Ever since Robert Rodriguez made El Mariachi, people have considered making upscale movies with pocket-sized gear. In the three decades since El Mariachi released, cameras, lights and post gear have all advanced faster than the demands of film and TV distribution. That should make things easier, but somehow, it took until The Creator, arguably, for someone to make a huge movie on a tiny camera.
Contents
Hidden figures
Nobody is under any illusions about the difference. El Mariachi famously cost $7000. The Creator came in at a cool eighty million. Never mind that $80m in 2024 is a bargain price for a globetrotting concept movie with named actors and extensive visual effects. Never mind that the oft-quoted El Mariachi budget barely covered Rodriguez running around with an Arriflex 16S for fourteen days. The film that most of us have seen is the result of a further post production effort which cost almost thirty times as much.
Movies can give a very misleading idea of what they really cost. In the world inhabited by most people, that’s the idea. Making things look like they cost more than they did is the sort of ability that makes crew popular with producers. Between the extremes of having next to nothing, and having practically everything, there are lots of ways to bring big ideas to small shows.
Making things look like they cost more than they did is the sort of ability that makes crew popular with producers.
Go large
Consider First Man. There’s almost no better example of a film that mixes big and small. It combines 16mm and 65mm film to represent personal tragedy and global fame. It went big in other ways, too.
The Luminys SoftSun derives from a sort of industrial lighting which was common in the eastern bloc during the latter years of the cold war. It’s a tube full of xenon gas—think of a photo flash gun, only six feet long. Apply electricity, and the gas glows. It’s not quite the same physics as a photo flash, since it runs continuously, and compared to LEDs, it isn’t particularly efficient.
But it does have one huge advantage: scalability.
Traditional SoftSun lights come in ratings around 3.5, 10, 25 or 100 kilowatts. For comparison, the biggest LEDs are currently about 2.5kW. For First Man, cinematographer Linus Sandgren, FSF, ASC, tried putting two 100kW SoftSuns side by side. The result was a very faint double shadow which didn’t quite look like sunlight. So, they had the company make a SoftSun capable of 200kW (that’s two hundred thousand watts) in a single tube.
It’s hard to prove that nothing bigger has ever been deployed on a film set. Beneath all these spectacular numbers, meanwhile, is a question: why did First Man need such a monster? More to the point, why does any of this matter to anyone who doesn’t have an eight-figure budget?
How to fake a moon landing
No matter how big a light we use, it always falls off to black at some point. If the phrase “inverse square law” has already sprung to mind, you can probably skip the next paragraph. If not, think of it this way. If the light is bright enough to illuminate the scene from a thousand feet away, we can move fifty feet closer or further and that’s only a tenth of the light-to-scene distance. The variation in light level is small. If we use a smaller light, we might have to put it two hundred feet away. Move fifty feet closer or further, and we’re now moving fully half the light-to-scene distance. The variation in light level is much larger.
And First Man needed to light its moon exteriors. In reality, the sun would be about 93 million miles away and the horizon on the moon is only about a mile and a half distant, assuming flat terrain. One and a half is a very small proportion of ninety-three million, so on the moon the light has basically zero falloff. That’s why big movies use giant lights on cranes even when the scene is set in someone’s lounge. That way, people can cross that lounge without the exposure changing much.
All this might seem irrelevant to people shooting short films. A single socket in 110-volt places is good for a kilowatt and a half. A 200kW SoftSun, let alone its rental price, consumes the equivalent of a hundred and thirty-three wall sockets. You need a generator, and a big one.
A more domestic arrangement
Most of us aren’t trying to depict the lunar surface. For more everyday setups, consider the Arri M18 HMI, which was more or less born for situations like this. It will (barely) run from the most capable American wall sockets, though keep a fire extinguisher handy in houses which haven’t been rewired recently. An LED at a similar power level is only a little brighter. It doesn’t really change what’s physically possible, but it’s probably a bit cheaper, so it does change what’s practically possible.
And if we know anything about working at scale, that’s useful. Anyone who’s tried to shoot a scene in which someone walks across a room, using only lighting that will fit entirely within that room, will know the problem. It’s too easy to pull out a camera capable of low-noise pictures at 2,000 ISO and realize that even a medium-sized LED is way too much. There’s probably some sort of law against running a light at one per cent, but it works. Or at least, it works until the actor moves a few feet away from a light that’s only a few feet away to begin with, and vanishes into the shadows.
This is why a lot of small productions feel so static. People put low-power lights an inch from the actor’s ear. That’s not making good use of high-sensitivity cameras and high-efficiency lights which should lend us more freedom, and we don’t need to be First Man to do better. It’s a logistical challenge to find more distant places to put lights, and get power to them, but, well, that’s the job.
Medium scale
Based on all that, it’s not hard to get the impression that the film industry is a place where bigger is always better. Movies seem to do well when they make cinema visits into a special occasion. Audiences seem to care that what they’re watching is a high-end example of traditional screencraft. It’s why ARRI built the Alexa 65. It’s also why cameras such as Fujifilm’s GFX 100 II have enjoyed so much attention.
One of them was rebuilt into a movie-oriented shell as the OFG Customs 65 and shown on Fujifilm’s booth at NAB 2024. The sensor in the GFX 100 II, at 43.8mm wide, is actually not quite as big as a 52.48mm 5-perforation 65mm frame. Still, the unmodified GFX 100 II can ride on a handheld gimbal and let anyone shoot almost-65mm for comparatively little.
Blackmagic announced a huge-sensor camera at the same show, and people have been shooting 4K on digital sensors about the size of a VistaVision frame for a while now. VistaVision, like 65mm, was developed to make cinema special enough to tempt audiences away from the small screen. The screens have gone from small to literally pocket sized, but big movies made on big formats are still drawing viewers. For anyone trying to squeeze 65mm ambitions out of a VHS budget, reaching for the nearest big-chip camera is almost second nature.
There are reasons, though, that might not be a brilliant idea.
Hard stops
There are cameras, such as Blackmagic’s 12K Ursa, which often have more resolution than the lenses mounted on them. We won’t go into the mathematics of diffraction limits here, but depending on f-stop, we’re nearing the point where cameras are routinely better than lenses can ever be, no matter how carefully made. That’s not really a problem, in a world where we don’t always prioritize sharpness anyway.
We’re nearing the point where cameras are routinely better than lenses can ever be.
The problem is coverage. Reach for a medium format camera, and kiss goodbye to your carefully eBayed collection of fungus-coated 70s photo glass. Even full-frame cameras can be a nuisance. Some of Panavision’s most popular lenses are the C and E series anamorphics used on all-time greats like Alien and Blade Runner.
They sit alongside the likes of Technovision—the 1970s vintage lenses used on Apocalypse Now (not to be confused with the modern Technovision 1.5:1 full-frame lenses made by P+S Technik). All are part of a group of much-loved anamorphics built for 35mm film that’s about the same size as an APS-C digital camera.
Peripheral vision
Few of those lenses will fill a medium format frame. Some of them don’t even fill the Super-35mm 1.78:1 frame. Anamorphics, in particular, were often designed for squarish 4:3 sensors. That can sometimes mean that a larger chip with more pixels makes sense, so that we can crop down to the area the lens was intended to cover. Because lenses are generally fuzzier around the edges, variable crop levels on a big sensor are also a way to achieve variable levels of optical grunge. Exactly how big a sensor a lens will cover might be a matter of how much fuzziness we’re willing to tolerate.
Even so, it should be clear that smaller sensors give us more lens options. It’s probably worth a quick digression here, because there are dangers. Much as we might take a moderate approach and pick a medium-sized, we might also take a middle-of-the-road attitude to lenses. Any approach which deliberately seeks less-than-perfect lenses needs to test very, very carefully. Glass that looks great at f/8 on a bright summer’s day might look a lot less great at f/2 if things run long and the sun starts to go away.
Adding character
If we want characterful lenses and full frame cameras, all of our equipment has probably some combination of bigger, heavier, more expensive, slower, fuzzier, and less compatible. That’s true with all lenses, but it’s especially true with zooms, and zooms are a huge time saver. There’s also a small controversy to address here. We touched on it in the context of Furiosa, where we learned that it’s normal for the biggest shows to carry Fujifilm’s zooms in a role often described as “just in case we need it.”
It’s hard to be confident about specific productions, but in reality quite a lot of big movies probably use them quite a lot.
If we’re interested in bringing big-movie looks to smaller productions, there should be no squeamishness about grabbing a zoom lens. The time saved is likely to be a bigger positive influence on a small production than any amount of clever optical fireworks. Filter, if you like, for a look.
Again, big-chip zooms, with manual focus, tend to embody all compromises of lenses for really big sensors. The best of them avoid the issues of speed and quality. Insist on zooms for really big sensors and again, the compromises are size, weight and price.
Not without benefits
Happily, the push for big chips does have some compensations. Zooms for Super-35mm cameras have become comparatively affordable. Fujifilm’s Fujinon XK6x20 and the Canon CN7x17 are, respectively, 20-120mm T3.5 and 17-120mm T2.95-3.9 zooms. They should cover most things for most people most of the time. They’re not tiny, though the Fuji will just about balance on something like an Ursa Mini. They’re also available with servo grips like a broadcast lens, making them incredibly convenient for under-crewed single operators. Crucially, both are available for rental almost everywhere, and even starting to sell, used, for under ten thousand units of some currencies.
Productions that can get away with using photographic lenses have a lot more options. Most people will quickly outgrow that solution when they encounter a need to do accurate focus pulls. In general, lenses have not improved as rapidly as cameras, and we’re still nudging up against the limits of optical physics. Which is a shame, because big-chip, high-resolution cameras have never been easier to get hold of. They rejoice in the gloss of the high end. And yes, in the end, more silicon has more performance.
They’re just not the right solution for everything.
Stepping down
Some of you may still be frowning at the claim made above that a sub-$10k lens is affordable, and that’s fine. Go used, reach for stills primes, and you can put together a very capable camera setup for far less. At that end of the market, setting up something with actors, costumes and locations is probably also a bit out of reach. Happily, there’s at least one way of shooting persuasive demo reel material which might not even require that you move from your seat.
If there’s a cheat code for small-scale camerawork, it’s probably macro. It’s been a favorite of names as big as David Fincher, whose work with cinematographer Erik Messerschmidt, ASC, led to the opening sequence for Mindhunter. In part, that sequence is a triumph of production design.
The Sony TC-510 tape deck is a particularly sculptural piece of late-seventies technological precision. The example seen opening Mindhunter was apparently rebuilt from the best parts of two to create the newest-looking device they could. Fincher’s attention to detail ran so deep that he had special tape reels made to create just the perfect look.
Fincher likes doing this. Compare the opening of The Girl with the Dragon Tattoo. While it probably involved more CGI than a James Bond title sequence, it’d be interesting to see what was possible if it were possible to find an actress willing to have some sort of skin-safe black fluid poured all over her.
OK, not all of that is necessarily macro, but it’s a fun example of the sort of extremely limited-environment photography that lets us set up on the kitchen table and play with lens effects all day.
A closer look
In principle, the word macro means that the image of an object projected on the camera’s sensor is at least as large as the object is in reality. Sometimes, the same term is used (inaccurately, in theory) to describe a very small object being shown in a screen-filling closeup. That might mean a lens with incredibly short minimum focus distance, often with very short depth of field. A similar look can sometimes be achieved with very long lenses that have a fairly normal minimum focus distance.
True macro changes the way lenses work. On a macro-capable zoom lens, the zoom and focus controls may change and interact. Zooming may be difficult. Depth of field is typically tiny. New optical effects, flares and aberrations may occur. If this is starting to sound like a playground for entertaining camera techniques, well, it is.
The best news is that it isn’t even particularly hard to get into. If macro gear is old news to you, skip ahead, but the simplest macro lenses simply place the glass further from the sensor than usual. That can be done with a simple, inexpensive extension tube. About the only limit is that they don’t work properly on some types of anamorphic lens, because the focus distances for the different horizontal and vertical geometry only coincide at one distance. For anamorphics, or any situation where you’d rather add things to the front of the lens than the back, consider a dioptre filter.
All kinds of subjects are great fun in macro, from food to technology to natural objects to people’s artfully-lacquered fingernails. Plants are an amazing subject, with their tiny, complex structures. An arty macro of a flower is so conspicuously easy that it sometimes costs a round of drinks among camera people. It combines spectacularly well with fundamentals like slow motion, speed ramping, playing things backwards, flipping things upside down, and moving light sources.
Shine a little light
The latter has become even easier since the dawn of LEDs. Before modern lighting, there really weren’t that many tiny movie lights. Early LEDs were often criticized for being too small—the earliest ones were under 100 watts. Dedo Weigert’s famous Dedolights were much-beloved of anyone doing desktop product shots, which were themselves often macros. Now Dedo has built LEDs and small, early LEDs from many manufacturers go for pennies on eBay.
In the end, tabletop stuff can be lit with desk lamps and flashlights. Pieces of paper become diffusion and flags. Kitchen foil becomes a reflector. Translucent candy wrappers create color. The largest piece of grip equipment is probably a magic arm and motion can be created simply by putting things on a record deck, and spinning it slowly, by hand. Objects swim in and out of focus and gleams of light become glows and glints.
Putting it into practice
It seems a little too easy to end on a technique as cheap, effective and low-resource as macro, but like anything in filmmaking, the real value of it is how it helps tell a story. Yes, it’ll look great as a showreel opener. What it really does, though, is concentrate the viewer’s mind on what’s in frame, just like proper lighting, proper lens choice, and all the other things we’ve discussed here.
But of course, we’re all aware of that sort of thing. We’re all learned film academics. We can all pronounce mise-en-scene after our time at film school—or perhaps our time in the “film school” section of YouTube.
We wouldn’t stoop to employing pretty tricks without a specific justification. We wouldn’t prefer a smaller-sensored camera just to make it easier to keep things in focus. We certainly wouldn’t shoot something in extreme closeup just because it looks amazing while requiring so few resources. We’re not using tricks. We’re working at an appropriate scale for the situation.
Featured image from First Man © Universal Pictures
