Thursday, April 15, 2010

The Third Dimension is Ennui: A Treatise on Cinematic Revisionism

Was this inevitable? According to the LA Times, storied producer Richard D. Zanuck has confirmed the fears of internet fanboys. From the article:

Zanuck and Spielberg spoke a few years ago about going back to [Jaws] with the digital paintbrush of CG effects to create a more horrific predator. They decided, for the time being, to leave the film alone, although Zanuck says he is intrigued by the notion of adding 3D effects to the 1976 classic for a theatrical re-release.

I don't feel the need to gripe about Hollywood adulterating, soiling, and repackaging our favorite celluloid memories just to turn a profit. In this age of remakes, we expect that, and any objection I could offer here would just be redundant cyberspace chatter. George Lucas has recut, amended, "enhanced," and digitally micturated upon -- then subsequently resold -- the Star Wars films so many times that even hard-core fans have trouble counting the number of versions in existence. Business is business. What I'd rather do is talk about the formal consequences to the work itself when films are manipulated.

The example of Star Wars is from one end of the spectrum, in which a film's primary author has been able to revise a previous work ad nauseam thanks to new technology or to the creative freedom that comes with financial success. On the other end of the spectrum, a film can pass through the lacquering hands of just about anyone before reaching an audience. In the bygone days of VHS, we watched films that had been pan & scanned (read: "butchered and mangled"). Most of us who watch silent films do so with musical accompaniment courtesy of Joe Schmoe's synthesizer, set to "organ." Then, of course, there's the abhorrent practice of colorizing monochromatic movies. But even the maestro, Ray Harryhausen, has spoken out in favor of colorization, with the caveat that he be involved, and James Cameron himself was famously a proponent of director-supervised pan & scan for standard TV displays. So cinematic revision practices run the gamut, and should never surprise us. And, as the old saying reminds us: "Only the projectionist has final cut."
Therefore, let's not waste time expressing our dismay that Zanuck & Spielberg would even consider a CG and/or 3D facelift for what is considered, with little exception, one of the greatest horror films of all time. Of course they would. If there's a market for rerelease, Hollywood will brainstorm how best to saturate it (I make no judgement call here; I'm in favor of good films being seen by new audiences). Let us instead engage the film, not as commercial property, but as film, and convince Mssrs. Zanuck and Spielberg that an altered Jaws would be an inferior Jaws, not simply because we fear change in principle, but because these particular changes would do the film disservice.

And, before proceeding, let's come clean and admit it: the movie isn't without its flaws, if for only one reason. The shark always looked fake. But that's by no means an Achilles heel. Obviously, it never hurt the box office, and more advanced shark FX films (Deep Blue Sea, e.g.) haven't come close to eclipsing Jaws in the public's consciousness, so a fake shark is no more necessarily a liability than a convincing shark is necessarily an asset. In fact, the shortcomings of the titular character in Jaws may be indirectly responsible for some of the film's strengths (more on this later).

Now, in the case of converting Jaws to 3D, there are two operating principles which govern our skepticism. First of all, there is the question of contextual anachronism -- that is to say, imbuing a work with technical devices not available at the time of the work's original completion (colorization is the simplest example of this). 3D technology has advanced to a form far-removed from the blue-and-red anaglyphics that would have been available to Spielberg at the time of production. Therefore, the process of converting mid-70's film stock to contemporary digital stereoptics would necessarily give the film a look of incongruence. As when we watch Casablanca in color, or we see Buster Keaton run at the wrong speed (because the frame rate of silent era cinema is rarely compensated for in projection), the most fundamental elements of the moving picture (what makes film film) are being sullied. Even if you're not a student of film, your eyes perceive that something is off; the textures are mismatched. Some people don't mind this. Some people just want color. Some people can't stand to see black bars at the top and bottom of their standard TV's. Some people would watch every movie in 3D if they could, because they think it's neat. Yes, and some people unironically believe that a velvet Elvis is a tasteful, expressive work of art that ties any living room together.

Lucas's Special Edition Star Wars films offer a striking example of this contextual anachronism phenomenon. Our eyes know how objects appear in 70's films. Subconsciously, we recognize a palette of light, color, grain, and a vocabulary of movement that is unique to any given feature we watch, or even to any given reel of film. In short, we know that a digital Jabba the Hut does not, in a purely mechanical sense, belong in the original Star Wars:
Even if we concede that the computer-generated elements in this sequence are of notably poor quality (especially by Industrial Light & Magic standards) the fact remains that the same FX seem less out-of-place -- at least, as concerns the visual mechanics -- in the more recent Star Wars Episode I, a film made in the digital age (check out Jabba at around 1:40):

As concerns these two films, I deliberately am not accounting for perplexing intangibilities like overall quality (cough, hack). I merely suggest that the effect of running analog film through a digital workflow -- for anything more than restorative or delivery purposes -- has the potential to create anomalous results, like running a piece of papyrus through a laser printer. It's a cinematic paper jam waiting to happen.

This anomalous effect can be put to use to create uniquely expressive effects, but then we're veering off into derivative art, not revision. By silkscreening different colored Mona Lisas onto canvas, Andy Warhol creates something which is -- love it or hate it -- a new work, one that could not be achieved with precision through the use of oils alone.
Now, if Zanuck were suggesting a mere reediting of Jaws, we could still gnash our teeth, but he would not be suggesting the use of a device that would contextually make the film seem the anomalous product of another era. Take, for example, James Cameron's Aliens: Special Edition, a mostly improved version of the original film that offers no more than a few new scenes and recut sequences here and there (Coppola gave Apocalypse Now the same treatment with Redux, to much poorer reviews). There are no "enhanced visual FX;" it's just editing. That may beg the question, "Well why couldn't Cameron and Fox get it released right the first time?" A fair point, to be sure, but we can't claim that anything feels treated, or subtly off. The medium of the film -- its celluloid footprint -- has not been altered. Our eyes remain content, even if our hearts might yearn for the more familiar versions.

3D conversion, on the other hand, will not fool even the least trained of eyes. We know that we are watching a corruption, however exhilarating it may be. If recent 3D film conversions like Alice in Wonderland and Clash of the Titans have seen outcry at having been put through the stereoscopic ringer, then what hope does a 35-year-old film like Jaws have? Even Hollywood brass is wising up. Dreamworks CEO Jeffrey Katzenberg told Variety, "If we as an industry choose this 2D to 3D post-production conversion, it's the end. As quickly as it got here, that's how fast it will go away." So even the most perfectly executed 3D conversion, it seems, will be contested at best. The only people who enjoy it will be those who go home from the theater to admire their velvet Elvises.

But let us forgive filmmakers their toys. Let us suppose that Spielberg always wanted to make Jaws in 3D. So what? The fact is, he made a 2D film; and this leads us to the more fundamental governing principle of why any major revision of a film is dangerous. That is: it is the inherent limitations of film itself which inform the director's process, and, therefore, the end result of any cinematic undertaking. In other words: any given film is the product of all that which is and is not possible in filmmaking. To revisit the process with a new set of rules is to destroy and recreate the original work, for better or worse.

It is beside the point to assert that had he the option, Orson Welles would have shot Citizen Kane in color; the fact is, it's gorgeously shot in black and white contrasts. Chaney's Phantom of the Opera never glowered at Christine with a scream; he was always eerily silent. The original King Kong isn't a motion-captured digital construct; he was a pioneering example of stop-motion visual FX work. All Quiet on the Western Front was not given a 5.1 surround THX-certified soundtrack full of explosions; it was expertly mixed to contain an entire war in one sound channel. One might say that changing such facts would be inoffensive technical enhancement. To that, I ask, "Well, if it's just minor adjustment, then why do you need it to enjoy the film?" More importantly, that stance diminishes the work of some great filmmakers and craftsmen to mere trifling. It implies that cinematographers who shot black and white were doing something less valid, expressive, or interesting than what can be done in color. Tell that to Karl Freund or John Alton. There is no need to digress here into a seminar on color theory, in which one could easily make the case that monochromatic imagery is far more interesting and explicative than color. Suffice it to say, the consideration of outdated filmmaking techniques as lacking, in any way, is as absurd as saying that all paintings became obsolete at the invention of the photograph. Tell that to Da Vinci... or Andy Warhol, for that matter.

Every tool in the evolving bag of tricks of filmmakers -- from actors' performances and lighting setups to pancake makeup and dolly tracks -- has seen evolution over the course of cinema's history. Whether it's a lens choice or a line of dialogue, every frame of film contains within it a thousand decisions on the part of a production team. But what director, in the history of film, has ever made one of those decisions, saying, "One day I will be allowed to revisit this scene with a computer, so for now, it's ok if it's not so good?" Even an early director who saw the horizons of sound or color would not allow that hypothetical prospect to affect their decision making on a current setup. For years, Hollywood studios made some major motion pictures in color, and others in black and white; even the Oscars had separate categories for cinematography. Were the directors of photography who shot black and white during those years lighting for future color conversion? I doubt it. They lit for what they were working with.

This all comes back to limitations. Limitations, in every form, are omnipresent in filmmaking, as they are in any art. A particular film stock may only be so fast, requiring so much light that you don't have. The dame in that big musical number may be unable to hit high C. The budget doesn't allow you to recreate the Hanging Gardens of Babylon. The catering blows and the crew is about to walk. The shark looks fake. The flat screen keeps the objects in the frame from bursting into the faces of the audience. You see where I'm going with this.

In fact, the most fundamental limitation of film -- the four sides of the frame itself -- is directly related to 3D and the audience's sense of self. If we were to take away the boundary of the frame, and have a film, photograph, or painting fill our horizontal and vertical field of vision entirely, then the composition of every shot would be ruined. In fact, we would have to redefine what composition means (and painters would have to use a lot more paint). By extension, a film originally intended to be exhibited in 3D is one thing, but when we convert a 2D film -- when we add an effectively limitless Z axis to what will still have a finite X and Y -- then what are we corrupting? In The New Yorker, Anthony Lane did a thorough and entertaining piece on the technical, sensory, and social history of stereoscopic imagery in the movies. He does well to discuss Edwin S. Porter's The Great Train Robbery (from 1903 mind you!), in which a gunman points his revolver at the camera and fires. The stories say that viewers fainted or ran screaming from the theaters. That's an image with 3D aspirations; in it, the filmmaker acknowledges the presence of the audience directly, incorporating the viewer and the film into one perceived spacial continuum. Jaws has its counterpart to Porter in the monster's "comin'-at-ya" leap to the lens. Whether it's a bullet or a shark, the Z axis between screen and us is invaded. One wonders, if filmmakers ever break the 2D bonds of a four-sided frame completely -- if they achieve true virtual reality -- how such a scene would play. Avatar on a 3D IMAX screen is certainly a step in that direction, but it's still a far cry from a completely immersive experience of being in a film. But does anyone really want to experience swimming with Jaws? Perhaps, but that's not the film Spielberg made.
Taken to its logical extreme, the 3D tech boom surely leads towards that fully immersive, 360˚ virtual reality. So where does upconverting classic films end? Ruined composition aside, what will the simple experience of a cut feel like when we're surrounded by image? How hi-fi can we stretch our relatively low-fi films? And why do we even need to? Dozens of action films from all throughout the 20th century have been remixed with 5.1 surround sound for home theaters (they even gave Disney's Snow White the treatment). But these sounds were never intended to fill such a space; are they big enough for their new britches? Does a technologically "upgraded" film enhance the original experience, or merely distract us from the filmmaker's intended vision? Check out the clip below, a commercial spec piece by Joseph Kosinski, director of the upcoming TRON: Legacy, in which we literally get to walk around inside a film (Kubrick's The Shining).

I think the technical term for that piece is "f#$%ing cool as balls," but can we, in our right minds, suggest that this is how Kubrick wanted us to watch The Shining? Lane, in his New Yorker article, cites Sergei Eisentein's clairvoyant prediction of 3D cinema and its advantages, and supposes what a number of classic moments in film history would look like if given the 3D treatment. But, he says, it doesn't matter:

There’s just one hitch. The scene works fine as it is... the posthumous application of 3D would not sharpen -- and might even vulgarize -- its moral thrust. Is that not, after all, how we have learned to read a painting since the time of Giotto? We know that perspective is a trick, and that a flat surface stands for a denser and more far-reaching world, but it is an illusion of which art... has availed itself with unstinting intelligence, relying on our instinct to decipher the code. What 3D movies say to us is: You have been fooled. You were duped, all this time, into thinking that a window was a world.

As long as films remain in that window, 3D will be just another bell or whistle -- not necessarily a gimmick, per se, but certainly just another tool in the kit. If Spielberg wants to add that tool to his Jaws kit retroactively, then he might as well go the full hog and recreate the shark digitally, because he will be asserting his freedom over the chains that bound him when he shot the film in 1974. But by doing this, he will be walking away from the fundamental battle he fought during production -- the lightning in the bottle that created such a memorable film -- namely, his difficult leading man:
Bruce, as he was affectionately named (after Spielberg's lawyer), never worked right. He went left when he was supposed to go down, up when he was supposed to go forward. He rusted. He malfunctioned. He was a diva. And even when he did work, he looked like what he was: a foam rubber shark. Frustrated, Spielberg shot around him. As Zanuck says:

In desperation, we came up with so many good ideas -- like the floating barrels, for instance, that were shown on the screen to suggest this shark beneath them, underwater -- and we did it because we didn't have the shark. In the script, the shark is on Page 1 when the girl gets eaten. It became more terrorizing than anything we could have hoped for. If we had CG then we would have had the shark in every frame.

So Zanuck admits that the limitations of the process drove them to a more effective result than they would have achieved if CG had been an option, and yet now he wants to bring digital technology into it? His own logic suggests that this would weaken the film. Jaws was made by embracing its limitations, rather than resisting them. Take the lesson from George Braque, who, along with Picasso, invented cubism: "Out of limitations, new forms emerge." On the one hand, this can be read as an excuse; Braque could no more paint like Rembrandt than Bruce the shark could accurately mimic a real great white shark. But that technical deficiency bred invention, and something new; cubism shows us something that Rembrandt can't, just as Jaws offers something that Shark Week footage does not. By the time Spielberg does allow the shark to play, we're so afraid of it that we can forgive his obvious shortcomings. Don't buy that? Then why are we still talking about Jaws 35 years later? Indeed, "don't show the shark," has become a staple tension-building device of monster films, so even if the shark had looked better, would seeing more of it have been a good thing?
The implication of revising any film with more advanced technology is that there is something to be gained in the process, i.e. that the original has room for improvement. But logic like that suggests that the 1983 sequel film, Jaws 3D, has something going for it that the original doesn't. If the film's marketing campaign is to be believed, then that something is "terror." But give me a person who finds Jaws 3D more terrifying than Jaws, and I'll give you a velvet Elvis.
This is, perhaps, unfair. Jaws 3D used technology put to shame by today's standards, and relied on chestnuts like severed arms and uncannily suspended jawbones for its thrills. Narrative stops in the name of cheap thrills. And, 3D aside, how can we even begin to compare this film to Jaws?We can't, of course. There's more to a movie than bells, whistles, and severed arms. It usually starts with something called a script, but that's another blog entry. My point is that technological revision like this is just varnish; what's underneath is far more substantial. In short: if it ain't broke, don't fix it. That's not to say that technique isn't important, or that shoddy attention to detail is ok, but, as John Waters says, "Technique is just failed style." Asserting that any film would be improved by converting it to 3D is to distract us from and discredit the work of the original film. Hitchcock's Dial M for Murder, while not his best film by a longshot, is still an effective thriller in 2D, though it was also released in anaglyphic 3D. Burton's Alice in Wonderland was converted to 3D as an afterthought, and released in two versions. It may be more nifty in 3D than in 2D, but is it a better film? Avatar was composed for 3D specifically, by a filmmaker with a savant-like understanding of the minutiae of the medium. Yes, I'd argue that's a better movie in 3D than in 2D. So yes, as concerns films made and released during stereoscopic periods in the history of major motion picture film distribution, I believe that there's room to question which version is better, with the answer being different from film to film. But Jaws was made well after the 1950's anaglyphic 3D boom, and well before the new 3D wave, so how does it figure in? I put forth that it doesn't. So leave well enough alone. 

When we make these sweeping revisions to films, we seem to say that our best work at the time wasn't good enough. We discredit our own film history, and the films themselves. And so I say that converting Jaws would damage it, if only by implication. Even if done well, even if it's an exhilarating experience, the very fact of the new film would be a spit in the face of the original. Of course, I have to admit that I'm no Spielberg. Who are we, the proles, to tell a filmmaker what to do with his or her baby? After all, could he not surprise us and make something better? Few would argue that George Lucas ever improved a film of his in one of his subsequent passes, but is it possible to enhance (using more than mere editing) a film at all, years after it's been released? Even Spielberg tried it, with a new cut of E.T., including CG effects where none had been before. I guarantee you, that's not the version of E.T. my children will watch.

Indeed, I can think of only one exception that proves the rule. Ridley Scott's Blade Runner: The Final Cut is more masterpiece than I thought a masterpiece could ever be. It's a cut of the film very close to the mislabeled Director's Cut, but it includes digitally cleaned up and enhanced shots, removal of continuity errors, and, briefly, some newly generated CG shots. Why does it work? Why does the new material feel so unobtrusive (so unLucas)? I can only credit Scott's restraint. Each and every decision he makes in the Final Cut seems so in tune with the original film's vision, so a-part-of that film's vocabulary, that we can only assume these truly were the shots he would have achieved with a little more time and money. He has only added in that which he felt the film needed in the first place. Case in point: the infamous Zhora snake dance scene. Chronicled in the making-of book, Future Noir, and on Blade Runner's DVD extras, this was to have been an elaborate stop-motion animated sequence in Taffy's club. Drawings survive, but the scene was deemed too ambitious to film, and unessential regardless. Given the opportunity to create and insert a CG version of the scene for the Final Cut, Scott passed. That was the decision he made in 1982, and he stood by it nearly three decades later. While the spectre of CG Jabba may bemoan the lost opportunity to realize the snake dance, I thank the heavens that Ridley Scott is not George Lucas. I can only hope that Spielberg follows Scott's lead on the subject of digital revision, 3D conversion, and dancing serpents. Really, what's next, techno remixes of the King? Aw, nuts.


Ann Horwitz said...

Loves it. But why are you picking on velvet Elvises? That shit is HOT.

MQA said...

what a fantastic essay! and i agree: balls are the pinnacle of coolness.