Shutter: The Thai Ghost-Photography Landmark

The great cursed-media horror films all understand the same thing: the terror is in the delay. Something is recorded. Time passes. The recording is developed, played, opened — and by then the thing has already happened, already been in the room, already been standing behind you while you smiled. Ringu made that structure world-famous in 1998 with a videotape. Six years later two young Thai directors, Banjong Pisanthanakun and Parkpoom Wongpoom, made it work with a camera, and Shutter has been the most exported Thai horror film ever since.

Tun (Ananda Everingham) is a photographer. Jane (Natthaweeranuch Thongmee) is his girlfriend. Driving home from a wedding, she hits a young woman in the road, and Tun tells her to keep going. Afterwards, Tun’s photographs start coming back wrong. There are shapes in the frames — over shoulders, in corners, in the fog of a long exposure — and Tun, being a professional, first assumes a light leak, then a lens flare, then a fault in the developer. He is a man in possession of exactly the technical vocabulary required to explain away the evidence, which is why the film gives him a hundred minutes to run out of explanations.

The real subject: spirit photography

The collector’s cross-reference here is older than cinema. Shutter is built on the practice of spirit photography — the nineteenth-century business of William Mumler in Boston and, later, the Edwardian British trade in “extras”, faint faces appearing beside a sitter that the studio would sell to the bereaved. Japan has a modern, vast, popular version of this: shinrei shashin, spirit photographs, a whole magazine-and-television subculture built on ordinary snapshots with something ambiguous in the corner. Thailand has its own equivalent. The film is not inventing a supernatural rule so much as filming one that a large number of its first audience already half-believed.

This is why it lands harder than a comparable Western ghost story. The scaffolding is pre-built. When Tun’s editor looks at a print and says the word for the phenomenon without hesitation, the film does not have to construct a mythology; it can go straight to the consequences. The one substantive invention is the film’s second idea, and it is the good one: the ghost has weight.

The neck

Tun’s neck hurts. It hurts from the first act, and it keeps hurting, and he attributes it to stress, and the film keeps quietly reminding you. He goes to a doctor. He is weighed. The scale is wrong.

This is the mechanic, and it is the most economical piece of horror design in the film. Everything supernatural in Shutter is also a symptom. The neck pain is guilt made structural — a man who did a monstrous thing in a car and drove away is carrying it, literally, in kilograms — and the film never says so out loud, because it does not need to. The idea does double duty: it is a metaphor that is also a physical fact, which is the only kind of metaphor horror can really use.

It also solves the ghost’s staging problem. A ghost that appears at the end of a corridor has to be shot, lit and timed, and the audience has seen it a thousand times. A ghost that is on you, at all times, in a position where you cannot see it and other people can, is available in every frame of the film for free. Every shot of Tun becomes a shot with an implied second subject. That is a startling amount of dread extracted from a design choice rather than a set piece.

Craft: the darkroom

The directors were in their twenties, working with a tight budget, and the film’s technique is entirely built around that constraint. The horror happens in the developing tray. A print emerging in chemicals is a reveal with a built-in timer — the image arrives slowly, from nothing, and there is no way to speed it up and no way to stop looking. Banjong and Parkpoom shoot these sequences long and mostly static, red-lit, and let the emulsion do the work. It costs nothing. It is more effective than most of what the same year’s American studio horror bought with real money.

The other economy is the sound. Shutter is very quiet for long stretches, and its scares are frequently delivered as an absence — a shutter clack in a room where nobody has fired one, a sound of a body moving that has no body attached. When the film does go loud it goes loud with a shape that has already been established in a photograph, which means the audience recognises it. The recognition is the scare. Novelty is not required.

Ananda Everingham is the film’s other asset. He plays Tun without a hint of the standard horror-protagonist decency — Tun is evasive, self-serving and technically fluent, a man who is very good at reframing what he has done — and the performance is what makes the last act tolerable rather than punitive.

The wave it arrived in

Some context for why Shutter mattered beyond its own merits. By 2004 the regional horror wave was six years old and visibly tiring. Ringu had detonated in 1998, Ju-on had industrialised the grammar, Hollywood had bought both, and the cycle had entered its imitation phase — a great many films with a dead girl, a lot of hair and a blue filter, made quickly because the previous one had made money.

Thailand was, at that point, a minor player. The Thai film industry had been rebuilding through the late 1990s and its horror output was largely folk-ghost comedy, a domestic form with almost no export record. Shutter changed the arithmetic in a single stroke, and it did so by being smarter about structure than the films it was competing with: while Japanese horror was chasing the mechanism of the curse — the tape, the rules, the seven days — Banjong and Parkpoom built theirs around a crime, and gave the audience a protagonist who was concealing it. That is a detective structure wearing a ghost’s clothes, and it gives the film a forward motion that most of its contemporaries lacked.

It also travelled because it is cheap-looking in a useful way. There is nothing in Shutter that depends on a budget. A camera, a darkroom, a corridor, a face. Films built that way age at a different rate from films built on effects, which is why this one plays now and a lot of its better-financed rivals do not.

The remake problem

Shutter was remade in America in 2008 by Masayuki Ochiai, with Joshua Jackson, relocated to Tokyo, and it is a textbook demonstration of the failure I’ve written about at length in why Asian horror remakes flatten the ambiguity. The remake explains. It offers reasons. It sands the protagonist into someone the audience is comfortably permitted to like, which vandalises the entire structure, because the original is engineered around how much weight a viewer’s sympathy will bear before it gives.

The Thai original was a substantial domestic hit and travelled through Asia and then, on DVD and torrent, everywhere else — which is how a lot of us met it. Banjong went on to co-direct The Medium with Na Hong-jin seventeen years later, and the through-line is clear: the same interest in a camera that keeps recording after the operator has stopped wanting it to. For the neighbouring tradition, The Eye is the Pang brothers’ version of the same idea with a cornea instead of a lens.

The case against

The middle act is soft. Once the rules are established — the photographs show her, the pain is her weight — the film goes on an investigative tour that is largely functional: find the school, find the friends, find the flat. It is competent and it is filler, and the picture idles for a good twenty minutes.

The ghost’s design has dated in the way that all long-black-hair ghosts of the 2000s have dated, through no fault of this film. And the scare grammar occasionally reaches for a jolt in a period where the whole regional industry was reaching for the same jolt.

The reason none of it matters is the last shot, which is one of about five images in modern horror that everyone who has seen the film can describe from memory twenty years later. It streams in most territories; find the Thai-language version and refuse the remake.

Spoilers below

The woman in the road was Natre, and Tun knew her. She was his girlfriend at university — a quiet, unfashionable young woman he was ashamed to be seen with, and whom he never publicly acknowledged. The hit-and-run is therefore the second time he abandoned her in a road, which is a structure the film builds carefully and never over-signposts.

Worse is what he permitted. Tun’s friends assaulted Natre and photographed it, and Tun was in the room, holding a camera, taking pictures. He did not intervene. He kept the negatives. The film’s cursed-media premise turns out to be exactly literal: the photographs are haunted because photographs were the instrument of the crime. The friends who took part start dying, one by one, and Tun’s investigation is really an attempt to reach the end of the case file before Jane does.

Jane finds out. This is the film’s best structural decision — the audience is placed with her, and she spends the last act discovering that the man she has been protecting is the reason a dead woman is following them both. Natre’s corpse, it emerges, was never buried; her mother kept the body in the flat, unable to let go, and the discovery of it is staged with a matter-of-fact grimness that the American remake could not bring itself to reproduce.

And then the last image. Tun is in a psychiatric ward. Jane comes to see him. He is not injured, exactly — he is diminished, hunched, and everyone keeps mentioning the weight. She takes a Polaroid of him, and in the developed print Natre is sitting on his shoulders, arms around his head, face pressed to his, smiling, and she has been there for the entire film, in every frame, in every photograph, since the first reel. The neck pain was her.

It is a perfect ending because it is retroactive. It rewrites everything you have already watched without a flashback, without a line of explanation, in a single still frame — the horror was in the room the whole time and the film had been showing it to you, and you were looking at the man’s face.