The Cinematographer’s Guide to Wes Anderson’s “Asteroid City”
We talk about gear a lot. So much, in fact, it’d be easy to think that for most of the history of filmmaking the main thing that stopped indie filmmakers doing wonderful work was technological sticker shock. In 2023, though, the main reasons people don’t shoot James Bond movies on mirrorless stills cameras are convenience, reliability and tradition.
“Choosing the very best gear might mean paying ten times the money for five per cent more performance.”
All right, so that’s probably a bit reductive. Sometimes the high end is worth it. Choosing the very best gear might mean paying ten times the money for five per cent more performance. That might mean enough spare image quality to get away with mistakes, which is one reason why nine-figure blockbusters so rarely have rough-looking shots in them. It’s not that people are perfect; it’s that they can get away with greater imperfections.
More with less
But modern cameras are starting to give us some of that even at the low end. People will raise objections about rolling shutter and fiddly menus, but even very elementary cameras could in principle be used to shoot major motion pictures without anyone batting an eyelid.
And that’s why absolutely all low-budget filmmaking now looks and sounds effortlessly wonderful.
No, wait, hang on.
Camera people are often terrible gearhounds, but most of us have always known that it’s not about the toys. To a much greater extent, it’s about associated gear like lighting and grip, and even more about production design, blocking, framing, camera operation, and editing. Hiring skilled people isn’t free, but learning new things can be, and the results can sell almost anything as way higher end than it really is.
So the problem we’re going to try to correct in this article—which could easily spin out to a series if it hits the mark—is that technique is often overshadowed by technology. We’re going to start with a look at how the equipment influenced the look of Asteroid City—and how it didn’t. This matters at the high end. It matters a lot more when our camera choice is limited, but our ideas are not.
The cinematography of Asteroid City
Anderson’s storytelling is an acquired taste, but his films have very identifiable visuals. Asteroid City is a brilliant example of filmmaking where the equipment isn’t intended, or allowed, to dictate the visuals. It was shot on a collection of gear that many independent filmmakers would sell a major organ to acquire, but the look is controlled independently from that. Some people might even call those equipment choices slightly perverse.
Director of photography Robert Yeoman, ASC, is a serial Anderson collaborator. The pair seem to like anamorphic lenses, which is a downright odd choice considering Anderson’s keenness for very square framing that lines up precisely with the surrounding architecture. Behind-the-scenes material includes interviews in which Yeoman describes using a measuring tape to ensure the camera is properly centered in whatever piece of architecture Anderson was interested in. Asteroid City is ostensibly about a play, and Anderson, as ever, likes his proscenium square and true.
But ask almost anyone to describe the characteristics of an anamorphic lens, and the response will undoubtedly involve references to softness, aberration, distortion, flare, and all of the ways in which a lens can be less ideal. That’s the sort of result we’re trained to expect by films like Blade Runner, which embraced all of those things to create its self-consciously grunge-laden images (even though some of the effects shots include some very visible circular flares, which are technically inconsistent with the live action unit’s work).
The problem is, we only have to watch a few lens demo reels to realize that only a small subset of anamorphics actually panders to popular expectation. Panavision’s C and E series exhibits the blue-pink flares, soft edges and other famous artifacts. Asteroid City was (mostly) shot on Arri’s Master Anamorphics, a design that’s decades more recent. They’re intended to be clean and precise. They’re so clean and precise, in fact, that Arri offers the Flare Set of interchangeable front and rear elements, which create more visual interest.
That’s not done here. Flares are rare, and Anderson’s predilection for putting horizontal and vertical lines (windows, doors, etc.) near the edge of frame would normally betray even a hint of distortion. There’s practically none, even though Anderson seems to adore wide angles. The 35mm Master Anamorphic is frequently seen in behind-the-scenes material. Wider lenses typically suffer more obvious distortion, and anamorphics intrinsically have a wider field of view in the horizontal axis. A daring interlocutor might speculate that it’s so straight that any remaining distortion was cleaned up in post.
The Sunny 16 rule
Yes, Master Anamorphics will produce elliptical out-of-focus artifacts, but the film isn’t really shot in a manner that emphasizes focus falloff. Wide angles tend toward more depth of field, and the film was shot outside in bright sunlight. We don’t know how bright it was in Spain when Yeoman was metering his shots, but it’s a sun-drenched movie. The Sunny 16 rule-of-thumb suggests that in bright sunlight, the shutter speed should be the reciprocal of the film sensitivity at f/16. Asteroid City was (mostly) shot on Kodak’s ISO 200 color negative film 5213, suggesting a 1/200s shutter speed at f/16.
Since Yeoman would invariably have been running at 24 frames per second with a 180-degree shutter for conventional motion rendering, creating a 1/48s effective exposure, he would have needed to close down the lens two stops from f/16 for reasonable results. A Master Anamorphic won’t actually do that, because it tops out at f/22, but there are other considerations. We could do the calculation for diffraction limits of the 35mm Master Anamorphic on that type of negative stock, but most people would shy away from f/16, let alone even smaller apertures.
As such, Yeoman probably had enough light to use any aperture he wanted, and would still have been using ND filters for proper exposure. A three- or four-stop ND, which Tiffen would call an ND0.9 or an ND1.2, would make an f/11 or f/8 aperture a plausible choice. The result of that, especially combined with the wide angle lenses, is a film which does not exhibit anamorphic focus artifacts that noticeably. It can’t. Most things are in focus.
Combine all that with the relatively insensitive, fine-grained film stock, and it’s no surprise that Asteroid City is an anamorphic movie shot on film that doesn’t obviously look like anamorphic or film. It’s almost as if Yeoman and Anderson were using anamorphics because they wanted the wider frame, not because of some much-vaunted and magical optical effect associated with the lenses. And that, after all, was why anamorphic cinematography was invented in the first place.
“That, after all, was why anamorphic cinematography was invented in the first place.”
That’s good news to those of us who are never likely to have the budget for Anderson’s camera setup. A particularly heretical analysis might conclude that it could probably have been shot on any number of other camera systems without looking jarringly different, and that’s encouraging.
And now the bad news…
What’s a bit less encouraging is that while Asteroid City doesn’t derive too much of its visual style from the camera equipment, it absolutely does rely on a lot of production design. In fact, it does things that wouldn’t work without careful production design.
Anderson (or rather production designer Adam Stockhausen, another frequent collaborator) built the titular Asteroid City from scratch in the Spanish desert. It looks new-built, and that’s fine. A little bit of visible artifice is entirely appropriate to the story. The background mountains and cacti look like something out of a Road Runner cartoon. The bird shows up, too. What’s crucial, though, is that almost everything is a pastel shade of red, blue, yellow or green, and in its final form, Asteroid City exists in a world of vibrant color, with the saturation noticeably increased.
It’s exactly that interaction between the production design and the grade which demands control. The same saturation control is available to anyone with the free version of Resolve, but trying to achieve Anderson results by cranking up the colors on a scene shot in someone’s beige-walled lounge will just highlight the inconsistencies.
That’s often the last thing that small productions have the option to do. Production design is often about color control, given a limited palette helps with visual consistency across a project. That’s notoriously difficult to do on a budget, because it means buying things that are the right color. As a result, it’s often tempting to desaturate less-than-ideal cinematography.
“It’s often tempting to desaturate less-than-ideal cinematography.”
Gray things out a bit, apply an overall color tint, and at least everything looks consistent. You can justify it by tinting blue and referring to the steely nights of Terminator 2, or recreate the opening of Top Gun by going for orange, or The Matrix by going green. Either way, it’s a technique with limited horizons. Controlling color properly, in production design, achieves so much more.
If this sounds like a reason to stop counting pixels and spend the time lobbying for better locations, it is. Not every film can build a world like Anderson does. Not every director wants every story to exist in an Andersonian world of cartoonish backdrops and block colors. Even so, taking a few hours to explore thrift stores for a couple of appropriately-coloured jackets might have more actual impact on the cinematography than doubling the camera department’s equipment spend.
A more natural approach
So we’ve got a camera system, some production design to point it at, and an approach to the cinematography that draws those things together in a way that makes sense for the story. The manner in Wes Anderson actually points the camera at things is so identifiable that a lot of people might actually prefer to do something slightly different. Nonetheless, Asteroid City was shot and lit in a way that’s really enticing to have enthusiasts of the old school—and surprisingly accessible.
Anderson himself said “we can be a bit like a student film” in an interview, referring to his small crew and gear footprint, treating the set as a location. A casual observer might assume that Asteroid City was shot largely under available light. A lot of sunlight is allowed to fall directly, or with minimal filtering, on actors.
That would ordinarily provoke frowns from appearance-conscious people (as well as sunburn). There’s certainly some use of bounce and flagging, presumably to even things out. Occasionally people sit under perforated canopies. Still, there are few fill lights visible in the published behind-the-scenes material.
That’s something that many indie filmmakers would do well to remember. High-tech lights are wonderful, but on a day exterior there’s only so much that a few hundred watts of LED can do against fifty thousand lux of midday sun. A big black flag or white bounce—or even some appropriately painted polystyrene insulating board—is cheap, easy and effective, and missing from many an indie moviemaker’s palette.
“There’s only so much that a few hundred watts of LED can do against fifty thousand lux of midday sun.”
Just make sure you’ve a few C-stands. Lots of things can be approximated. C-stands can’t.
To some extent, difficult day exteriors on location can also be mitigated with proper blocking and framing, even if you’re not attempting Anderson-style architectural camerawork. If you are, on the other hand, traditional grip technique still works. It also highlights the problems with modern toys. Gimbals smooth out handheld camerawork, but they don’t stabilize where the camera is—just where it’s pointing.
For comparison, Anderson’s approach is so determinedly geometric that for one scene, the camera rides on a combination of four dolly tracks, arranged to create a two-dimensional, precision-movement system like a pen plotter.
The man is… interesting. You could try to do that with a gimbal, but it might not be a good idea.
Perhaps that’s the biggest takeaway here. Asteroid City cost about US$20 million, which isn’t nothing, but it’s modest for modern American cinema. But that’s not really why some of Anderson’s technique is so accessible; it’s accessible because it’s fundamental filmmaking.
Either way, the point is that the film refuses to allow the gear to dictate the approach. That tells us a lot about our approach to gear in general. It suggests that pixel-peeping, while valid, is ideally about keeping the gear out of the way of the craft, and making the conjunction of the two at least seem like a happy accident.
Filmmaking is a very technical artform, and it’ll never be possible to separate those things entirely, but Asteroid City seems to suggest we should try.