Archive for the 'Film technique: Cinematography' Category
The Limey (1999).
Here’s a story, so far not finished.
Chapter 1: You can’t put ten pounds of mud in a five-pound sack (Dolly Parton)
Once upon a time, before home video and cable, movies were broadcast on television. They might be drawn from local TV stations’ 16mm collections or transmitted from the networks on “movie of the week” shows. When the film was a widescreen film, especially in anamorphic processes, it was adjusted, as we now say, “to fit your screen.”
That TV screen was in an aspect ratio of about 1.33:1 (4 x 3), like pre-1950s commercial films. But the film to be shown might be 1.75, 1.85, or 2.35. How to show it?
You could retain the whole original frame with letterboxing at top and bottom. This was done more often in Europe than in the US, I believe. Our TV system had 100 fewer lines, so the tiny picture area became quite degraded. And then (as now) there were people who persisted in believing that those black bars at top and bottom were somehow taking away part of the movie.
Far more common was the tactic of somehow making the wider image fill the box. This tactic came to be called “pan and scan.” It gained the name because in preparing the TV version, an engineer would sometimes swivel the TV frame across the original image, carving into it and sliding to and fro. The term pan and scan covered another tactic, simply making two shots out of one. We might call it “cut and scan.” Here’s an instance from an old VHS tape of Advise and Consent, one of the most daring American widescreen films. The slightly fatter faces are due to the distortion of the CRT monitor I shot from.
Sometimes there was neither scanning nor cutting, just simple cropping. The engineer would find a 4×3 chunk and extract it from the bigger image. To keep things simple, that chunk was usually around the center. This “center-cutting,” as it was called, could yield some funny results, as in the protruding nose in a 16mm TV print of Tarnished Angels.
Chapter 2: Shoot, protect, screw up
Some widescreen processes from the 1950s onward didn’t frame the image wide during filming. The image would be captured “full frame.” The result might later be “hard matted,” or letterboxed, during printing, as in the shot from The Limey surmounting today’s entry.
Sometimes, the 35mm theatrical prints would retain a 4:3 image, or something even squarer. The frames in a 35mm print of Do the Right Thing are about 1.19:1. Note the extensive headroom in the left image, from a 35mm print. For screenings the projectionist would mask the image to the proper proportions—typically 1.85. Here’s the version on the Criterion DVD, at 1.75 for widescreen monitors.
The full-frame image was available for cropping to 4:3 TV viewing. Many cinematographers used the “shoot and protect method” by composing for both 1.85 and 1.33.
There were problems with this “open matte” process in theatres. I recall a mis-framed version of Jerry Maguire in which the locker-room scene showed more of Cuba Gooding’s goodies than was intended. And a full-frame print of Godfather II reveals the duct-taped marks on the set floor (as if Pacino would ever hit them).
The same problem would appear in full-frame TV versions that were carelessly transferred from film. Microphones, unfinished stretches of the set, and other production elements might jut in. Some cinematographers didn’t consider TV versions important enough to worry about. “Fuck TV,” was heard occasionally.
More basically, the full-frame result wasn’t faithful to the “original” theatrical version, which was designed to be shown wide. Purists objected that TV versions, even if shot full-frame, didn’t fulfill the intention of the director.
Chapter 3: 16 into 9 won’t go
Actually, pan-and-scan hung on quite a bit. Many laserdisc editions weren’t letterboxed. Airline versions offered cropped images, and still do. (Those tiny screens, after all.) Some DVDs were released in pan-and-scan versions, while others offered a letterboxed version of the film on one side of the disc and a full-frame version on the other. But serious cinephiles could assume that the general public was starting to understand that films intended to be seen at a certain ratio should be shown that way.
As the new disc formats emerged, so did a plan to standardize High Definition video for a 16 x 9 display. The European Union promoted it during the mid-1990s, and over the next twenty years it became accepted for both computer monitors and video displays. The new format developed in sync with the gradual replacement of analog broadcasting by HD transmission.
The problem is that 16:9 is a ratio of 1.77:1, close enough to 1.85 for most purposes but far off the ratio of 2.20:1 (many 70mm films) and 2.40 (anamorphic widescreen after about 1970). Yet maybe this isn’t a problem. Why not just letterbox those wider images within the 16:9 format?
If only things were so simple.
Chapter 4: Slightly off, and more
Today, at least 77% of US households have HD monitors. (Surprisingly, 41% of TVs still in use are Standard Definition, many presumably comforting children and old people.) Broadcasters make concessions to the 4:3 ratio by keeping the key action centered and cropping the 16:9 image.
Nature abhors a vacuum, and TV monitors apparently can’t bear blank space. So some purveyors of movies on video have updated pan-and-scan for the HD age. Cable, for instance.
It’s been years since I clicked my cable remote to the Sundance Channel and the Independent Film Channel, now known as IFC. Seeing them a couple of weeks ago was a mild shock. Now each boasted a bug in the lower right corner, and swarming over the image were lots of texts plugging other programs. Worse, there were commercials for weight-loss scams, Burger King, and Portlandia. More to the point here, these services give us a new version of pan-and-scan.
Thanks to the center-cutting method, I found plenty of examples of anamorphic 2.40 films sliced up to fit our monitor. Here are a couple of examples from IFC’s version of Jaws. In the first pair, the wider image gives Bruce room to move, and Brody is more salient. In the second pair, Hooper is speaking, but in the IFC image he’s even more peripheral than Mr. Nose in Tarnished Angels.
In rare cases, IFC seems to run full-scope prints; I spotted one of Chinatown. But it doesn’t seem to be the norm. Maybe IFC’s New Age pan-and-scan is trying to live up to the channel’s slogan: “Always on. Slightly off.”
Many films of the early widescreen era are compromised even at 16:9. The nose shot from Tarnished Angels would be a bit cramped in a 1.77 format, and many other shots would suffer as well.
When the wide format was still new, filmmakers were exploring what they could do it, and that included scattering important information throughout the frame. Preminger was all over the place in the Senate and Committee-Room scenes of Advise and Consent. What would you do to fit these shots into 16:9? I make a silly stab at center-cutting.
Films like this remind you of how the technology of home video has made our filmmakers less daring in their compositions. Would anyone today risk the bold composition that Richard Fleischer uses in the mortuary scene of Compulsion? As you can see from the image at the end of the entry, the crucial pair of eyeglasses rests on the lower frameline, as if on a shelf. ’Scope nudged some directors toward quite intricate shot designs.
By the time Tootsie was made, with home video on the rise, filmmakers were pressed to shoot and protect better. So here there’s plenty of unused space on the sides, and most action will fit into any old frame proportions. In the shot below, the 1.33 version (available on the flip side of the DVD) is extracted from the Scope version (and slightly squeezed as well). The 1.77 version I grabbed from the Sundance Channel is also extracted from the wider framing. Both reframe the image to emphasize Michael and Jeff, the comic duo. The original 2.35 image centers the food counter and for some reason gives equal space to the inconsequential aisle and stove on the right.
Obviously, composition isn’t particularly important here. Solid as Tootsie is at the level of writing and performance (“Oh, God, I begged you to get some therapy”), we wouldn’t point to it as a model of pictorial finesse.
Moreover, pan-and-scan isn’t just a matter of retaining the action in different framings. Recasting the image changes the scale of the things in it. Old TV pan-and-scan, as in my first Advise and Consent example, makes the figures chopped out of the composition seem closer to us. Sometimes today’s pan-and-scan makes the figures seem smaller. That’s because the full-frame original image offers a bit more space at the top or bottom than we see in the anamorphic version, and that gets incorporated into the 1.77 or 1.33 version. (You can see it in the Jaws example above, with the space above Quint’s head.)
The push-pull of different versions is particularly noticeable in close views. The center-cut (and perhaps slightly squeezed) IFC version of The Matrix keeps Cypher prominent in the foreground, with the car crash in the rear. Yet do you see him and the crash as closer to us in the wide original? I do. Maybe it’s because both Cypher and the background plane are larger in the second frame.
By contrast, several shots in the IFC version of The Matrix have areas at the top and bottom not visible in a wider-frame DVD. In any case, extracting from an anamorphic frame or here, pulling the 1.77 version out of a full frame, can subtly alter the scale and perspective relations within the shot.
In the old TV days, networks would show credit sequences in full widescreen, as they were obliged to make all the contributors’ names visible. So you often had the odd situation of a widescreen credits sequence that would end a movie shown in 4:3. Sundance has updated the technique. Tootsie’s penultimate shot of Julie is at 1.77, but as she and Michael walk off, we cut immediately to a final shot at 2.35 letterboxed, over which the credits roll.
Another old trick has been revived. Back in 1960s TV, you’d have an obligatory widescreen image with credits information, but the shot might go on to show story action. Then the TV framing would close in on the wide image and gradually fill the screen with a 4:3 composition. Damned if I didn’t find the same thing happening last weekend on Sundance. The opening credits for The Graduate were in full ‘scope. That shot dissolved to the famous shot of Benjamin sitting in front of his aquarium, blankly facing the camera.
To retain the credits and keep the dissolve smooth, the engineer slowly enlarged the original shot to 16:9 proportions as Ben’s father comes in. (You can see the Sundance bug lift slowly into the lower right of the image.)
The result is a camera movement, a zoom in, that isn’t in the original film. In the original framing below, the shot scale on Ben isn’t much different, but the little scuba diver remains prominent in a way he isn’t in the 16×9 image.
Chapter 5: VOD = Variants On Demand
On cable TV we Americans can go to TCM for a purer experience of widescreen from the old days. But the more recent films shown on IFC, Sundance, Pivot, and other cable channels are likely to be hacked up in the new way. In this respect as well as others, cable has become the network TV of the new age. We get shows of all types, not just sports and movies but original series, except that now we get to pay to watch commercials.
Surely things are better with streaming?
In the spot-checking I did on Amazon and Hulu Plus, I didn’t find instances of cropping. In particular, the Criterion films I checked retain their proper aspect ratios. Netflix is another story. At this point we must thank the anonymous genius behind What Netflix Does. This site exposes a great many ways that the most popular streaming service has relied on pan-and-scan, with different crops for different markets.
In fairness, once the anonymous genius contacted Netflix, some deficient versions were replaced by proper ones. But several of the more recent examples have not been changed. Who knows how many others remain panned and scanned because the alterations haven’t yet been detected?
In the light of all this, I’m wondering if filmmakers have been protesting this mangling of their work. During the early days of TV broadcasts, directors complained about cuts and commercials, and in more recent years we’ve learned that directors were won over to digital projection in theatres because then “the film would be shown exactly as the director intended.” With many more people seeing the film on video platforms than see it in theatres, you’d think we’d have heard more from the creative community. Once more their movies are being jammed and lopped to fit whatever box they’re put in.
I’m tempted to say that when we want to get the movie in a form close to the ways its makers wanted it to be seen, we need to see it in a theatre or get it on DVD/Blu-ray. There are exceptions, of course. Theatres can botch aspect ratios today just as they have in earlier decades, chiefly by using the wrong masking. And we we have filmmakers who alter the aspect ratio on DVD, usually to expand the field from 2.35:1 to 1..77 in hopes (vain) of evoking the Imax effect (Tron: Legacy, The Hunger Games: Catching Fire).
Without getting into technicalities, some of the 1.77 versions you’ll see have a bit more area at the top and the bottom of the frame than we find in the full anamorphic image. In many cases, like the shot of Cypher in The Matrix, it’s because the film was shot in Super-35. This very full-frame capture format allowed filmmakers to extract a 2.40-proportioned image, or a 4:3 image, or images in other aspect ratios. The Wikipedia entry on Super-35 is very helpful on this and other aspect ratios, both for cinema screens and TV. Some shots in the IFC Jaws had a little more vertical area as well. I assume that’s because the ‘scope image on the DVD cropped a tad off the top and bottom of the full-frame image on the film.
Thanks to Jonah Horwitz for pointing me toward the site What Netflix Does. More general thanks to James Quandt for our long-running conversation about aspect ratios. Other entries on aspect ratio on this site involve Fritz Lang, Jean-Luc Godard, and Wes Anderson. I set out some ideas on CinemaScope aesthetics in this video lecture.
The Little Foxes (1941).
For me, shooting is a struggle where you only get to be happy for five minutes before you start thinking about the next problem to solve.
Ruben Ostlund, on Force Majeure
One of the most famous shots in American cinema occurs at a climactic moment in The Little Foxes (1941). Regina Giddens has just learned that her sickly husband Horace has let her brothers get away with a business deal that double-crosses her. They will reap all the rewards of bringing a factory to town, while she, who engineered the deal and expected Horace to fall in line, will get nothing. Horace is already not far from death, and their quarrel in the parlor precipitates a heart attack. He spills his bottle of medicine and needs some from his upstairs supply.
Regina refuses to go fetch it, and instead Horace must stagger up and out. While she sits, fiercely waiting, on the sofa, he tries to pull himself upstairs, but he collapses on the steps. Once he has fallen, and perhaps died, she stirs to action and rouses the household.
Lillian Hellman’s original play had been a Broadway success, and this was one of the most notable scenes. How did Wyler stage it? Very oddly, as the frame up top suggests. We can’t really see Horace’s struggle on the stair. Not only does the camera put Regina in the foreground, but Horace is out of focus in the rear, at least until she rises whirling and runs to the background, the damage done.
Why did Wyler stage it this way? It depends, as Bill Clinton might say, on what your definition of why is.
1941 was the breakout year of deep-focus filmmaking in Hollywood. Citizen Kane, The Maltese Falcon, Kings Row, Ball of Fire, I Wake Up Screaming, How Green Was My Valley, and several other films set the pace for a new stylistic option. In this style, the action is staged in depth rather than perpendicular to the camera, as most scenes in Hollywood cinema were. And the camera lens creates depth of field, in which even fairly close foreground planes are just as sharp as the action in the rear. Such images weren’t unknown before; we can find them in silent cinema. But from 1941 on, depth staging accompanied by depth of focus would be increasingly common in Hollywood dramas, from thrillers and melodramas to film-noir exercises. Not all shots would be designed for maximal depth; continuity editing and closer views would still be used. But we do find such imagery becoming more common, particularly at moments of tension.
Cinematographer Gregg Toland is usually cited as a main source of this trend, and his work on Kane and Ball of Fire, as well as Ford’s Grapes of Wrath (1940) and The Long Voyage Home (1940), became models of the new look. Toland also worked with Wyler on several films, including The Little Foxes. But even without Toland, Wyler had in some films cultivated a deep-focus look (as had Ford). Coming when it did, The Little Foxes proved a powerful demonstration of the deep-focus style.
Three aspects stand out. First, there’s a certain economy of presentation. As Wyler and others pointed out, depth imagery permits directors to minimize editing. Instead of cutting from action to reaction, we see both at the same time.
Wyler suggested in publicity of the period that this gave the viewer more freedom of where to look, and André Bazin seized upon this rationale as part of his aesthetic of realism. Just as in the real world, in some films we must choose what to pay attention to.
But The Little Foxes went beyond the moderate deep focus of Stagecoach and other films to create very aggressive images. This is the film’s second novelty. Several shots place the foreground very close to the camera. As a result, we get looming faces or objects in the front plane, and we still see well-focused dramatic elements behind.
A third source of power is less noted. In The Little Foxes, Wyler found ways to make deep shots comment upon the plot. For instance, the action offers Regina’s daughter Alexandra, usually called Zan, a choice of being more like her mother (tough and vicious) or her father (tolerant and gentle). At other points Zan is paralleled to her ineffectual, alcoholic aunt Birdie. At one point, Birdie has predicted that Zan may wind up like her.
In a theatre production, there would be many staging strategies that would create these parallels, but Wyler uses a particularly striking one. One evening, while Regina and her brothers plot their scheme, Birdie has been relegated to a chair far from the discussion.
The composition diagrams Birdie’s situation in the scene and her place in the family. Then Wyler cuts in to her.
This might be seen as a bit of heavy-handed emphasis, but actually he’s doing two things. He’s making manifest her reaction, a numb resignation to being excluded. He’s also setting up, thanks to another depth composition, the chair in the hallway by the staircase. At the climax, it’s Zan, as beaten down as Birdie, who slumps in that chair.
Thanks to depth staging and deep-focus cinematography, the second image emphasizes Birdie’s solitude and prophesies Zan’s.
Which only makes my first question more pressing. Some shots of the quarrel leading up to Horace’s collapse on the stair exhibit flagrant deep focus.
We know from other shots in the film, like the Birdie/Zan comparison, that Wyler could have simply shown us Regina on the sofa in the foreground, in long shot or medium shot, while keeping Horace in focus in the background. In fact, Wyler tells us that Toland said, “I can have him sharp, or both of them sharp.” Why opt for shallow focus that makes Horace’s staircase seizure blurry and hard to see?
Fun with functions
Asking why? about something in an artwork actually veils two different questions.
The first is: How did it get there? The answer is a causal story about how the element came to be included.
The second sense of why is: What’s it doing there? That’s not a question of causes but of functions. How does the element contribute to the other parts and the artwork as a whole?
Take the second question first. You can imagine many functional reasons for Wyler’s choice. Exactly because the rest of the film keeps image planes sharp, this moment gains a unique emphasis. Horace’s collapse is marked as a major turning point in the plot. In an ordinary film, we wouldn’t notice an out-of-focus background. Here, by reverting to the more traditional choice, Wyler makes shallow focus stylistically prominent. For once in a film, a dramatic high point isn’t given to us with maximum visibility.
Another function is character revelation. In the film as a whole, we haven’t been consistently restricted to any one character. Here, Wyler could have concentrated on either Horace or Regina, or he could have given them equal treatment. An obvious choice would have been intercutting shots of Horace crawling up the steps with shots of Regina, impassively turned from him. Probably most directors would have done it that way.
Alternatively, we might have been attached to Horace, letting us see Regina in the distance. That would have diminished her reaction and played up Horace’s suffering.
Wyler’s choice puts the emphasis not on the action—thanks to the distant framing, Horace’s collapse can almost be taken for granted—but Regina’s reactions, or rather non-reactions, moment by moment. We’re made to see her turning slightly to listen to his struggles, while her staring eyes suggest that she’s visualizing the action with a horrified fascination. It’s as if her denying him the medicine was an experiment in seeing how far she could go. Now she knows. Her straining face is virtually willing her husband to die.
Keeping both this monstrous woman and her victim in focus would have divided our attention, then, and Wyler wants it squarely on Regina. He seems to have said as much in interviews.
We said we’ve got to stay on Bette all the time and just see this thing in the background, see him going in the background, but never lose her.
I wanted audiences to feel they were seeing something they were not supposed to see. Seeing the husband in the background made you squint, but what you were seeing was her face.
The second remark suggests another functional result of Wyler’s choice. By making the collapse almost indiscernible, we become very aware of what we can’t see. Thanks to selective focus, Bazin remarked, “The viewer feels an extra anxiety and almost wants to push the immobile Bette Davis aside to get a better look.” The dramatic tension of the scene finds its counterpart in our frustration to see what any other film would show us.
Finally, we should note that the staircase is an essential element in the film’s drama. Horace’s collapse is only one major incident taking place around and on it. Significantly, when Zan finally breaks free of Regina and the rest of the family, the matriarch learns of it standing on the stairs. Having all but murdered her husband there, now she sees her daughter abandon her.
Shoot my good side
The Bishop’s Wife (1948).
There are other functions we, as good critics, might seek out. For all of them, there is probably a loose causal story we’re relying on: Wyler and his colleagues made some choices that bore fruit. Some of those choices may have aimed at fulfilling the functions we notice. Other functions we notice may come along as bonuses—unintended but still benefiting the scene. Unintended consequences, good or bad, come up in art as elsewhere.
There remains the other implication of why-did-they-do-it questions: the one that seeks out quite specific causes that govern the scene. How do we tackle that?
In my book On the History of Film Style, from which some of these Little Foxes observations are drawn, I argued that we can make stretches of stylistic history intelligible by thinking in terms of problems and solutions. Art historians have done this for a long while. Assuming that you want to suggest that something in the picture is farther away than something else, how do you do it? One way is through overlap, as in Egyptian art. Here the fishermen overlap the background, their legs overlap each other’s, and the strings of fish that one is carrying overlap some legs.
Later image-makers suggest variable distances through size variations, placement in the format (a little bit of that here, with the river above/behind the men), tonal contrast, atmospheric perspective, linear perspective, and other techniques. These can be considered solutions, available to artists of different times and places, to the problem of suggesting three dimensions on a flat surface.
A problem/solution way of thinking can clarify some developments in the history of filmmaking too. If you have to represent two actions taking place simultaneously, how can you do it? Crosscutting, as Griffith and others showed in the 1910s, solves that problem. It offers spillover benefits too, such as controlling pace. Similarly, there’s the problem of representing spoken dialogue. Silent films solved this in various ways—through a commenter in the theatre (the benshi in Japan), through actors voicing the roles behind the screen, and most commonly through intertitles. Later, synchronized sound solved the problem in a more thoroughgoing way.
These are very general answers to the how-did-it-get-there question. Occasionally we get more concrete information about problems and solution. For example, some Hollywood stars believed that one side of their faces was more appealing than the other. The stars with the most power could insist on being filmed on their good side, which led directors to make particular staging choices. (Claudette Colbert insisted her left side was her good side, so she’s usually positioned on screen right, with her face turned toward screen left.) David Butler knew that Edward G. Robinson likewise favored his left side, so Butler needed to stage Robinson’s one appearance in It’s a Great Feeling (1949) with him entering a scene from right to left and playing in that position.
One vain star is problem enough, but what happens when you have two who prefer being shot from the same side? According to Henry Koster, the demands of Cary Grant and Loretta Young led to the staging of the scene shown at the top of this section. (For my reservations, see the codicil to this entry.)
The Little Foxes production provides evidence of another very specific problem. In staging the staircase collapse, Wyler faced an unusual difficulty. The actor playing Horace, Herbert Marshall, had a prosthetic leg.
Marshall lost his right leg, from the hip down, in World War I. Through practice he managed to stroll quite smoothly nonetheless, and he became a significant star and featured player in theatre and films. He doesn’t need to walk much in The Little Foxes because his character is rolled around in a wheelchair. But the parlor-and-staircase scene was very demanding. As Wyler explains:
Now there was another problem involved with that, and that was the fact that Herbert Marshall has a wooden leg and couldn’t make the stairs, you see. This is a trade secret. I had him stagger in the background, get behind her and just for a moment when he gets to the stairs he had to go to a landing over there, and just for a moment went out of the picture. And a double came in and went up the stairs, staggered way behind out of focus.
Here you can see Marshall leave the foreground.
An axial cut in to Regina shows him stumbling behind her and going out of shot in the distance. This much Marshall could manage.
At that point the double stumbles into the frame and starts to crawl up the staircase.
Regina leaps up and runs to the rear, and the camera racks focus to the stair, but by now the double’s face is out of frame.
So the director solved the problem of the actor’s disability by a combination of deep staging, the use of a double, and shallow focus. This “trade secret” yielded a range of effects that, I think most viewers would agree, were vivid and exciting.
But there’s always more than one way to do anything. Given the constraint of Marshall’s artificial leg, or a player’s insistence on being shot from one side, or the leading lady’s overnight pimple, a director can work around it in several ways. One of the few critics to notice the implications of Wyler’s choice was Raymond Durgnat, a critic very sensitive to style.
Given a “pimple” or a “wooden leg,” different stylists will find different solutions. One changes the camera-angle; another introduces a last-minute panning shot; another will retain the original set-up, but throw heavy shadows to conceal the offending detail; another will interpose a pot of flowers or a table-cloth to conceal the trouble spot from the camera. The director has ample opportunity to maintain his style in the face of “accident.” And it’s no exaggeration to say that such stylists as Dreyer and Bresson would imperturbably maintain their characteristic style even if the entire cast suddenly turned up with pimples and wooden legs.
I’d add only that the director’s choices are further constrained. Beyond the immediate problem, the broader pressure of norms will kick in. The norms of classical studio lighting, cutting, and performance limit the ways Toland and Wyler can cover up Marshall’s infirmity. The norms of quality A-picture American filmmaking of the period militate against, say, editing the scene so that a dummy is substituted for Marshall on the stair. (We might get that in a serial, though.)
There are also the intrinsic norms set up in The Little Foxes as a formal whole. These favor handling the scene in depth in some way. Wyler reports the decision: “We said we’ve got to stay on Bette all the time and just see this thing in the background, see him going in the background, but never lose her.” Wyler’s earlier choices in the film created a kind of path-dependence for this critical moment. Deep-space staging could stay in tune with the rest of the film; but because of his actor’s infirmity, he could give up deep-focus cinematography. This solution created a vivid variant on the film’s intrinsic norm.
You can also argue that by deciding to call our attention to a distant plane in soft focus, Wyler fell back on something he had tried before. In the extraordinary late silent The Shakedown (1929), he showed a pie being stolen in a diner. First, there’s a close-up, then a shot of the main couple looking to the background. In the center, out of focus underneath the coffee urn, the pie is slipping away.
The action isn’t very discernible in my image, which is from a 35mm print; but the scene is shot quite soft anyway. I think audiences notice the gesture, slight as it is, because it’s centered and nothing else is moving in the frame. More visible is the background action in a shot Wyler and Toland used in Dead End (1937). Two gangsters are sitting in a bar debating kidnapping a child. In the out-of-focus background,we can discern a woman wheeling a baby carriage along the sidewalk. She isn’t the target, just a sort of reminder of children’s vulnerability. As in The Little Foxes, a centered background action attracts our attention and makes us strain to identify it.
Faced with a similar problem in The Little Foxes, Wyler had the chance to dramatize a soft-focus background to a much greater extent than in these films.
One more causal factor might have shaped Wyler’s decision. Lillian Hellman’s original play takes place wholly in the Giddens’ parlor and the hallway behind. The play text indicates that the staircase is in the rear of the set, with a landing offstage. The furniture sits downstage, closer to the audience. The foreground/background interaction in Wyler’s staging is already there, in a rougher form, in the play’s set arrangement.
And how does the play handle the moment of Horace’s collapse? When Horace’s medicine bottle breaks, Regina doesn’t move. Calling for Addie the maid, Horace leaves and staggers to the rear playing area.
He makes a sudden, furious spring from the chair to the stairs, taking the first few steps as if he were a desperate runner. Then he slips, gasps, grasps the rail, makes a great effort to reach the landing. When he reaches the landing, he is on his knees. His knees give way, he falls on the landing, out of view. Regina has not turned during his climb up the stairs. Now she waits a second. Then she goes below the landing, speaks up.
REGINA: Horace, Horace.
The foreground/background dynamic, as well as the frozen indifference in Regina’s performance, are written into the scene’s stage directions. Hellman’s instructions yield a further hint: Horace “falls on the landing, out of view.” Within the norms of the deep-focus aesthetic, Wyler and Toland found a cinematic equivalent for this barely-offstage action–one appropriate for their film’s particular style. They make Horace present, but he’s “out of view.”
Somebody may say: “See? You don’t need all this fancy analysis. At bottom, Wyler was forced to shoot the scene this way because of Marshall’s bum leg.” This retort assumes that causal factors always trump functional ones. Instead, I think that by considering causal factors, insofar as we can know them, alongside functional ones, we can better understand filmic creativity in history.
Durgnat’s point shows us how. Even when contingent circumstances “force” a filmmaker to change course, there are always several ways to do that. Picking any option brings in a cascade of other constraints and opportunities. Once Wyler has decided to double Marshall and sustain the take on Davis, soft focus is more or less necessary so we don’t spot the stand-in. But the soft-focus provides a nifty opportunity to create the sorts of functions and effects we’ve already noticed.
Like everybody else, filmmakers choose within constraints—some apparent, some less visible, many just taken for granted. Those constraints limit what can be done, but they also enable other things to happen, perhaps things that the filmmaker couldn’t have planned in advance. Once other filmmakers realize the results, they can plan in advance. A moviemaker today can try out Wyler’s solution, free of the pressures that drove him to it. A significant part of filmmaking’s traditions may consist of workarounds.
The Ostlund epigraph, apparently not available online, is taken from Hollywood Reporter’s December awards issue, p. 13. My Egyptian picture comes from the Metropolitan Museum of Art. It’s called Fish Preparation and Net Making, from the Tomb of Amenhotep (1479-1458 BCE), as rendered by Nina de Garis Davies. I draw the stage directions in The Little Foxes from Lillian Hellman, The Collected Plays (Little, Brown, 1971), 195.
My quotations about It’s a Great Feeling and The Bishop’s Wife come, respectively from two books by Irene Kahn Atkins, David Butler (Scarecrow, 1993), 227; and Henry Koster (Scarecrow, 1987), 87. Koster’s memory fails him in his account of the Bishop’s Wife window scene. It seems likely that Loretta Young favored her left side, which is her dominant orientation throughout the film. But there’s no evidence in the film that Cary Grant favored that side of his face too. The scene at the window is too brief to count as an instance of much of anything.
When I wrote On the History of Film Style in the mid-1990s, I had the nagging memory that Marshall’s artificial leg played a role in Wyler’s staging, but I put it down as legend. (It’s a pity I didn’t pursue it, because the information would have fitted snugly into my sixth chapter.) Only when I discovered a 1972 interview with Wyler, with the “trade secret” mentioned above, did I realize there was something to the story. That interview was once online, but seems to have vanished. It’s available at Columbia University. Durgnat’s discussion is in Films and Feelings (MIT Press, 1967), 41. My other quotations from Wyler come from Axel Madsen, William Wyler (Crowell, 1973), 209.
Otis Ferguson reported on the filming of a different scene in The Little Foxes; I discuss that here. More generally, on the Bazin-Wyler connection, see this entry. Other Wyler-related entries can be canvassed here. For more on Hollywood’s development of deep staging and deep focus, see not only On the History of Film Style but also Chapter 27 of The Classical Hollywood Cinema: Film Style and Mode of Production to 1960. As for Bette Davis’s eyelids, much in evidence here, there’s this entry.
In The Sound of Fury (aka Try and Get Me!, 1951), Howard Tyler has drifted into crime under the guidance of a breezy sociopath. They commit a string of holdups, culminating in a kidnapping. Howard’s partner bashes in the skull of their young captive. Wandering drunk and despairing, Howard ends up in the apartment of Hazel, a lonely manicurist. As Howard lolls on the sofa, she turns away to switch off the radio.
The next move is up to us.
If we’re alert, we can spot, on the end table in the corner of the frame, a newspaper with a headline that may be announcing the police investigation.
At first Hazel takes no notice. Will she? She does. She lifts the paper and is appalled.
Hazel turns toward Howard. Now we can see the entire headline as she reads aloud: Police are intensifying the search. She hasn’t made the connection between her guest and the boy’s disappearance.
Panicked, Howard lunges at her and crumples the newspaper.
Will this display of shattered nerves tip Hazel off?
As in the bomb-under-the-table model of suspense, at the start we know more than both characters know. She’s unaware of the kidnapping, and he’s unaware that the cops have found the victim’s car. In addition, the arc of suspense around the headline is quite small, though it leads on to something larger: Will Howard give himself away to the unsuspecting Hazel?
I’m impressed by the economy of presentation. Hitchcock might well have treated this moment in point-of-view shots, and a fairly protracted series of them. Or imagine how several filmmakers today would have handled this scene. There’d be a slow a track-in to the headline, then a circling camera movement that first concentrates on the woman picking up the paper, then racks focus to Howard on the sofa in the background.
Instead, director Cy Endfield makes very small changes of framing and staging matter a lot. The camera simply swivels, the actress simply comes to the foreground and pivots. The entire action, crucial as it will prove in what follows, consumes only twenty-five seconds.
Some stretches of a movie tend to be simply, barely functional: connective tissue or filler. Shots show cars driving up to places where the real action will take place, or characters striding down a corridor before going into a doorway. Other images want to engage us more deeply, but they do it through immensity. They try to awe us with majestic swoops over the sea or into the sky. (Recent example: Interstellar.) But other films engage us through detailing. They train us to notice niceties.
The Sound of Fury moment creates its detailing through visual space. What about time? And what about auditory factors? Our old friend, the telephone call, can furnish some examples.
Clay Pigeon (1949).
Filmmakers must always decide how much of any action to show. Sometimes that allows the director, the cinematographer, and the editor to create fine-grained delays. These might not build up a lot of suspense but they can make us uneasy, and prepare us for a surprise later down the line.
As we mention in Film Art, and discuss in a related blog entry, a telephone scene forces the filmmakers to choose among clear-cut alternatives. Do we see both parties? Do we see only one and simply hear the other? (And is the voice of the one we don’t see futzed?) Do we see one and not hear the other at all? Most films don’t ask more than simple functionality, but even a B man-on-the-run feature like The Clay Pigeon (1949) shows what can be done with details of timing in setting up a phone call.
Jim Fletcher has war-related amnesia. He doesn’t know why he’s about to be court-martialed for treason. After escaping from the hospital, he learns that he is accused of betraying his best friend during their time in a Japanese POW camp. After convincing Martha Gregory, the friend’s widow, that he’s innocent, he searches for proof. The Clay Pigeon sticks mostly with Jim, but like most suspense films it slips in bits of unrestricted narration as well. Jim’s quest is tracked by mysterious men, and brief scenes give us glimpses of the forces pursuing him: agents of Naval Intelligence, and a gang of counterfeiters protecting the Japanese soldier who tortured Jim in the Philippines.
It’s the familiar structure of the double chase, dosed with minor mysteries. For example, when Jim gets a lead from a management firm, he leaves the office but the narration stays with the secretary who notifies her boss that Jim has been asking questions.
Cut to the executive’s office, where the camera reveals many stacks of wrapped bills on his meeting-room table. Something sinister is going on here, but what?
The decision to insert information addressed to us alone has more subtle consequences in two telephone scenes. Jim calls Ted Niles, another veteran of the POW camp. During these scenes, the filmmakers had the option of showing only Jim and never revealing Ted at the other end of the line. That tactic would have enhanced mystery, but it would have thrown suspicion on Ted. If he’s Jim’s friend and ally, why not show him?
So the filmmakers show Ted replying in his apartment. But later it will be revealed that Ted is working with the gang. The task is to introduce this important character in a way leaving open the possibility of his treachery. The solution the filmmakers hit upon is to show Ted just before he picks up the line. Here is the first instance, when Jim cold-calls him.
The camera shows Ted innocuously answering the phone and learning, to his surprise, that Jim has tracked him down.
At first Ted seems annoyed, but then he smiles and agrees to help.
The scene ends on Jim hanging up. If we wanted to plant more suspicion of Ted, we’d show him hanging up too and reacting to the call.
A later scene starts much the same way, with Ted coming in to answer a ringing phone and getting a message from Jim.
Both scenes show Ted answering the phone in a completely innocuous way. Yet the very fact of dwelling on his action of coming to the phone can be seen as planting uncertainty. In the second scene, for instance, where is he coming from? And in both scenes, Ted frowns at certain points. Perhaps he is pondering ways of helping Jim, but the expressions leave open the possibility that he is plotting against him. Ted’s duplicity is fully revealed only at the climax. (See image surmounting this section.)
In a mystery situation, a few seconds showing Ted alone gain a force they wouldn’t have in another genre. Some viewers will be surprised, some will say they knew it all along, but either way the detailing of a moment here and there has opened the possibility.
The Clay Pigeon telephone scenes show the speakers in alternation. The give-and-take of the conversation is presented by cutting back and forth. Another option is simply to show one speaker and let us hear the other without seeing him or her. As we’ve noticed, though, that would tend to make Ted a more mysterious figure.
Yet another possibility is the silent treatment: One speaker is shown talking, and we don’t hear the other at all. This option forces our attention wholly onto the reaction of the person we do see, and keeps us in the dark about the words and tone of voice of the person at the other end of the line. If the Clay Pigeon telephone calls presented Ted this way, that would be another tipoff.
Still, suppressing one half of the conversation can pay dividends when we already know the characters. At the climax of Humoresque (1947), detailing involves not a prop or a passing moment. Instead, a simple cut accentuates the shift from one sound space, that of violinist Paul Boray’s dressing room, to another, the luxurious living room of his lover Helen Wright. When he gets her call, he can’t understand why Helen isn’t at his big concert. But she is distraught because her own worries about keeping Paul’s love have been reinforced by Paul’s mother, who insists that she’s no good for him. And Helen is drinking again.
The scene’s tension is ratcheted up by first presenting only Paul’s angry questioning. We don’t hear Helen’s replies. When the dramatic momentum shifts to Helen’s desperate excuses for missing the concert, we concentrate on her meltdown more intently because now we don’t hear Paul’s replies. Her emotional response is magnified by the yearning climax of Tchaikovsky’s Romeo and Juliet Overture on the radio broadcast–another reason to suppress Paul’s voice.
The scene has been split between Paul’s end and Helen’s. By not seeing Helen’s reaction to his urgent questions, we wonder what is keeping her away. Like Paul, we’re unaware of her torment. But then we see and hear her, and our inability to know what he’s saying makes his pleas seem ineffectual. Whatever he’s saying doesn’t seem to matter. A simple speaker/listener cut raises the scene to a new pitch, which will build still further when we follow Helen out onto the terrace. One more detail, brutal: We don’t hear Paul’s voice, but we do hear the click when he hangs up.
One thing that links all these Little Things: What the filmmakers did not do. Cy Endfield did not indulge in camera arabesques or POV cutting. Richard Fleischer and his colleagues did not suggest Ted’s duplicity with music or a noirish shadow. Jean Negulesco and company didn’t yield to the temptation to crosscut furiously between a panicked Paul and an anguished Helen. These directors did something rare today. They presented the situation with stylistic simplicity. That way the big moments–the revelation of Ted’s treachery in the train, the frenzied mob in The Sound of Fury, the all-enveloping climax of Helen on the beach–become more vivid. Big things need little things to seem bigger.
Thanks to Jim Healy, who introduced me to The Sound of Fury and The Clay Pigeon.
For more on the bomb under the table, see the followup entry here.
Lest someone think I’m dumping on Nolan, let’s just note that he can, when he wants, summon up niceties. (By the way, thanks to readers for hustling to our Nolan vs. Nolan entry, but they should read the one on The Prestige and our Inception series here and here to get a fuller sense of our estimation of him. All of these are put into reader-friendly order in an insanely inexpensive ebook…..)
Several other blog entries consider detailing in performance: Henry Fonda’s hands, Bette Davis’s eyelids, and the facial expressions in The Social Network. I’m still mulling an entry on eyebrows, which are terribly underrated. For another Joan Crawford tour de force, there’s this.
The Grand Budapest Hotel (2014).
Be shot-conscious! I urged in a blog entry some years ago. I illustrated the point with a tradition of staging and shooting that seemed simple and modest but was actually quite flashy, and even fashionable. Although many filmmakers resorted to it, either often or occasionally, critics hadn’t attended to it. Wes Anderson’s work yielded one of many examples of what I called (swiping from art historian Heinrich Wölfflin) a “planimetric” style.
Ideally, you should look at that entry before reading this one. (To encourage you, I link it again. Not for the last time.) Very briefly, this style involves a frontal presentation of the action. You frame people against a perpendicular background, as if they were in a police line-up. Usually you face them to camera, as in this shot from Godard’s Made in USA.
As we’ll see, sometimes you can frame the characters at right angles to the camera, or turned directly away from the camera. Here are examples from Napoleon Dynamite and from Angelopoulos’ The Traveling Players. (Is this the first time these two movies have been mentioned together?)
The key idea is that the people and the setting aren’t observed from an oblique angle; if the background is perpendicular, the people will stand or sit at 90 or 180 degrees to that.
You can arrange them in some depth too, but again, they are stacked in perpendicular fashion, making each area a pretty strict plane. Here’s an example from Pulp Fiction.
One point of my earlier entry is that this is a surprisingly old strategy; Keaton used it occasionally, and Godard was using it heavily fifty years ago. Here are two shots from Contempt (Le Mépris, 1963).
It has endured in some surprising places. It’s now a go-to option for one-off effects in mainstream cinema. Here are examples from Shutter Island and The Secret Life of Walter Mitty (2013 version).
A few filmmakers make it the basis of an entire film, as I indicate in this entry on Oliveira’s Gebo and the Shadow. And since I wrote the original entry, I’ve drawn on other examples from time to time, particularly from directors who are pastiching Ozu to some degree or another.
Still, Anderson is today the most widely visible example of the style, partly because while others use it sporadically, he is single-minded about it. He has made people shot-conscious (at least when they watch his movies). So after seeing his newest film, I thought it would be fun to think about what distinguishes his approach.
Playing with planes
With the release of The Grand Budapest Hotel, several bloggers have pointed to recurring compositional features, most obviously bilateral symmetry. I’d just add that such symmetry is often used by practitioners of the planimetric approach, with results that sometimes exceed Anderson’s. Here are two shots from Angelopoulos’ Weeping Meadow.
When you think about it, it takes a brave filmmaker (e.g., Godard) to use this approach and not deploy symmetry.
Anderson has used the planimetric approach more extensively in recent years, and he modifies it some distinctive ways. I think particularly of his habit of crowding people together in layers rather than stretching them along a single line. He makes some images look like group portraits or over-posed highschool yearbook shots (The Royal Tenenbaums; Fantastic Mr. Fox).
By employing the planimetric strategy, Anderson gains a somewhat awkward formality, a sense that we are looking from a distance into an enclosed world that sometimes looks back at us. There are as well the sort of comic possibilities that Keaton recognized in Neighbors and The General. A rigid perpendicular angle can endow action with an absurd geometry.
These apparently simple framings often evoke a world of childhood. Just as Kitano Takeshi shows us gangsters behaving like little boys, Anderson’s dollhouse-room frames make adults seem to be toy people arranged just so–like items laid out in a Joseph Cornell box. It’s a style suitable for magical-realist premises like The Life Aquatic with Steve Zissou, and in Moonrise Kingdom it finds its echo in children’s illustrated books.
All in all, then, I have to salute an American filmmaker who thinks about his images carefully and has incited sensitive viewers to notice them. I think we should go further, though. We can ask: How does Anderson, staying loyal to this tradition, vary the look of the shots? And how does he cut them together?
Consider the editing option first. Unless every scene is to consist of only one shot, the question comes up: How do you maintain the style while cutting? Either you make all your cuts axial, straight in or straight back.; or you create a sort of compass-point editing. This can involve cutting 180 degrees, to what’s “behind” the camera in the initial shot. So if characters are confronting one another, the camera is in effect sitting between them as each looks over or through the lens at the other (Ozu’s Late Autumn).
In effect, this option respects the classic 180-degree line, or axis of action, between the characters. It’s just that the camera sits right on that line. Parking the camera on the axis is a common tactic for subjective cutting, showing us first a person looking, more or less at the camera, then what she or he sees from their vantage point. Our example in Film Art: An Introduction comes from Rear Window.
Ozu used this 180-degree reversal often, but not absolutely; he had a more complicated way of conceiving space, and the 180-degree frontal cuts were only part of it. Kitano made a simpler variant central to his early films.
When I asked Kitano why he did it, he explained that it was exactly the way people saw each other in ordinary life. We face each other. He then added that he was such a naive director when he started that it was the only way he knew to set up scenes. We get kindred images in Terence Davies’ work; his frontality may owe something to the Hollywood musical.
Compass-point editing offers another possibility, that of cutting at 90-degree angles to the background plane or the figures’ position. Chantal Akerman does it throughout Jeanne Dielman 23 Quai du Commerce 1080 Bruxelles (1975).
Anderson exercises all these cutting options inThe Grand Budapest Hotel. Here a planimetric profile 2-shot yields two frontal shots; we shift 90 degrees and then 180 degrees.
Now here’s a 90-degree shift for the reverse shot.
In the passage below, the first cut rotates 90 degrees, and the second cuts in right on the lens axis. In this tradition, an axial cut respects the perpendicular layout of the space.
In such cutting patterns, the compositions keep the action in the same upper zone of the frame from shot to shot. As a result, our eye doesn’t wander much. In long shots, Anderson sometimes follows the classic Hollywood practice of allowing some decentering, as long as the cuts balance one off-center composition against another. Here the changing angles obey the compass-point principle across three shots, and they crisply shift the emphasis from the right side of the frame to the center to the left.
Someone who wanted to deflate Anderson’s visual ambitions could say that his shots are monotonous. Having imposed a big constraint on himself, he’s now obliged to show us that this approach can be varied–in obvious or subtle ways.
One way is through lens length. Most planimetric filmmakers use long lenses, which flatten the space even more. The figures can look like clothes hanging on a line. But Anderson favors quite wide-angle lenses (often 40mm). These make horizontal lines bulge, as in early CinemaScope films (Rushmore, The Life Aquatic).
You can see similar distortions in the straight-on shots of the hotel desk in Grand Budapest, above.
Another way Anderson varies his images is by departing from straight-on angles. As long as the framing maintains a planimetric geometry, we can look down or up at the action. In this passage, again the camera makes 180-degree reverses. This contrasts with the more orthodox shot/reverse shot framings in a comparable scene in The Little Foxes.
In this spirit, Anderson can give us bird’s-eye views, as Matt Zoller Seitz points out in his sumptuous book-length interview with the director. It’s rare, but there are precedents, as in the work of the Coens. In one shot of The Hudsucker Proxy, a movie with an inordinate number of straight-down angles, the inflexible framing creates a joke.
Grand Budapest Hotel has room for some classically funny framings. If you want somebody to look lonely, common practice says, frame the figure off center in a long shot. Here Anderson seems to be having a joke on the convention. He presents it as a POV, although presumably if the Writer were looking at the mysterious man he would put the object of attention in the center of his field of vision.
I think that Anderson’s earliest films weren’t quite so strict in obeying the planimetric and compass-point strategies. Those options were often slipped in as alternatives to more orthodox framing and cutting. But as he’s become more rigorous about using them, he has found ways to put his stamp on some common techniques. Like Ozu incorporating devices of classical continuity into his unique stylistic system, Anderson can recruit certain conventions while staying faithful to his basic approach.
For instance, Anderson sneakily brings in the OTS–the over-the-shoulder framing standard in shot/ reverse-shot dialogue scenes. In one prison scene, Harvey Keitel’s Ludwig is granted an OTS that varies subtly from the more purely straight-on views.
Much the same thing happens with in the punching scene at the reading of the will, when frontal characters are assaulted by fists coming in as if in reverse angles.
Anderson has figured out another way to vary his compositions. I learned this before I saw the movie, thanks to some comments by the cinematographer Robert Yeoman (great name).
High or wide, and handsome
To get the criticky part of this entry out of the way: The Grand Budapest Hotel has all the charm, fussiness, and intricate whimsy typical of Anderson’s work. As often in his films, it cuts its preciosity with moments of offhand brutality (sliced-off fingers) and flashes of naughty sexuality (fellatio, the lesbian painting). With its ensemble cast, sometimes deployed in cameos, it suggests a PoMo remake of those sprawling, self-congratulatory spoofs of the 1960s like The Great Race, Those Magnificent Men in Their Flying Machines, and It’s a Mad, Mad, Mad, Mad World. (The film’s title evokes those all-star films set in hotels, like Grand Hotel and Hotel Berlin.) It’s much better than those, partly because it engages in an oblique way with history, creating a comic-pathetic alternative account of Nazi imperialism. It imagines the collapse of Europe in operetta terms, filtered through Anderson’s pawky humor and distinctive style. I admired the film but don’t feel able to analyze it much after only one viewing. Fortunately for me if not you, its stylistic aspects fit today’s sermonette.
The Grand Budapest Hotel is set in several time periods, and they’re presented via The Blog’s old friend, the device of flashbacks within flashbacks. One character recalls the past or tells a story, and inside that line of action another character recalls or recounts a story, and so on. In Grand Budapest Hotel we move from the present, more or less, to events in the 1980s, then the 1960s, and eventually the 1930s, which constitute the central episodes.
Anderson has shot the frame stories in different aspect ratios. It’s 1.85 for the near present and the 1980s, when the Author recounts meeting the hotel owner. That meeting, set in the 1960s, is shown in 2.40, the anamorphic aspect ratio. The central story, taking place in the 1930s, is presented in classic 1.37, or 4:3 imagery. With typical Anderson butterfly-collector wit, each era gets a ratio that could have been used in a movie at the time. It’s remarkable that Anderson could persuade Fox Searchlight to let him do this.
Most commercial releases in the 1950s and afterward were filmed in some widescreen ratio. In the early days, a popular option was a sort of clothesline staging, centering a single character or balancing others around the central axis: two side by side, three across, four as a pair of pairs, and so on. These shots are from Demetrius and the Gladiators and How to Marry a Millionaire.
Thanks to the widening of the frame, there is less air above the characters and less ground below them. The empty spaces are typically on the sides, particularly in the anamorphic 2.40 ratio. The problem of filling that up was solved, at least for some directors, by moving the camera very close to the actors. Spielberg remarked that he began shooting more close-ups when he filmed in anamorphic.
If you’re inclined to the planimetric approach, it fits the wider format nicely. Anderson wasn’t worried by the extra acreage; he just used the set or empty areas to balance one side against the other. Shots of only one character could be centered, as if posed, and shots of groups could be arranged more or less symmetrically, as in this passage from Moonrise Kingdom. Central perspective helps drive your eye to the main items.
In Grand Budapest, Anderson’s signature framings fit snugly into the scenes shot in 1.85 and 2.40. (The latter has been his favored ratio over the years.) But what about the 1.37 scenes? This brings me to Mr. Yeoman’s remark.
Explaining why he and Anderson watched a lot of films from the 1930s, especially by Lubitsch, Yeoman notes:
We looked at those more to familiarize ourselves with the 1.37:1 aspect ratio, which Wes wanted to use for the 1930s sequences. This aspect ratio opens up some interesting compositional possibilities; we often gave people a lot more headroom than is customary. A two-shot tends to be a little wider than the same shot in anamorphic. It was a format I’d never used before on a movie, and it was a fun departure. You can get accustomed to 1.85 or 2.40 to the point that the shots become more predictable.
Put it another way: Anderson’s penchant for centering and symmetry inclines him toward widescreen compositions that could be simply cropped right and left to fit the 1.37 ratio. His single characters and huddled groups could remain much as before. But in more distant framings you might get a lot of extra space at top and bottom–areas that simply aren’t there in the wide ratios. In other words, Anderson’s multi-format strategy gave him a new problem in maintaining his signature style.
How did he solve it? Many Budapest Hotel shots do leave considerable headroom, as you see in most of the 1.37 examples above. But other shots show Anderson filling his 4:3 frame in varied and engaging ways.
As Ozu showed, for instance, the planimetric option can fill the frame’s upper area when the camera height is below eye level. During the conversation in the car, above, Anderson gets the head of M. Ivan (Bill Murray) in the top of the frame thanks to a low angle. Here are two more examples of filling the upper reaches of the format by use of a lowish camera position.
In the elevator shot, the headroom becomes comic, with M. Gustave and Madame D. seated on the right, the morose bellboy filling the vertical area on the left, and Zero in the middle. The empty space above the couple creates a lively imbalance emphasizing them in a way different from the very balanced framing that centers Henckel among his men.
The set can cooperate. In the first shot below, Zero’s and Agatha’s centered embrace leaves lots of headroom, but the slightly disheveled stack of pastry boxes in the upper background contributes to the sense that they’re engulfed. In the second shot below, part of its humor comes from the rigid geometry of the grid and the way M. Gustave and his colleagues fill in the matrix with their intent faces and busy hands.
In all, Anderson seems to me to find intringuing ways to create visual interest in the 4:3 format. But as with any severe style, you wonder about what’s been lost.
Most obviously, Anderson loses some of the intimacy that comes with more angular and less strict approaches to the classic ratio. We like to see people from 3/4 views too. We also like depth shots that plunge us into a dynamic, diagonal playing space. Here’s a shot from John Huston’s In This Our Life, as precious in its own way as Anderson’s imagery.
As Hogarth pointed out, with the serpentine line in painting and drawing, such shots can lead our eye on “a wanton kind of chase.”
Because directors of the 1920s-1940s accepted a wider range of compositional options than Anderson embraces, headroom simply wasn’t an important problem, as in the Huston shot. Even in simpler shots, classical uses of the 4:3 ratio permitted a flexible organization of figures.
Centered symmetry against a flat ground is a fairly easy compositional strategy, after all. It wasn’t used much in the mainstream tradition because it looks artificial; perhaps only with the rise of art cinema was this sort of self-conscious composition welcomed. In any case, sticking with symmetry sacrifices the more delicate spotting of figures and faces around the frame.
A lot of visual art tries for more supple and subtle twists, torsions, and counterbalancing. Apart from organizing your space along the horizontal and vertical axes, you can try to set figures in delicate array along diagonals. This is why some old-time cinematographers argued that the 4:3 ratio was the best suited to the human body: it can flatter it from any angle.
To get a sense of these possibilities, I’ve compiled a little collection of images from a film that doesn’t boast any outrageously pretty shots: Otto Preminger’s Angel Face (1953). It’s typical of the unassertive approach we find in Preminger’s work of the 1940s and 1950s. He avoids the flashy depth of the post-Kane directors and offers something less aggressive but no less fascinating. Composition and staging integrate expressions, posture, glances, and gestures to create a smooth flow of action. My samples also indicate how rare straight-on views of faces and bodies are in American studio cinema. The 3/4 angle rules.
As with the American films of Lang, Preminger’s work displays a style that’s tough to analyze because the technique isn’t obvious. There’s a marvelous variety in the ways that the 4:3 ratio can render a single figure or two figures, or three, shifting them not around the perfect center of the picture format but around curves and diagonal axes–that yield interest in their own right.
This last comparison isn’t a slam on Anderson. I think well of many of his films, particularly the most recent ones, and I appreciate anyone who takes on a challenge of narrowing his range of creative choices. Once you narrow that range, it turns out there’s a host of new possibilities that pop out. Call it the Ozu strategy: refine your means and you discover nuances nobody else notices.
Still, in art as in life, every choice is a trade-off. It’s worth remembering what one loses by pursuing a particular path. By sticking to his signature look in working with 4:3, Anderson gave himself a problem that didn’t exist for directors of an earlier time, the problem of maintaining a frontal style in a squarish format. I’m glad he faced it and solved it. But I’m also glad that classical filmmakers, quite intuitively, showed how much you could do with an alternative option.
Iain Stasukevich’s American Cinematographer article on the making of The Grand Budapest Hotel is well worth your attention beyond the technical matter I latched onto.
The Huston image came to hand because of the previous entry. Go there for more instances of the sort of framing and staging that Anderson and his planimetric colleagues don’t aim at.
I survey the planimetric style in On the History of Film Style and in Figures Traced in Light. A search of this blog’s archive will bring other instances to light. I analyze Ozu’s style in Ozu and the Poetics of Cinema, available as a pdf here. For more on CinemaScope, you can visit my online lecture.
P.S. 27 March (Hong Kong time): Jonah Horwitz writes with a useful point:
One thing I would add to your summary is that as early as Rushmore, most notably in The Darjeeling Limited, Anderson purposely inserts into his limited stylistic palette selected, isolated “foreign” devices like loose framings, handheld camera, and relatively aimless zooms (as opposed to his more common precise shock-zooms). In some cases, as in the drama-club staging of “Serpico” in Rushmore, these devices serve as citations, in that case to “realist” New Hollywood cinematography. But they also feel very much like the exceptions that prove the rule: they stand out from his usual stylistic register so much that they effectively reinforce the latter. I’m looking forward to seeing Grand Budapest to see if this continues, or if he emphasizes instead a further refinement of his typical gestures.
I agree with Jonah that importing foreign devices often throws into relief a filmmaker’s signature style–a matter of a film’s intrinsic norm getting reinforced by some marked deviation from it. I think of Ozu’s pans or tracking shots, which occur in all his black and white films, and which often just remind us how narrow the style is in the rest of the movie. And sometimes, as in The Flavor of Green Tea over Rice, those camera movements are hybrids or compromises with with his static style. Thanks to Jonah for corresponding.
P.P.S. 27 March: This entry has been revised to eliminate an error. Originally I had said that the play with aspect ratios in the film wouldn’t have been possible before digital projection. Bryce Utting wrote to point out that it was indeed possible on film, since Peter Greenaway’s Pillow Book used both 1.85 and anamorphic widescreen. I had even seen the film and forgotten that! Thanks to Bryce for the correction.
P.P.P.S. 30 March (Hong Kong time): Jim Healy, impresario of our Wisconsin Cinematheque, writes to point out several other films that mix aspect ratios:
The first hour of Redford’s The Horse Whisperer, the urban-set part, is in 1.85. When the characters make it to the open horse country, the image widens to ‘scope. . . . The 2002 Disney animated feature Brother Bear (which isn’t so bad) is 1.85 for about the first 20 minutes and when the principal Inuit character (voiced by Joaquin Phoenix) is transformed into a bear, the picture goes to Scope.
The Grand Budapest Hotel (2014).