I just want to say: Wow! Here are frame grabs from three shots in Kenji Mizoguchi's 1936 "Osaka Elegy" ("Naniwa erejî") that strikingly illustrate what in-depth staging and deep focus are all about. In the first one, we see Ayako (Isuzu Yamada), a switchboard receptionist at Asai Pharmaceutical, in her glass operator booth. The image has an Art Deco sheen to it (highlights on shiny objects twinkle with little starbursts) and the frame is soft around the edges, but in the center the focus is deep. I don't know if the blurred edges are a separate artistic choice or simply an optical property of this particular lens.
Anyway, this is our introduction to Ayako, who appears in a close-up profile on the left side of the frame. She notices a woman enter the office to speak with Susumu Nishimura (Kensaku Hara), a young man with whom, we will learn, she has an undeveloped romantic friendship. Ayako places a call and we realize who she's calling when Susumu -- that's him in center frame -- answers it. (He may look a little out of focus in this particular frame, but that's only because he's moving his head.) She turns her face away from us to make eye contact with him across the room, and as she does so we see her face reflected in the glass of her booth. Fantastic! It's a spectacular composition, but it develops without seeming contrived, leading your eye from place to place in the shot quite naturally as you discover what's going on.
The second shot, from a little later in the film, shows Ayako back at the switchboard, but this time there's no action of interest in the outer room, so Mizoguchi doesn't call our attention to it. He leaves it out of focus and we stay behind the glass with Ayoko as she distractedly flips through a magazine. (One headline reads: "A Woman Corrupted By Money.")
In the third shot from the switchboard booth, the entire frame is in focus, emphasizing the proximity of the two characters. Again, she has called him -- this time to set up a rendezvous. She's about to ask him if he will lend her some money to pay off her father's debts...
These shots had a particularly strong effect on me since I've been thinking and writing about James Cameron's 3-D staging and depth of focus in "Avatar" -- and why his persistent choices strained my eyes and gave me a headache. More about that later, but the principle Mizoguchi demonstrates here is worth noting: Let the action and the composition lead the eyes where they want to go. If you're going to put interesting or relevant things in the background, let the audience see them. In focus. If not, don't stage and frame the shot with distracting action in the distance (or the foreground) that competes for the viewer's attention, especially when the 3-D effect tricks the eyes into believing they can be brought into focus.
Our eyes are naturally drawn to faces and eyes (and you can't miss the golden, saucer-sized peepers on Cameron's Na'vi), so a big face in a shot is going to capture our attention no matter what else is going on around, in front of, or behind it. The skill -- especially when working in 3-D -- is in composing the shot so that we in the audience see what we need to see effortlessly. The laziest, least creative way to do that is to simply rack focus, so that we're forced to look at only one thing at a time -- usually at the face of the person who happens to be talking at the moment. A movie may contain many different kinds of shots -- deep focus, shallow focus, telephoto, extreme close-up, extreme long shot, "plan américain," and so on. They're all part of the cinematic vocabulary and a director's job is to figure out how and when to use them most effectively.
19 Comments
Brilliant.
THIS IS WHY I READ YOUR BLOG!!!
This is a fantastic scene to demonstrate an effective use of deep space staging, and also the fact that Mizoguchi knows when he doesn't need the extra information in the background to distract us. The third shot is a uniquely inspired one when it comes to staging, I think, in the way that he manages to get both of the characters talking on the phone within the same frame, and that he can emphasize a character far in the distance with his back facing to us by placing him in the center of the screen.
However, what you have addressed about putting the important information in focus and the unimportant information out of focus is something that can be applied to much simpler (but still effective) setups. It's quite simple; if the character at a given moment is significantly unimportant to us that you need to put her out of focus, why on earth would you have her in the foreground, facing the camera? This is a bizarre staging choice even disregarding the fact that Cameron was making this movie to be seen in 3D. There are, of course, reasons that a director might want to have something out of focus but would want to attract our attention to us; maybe he wants us to notice some moving blob in the background but wants us not to understand what it is. However, no such inspired reason seems to be informing Cameron's staging choice here.
There is another logistical problem with shooting a 3D movie with a telephoto lens, though. As you said, it does narrow the part of the screen that's in focus. However, it also flattens everything that is in focus onto a single plane, which is the exact opposite of what you want to do if you're making a 3D movie.
There are, of course, many situations in which this type of staging and focus would be appropriate.
For example, if the shot is a point of view shot and you want to show that the offscreen character is looking past the person in front of her to see someone in the back of the room. Or, if it is part of a rack focus: the foreground character speaks, and then focus shifts to the background character for a reaction or for a line of dialogue.
I haven't seen Avatar, and I understand that this shot is here as part of a larger argument that Jim is making (and that 3-D adds a whole new component to consider.
But based on this screenshot alone, this does not seem like a bad choice. In fact, it seems to accomplish in one shot what most directors would need three to do: By keeping the male character in selected focus, we identify with him and what he's doing. And what is he doing? He's watching the female character and her actions, which we (the audience) also get to see (because she's out of focus, but not a total blur). Of course, we are looking at her from a different angle than the male character, but it is clear that he is seeing the same actions that we are seeing.
It seems to me that this is a more efficient shot than the typical medium shot of guy, cut to guy's POV of girl, cut back to guy's reaction that most directors would give you.
Like I said, I haven't seen Avatar (despite my defense here of Cameron's direction, I dread having to sit through his writing), so I'm basing this solely on this screenshot, but I don't quite understand what's so bad about this.
You're absolutely right -- shallow depth of field and rack focus can valuable tools. As you suggest, if the goal is to compose a two-shot in which one character is in front of the other (a composition often used when one is hiding something from the other, with one character hiding his/her face from the other, but not from the camera), the director may want to focus on only one face at a time to suggest a disconnect of some kind between them. In the scene above (leading up to the lovemaking scene under the tree), Neytiri is trying to find out how Jake feels about her by suggesting other mates for him. If both characters were in focus we would have the opportunity to choose at any given moment whether to watch the speaker or the listener, to gauge the interplay between them. And in 3-D, your eyes would be changing the focus between foreground and background -- which is why I say that doing it in the camera (or, in this case, the computer) is redundant. You're right: at least he doesn't deprive us of watching the shifting dynamics between the characters entirely by breaking up the scene into single reaction shots. Cameron gets us halfway there by staging the shot in depth this way, but then unnecessarily underlines/highlights the effect by rack focusing when we're perfectly capable of doing that with our own eyes. It's like any shot where, say, there's a conversation between characters in the front and back seats of a car, and we can choose where to look. By racking focus like this in 3-D, though, the director is going to give some in the audience headaches and eyestrain -- as has been the case with substantial numbers of "Avatar" viewers.
Thanks for responding, Jim. It seems like we're on the same page, and I agree with you that a director shouldn't be afraid to empower the viewer with the choice of what to look at in situations like this.
Imagine, for example, how terrible the last shot of Caché would have been if the two characters we recognize had been selectively in focus. Or if the shot of young Kane playing in the snow outside the window had featured rack focusing between foreground and background (probably not technically possible given the lenses Toland used, but you see my point).
Perhaps I should bite the $12 bullet and see Avatar, if only to participate more in the conversation.
My point wasn't that it was an ineffective way of communicating the scene, but mostly that his use of a narrow depth of field is not really conducive to working in 3D because 1) the 3D makes her face stick out far more than anything else on the screen, even though it's out of focus, and 2) telephoto lenses have a tendency to flatten space, when the entire purpose of 3D is the opposite of that. For 2D purposes, it's not bad, but for 3D I think that if he could have done this shot with both characters in focus and drawn our emphasis on them through staging, it would be a more effective use of the technology he's using.
You might need to get Mr. Cameron a copy of Dogme 09.8:
"9. Fit the format to the film. Your reasons for shooting on video, or 35 mm, flat or scope, [Or 3-D] should be apparent in the film itself. Take advantage of your chosen format."
Thanks for publicizing Mizoguchi (and for linking our blog, somehow invisibly: People are coming there from here!). This scene is especially interesting because so much of NANIWA ELEGY is shot at an unusual distance from the figures and employs Mizo's characteristic "distant depth"--the foreground plane is quite far from the camera, and lots is going on way, way back. In a way, this is a more daring option than the proto-Welles and -Wyler stuff we can find in his early 1930s films.
Plug coming up: If people are interested, I trace the history of depth staging in Chapter 6 of ON THE HISTORY OF FILM STYLE. Chapter 3 of FIGURES TRACED IN LIGHT is devoted to a close analysis of Mizo's staging techniques.
Our only difference of opinion is that I would defend rack-focusing between planes...when it's well-motivated, as in McTiernan's DIE HARD and HUNT FOR RED OCTOBER. Mizo does this sometimes as well.
Keep up the great work! In an era when people are gaga for Cameron's overblown look, we need to be reminded of what pictorial nuance looks like.
Thank you for putting Ozu and the Poetics of Cinema online for free.
Thanks for that, DB. I tried to suggest in my last two sentences that ANY technique can be effective, depending on the context and what the director is trying to do. The rack-focusing between planes in "Avatar" worked fine for me when I saw it in 2-D -- but not in 3-D, where (as in the image above), my eyes felt they could focus on Neytiri's interaction with the seed in the foreground, but they actually couldn't. Result: eyestrain and eventual headache. In 2-D, of course, the image remains out of focus, but your eyes are still focusing on the flat screen and not some illusory focal depth within the frame. No eyestrain from that.
I'm sorry, but just because you got a headache doesn't mean the rest of us did. I had no problem with the depth of field both times I saw the film in 3D. In fact I thought it was quite brilliant how Cameron managed to push certain things forward and other things far far back, establishing an even greater field and an even more realistic, immaculate sense of space.
If I was the only one who had this experience I wouldn't have written about it. But many "Avatar" viewers have had complaints. As I've said, if you're shooting in 3-D, then you are using that dimension of depth to guide viewers' attention. There's no need to use focus to do that, when you can use the properties of 3-D and let the viewers' eyes do it naturally. Rack focus demonstrates a limited depth of field, not a greater one. The eye can normally focus up close or far away at will; to use rack focus in 3-D is unnecessarily restrictive and misleading to the eye, and the cause of much eyestrain. There are plenty of good uses of shallow depth of field and rack focus, but in general, for the reasons I've been illustrating, I don't think Cameron's decisions were always the smartest or most effective.
Side note: Do you ever teach classes? Because maybe you should.
I work in production, as a documentary filmmaker with a company in the Midwest. I also do freelance video, and studied cinematography and photography as a student. And from that experience, I can tell you that the degree to which people working in this industry have NO GRASP of depth of field would appall you.
The current craze in the field of digital high definition video are lens adapters. What these do is allow a cameraman to mount interchangeable 35mm lenses-Canon, Pentax, Nikon, etc-in front of the primary lenses for their camcorders, which are often fixed and not interchangeable. This is done not because a person wants a lens for its sharpness or quality (indeed, mounting one piece of glass in front of another reduces light and sharpness). Instead, this is to recreate the shallow focus "Look" of film. Also, to mimic film!
And when I've talked to people who use these adapters, they invariably say the same thing: "This adapter is great! I can mount a 35mm lens and get that great depth-of-field look."
Think about that statement. They clearly have no understanding of what depth of field means, let alone are able to make an educated decision about what they want to keep in focus, and keep out of focus. They're suckers for spending money on these expensive adapters (they cost hundreds, even a thousand dollars!) when to get the look they want, would be as simple as manipulating the position of the camera relative to the subject, and zooming in, or using ND filters to open up their camera's iris more.
Many who I have encountered actually shun deep focus, not for artistic or stylistic reasons, but because, they say, it looks too much like video. They ACTUALLY believe that to achieve shallow focus is achieve professional calibre imagery.
It's not all their fault. I have to blame Hollywood for much of this. I find many films aggravating in how lazy their direction is, how damned reliant they are upon awful rack focus shots to tell their stories. "Spiderman III" was particularly egregious, perhaps because I saw it in Imax, where everything was magnified. Constantly, rack focus on one thing to another. I soon was able to predict when such a shot would happen, and it felt uncannily like the director was sitting behind me, nudging my shoulder, whispering, "Oooh, no, wait, look at that. Now look here. See?"
But some are learning. I for one practice religiously with 35mm and medium format still photography, to keep my skills sharp, to continue to learn to understand depth of field and composition. I've grown to love deep focus, and its ability to allow the viewer to form their own narratives. And by practicing deep focus in my work, it has helped my composition immensely. And best of all, by going against the tide, those who do will stand out against all the mediocre people wasting their money on lens adapters which are nothing more than snake oil for the cameraman.
BR
You write: it felt uncannily like the director was sitting behind me, nudging my shoulder, whispering, "Oooh, no, wait, look at that. Now look here. See?" That's exactly it! I find it exceedingly annoying when the director isn't skilled enough, or doesn't trust the audience enough, to shape the action without (metaphorically) circling it and drawing arrows to it. I've talked to many directors, and seen them at work, and I agree that the extent of the ignorance, even among so-called professionals, is appalling. Unfortunately, lucrative montage-heavy, short-form popular formats (commonplace styles for music videos, TV commercials) emphasize this approach of breaking things into little pieces, each of which offers one thing at a time to look at. Direction is largely about framing the action and guiding the eye. That's one of the things that distinguishes a good or great director from a mediocre or lousy one.
Excellent piece, and a big thank you for using Mizoguchi as your example.
This whole discussion of focus and the general decline of shot composition in favor of slash-everything-to-bits editing was put into, uh, focus (sorry, no better word comes to mind at 12:29 in the ayem) by two things. One was watching Götz Spielmann's "Revanche", which nailed the camera down and used good composition to guide the eye and fill the frame. No focus games, no razzle-dazzle crosscutting.
One day later I sat through the interminable remake of "Pelham 123" and while the DOF wasn't the only shallow thing in the film, it was sure one of the most obvious. Tony Scott seems to have this obsession with building toy traintracks around his characters and mounting the camera on them so they can be circled obsessively. The idea, such as it is, is to give us a 360° view of a conversation, but it's indulged in to such excess that it becomes sickening and not enlightening.
The more someone needs to move the camera or edit, the less comfortable they seem with just sitting and looking -- or effectively directing their actors within the frame, for that matter, which seems to be falling out of fashion since their performances don't seem to exist for any reason except to be slashed to ribbons by the editing anyway.
Sigh.
Any number of the films of John Frankenheimer could be used as examples too. His favorite shot seems to be a face in extreme closeup on one side of the screen and a full body shot of another character on another. There's also a shot like that in Ali of Mario van Pebbles as Malcolm X in extreme closeup and Will Smith as Clay behind him.
YES! this blog just basically saved me $100.000 from going to film school and learning this stuff. You truly know what you are talking about.
Leave a comment