Werner Herzog

 

I keep coming back to the idea that Werner Herzog is one of the purest creatives we have, in the absolute, unqualified sense of what an artist is supposed to be. Not a stylist, not a brand, not a guy managing a filmography like a stock portfolio. Herzog treats the making of a thing as inseparable from the risk of the thing, and he has spent sixty years proving it in ways that would end most people's careers or their spines. He does not perform eccentricity for the press. The eccentricity is load-bearing.

It starts, depending on how generously you want to phrase it, with a theft. At nineteen, unable to get the Munich Institute for Film Research to lend him equipment, Herzog broke into an unlocked cabinet and walked out with a 35mm Arriflex that would become the camera he shot his first films on. He has never called it stealing. He's called it a natural right, the position of a man who'd already decided the work was going to happen whether or not anyone gave him permission, and that decision is really the whole engine behind everything that follows.

Start with 2006, a BBC Culture Show segment, Herzog on a Los Angeles street talking to Mark Kermode about Grizzly Man. Mid-sentence there's a crack, and Herzog glances down the way you'd glance at a fly. He'd been shot with an air rifle, pellet lodged near his stomach, and his entire reaction amounted to lifting his shirt, confirming the bruise, and saying it wasn't a significant bullet. Then he kept talking. No ambulance, no reshoot, no framing of the moment as anything other than mildly interesting weather. This is a man who spent years in the Amazon shooting Aguirre and Fitzcarraldo through dysentery, mutiny, and a lead actor in Klaus Kinski who was, by every account including Herzog's own, a legitimate danger to the crew, so a stranger's air rifle barely registers as an event worth flinching at.





Then there's the shoe. In the late seventies Herzog told a young, stalled Errol Morris that if he ever actually finished the documentary he kept talking about and not making, Herzog would eat his own shoe. Morris finished it. Gates of Heaven premiered at UC Berkeley in April 1979, and Herzog showed up, boiled his boot for five hours in duck fat, garlic, and stock with help from Alice Waters at Chez Panisse, and ate a portion of it in front of the audience, dedicating the meal to anyone too scared to start making the thing they wanted to make. Les Blank filmed the whole ordeal as Werner Herzog Eats His Shoe, which is now a more widely quoted document than most actual film theory. The bet wasn't a bit. It was a debt, and Herzog does not seem to be a man who leaves those unpaid, however literal the collateral.

A few more, because they earn their place. For Heart of Glass in 1976, Herzog hypnotized almost the entire cast before nearly every scene, sending untrained Bavarian actors through a story of a village descending into collective madness while they were themselves in something close to a trance, memorizing dialogue under hypnosis and improvising gestures no one, including Herzog, had planned. During Fitzcarraldo, Herzog decided the film's whole premise, a man dragging a steamship over a mountain to reach a new river, wasn't worth shooting as a miniature or a matte painting, so he had a real 320-ton ship hauled up a real Peruvian hillside using pulleys, a makeshift railway, and roughly a thousand indigenous extras, with no visual effects standing in for any of it. Kinski's treatment of that same crew got bad enough that one tribal chief offered, in complete seriousness, to kill him on Herzog's behalf. Herzog turned it down, not out of any particular tenderness toward Kinski, but because losing his lead that far into production would have meant reshooting forty percent of the film. And then in 2006, on a Los Angeles canyon road, Herzog happened to be driving past when a car flipped upside down in front of him. He got out, found the driver dangling in his seatbelt and fumbling for a lighter with gasoline dripping around him, snatched the lighter away before asking any questions, and only then recognized Joaquin Phoenix. When Phoenix climbed out and tried to thank him, Herzog was already gone.

The one that gets me most is quieter than either of those. In November of 1974, Herzog learned his friend and mentor, the critic Lotte Eisner, was dying in a Paris hospital. Instead of flying, he took a compass, a duffel bag, and a new pair of boots, and walked. Munich to Paris, three weeks, five hundred miles, through a winter storm, sleeping in whatever shelter he could break into, convinced with the kind of certainty that sounds insane on paper that his arrival on foot would keep her alive. He wrote it all down and published it as Of Walking in Ice, and it reads less like travel writing than like a man negotiating with death using his own legs as collateral. He got to Paris. Eisner was still alive. She lived another nine years.






What makes all of this harder to file away as pure showmanship is that the work itself refuses to sit still long enough to be a brand. Herzog has directed something like seventy features and documentaries since the early sixties, and there is no single Herzog thing the way there's a single Wes Anderson thing or a single Kubrick thing. Aguirre and Fitzcarraldo are hallucinatory colonial epics shot with Kinski screaming through the Peruvian jungle, dragging an actual riverboat over an actual mountain because the metaphor wasn't worth making if it wasn't also physically true. The Enigma of Kaspar Hauser and Woyzeck are quiet, interior studies built around outsiders and the mentally fragile. Lessons of Darkness turns the burning Kuwaiti oil fields after the Gulf War into something closer to science fiction than reportage, almost no commentary, just fire treated as landscape. Grizzly Man is a documentary that quietly disagrees with its own subject's romanticism about nature the whole way through. And then there's Herzog the actor, showing up in The Mandalorian, in Jack Reacher as a villain, voicing a character on Rick and Morty, which sits so strangely next to the man who directed operas at Bayreuth and La Scala and published more than a dozen books that it stops looking like range and starts looking like someone who genuinely does not distinguish between a jungle epic, a puppet show, and a pop culture cameo, because to him they're all just more chances to point a camera, or himself, at something true. That's the part that actually earns the word purest. Not that everything he touches is good, because it isn't, but that none of it is calculated for a lane.

None of this is stunt work, or if it is, it's stunt work in the oldest sense, where the stunt and the belief behind it are the same object. Herzog calls what he's after "ecstatic truth," a truth you can only get to by going further than the facts will comfortably carry you, and everything above is just that idea walking around in the world unsupervised. Most working artists manage risk. Herzog spends it, on purpose, like it's the only currency that was ever real.

He is going to outlive all of us just to see what happens next, and I don't think he'll even mention it.



Rewindpix - A first impression review of a cool kickstarter digital screenless camera










I've been following this Kickstarter since it launched, which is its own kind of slow-motion commitment — the sort where you watch a project accrue updates and stretch goals for so long that by the time it ships you've forgotten what convinced you in the first place. I pulled the trigger at four in the morning, half asleep, in that specific window where a purchase feels less like a decision and more like something that happens to you. I remember thinking, almost immediately after, that I'd done something dumb. That feeling didn't fully go away; it just got quieter.

What followed was the modern pre-order experience in miniature: downloading the companion app before there was anything to pair it with, joining the Discord, and finding a community already mid-argument. Some of it was garden-variety impatience, some of it was people who'd clearly been burned by crowdfunded hardware before and were bracing for it again. I sat with that ambient frustration for a while without much to show for my own money except a countdown. In the meantime I did the thing you do when you're waiting on gear instead of using it — I developed film recipes and tested them on whatever pictures were lying around, building a little library of settings for a camera I didn't have yet.





Then it arrived, courtesy of Evri, in a box that had clearly had a longer journey than the tracking number suggested. Banged up, but intact, which is about the ceiling of what you can ask from that courier. First impression was good — better than the online noise had prepared me for — with the one immediate oddity being the absence of a basic cap in the box, which feels like the kind of omission that saves a few cents per unit and costs a proportionally larger amount of goodwill.





Firmware update went fine, technically, though "fine" came with an asterisk: it didn't apply until I unplugged and replugged the USB-C myself, no automatic restart prompting me through it. Getting a card ready was its own small errand — I used a SanDisk Extreme, and since Windows refuses to format anything over 32GB as FAT32, I had to go through Rufus to get there. Nothing difficult, just a reminder that this is still a product with rough edges in the setup path, not just the shooting experience.

The shutter button was the first physical surprise. There's a loud clack to it, loud enough that on the first few presses indoors I half expected something to have actually broken. I'm hoping it reads as less alarming outside, away from walls that turn everything into an echo chamber. I already added a bit of foam after opening up the camera as suggested by other users but the difference is basically null - 

The workflow around getting images off the thing is its own adjustment: it doesn't hold a wifi connection while you're shooting, and because it's wifi rather than Bluetooth, pulling photos means disconnecting my phone from my home network every single time. I also haven't found a way to preview a roll's thirty-six exposures fullscreen without first moving them into the gallery, which might be a gap in my understanding of the app rather than a genuine limitation, but either way it's friction.










And then there's the stuff that isn't a mystery, because it's the stuff people online had already flagged before mine even shipped. Under thirty shots in, I've had three misfires while winding, a viewfinder inaccurate enough that I've mostly stopped trusting it, and focus that's soft in a way I'm choosing to describe generously. There's also a real gap between pressing the shutter and the green confirmation light, more than I'd chalk up to my own reflexes. None of this is a surprise, exactly — I went in with expectations calibrated by other people's early units — but there's a difference between reading about a problem and winding into your fourth misfire of the evening.

The good part, unambiguously, is the community — watching other people work through the same app quirks and firmware growing pains in real time, comparing notes, occasionally commiserating. The bad part is that the thing currently feels half-assed, in the specific way that products do when the hardware shipped slightly ahead of the software and the documentation both. It's easy for me to say that with some distance and a functioning sense of humor about it; less so, I'd guess, for the backers who've been more patient than the timeline deserved. I'm only a few hours and thirty exposures into a relationship I plan on having for a while, so I'll withhold the real verdict. For now: it clacks, it misfires, and I already know I'm going to keep shooting it anyway.










you can get yours here https://rewindpix.com/ 

Game Review - Better Than Dead




I went in expecting jank and got something meaner than that.

Better Than Dead is the second game from a small studio called Monte Gallo, published by MicroProse, and it launched into Early Access on May 12 wearing its gimmick right on its sleeve: you're playing back a bodycam recording, fisheye lens and all, of a woman who got out of a trafficking operation with nothing but a pistol and decided filming her own revenge was part of the plan. The Hong Kong setting is doing a lot of work here — neon corridors, cramped apartments, the kind of seedy geography that's clearly been raised on 80s and 90s HK action cinema rather than any modern shooter template, rendered through photogrammetry so it all has that slightly-too-real documentary sheen. Nice touch for gun nerds, the main character is using a Type 54 Norinco Pistol, a Chinese copy of the Russian Tokarev TT-33, also known as Blackstar - 





Mechanically it strips down to almost nothing. One gun, unlimited ammo, no HUD, no reticule — you line up shots by eye through an ADS that never quite promises you'll land them. Get hit once and your vision blurs and the gun starts swaying harder, which turns every subsequent shot into a bet against your own hands. Slide and you get a half-second of bullet time, just enough to feel like an advantage and just short enough that you can't actually plan around it.

That's the part that's been growing on me. It's quick and it does not forgive you — two hits and you're rewinding the whole encounter — and my instinct going in was to want more of a cover-and-line-up-your-shot structure, something closer to tactical shooting than to whatever this is. But the game clearly doesn't want that from you. The fragility is the design. It keeps shoving you forward into rooms instead of letting you post up at a doorway and pick people off, and once I stopped fighting that and started playing aggressive — moving through instead of around — it clicked in a way cover-based shooting never would've let it.





It's landed, predictably, in a weird spot with people — press comparisons keep reaching for Hotline Miami crossed with Kane & Lynch 2, which tracks on rhythm and setting: same trial-and-error loop, same soaked-in-noir Asian crime backdrop. The camera's not really the same trick, though — Kane & Lynch 2 slapped a broadcast-TV filter over third-person carnage, nobody in the fiction is actually holding that camera. Better Than Dead's is diegetic, it's strapped to the woman doing the shooting, which changes what the shakiness means. I'll admit I'm not a neutral party in this comparison — I loved both Kane & Lynch games, dog days and all, and Hotline Miami is one of those titles I still think about years later, so anything willing to work that same seedy, trial-and-error vein gets goodwill from me before it's even earned it.




Might be worth doing proper pieces on both of those at some point — there's more than enough there to justify it. Steam's got Better Than Dead sitting Mostly Positive, and the split in the reviews is basically the same split I have with myself about it — people either bounce off the sway and the fisheye immediately, or they lean into it and start treating every death as data instead of punishment.

It's incredibly weird how games that I start as a "let's see what this is, i'll probably hate it" are usually the ones that stick with me . Played this because I got tired of waiting for Unrecord (https://store.steampowered.com/app/2381520/Unrecord/) - I still hate the promotional image (the one on top of this post) 

I'm still not sure it's what I wanted. I'm fairly sure it's smarter than what I wanted.

The Fiat X1/9 and the pettiest rivalry in Italian car design

 



The X1/9 is proof that a wedge doesn't need a Lamborghini badge to be terrifying to look at.

It started life as the 1969 Autobianchi A112 Runabout, a barchetta with a showy roll bar that Gandini built at Bertone as a sportier pitch for the A112. It did well enough at the Turin show that Nuccio Bertone turned it from an open barchetta into a targa, on the theory that a roof, even a removable one, is easier to sell to an actual factory. Giotto Bizzarrini wandered in around the same time with a rear-engined Fiat 128 he'd built for racing, which fed a few more ideas into the mix. By most logic the prototype should have died quietly there, another Bertone one-off with nowhere to go.

Then Gianni Agnelli walked through the Grugliasco plant, saw the little targa sitting around, and decided it should replace the aging 850 Spider in the Fiat lineup. The internal project code was X1/9. Nobody bothered changing it for the showroom.



What happened next is the part nobody puts on a spec sheet.

Nuccio Bertone sent the prototype out to a small outside body shop for cleanup and repaint, to keep it off the main production line — and somehow Alejandro de Tomaso got wind of it. De Tomaso was mid-feud with Fiat over European market share, so he had Tom Tjaarda sketch a copy on the spot, which Ghia then built into a running prototype on De Tomaso Vallelunga mechanicals. It got shown at Turin in 1971 as the "1600 Spider," months ahead of the X1/9's planned reveal, torching the surprise. Ford, entertained by the whole mess, bought De Tomaso's copy outright with the stated intention of putting it into production. None of it went anywhere legally, and none of it stopped Fiat. The X1/9 was finally shown to the international press on December 22, 1972, on the Targa Florio's own circuit through the Madonie mountains, alongside the Fiat 124 Abarth Rally.

The mechanical layout is a Fiat 128 turned inside out — engine and gearbox pulled from the nose and dropped in transverse, ahead of the rear axle, with the 128's independent suspension and four-disc brakes carried straight over. The chassis, though, is a direct relative of the Lancia Stratos — same Gandini pencil, same era, sharing body components with the car that would go on to win rallies for a living.




That kinship got tested properly. In 1974, Fiat's competition department built an Abarth-tuned X1/9 rally prototype around the 124 Abarth Rally's 1.8-liter engine, good for 200 horsepower at 7,600 rpm, in a car weighing 750 kg — lighter than the 124 Sport Spider and lighter than the 131 Abarth Rally that came after it. Drivers who tested it said it was easier to handle than the Stratos itself, despite sharing almost the same underlying architecture, which is either a compliment to the X1/9 or an insult to the Stratos, depending on who you ask. A turbo version got as far as being discussed and no further. The whole program was shelved to clear the way for the X1/20, which became the Lancia Beta Montecarlo, so the X1/9's competition career amounted to a rough debut at the Sicily and Four Regions rallies followed by three outright wins before it got quietly cancelled. Separately, Dallara built a run of X1/9s into full Group 5 silhouette racers under the name Icsunonove — Italian shorthand for "x-one-nine" — with a 16-valve head making up to 200-plus horsepower, widened arches, a rear wing, and the pop-up headlights deleted entirely. They were good at hillclimbs.




The road car got one real update in its eighteen years on sale. The 1978 Five Speed brought a bigger 1,498 cc engine, an actual fifth gear, a top speed of 180 km/h, and — because U.S. regulations demanded it — a set of aluminum energy-absorbing bumpers that gave the nose a heavier, more American look than Gandini probably intended.

Fiat handed full production to Bertone in 1982, who'd been building the bodies from day one anyway and just put their own name on it. The Bertone years got you special editions like the red-leather "In," nine-layer anticorrosion paint the Fiat era never bothered with, and eventually the 1989 Gran Finale, with alloy wheels and a small rear spoiler, before the line closed for good. Total production across both eras runs somewhere around 160,000 cars.



It was also the only mid-engined car Fiat ever put its own name on, and it stayed the last two-seat open car in the Fiat lineup until the Barchetta showed up years later. All of that, and it still gets remembered as the budget Lambo rather than the car that quietly out-cornered a factory rally weapon that shared half its DNA.

I think that's the appeal. It's the wedge era, and a surprisingly petty piece of Italian corporate history, distilled into something you could actually park outside your flat.







DEATHDISCOUNT MERCH!








So yeah. The blog finally has merch. Nothing polished, nothing corporate — just rough prints, shirts, and junk made the way I like it - this just for supporters of the blog, the usual design work is still available on Redbubble an on other collections on 4thwall - 

Check it out: 👉 Death Discount – Cans1 Shop

Catch you on the flippity-flip

The Ghost in the Shell (2026)



If there is something I usually don't care about much when watching a cross-media adaptation, it's coherence with the source material. If the tone, vibe, and overall style are there, I couldn't care less if a character's jacket was blue in the source material and white in the new adaptation. For some people it's like religious dogma. Not for me. Every medium has its own language, and I think we need to trust authors with their craft and skill to tell a story. A shot in a manga panel and a shot in an anime frame are doing different jobs even when they're depicting the exact same moment, and demanding one behave like the other has always struck me as a category error dressed up as fan loyalty.



That said, I couldn't be happier that the latest The Ghost in the Shell takes directly from the pages of the original manga — not because it was necessary, but because up to now the gap between the manga's vibe and the more pretentious animated movies was huge. Shirow's original book was funny, horny, digressive, occasionally sloppy in a way that felt lived-in rather than unfinished. The films sanded all of that off and left something closer to a cathedral: beautiful, hushed, and a little exhausting to spend two hours inside.

The technical side of this is worth sitting with for a second. This isn't a passion project bolted onto an existing IP by whoever had a release slot open — it's animated by Science SARU, the studio behind Devilman Crybaby and Keep Your Hands Off Eizouken!, and it marks the directorial debut of Mokochan, after his work as assistant director on DanDaDan and on Scott Pilgrim Takes Off. The script comes from novelist EnJoe Toh, which tracks — you don't hire a prose author known for digressive, idea-dense sci-fi if you want another two-hour lecture on the nature of the soul. Character design is handled by Shuhei Handa, who previously worked on Scott Pilgrim Takes Off and Spriggan. And crucially, this is the first adaptation to work directly from Shirow's manga rather than reinterpreting it or slotting into its own corner of the timeline, which is the whole reason it can afford to be funny and horny and digressive instead of solemn.




What's telling is that the director isn't shy about naming the thing  that I clocked as the difference-maker. Ahead of the series' Annecy premiere, Mokochan talked about chasing a "human feel," about wanting to get close to the texture of Shirow's original manga rather than its philosophy in the abstract. Even the choice to lean on hand-drawn animation was framed as the most honest way to represent a manga that was itself drawn by hand, on paper — analog. That's not incidental to the slice-of-life quality I keep coming back to. A show that treats "human feel" as a production value, not just a theme, is going to land its comedic beats and its paperwork-complaining scenes differently than a show that treats humanity purely as a thesis to be argued.




It's worth remembering how crowded this lineage already is before getting to the new series. Mamoru Oshii directed the original 1995 theatrical film and its 2004 sequel, Innocence. Kenji Kamiyama took the TV route with 2002's Stand Alone Complex, which is still, for a lot of people, the high-water mark of the franchise's episodic side. Kazuchika Kise handled 2013's Arise, a prequel that rebuilt the cast around a younger Motoko and, depending on who you ask, either freshened the aesthetic or sanded off its edges. Kamiyama and Shinji Aramaki co-directed 2020's SAC_2045, which swapped hand-drawn animation for full CG and split fandom right down the middle. And then there's the elephant everyone pretends not to see at the party: the 2017 Hollywood live-action film with Scarlett Johansson, which took the aesthetic and none of the philosophy, and is mostly worth mentioning now as a cautionary tale about what happens when you adapt the imagery of this story without adapting the ideas underneath it.

The games are a smaller, weirder footnote, but worth a line. The first one, released on PlayStation in 1997 on the back of the Oshii film's success, didn't even put you in Motoko's shoes — you piloted a Fuchikoma tank through a third-person shooter, which in hindsight is a very on-brand way for this franchise to sidestep its own protagonist. A second game under the Stand Alone Complex name followed on PS2 in 2004, with a PSP sequel in 2005, and neither one is remembered as anything more than a curiosity for completionists - like myself. None of them mattered to the franchise's ideas the way the anime and manga did — they're here for the sake of thoroughness, not because they're doing any of the heavy lifting.

The core elements are always the same, and the depth of the conversation about the very nature of humans — their souls and their minds in a transhumanist reality — was always present across both media, at the core of the story. But what the animated productions kept missing was the thing that made me feel like I was actually reading about someone living in that future world. Instead of high-concept stereotype characters spouting philosophy between one shootout and the next, we finally get natural, slice-of-life, almost comedic moments that make this reality feel like the norm for the characters, not some theatrical sci-fi opera where every interaction exists to litigate the human soul. Section 9 bickering about paperwork, or someone complaining about a cybernetic limb the way you'd complain about a bad knee, does more to sell a lived-in future than another monologue about the Ghost ever could. Philosophy that has to be earned through the texture of daily life lands harder than philosophy that announces itself, and that's the trick the films kept missing and the manga always understood.







That's one of the main things this new series gets right. And it's paired with an art style that is simply gorgeous in every single shot, without falling into the muted-tones cliché we see in every sci-fi piece — the world isn't bleak because it lost its colors. Cyberpunk as a genre has trained us to read desaturation as a shorthand for despair, to the point where color itself starts to feel like an act of optimism the material hasn't earned. This series refuses that shorthand entirely. The world is bleak because of the fate and choices of a humankind so disconnected from its own nature that it's just ghosts living inside shells now, and letting that bleakness live in the writing instead of the color grade is, frankly, a harder and more honest way to make the point.














HXR-K1S - pics from my toy camera

 



Follow-up to my previous entry about photography and lo-fi tools — here are some pics shot with this cheap, unbranded, unnamed camera I got from a certain Chinese reseller for dirt cheap.

It looks way better than a similarly-priced Kodak Charmera, and it's more fun to use: no display, 5 ugly filters, 8 megapixels — probably less — and a cool total-blackout stealth look. Also no way to switch off the flash, so I added some gaffer tape, since I love shooting at night, as you can see. I renamed the camera after the name that shows up when you connect it to a PC — even though I'm pretty sure that's not its actual name — so please meet HXR-K1S. K1S for friends and family.

I love how using a camera like this solves two, if not three, issues at once:

  1. I don't have to worry too much about losing it, having it stolen, or dropping and breaking it — it's cheap enough that it's almost refreshing. I'd still be pissed if something happened to it, but because I like it, not because my wallet would hurt.
  2. The images are going to be bad — you know that going in, and you shoot anyway. Most of the time you get a nice surprise. I honestly wasn't sure I'd get anything at all shooting at night with no flash.
  3. There's no screen, no way to review and get distracted — and that matters more than I expected.

EDIT — Jumping down the EXIF/metadata rabbit hole, I found out the camera actually reports:

  • Camera Maker: HONGXR
  • Camera Model: K1-S

Technically 5 megapixels, but I suspect it might really be less, interpolated up. I'm now planning to open it up, figure out what chip is inside, and maybe try reflashing it.

















Na Hong-jin











 Three films in eighteen years. Na Hong-jin makes movies the way other people have chronic illnesses — infrequently, painfully, and with a level of bodily investment that shows in the finished product whether he wants it to or not. He debuted in 2008 with The Chaser, a film so viscerally unpleasant that it's easy to forget how disciplined it is, and hasn't managed more than one feature per half-decade since. I've made my peace with the fact that this is simply the rate at which a Na Hong-jin film gets built. What I haven't made my peace with, not fully, is how a director this rigorous keeps letting his third acts run wild on him — and how often that wildness is exactly the thing that saves the picture from being merely well-made.



The Chaser announces the whole project in its opening minutes and never really deviates from it: a disgraced ex-cop turned pimp, played by Kim Yoon-seok with the specific self-loathing of a man who knows he built his own trap, chasing down a client who has been quietly killing his girls. Na based the killer loosely on Yoo Young-chul, the "Raincoat Killer" whose crimes were still raw in the Korean public memory when the film went into production, and he shot several sequences on location around Mangwon-dong in Seoul's Mapo District rather than dressing a backlot to look poor and wet. That specificity is the whole engine of the film. It's not interested in the killer as psychology, and it isn't especially interested in Joong-ho's redemption either — what it's interested in is bureaucratic failure as a physical, spatial problem, cops and captors separated by hallways and doors and the four minutes it takes to fill out paperwork while somebody dies on the other side of a wall. The film premiered at Cannes and picked up Best Director and Best Screenplay at the Chunsa Film Art Awards, and I understand why critics reached for "instant classic" — the control is real. But I'll say what I always say about The Chaser to anyone who'll listen: the moment that has stayed with me longest isn't a violent one. It's a single wide shot of Joong-ho standing in the rain outside a house, having arrived exactly late enough, doing nothing, and that stillness does more damage than the film's actual gore ever manages to.



The Yellow Sea is the film I recommend least and think about most. Na scaled up — this was the first Korean production to fold in a Hollywood studio as co-producer, with Fox International Productions coming aboard — and the ambition shows in ways both thrilling and undisciplined. Ha Jung-woo, who played the killer in The Chaser, flips to protagonist here as Gu-nam, a Korean-Chinese cab driver from the Yanbian border region whose wife has vanished into South Korea for work and whose gambling debts get him recruited for a contract killing across the water. Kim Yoon-seok flips the other direction, from hero to monstrous crime boss. The whole structure is a mirror of the first film, and Na clearly knows it, using the inversion to say something sharper than either film says alone: that in his moral universe, position determines virtue more than character does, and nobody who crosses a border — literal or social — gets to keep whatever decency they walked in with. Gu-nam is Joseonjok, ethnically Korean but raised Chinese, belonging fully to neither, and the film's structure punishes him for that in-betweenness with a relentlessness that borders on the theological. There's a genuinely notorious behind-the-scenes fight buried in this film's history too — Na cut a leaner 140-minute international version after the rushed 157-minute Korean theatrical cut, and there's still argument among people who track this stuff about how much of that "director's cut" was actually his hand versus the studio's. I don't know that it matters for the viewing experience, except that it explains why the back third feels like it's sprinting to catch up to itself — hatchet fights spilling into hatchet fights, gang war swallowing the personal story whole. It's the least controlled thing Na has made and also contains some of the most purely kinetic filmmaking of his career, extended chase and slaughter sequences that abandon geography for velocity in a way that shouldn't work and does, right up until it doesn't. In the Ending there is a whole mute scene that is really a masterclass of storytelling. 



Which brings me to The Wailing, the film that made me a Na Hong-jin person rather than someone who'd seen two of his movies. It premiered out of competition at Cannes in 2016 and had already done enormous business at home — over two million admissions within days of its Korean release — before the international press got hold of it, and I think that commercial appetite tells you something true about the film that the festival framing doesn't: this is a movie that wants to be popular entertainment and prestige dread at the same time, and mostly succeeds at both without ever quite reconciling them. It's set in Goksung, which is not an invented backdrop but Na's own grandmother's hometown, a place with a specific and, per Na himself, personally lived-in history — persecuted Catholics who fled south and were eventually martyred there, a community his family was part of. That personal thread matters, because the film's engine is less "is this village possessed" than "what does a father do when every authority he could appeal to — police procedure, folk shamanism, the church, his own instincts — turns out to be equally useless and equally plausible." Kwak Do-won's Jong-goo spends the film's back half getting steered by three separate figures, a shaman, a Japanese stranger, a woman in white, each insisting the other two are the demon, and Na refuses with real discipline to ever let the audience get more certain than Jong-goo does. That epistemological cruelty is the film's actual subject, dressed as folk horror.


It's also, structurally, the most extreme example of the thing that both draws me to N. The first act plays almost as procedural comedy — bumbling cops, a doughy protagonist, small-town rhythms — before the film tips into a register so operatic it risks tipping the whole thing over, screaming exorcism sequences, a hexing ritual staged with the pacing and cross-cutting of an action climax, grief expressed at a volume that occasionally reads more as spectacle than as feeling. I won't pretend those stretches don't test me. Na has never fully shaken a taste for the overstated gesture, the moment held two or three beats past where restraint would have hit harder. But I've come to think that's the tax you pay for access to what he's actually good at, which is not restraint at all — it's pressure.

Every film builds toward a space, physical or moral, where there is no correct decision available anymore, only worse and slightly-less-worse ones, and he holds you there in real, elapsed time rather than cutting away to spare you. When the excess lands, and often it does, it's because he's earned the release valve. When it doesn't, you're watching a very serious director briefly lose his nerve about how serious the material already is.


What actually keeps me coming back isn't the violence, which gets most of the ink in every piece written about him, English-language coverage especially. It's the rain, honestly — Na shoots weather like it's a character with its own motive, sheeting down over Mangwon-dong, over the strait between the Chinese and Korean coasts, over a mountain village that can't tell divine punishment from bad luck — and it's the way his camera keeps finding people standing very still in the middle of catastrophe, arriving exactly too late to matter, which might be the truest through-line in all three films: not that evil wins, but that by the time you understand the shape of the thing you're fighting, the fight was already over.


And now, I'm so damn excited for his next film - "hope" 


A Short History of Bad Cameras I've Loved



I started shooting because a high school class required a camera and I didn't have one, which is a less romantic origin story than most photographers give you but probably the truest one. The camera I bought with my own money was a Canon PowerShot A580, a plastic point-and-shoot with none of the qualities anyone associates with a serious tool, and I used it because it was what I could afford, not because I loved it yet. That came later, and it came sideways, the way most real obsessions do.

The Canon lasted exactly as long as it took me to understand what I was missing, which is when I picked up a used Nikon D70s and ran it until there was nothing left to run. I shot with that camera the way you'd wear a single pair of boots into the ground, no ceremony about it, just use until failure. Then came my first workhorse, the D7100.

Somewhere in that stretch I started paying attention to lenses, which is the moment every photographer's bank account starts quietly dying, and somewhere after that I discovered film, which is the moment it dies loudly. Film photography is not a hobby, it's a leak. Money for cameras, money for stock, money for development, money for cameras you didn't need but wanted because they were beautiful mechanical objects with a history attached to them, and before long I wasn't just shooting film, I was collecting the machines that made it, studying their internals, their lineage, the decade-by-decade evolution of a technology that peaked before I was old enough to buy it new - 




I liked the pictures. I liked the objects more, maybe, or at least equally, which is its own kind of tell. What I was really chasing, without having a word for it yet, was the fact that every one of those machines had a personality baked into its flaws, a light leak here, a soft corner there, a specific way of rendering skin or blowing highlights that no two bodies did quite the same. The imperfection wasn't a bug I was tolerating. It was the whole point.





Eventually the hobby became a job, one of roughly thirty creative roles I've cycled through in a career that resists a clean job title, and the job meant professional gear, for photo and increasingly for video too. This is where the story usually turns upward, more skill, more clients, more capability, and for a while it did. Then it turned. I can point to the exact hinge: I went from a Fuji X-T3 to a Sony A7III, a technically correct upgrade in every spec that matters, and something in me broke on the way. The more resolution and dynamic range and computational cleverness got packed into the body, the less I felt anything holding the camera. Work slowed too, which didn't help, but the bigger wound was aesthetic, not economic. I stopped knowing which bag the A7III lived in. Not because I'd stopped shooting. Because the images it made looked like plastic, correct and lifeless, and I couldn't locate the version of me that used to care. A camera that corrects for everything ends up saying nothing on its own. All the character I'd spent years chasing in old mechanical bodies had been engineered out in the name of accuracy, and accuracy, it turns out, isn't a point of view.





What survived was smaller and older. My X100 never left my side through any of this, the one camera I never had to talk myself back into loving, and it's still the closest thing I have to a permanent creative partner, mechanical enough to still feel like mine even with a sensor inside it. And what's pulled me back into the rest of it wasn't better gear, it was worse gear, deliberately: a Game Boy Camera, a Casio wristwatch camera, tools built with resolutions measured in the low thousands and interfaces that seem actively hostile to the idea of getting the photo off the device and onto a screen. I still haven't solved how to pull images off the watch. The Game Boy Camera is, charitably, impractical. None of that matters. What matters is that the grain and the smear and the total technical incompetence of these things put something back that a full-frame sensor had quietly removed, and from there I fell into toy cameras properly, chasing that same lo-fi imperfection on purpose, the same thing that had quietly hooked me on old mechanical bodies years before I could name it.








So this is where I've landed, kit-wise, for now. The film rotation is an Olympus XA3 or a Lomo LC-A, and I only swap between them when I finally finish a roll, a rule I hold to so I don't end up with four half-shot rolls sitting around unaccounted for. Alongside that there's a cheap, unbranded 8-megapixel CMOS thing that shoots like garbage and has no screen, no settings, nothing to adjust or second-guess, which is precisely why it's fun. The X100 comes out when I want something that actually has to look good, or when I'm shooting something specific enough that I need control back. The Sony sits somewhere in the bag collecting dust, mostly waiting for someone to ask for a portrait so I have an excuse to put the 80mm on it and pretend I still care about that kind of precision. But even then I'll still grab the pocket digital shitbox and the Chuzhao mini camera on the way out, because those are the ones that still make me want to take the picture in the first place. None of them are technically good cameras. That's not incidental to why I love them, it's the reason. A flawed image has a fingerprint on it, a light leak, a soft edge, a color that's wrong in a way only that specific broken plastic lens gets wrong, and that fingerprint is the closest thing to a voice a photograph can have. Perfect images all start to look like nobody took them. The bad ones sound like me.