the University of Wisconsin—Madison last fall. He toured our new Hamel 3-D media facility, met with faculty and students, and gave a talk, “Editing Digital Pictures.” Bill is Director of Editorial and Post-Production, a position that gives him an encompassing view of the Pixar process as he champions the efforts of the editors and their teams as key creative contributors.

A graduate of Brown, where he studied with our old friend Mary Ann Doane, Bill is like Pixar movies—intellectual, good-natured, energized, and adept at connecting with people. He started his career in news-gathering and TV editing before moving to work at Francis Ford Coppola’s American Zoetrope, in the days of Jack (1996) and the uncompleted Pinocchio project. He joined Pixar in 1996, while they were finishing Toy Story. When the success of A Bug’s Life enabled Pixar to move to a purpose-built facility in Emeryville, Bill went along.

Whittling vs. building

I’ve always been uncertain about what an editor does in the animation process. Since every shot is planned and executed in detail, what can be left for an editor to do? Bill started from that question. No, editing digital animation isn’t just a matter of cutting off the slates and splicing perfectly finished shots together.

As in live-action filming, the animation editor is working with dozens of alternate versions of every shot. The reason is that at Pixar, there are roughly five phases of production: storyboarding, layout, animation, lighting, and effects/ rendering. Each one generates footage that has to be cut together.

The static storyboards, for instance, present poses, expressions, and movements against a blank background. They are assembled in digital files that can be played back as if they were a movie. In order to plan the next phase, the resulting “footage” has to be edited, and choices are made at every cut. And each scene is storyboarded at least five different ways, with many variations of action and timing. A single film uses up to 80,000 boards!

At the next phase, layout, the scene’s overall action is planned. Layout artists develop the staging of each shot, testing different backgrounds and camera angles with the editors. Again, the alternatives have to be assembled and cut in various combinations.

Whittling versus building, Bill called it. The live-action editor gets a mass of footage that has to be triaged, but the animation editor is building and tuning the film from the start. Editing operates at each phase, from storyboarding to final rendering. This “almost overwhelming iteration,” as Bill called it, demands that the editorial department hold all the alternatives in its collective mind at once. Add to this the fact that Pixar can take up to five years to produce a film, maintaining several editorial teams to cover projects at different degrees of completion. When you realize that all this brainpower and bookkeeping are necessary for even the simplest shot, you appreciate the felicities of the finished product even more. These people make it all look easy.

Continuity and the viewer’s eye

Bill explained that digital animation occasionally requires something like live-action coverage. (1) Action sequences with fast cutting need to be spatially clear, and “chase scenes can be hard to board.” So sometimes the layout artists create master shots and closer shots from different angles that the editor will pick out and assemble, live-action fashion.

Like live-action editors, Pixar editors have to keep an eye on continuity of the objects in the frame. Because each shot is reworked across many phases, items of the set, lighting, color, atmosphere, effects and rendering have to be maintained, on many layers or levels of the program. (I gather it’s like the layers in PhotoShop.) Sometimes a layer, whether a prop, character, or set element, fails to “turn on” and so a discontinuity can crop up. A finished Pixar film typically has 1500 shots or more, so there’s a lot to keep track of.

In another carryover from live-action features, Pixar plots are conceived and executed in three discrete acts. It’s not only a storytelling strategy but a convenience in production. Rather than waiting until the entire film is done to examine the results of the different phases, the filmmakers can finish one act ahead of the others in order to troubleshoot the rest.

I’ve studied how filmmakers compose the image in order to shift our attention (2), so I was happy to hear that this process is of concern to the Pixar team. “Guiding the viewer’s eye,” Bill called it. He explained that in looking at storyboards and animated sequences, his colleagues sometimes use laser pointers to track the main areas of interest within shots and across cuts, especially when characters’ eyelines are involved. Nice to see that sometimes academic analysis mirrors the practical decisions of filmmakers.

The auteurs of Pixar

What makes Pixar films so fine? Bill supplied one answer: It’s a director-driven studio. As opposed to filmmaking-by-committee, with producers hiring a director to turn a property into a picture, the strategy is to let a director generate an original story and carry it through to fruition (aided by all-around geniuses like the late Joe Ranft). Within the Pixar look, John Lasseter’s Toy Story 2 and Cars are subtly different from Brad Bird’s The Incredibles and Ratatouille or Andrew Stanton’s Finding Nemo and upcoming Wall.E.

Bill covered many other fascinating topics, including the importance of sound (“the animated film’s nervous system”). But I’ll end with some pull-quotes from Bill’s talk.

*Francis Ford Coppola: “No film is ever as good as its dailies or as bad as its first assembly.”

*Gary Rydstrom: “Film sound is the side door to people’s brains.”

*Bill himself: “Editing is just writing, but using different tools.”

We’re grateful to Bill for his visit and look forward to seeing him again. Goes to prove what we’ve said before: Popular American filmmaking harbors many of the most intelligent, sensitive, and generous people you’ll ever find.

(1) In live-action production, coverage involves shooting a master shot of a scene that shows the entire action. Then parts of the action are repeated and filmed in closer views. This allows the editor several options for cutting the scene together.

(2) I talk about this in Chapter 6 of On the History of Film Style and throughout Figures Traced in Light. See also Film Art, pp. 140-153, and this blog here and here.

PS: Another glimpse into the kitchen: Bill Desowitz reports on Wall*E at Animation World. Now Pixar is trying to emulate the look of 70mm. And there’s footage from Hello, Dolly! in there? All the signs point to another nutty, dazzling achievement.