65 comments

[ 3.6 ms ] story [ 129 ms ] thread
I think they should have started with a more achievable goal to make it to 10 years, and then re-evaluate afterwards. Maybe 100 is achievable to some extent? 1000 seems like pure fantasy to me.
Or a series of cameras. With one being opened every 10 years or so.
I thought the same. I wonder if they did already? If so I would love to see the picture(s).
It's an art project whose purpose is to get people in the present to consider the passage of time, rather than a project whose purpose is to produce a photograph. To spark discussions like this about its feasibility in the face of the uncertainty of the future is as much a part of the project as the camera itself.
Saying something is art is really not an excuse to not put in extra effort. Proof of engineering makes a statement in itself.
> To spark discussions like this about its feasibility in the face of the uncertainty of the future is as much a part of the project as the camera itself.

Nobody's talking about the future of the world as a result of this project, they're mostly talking about the future of the project itself, and how it'll end up a wet fart because of poor planning. If the goal is to get people slapping their foreheads then it's a smashing success. Beyond that, I'm not sure what the effect will be. All press may be good press, but not all press amounts to art.

Yeah, it seems purpose built as an untestable thought project. Might as well have had some kids put hand prints in concrete at the base.

Imagine what the time value of the money spent on this will be for this in 1000 years. A market investment of $1000 that could return 2% over inflation over that time period would be equivalent to $400B in today's dollars.

Problem is that’ll get you a medium Diet Coke in 3024.
Note that I said beat inflation, but hey maybe 2024 Diet Coke cans will be the last remaining currency in 3024.
Beating inflation by 2% over 1000 years? Physically impossible
Well, I agree $ is unlikely to be a thing by then, but the broad stock market has beaten inflation by 6% (almost 7%) over the last 150 years. That's already over 1000x.

So, no more impossible than having a random stick take a photograph for 1000 years. Perhaps, they'll sell 0.0000001% of the stock of a corporation that owns the photograph for $4, or it will contain the last unburnt bitcoin.

"over the last 150 years"

I believe that the next 150 years are not going to be like that. If for no other reason, demographic collapse will destroy the consumer base for companies.

1000 years? That bug ridden pinhole camera will be around, we won't.

I took a photo like that with a ~ 1 year exposure time. The patterns the sun makes are really cool, you can make out the moments with and without sunshine throughout the days.

My gut feeling of doing this for one year: 1000 years is just plain impossible. Not with something made out of metal with an actual hole in it. Maybe some glass capped ceramic to avoid water ingress and corrosion... but after one year, my camera was already full of weird goop.

https://files.rombouts.email/IMG_6500.jpeg

That's a great image. Thanks for sharing.
Neat idea, what kind of camera did you use?
Stick a piece of photo sensitive paper in a pinhole camera and you're good to go.

People typically make them out of black film canisters and such. Google "solargraph"

I got one pre-made[1], but as another commenter said, it's "just" a box with a tiny hole and some photo sensitive paper, a pinhole camera

[1] https://solarcan.co.uk/

Depending on the chemistry of the image production, it may not need air inside, and could work similarly if it was a solid block of glass, so maybe that would be an option. Attaching the "film" securely in a removable way might be tricky, so maybe you don't remove it, just look at or scan the image through the glass.
How does the sensor work for capturing light for that long? Wouldn't it produce a solid white image after a single day? Is there any difference between taking a 1000 year exposure and taking a photo every day for 1000 years and laying them ontop of eachother?
It's a pinhole camera, a tiny hole creates a faint projection of the scene on the inside of the "camera". You can make one yourself with a suitable cardboard box, a bit of foil, a needle and a sheet of photographic film. Exposure time ranges from a few seconds to a minute or two.

In this case the artist used a layer of paint instead of a more traditional photosensitive material. The projected light will cause the paint to fade and change over time causing an image to appear over the course of centuries instead of minutes.

I've wanted to make a pin hole cameras since I was a kid. I didn't realize the photographic film part .. now that makes sense! Where do you source that in North America these days? How do you handle it? I recall loading film into cameras as a kid .. that was exposed to light but then you moved over that part of the film? Presumably the part exposed to light is junk?
It's not film, it's the paper would project the negative onto to make a picture using an enlarger.

It's not as readily available in this digital age but you should still be able to find it at a "good" camera store. I'd look for maybe single owner dedicated to craft type places vs chain dedicated to tech.

You'll be wanting the B&W stuff, easier to develop and you can work with it under red lights instead total darkness. You'll also need 2 chemicals called "developer" and "fixer" suitable to the paper and distilled water to bath it in.

Pinhole cameras are fun because you can build them in less then an hour for nearly free and it gets you hands on with the underlying principles of photography.

> Is there any difference between taking a 1000 year exposure and taking a photo every day for 1000 years and laying them ontop of eachother?

There is a difference between taking a photo every day for 1000 years, what you end up with is, say, a 365,000 second exposure. This is an actual 1000 year exposure.

The algorithm where you take 60 1/60th second exposures over the course of a second and then stitch them together is popular. There are many mobile apps that do this and try to use edge detection / "AI" to make them line up, so you can take handheld long exposure shots. (Spectre is an iOS app, but I'm sure there are others.) You can also do it yourself with a tripod and photoshop. The results are good, and it lets you take long exposures when you are constrained by having too much light and no ND filter / tiny aperture on your lens / etc.

> Wouldn't it produce a solid white image after a single day?

In general, reciprocity holds in photography. A 1 second exposure at f/8 is the same as a 2 second exposure at f/16 given the same sensor. (Films have a technical limitation called "reciprocity failure" for long exposures. Silver halide requires TWO photons to hit each atom, and the statistics of that aren't linear. This means that film exposures depend both on the shutter speed and ISO sensitivity; a properly exposed long exposure on film will often be longer than what the math dictates. The box that your film comes in will often give you a number to multiply your calculated exposure by.)

So what they need is exceedingly low ISO film. A picture of a sunny scene would generally be exposed at f/16 and 1/<the ISO of the film>. A pinhole lens can't open that wide, they're more like f/180. That is "7 stops" less light than f/16, so would need a shutter speed of 1.3 seconds to capture the same image. Now you can see how insensitive the "film" in this camera is; it needs 31,536,000,000 seconds of exposure. That means it's 3.17e-11 ISO film they have in there. What quantum effects happen to film that insensitive, I don't know. They say it's paint fading, and I don't know what the rate is. It might end up being overexposed. (Film is usually OK being severely overexposed, the opposite of digital. No idea if paint fading behaves the same way as silver halide. Probably!)

One other thing to think about; films often have differing sensitivity to different colors of light. Early film was overly sensitive to UV light, because it contains so much more energy than red light. Modern black and white film is pretty linear over the colors of light, so you end up with a picture that looks right to your eye. Paint does not have this carefully-designed chemistry, and is probably only sensitive to UV light, so that reduces the effective ISO of this camera setup even further. It won't respond to light you can see, only UV light. But, that works well enough to yield a recognizable image. (The problem in photography is that the sky contains too much UV, and you end up with an ugly white contrast-less sky. In this case, I don't think they care.)

But reciprocity breaks down eventually, and it's not obvious to me that the paint will fade at all.
I think there's a pretty significant amount of UV coming in even through a pinhole. There are indoor-only things that are damaged by UV (old computer cases) even though windows block a lot of the UV.

I think what's hard to balance is what effects other things will have. Humidity will get into the camera and the water will attack the paint. Air will be oxidizing the pigments. Bugs can probably get in and eat the paint. 1000 years is a long time! You have to hope that other effects don't destroy the image more quickly than it can be created.

Imagine something which slowly fades with light, like paint.

Taking a photo every day is work, stuff can break. A passive sensor like paint would just sit there.

But how can they be so sure the paint won't fade completely after 1000 years? It could be gone after 10 years. Perhaps you could check on it as long as you don't move it even a millimeter.
the chemistry of photography is well-researched and involves simple calculations. it comes in many flavors, the product is repeatable and testable, and can be made very stable.
We have 1000 years old buildings around. You can study parts which were in the sun versus shadow.
That steel pole definitely looks like it will last 1000 years. /s

As another posted stated, they should have started small: 1 yr, 10yrs, then maybe a 1000. Iterating each time with a better design.

Also, dump the steel. You need something more inert.

It's art. Doesn't really need to work. But yeah I would be more impressive art if it at least appeared like it could work.
Art may be useful (or at least usable and realistic), e.g. imagine providing instructions how to get the photo from that camera for (hopefully) people in 1000 years, or even describe what that cilinder on a pole is in the first place.

Science experiments demonstrating discoveries of the past centuries (magnetism, electricity, radio) are a kind long-term useful art.

I'm not sure I like this kind of pretend art. Surely many people will believe that it is a serious project that's intended to work. I visited Expo once and it was full of this "pretend stuff" and it was unsettling how it mislead many people.

But, it's ok, I'm certainly not saying it's bad art just because I'm on fence about it.

Art is subjective so everyone's entitled to their takes on it.

Personally, I don't like when art makes me contemplate the delta between _intended ideals_ vs. actual execution.

Especially for this piece... it's just fake. "We made a camera that's going to take a 1,000 year exposure!" - and it's like... maaaan. No, no you didn't - and it's unethical to market that you did.

Steel can work. Just not that steel.
I would have expected a hole drilled into a cave or something like that. A piece of steel attached to a deck probably won't survive 100 years, simply because someone will want to make the deck bigger or better in the intervening time and then the camera will move.

That said, that whole pitch drop experiment is going pretty well. It's not quite at 100 years, but it's getting very close. (2027). So that's 1/10th of a millenium. (I will note that the pitch drop experiment has been moved before. It's not nearly as sensitive to movement as a camera.)

What it some bug lays a nest there? Or some dust flies into it? There's so many ways to disable this even without discussing the types of materials used.
> an experimental philosopher at the University of Arizona

Clearly I am in the wrong line of work.

How does this fit into data protection law? If a person stands in front of it for a moment, (a tiny shadow of) their likeness will be preserved in the image. What if that want that removed?
You have a programmers view about how the law works. Such a claim would be dismissed immediately as frivolous.
Depends. Can you provide your date of exposure and a equivalent exposure of yourself against a neutral background so we can average you out?
There is no data protection required if you don't give them your data.
Unless your stand in front of it for 10 years in the same pose, you will not be recognisable in the least
I wonder if the paint used as exposed layer will see its “printability” property decay with tile. I have experimented pinhole exposures length up to 1 day with and photographic films and the light capture decay drastically with time: what’s happening in the beginning of your scene is much more present than what was present in the end.
A 1k year-long exposure will result in … a gray blob with no useful information. Now a 1k-year timelapse of one photo per day made into a film would be amazing!
If the probability of the camera being damaged, destroyed, vandalized, or stolen in any given day is one in a million, then this thing only has a 70% chance of surviving a thousand years. (And I think that's high...)
Or moved, which will ruin the photo
The photographer Michael Weseley did a number of long exposures (>12mths iirc) of Berlin during the 1990s as it was being transformed after the fall of the Wall and subsequent reunification. Viewed as huge prints in a gallery they were impressively detailed, as layers of various rebuilding and urban planning projects were revealed in varying degrees of transparency, depending on how long they took and how recent they were:

eg https://www.uncubemagazine.com/blog/14374185

Beautiful work and great magazine. Thanks for sharing
It would have been better to attach the camera to some ancient historic landmark that has been around for hundreds of years or even thousands of years and will likely still be preserved to be around for a thousand more years. Several exist.
(comment deleted)
I can’t wait to see how the picture comes out!
So after a millenium, how do you remove the film to look at it without it being ruined when it's exposed to light?

Does it need to be developed or something?

Or does the material stop being photosensitive after a few centuries?

(I mean I know this is an art project rather than something genuinely meant to last a thousand years in practice, but it seems like they're trying to make it work at least in theory.)

It does not seem to be using a material we would normally consider appropriate for photograpic film. The article mentions: "multiple thin layers of an oil paint pigment called rose madder"

If i understand the principle correctly, the captured light would simply slowly fade exposed portions of the "film" over centures, eventually resulting in a faint negative of the sum total of the scenes witnessed.

Exposing the film to light at that point, one kiloyear from now, would probably do nothing, unless you left it exposed for a couple decades to direct sunlight.

If you could’ve taken a 4.5 billion years long-exposure photo on our planet, I believe it would’ve been mostly a water world for most of its existence. So you’ll see a smooth water on the photo. But about the sky I’m not sure: will it be completely white because of the stars?
With a steel pole that looks like it's already rusting, I can't imagine it lasting 1000 years, and good luck putting it back in the same position if you replace the pole!

I don't think many buildings would show up, if any. I once took a 10 second pinhole camera exposure of a busy playground, and there's no sign of the children in the image at all. One kid standing still for one second would be like a building that lasts a century for this image, but they weren't visible because of the other 90% of the time.

A smarter project is to make a video then merge the frames.
I think the approach was chosen partly for its lack of moving/digital/energy-needing parts.
My approach saves more energy. This one has the camera continuously running and feeding light to the sensor or film.

Mine is just taking a picture once per day.

What camera - the hole in the paper/...? There's no active electronics here, also no timekeeping ("once per day")
It seems a symbolic gesture more than anything, though I can see where it would appeal to the kind of Hackernews who always writes a year with five digits (like 02024).