Posted by David Kalat on April 20, 2013
Once upon a time, I helped ruin a movie. This is overstating things a bit–my contribution to this movie was so slight that I didn’t even appear in the closing titles, which are otherwise so detailed as to run on to hundreds of names, including credits for “snake wrangler” and “additional typesetter.” But even if the makers of Even Cowgirls Get The Blues didn’t think I was worth crediting in the titles, there is no question that what I contributed was absolutely crummy. Or that my lousy work was central to the film’s defining metaphor. What it really comes down to is that even if I didn’t ruin this movie, it was doomed to be ruined because its makers allowed any of my contributions to appear onscreen at all.
A little backstory: I attended the University of Michigan, as one of only the second cohort of students to pass through the newly minted Film and Video Studies program, led by William Paul. Professor Paul had come from the English department, and approached the whole endeavor of film studies the way one would English studies: as a critical exercise focused on watching movies and writing about them.
At the time, I had it in my head that I was going to make movies, and I aligned my interests with the various underdogs and iconoclasts within Paul’s program who wanted to put the emphasis on technical knowledge of film production. I managed to emerge from those years with a chip on my shoulder, seething at some ludicrous schism between film theorists and filmmakers that existed mostly in my own mind. Anyway, I went looking for work in the technical side of film production, and landed at a film lab in the Washington, DC area.
It was one of the happiest periods of my life. I loved my job, I loved my coworkers, and I was part of the industry, in the loosest possible sense. I was a “color timer,” a job that may not even really exist anymore in this age of digital production.
For those of you who don’t know what a color timer is…I didn’t either. My mentor spent a lot of time drawing little graphs for me on napkins and scrap paper (“look at this bit of the S, David, this is where you want to be”) and I would nod my head, unwilling to admit I didn’t understand what he was talking about. Man, I was a stubborn git.
Eventually, about two years later, I’d left that job and that city and was living half a continent away while my wife attended law school. I set about building a darkroom in my basement, and as I played around with printing B&W prints in my own darkroom, I had one of those “eureka!” moments and suddenly realized what that whole S-curve thing had been about.
Film is a chemical medium, and its responsiveness to light is imperfect. That imperfection is what lends film its unique properties–why some people prefer the look of real film to digital, why digital filmmaking usually tries to artificially ape the chemical characteristics of film.
For the purposes of illustration, let’s just talk about black and white film. Now imagine that you shoot two scenes–one where you’ve aimed the camera at the sun and filmed the brightest light you can manage, and then you put the lens cap on and film the darkest nothingness. These two shots form your boundaries, the outer edges of your contrast range of white to black.
Between those two extremes, every individual gradient that the film is capable of resolving represents your contrast range. A very high contrast stock that was only capable of reproducing either complete white or complete black, but nothing in between, would be very poor at resolving realistic images (there are such film stocks, mind you, for specialized SPFX purposes). Alternatively, a hypothetical perfect film stock would match in perfect synchrony every incremental increase in luminosity from black to white.
With me so far? Now imagine that you graph the responsivity of your film stock. On the x axis you mark out each incremental increase from black to white, and on the y axis you plot how the film responds at that light level.
If you were plotting a hypothetical perfect film stock, your graph would just be a straight line at a 45 degree angle, with each increase in luminosity in the world perfectly reproduced by an increase in luminosity of the film image.
But films are not perfect, and some might overreact to increases in light, or underreact. More to the point, they might vary their reactions at different light levels. In fact, this is what happens: at the far left end of your graph, around the zero mark of total blackness, there’s a stretch where increasing the light level doesn’t help much for a long time–anything in this zone is going to be underexposed. Then past a certain point the film starts to respond correctly and you get into a stretch where the graph does chart out a 45 degree slope, or something close to it, for a while. Then as you get to the far right end of the graph, it levels off again as the overexposed image starts to be unable to respond to differing light levels anymore.
If you graph it, it looks a bit like an S.
And every film is different–each one has its own unique graph, its own “characteristic curve.” But for each one, that middle zone where the line approximates the 45 degree straight line of a hypothetical perfection is the zone where that film is at its best–and the job of the cinematographer is to expose the film so that you’re in that sweet spot of the “straight line portion of the S.”
And the thing is, this whole S-curve response happens more than once. It happens when the film is first photographed, creating the negative–but it also happens when that negative is printed to a positive. The cinematographer is charged with making the creative and technical decisions that manipulate those characteristic S-curves into the images of the negative, but it is the color timer who handles the creative and technical decisions that influence how the negative is turned into a positive.
My job, although I barely understood it at the time, was to convert the hard work of cinematographers into what audiences would actually see.
Well, not really. For anything that audiences were going to see, the lab tasked its professional experienced timers. I was the guy who made provisional copies for working purposes–before it ever went “live” someone who understood S-curves would have at it. Or so the theory went.
One day I had the task of making copies of some archival nature footage supplied by National Geographic–specifically footage of cranes. The client was making a film version of Tim Robbins’ novel Even Cowgirls Get the Blues, directed by Gus Van Sant (whom I have never met). They had asked for a batch of crane footage to be copied. I hadn’t read the book, but my wife had, and I gathered that cranes served as a central motif in the story. I guessed that Van Sant and his fellow filmmakers wanted to use the National Geographic material as placeholders during the editing, until they narrowed in on exactly what they needed for the individual scenes. Then, as the editing settled into a final cut, they’d come back to us for the proper interpositives and dupe negatives of the footage they chose, and at that time one of the more experienced timers would handle that assignment.
Except they never came back. And the movie came out–with my inexpertly copied nature footage plunked unceremoniously into place. It was appallingly sloppy. If the cranes were an important motif, why wouldn’t they want those shots to look their best, and fit as closely as possible in with their own footage of Uma Thurman and the other stars? We could absolutely have matched the cranes better to conform with the Van Sant-shot material. That was the whole point of color timing–and even before I understood S-curves I understood that much.
There is another irony here. Around the time that I built my darkroom and had my epiphany about what S-curves meant, I was hired to as a film reviewer for an arts journal. My work writing for that journal led directly to my first book contract in 1997, and in turn to everything that has followed since: writing books, my stint on the writing staff of Video Watchdog, my work as an audio commentariat, and being asked to join TCM’s Movie Morlocks. After all my haughty nonsense about being a filmmaker not a film writer, I never did become a filmmaker but pretty quickly did become a film writer. And what was the very first film I was asked to review?
Even Cowgirls Get the Blues.
Which I tore into–it’s a self-indulgent mess that seems to be constantly winking at the audience, “Look at me, aren’t I outrageous?” I tried to note that I had actually provided the images of cranes that are threaded through the film, that I was personally responsible for how grainy and washed out those stock footage shots are, and that it was indicative of an overall lack of care that my craptastic stock footage transfers were left in the finished product–but my editor snipped that section out.
In the end, almost everything I wrote for that journal got edited out. Which is right and good and proper–I was an inexperienced writer first learning the trade and I wrote a lot of nonsense that wasn’t worth publishing. Every time I turned in a piece that then got junked, I went back to nurse my wounds and tackled the next assignment more carefully, and more carefully still, trying to fine tune what I ought to be doing if I wanted to get published.
Because that’s the thing about good editing–you throw out the crappy bits and encourage your creators to try harder and get better. If someone gives you a duff essay, it shouldn’t be published–and if someone gives you a grainy pile of off-color shots of cranes, maybe that shouldn’t end up in a major motion picture either.
Streamline is the official blog of FilmStruck, a new subscription service that offers film aficionados a comprehensive library of films including an eclectic mix of contemporary and classic art house, indie, foreign and cult films.
Actors Alfred Hitchcock Bela Lugosi Bette Davis Boris Karloff British Cinema Buster Keaton Cary Grant Charlie Chaplin Citizen Kane Comedy Criterion Dracula DVD Elizabeth Taylor Film Film Noir FilmStruck Frankenstein Fritz Lang Hammer Horror Horror horror films Horror Movies Humphrey Bogart James Bond Joan Crawford John Ford John Huston John Wayne Joseph Losey MGM Movie movies Night of the Living Dead Orson Welles Peter Lorre Psycho Roger Corman Screwball Comedy Steve McQueen TCM The Exorcist Warner Archive Westerns