My favorite class to teach is a seminar on how to make 3D movies. When I lead a course in this subject, I try to encourage students to explore the outer limits of the form. We begin with a description of what 3D movies are, and what they could be.
To start, I explain that a 3D movie is nothing more (and nothing less) than two movies that a viewer happens to watch at the same time. A 3D movie's most fundamental property—the thing that makes it different from a regular 2D movie—it that it delivers separate streams of images for a viewer's left and right eyes. Most of the time, the "left eye movie" and the "right eye movie" are very similar to one another. 3D moviemakers usually manipulate the two-eye delivery system to create harmonious stereoscopic illusions, which give the viewer a sense that a movie's frame extends behind and in front of the plane of the screen. If a stereoscopic illusion is made in the conventional fashion, we combine those two slightly different views of the same thing in order to activate the brain's sense of stereopsis or "solid seeing."
This is the main appeal of 3D for most people who enjoy (and who can actually perceive) these movies. It recreates an important aspect of the way we see things in real life. It's a great trick with wide expressive potential, and I'm all for it. But there's no reason why the two movies that compose a 3D movie must work in the service of this kind of illusion.
I want my students to understand that mismatched (or "non-stereoscopic" or "dissonant") 3D is the unexplored frontier of 3D cinema. Many new 3D filmmakers discover this type of 3D by accident. Their stereo cameras might be misaligned or synchronized improperly, or their left and right eye views might get swapped in post-production. With smaller 3D rigs, a student's finger might accidentally obscure one of the lenses, etc. As you'd expect, viewing mismatched or unsynchronized left and right images can be difficult, or even painful. James Cameron calls the feeling "brain shear," defined as "the brain's inability to reconcile the images received by the left and right eyes into a coherent stereo image… which leads to headache and sometimes nausea." But if we decide to embrace brain shear, we can open ourselves up to new intellectual discoveries and perceptual pleasures.
Students are pretty good at thinking up new ways to mismatch 3D image pairs. I've had students decide to shoot totally separate scenes for the right and left eye, then merge them in the editing process. Another student created the visual equivalent of binaural beats, measuring a brain's alpha waves and translating this data into flicker films with alternating patterns for the left and right eyes. I've also had students play with 3D in more subtle ways, for instance filming the same scene from the same vantage point in consecutive takes, but swapping out one important detail (say, a different actress sitting in a chair toward the back of the room). Then they composite those takes into a 3D movie. The end result is a movie that looks mostly normal, but for that one detail that was swapped out on the second take. (The left eye see actress A, the right eye sees actress B.)
While many new possibilities for this type of 3D remain, mismatched image pairs have already been used to good effect. As Keith Sanborn demonstrated with his "retinal rivalry" films from the 1980s, and Godard demonstrated more recently in the most widely discussed shots from Adieu au langage, brain shear can hurt so good. When a viewer's left and right eyes are forced to go blatantly out of sync, the he or she experiences a novel way of seeing. The oscillation between what the left eye and right eye see is more complicated to reconcile, and more viscerally intense, than viewing a simple 2D "double-exposure." And in that experience, we gain a deeper understanding and appreciation for normal stereo sight, the ability of our brains to synthesize an image from two separate vantage points, like it's no big deal at all. (The fact that those of us with two functioning eyes are accustomed to seeing the world from multiple and simultaneous perspectives suggests that we are all visual relativists.)
Again, this is not new exactly. Intentional brain shear actually showed up in the first 3D feature film ever made, 1922's Power of Love, which supplied two simultaneously viewed alternate endings—one for the left eye, another for the right. Brain shear is not and never has been the enemy of traditional 3D; it is an expansion of the technique's uses.
Though I attempt to generate enthusiasm for new forms of brain shear, many students elect to stick with 3D's original use—to simulate the way our eyes work together to perceive stereo depth. This use has been with cinema from the beginning, and it continues to fascinate. The film theorist André Bazin famously saw 3D (or "relief") as one of the components of mankind's idée fixe that gave rise to a string of mimetic technologies—apparatuses that that would seek to create a total simulation of life, a "total cinema." The technologies of proto-cinema took a while to coalesce and mature, but it's clear that 3D was always part of an idealized notion of what cinema should be, even before it was a mature technology. There is historical evidence that 3D was a major concern of many of cinema's pioneers. For examples (including cool illustrations from patent applications), check out Ray Zone's excellent Stereoscopic Cinema and the Origins of 3-D Film, 1838-1952. Zone describes how many early and proto-cinema inventors (including Muybridge, Edison, the Lumières, etc.) were heavily invested in stereoscopic technologies, still and moving.
Of course, the same can be said of other components of cinema. Sound and color, for example, were also early goals of cinema, and they, too, took a while to get figured out. But unlike 3D, they never faded in and out of style: Once they arrived, they were here to stay. Stereoscopic cinema, on the other hand, has existed since the first decade of the 1900s, but its popularity has waxed and waned many times. See this Sony-sponsored infographic for a lovely visualization of 3D's peaks and craters, through 2011:
Why is this? Perhaps, every time there's been a 3D phase, audiences have tired of the physical discomfort that 3D films produce…or perhaps they start to feel ripped off by greedy theater owners charging too much for tickets to a crappy slate of 3D films. But I want to convince my students that there is an equally important reason why 3D has not taken hold as a universal part of the moviegoing experience—and it has nothing to with headaches or dorky glasses. The real reason 3D doesn't stick is that it simply isn't suited to the types of films which succeed at the box office. I truly loved the 3D conversion of Titanic, but most people would not consider the 2012 version that much different (let alone better) than the original. In fact, people will happily watch Titanic on a tiny 2D airplane seat-back screen, and the tiny screen does fine.
Which leads me to my grand analogy: Imagine cinema as the Earth, and 3D as its moon. In this analogy, the earth and moon were once part of the same mass, a big and primitive "proto-earth." But in the early phases of cinematic history, something happened, and 3D split off into orbit as a satellite. The historical trauma that sent 3D apart from mainstream cinema was not as abrupt and dramatic as the one that created the actual moon (a Mars-sized planet that crashed into us 4.5BN years ago). Instead, the trauma that detached 3D from Earth took place over the first few decades of cinema, as narrative conventions were being codified. 3D spun out into orbit when cinema collided with advanced narrative codes, during an extended moment that film theorist Noël Burch has referred to as the eclipse of the "Primitive Mode of Representation" by the "Institutional Mode of Representation."
Burch, like Bazin, argues that three-dimensionality was an aspiration of the early cinema. But the Primitive Mode of filmmaking (think fixed camera position, tableau style framings, a series of "attractions" presented in real-time continuous takes) was soon replaced by all the sophisticated shooting and editing innovations—the 180-degree rule, continuity editing, etc.—that are still in use today. Burch explains that the development of advanced narrative codes gave its own sense of multi-dimensionality that satiated viewers "more than all the ephemeral re-appearances of red-and-green or polarizing spectacles, raster screens, etc." ever could. 3D, like the Primitive Mode, didn't exactly disappear. But it has remained in orbit ever since the IMR collision. It remains bound to cinema by invisible forces, remote but not quite inaccessible.
The 3D Moon is haunting, beautiful, magical, romantic, enigmatic, and bumpy. Every now and then, people go outside and gaze upon it through special lenses. Sometime we even pay a visit to its alien surface. It may be that the current interest in 3D marks the beginning of our final colonization of the moon, a process by which 3D becomes the home environment for all of cinema. But if you ask me, it's more like we're in a period of increasingly accessible moon tourism. (And a little bit of scientific exploration.) We don't live on the moon yet, because our Earthly cinema of narrative conventions doesn't really need it and probably never will. Most people would rather stay where they grew up.
I urge my students to suck it up and accept the fact that 3D adds sadly little to the Hollywood narrative. It's time we stopped harping on what 3D can do for "storytelling." It does something different, and better. 3D is ideal for unusual cinematic forms, forms that veer away from the familiar Institutional Mode of Representation. It's no accident that some of the most enduring 3D movies of the current era are documentaries, dance films, musicals, and formally unusual narratives such as Gravity and Adieu au langage. Good 3D films go full lunar. We owe it to the future to continue the exploration of this alien surface.
Putting it another way, we should go ahead and enjoy the moon for the poetry it inspires. We shouldn't try to drag it down to Earth.