Night-vision lenses so thin and light that we can all see in the dark | Hacker News Reader

Night-vision lenses so thin and light that we can all see in the dark

newatlas.com

180 pointsbookofjoe94 comments

While the experiment is certainly cool, it should be pointed out that this is not a passive process and the demonstration needed a femtosecond laser operating at 830nm with an average power of 10mW. While it might be possible to make integrated lasers with those properties in the future currently they use a Ti:Saphire laser which occupies a space of about 1.5m x 30cm x 30 cm. So clearly not portable.

That would fit and work on vehicles though, which would let them drive at night without visible lights. So still some good military applications, which means it'll probably still get a lot of funding and someone will probably manage to shrink the laser down quite a bit more. At least with a backpack-able version anyway.

Military vehicles can already drive at night without visible light. Except in the most extreme case (new moon and heavy cloud cover) - and then they can illuminate using infrared. Such a device costs <2,000 USD and is already wearable. More expensive and/or vehicle-mounted systems increase that capability to the point where you basically never even need the IR illumination, unless the vehicle is fighting inside a building (rare).

The technology in the article, aside from being heavily editorialized, will remain inferior to that for a long time. However, one possible application would be the use of higher-wavelength infrared for the active illumination, so that other militaries with night vision are unable to see your infrared headlights and then blow you up.

IR illumination (no matter the wavelength), or this laser-based night vision, would light up a military vehicle like a big bright target. For night vision to be viable in a military setting, it needs to be passive.

Is it possible to design ultra wide spread spectrum light pulses that are below the signal to noise ratio if you don't know the code? At least some RF signal sources are at least difficult to detect I guess.

Not too long ago, some luxury cars had an optional built-in night vision display, but as far as I know, they've stopped offering that feature, not sure if it was just ineffective or too inexpensive. As I age and my night vision degrades, I'd be really happy for a car with night-vision built-in, even if it meant wearing a headset or goggles.

https://www.autobytel.com/10-vehicles-with-night-vision

I'm curious as to the potential harm concentrations of light can be, even if outside the human visible spectrum. Not just to people, but wildlife and plantlife.

The problem with illumination is that it’s easy to pick up stray rays from the light source. Hence why infrared is preferred.

True currently, but laser form factors are shrinking quickly. Reminds me of this article from a couple months ago where a team built (among a host of other sizes/use cases) a 20W(!) laser using a 500um PCSEL which they describe, mounted, as "roughly the size of a webcam". Their lasers are also at comparable wavelengths

https://spectrum.ieee.org/pcsel

The laser doesn't seem to be a core requirement of the system, it's mainly used as a very short bright illumination source and isn't part of the up conversion system.

The full experimental setups are available in the supplemental materials: https://onlinelibrary.wiley.com/action/downloadSupplement?do...

Full paper: https://onlinelibrary.wiley.com/doi/10.1002/adma.202402777

Seems like it could be done with mode-locked fiber. It’s just easier to get high-Q with a tabletop setup. I’m bullish on hollow core to solve that too.

IIRC this is how cheap green laser pointers work too. They use a near-ir diode and then use magic crystals to change the frequency into the green range. And if the manufacturer is really naughty, they won't put an IR filter on the end to stop you blasting your eyes with any unconverted IR

Ok but you shouldn't blast your eyes with the green light either?

Yes, but at least you have a chance at seeing it. IR? Invisible beam. Stuff of nightmares that this is so easily and cheaply available

If you want to learn more ...

Vertasium has a good video on the 3-different types of Night Vision.

https://www.youtube.com/watch?v=UAeJHAFjwPM

1. image intensification

2. active illumination

3. thermal imaging

>Enhanced Infrared Vision by Nonlinear Up-Conversion in Nonlocal Metasurfaces

https://onlinelibrary.wiley.com/doi/10.1002/adma.202402777

Thanks for the actual paper they published

This is great. Everybody can get them if they want and then we can turn off all the streetlights and have a starry skies again.

Yep. Plus a whole new class of traffic accidents: Electric cars crashing into pedestrians and each other at midnight that were never heard nor seen because a driver wearing these glasses thought it would be cool to drive with his lights off.

People already do this.

Heck, I remember a decade or so ago in the town I was living at two cops killed each other this way. One was parked sideways in the middle of the road with his lights off at the top of a hill, waiting for speeders down the hill and another cop crashed into him, lights off, also looking for speeders on the same road. All at around 3am or so...

>parked sideways in the middle of the road with his lights off at the top of a hill

Er, whut?! This sounds a bit Keystone Cops, to put it politely. I hope the reality was not as dumb as this sounds.

No, it was exactly that dumb. It was real "country" out there...

The guy driving with his lights off hit the parked cop car at about 80mph...on a 45mph limit road. Those roads are usually filled with deer at night too. Total darwin award winners.

I think a lot of people don't understand the lights are not only for you to see, but so others can see your car.

Once someone from the family turned of my headlights from auto to off. I don't know for how many time I drove at night going to the squash club and back, grocery shopping, until the day I was stopped by police. Thankfully, got no ticket, but I got unease thinking about how many intersections I had driven at night with headlights off, being essentialy invisible for other drivers until too late.

Now I make a point of checking the lights are on auto every time I drive.

I've done the same thing several times. Especially on my car with popup (non-auto) headlights.

Being stuck one the side of a pitch black road because your eye glass batteries run out or you dropped them.

Ehhh… I dunno, man. It’s a great first step and an awesome demonstration of metasurfaces, but NewAtlas is heavily editorializing the outcome, to the point of making the paper’s authors look foolish (if you didn’t know that the authors aren’t making the outlandish claims)

Among NA’s claims: night vision without cooling, and comparison to MCP-based image intensifiers.

The authors successfully upconverted ~1500nm photons to 550nm with high efficiency.

So, not really night vision in the sense of thermal IR (yet) That’s why they don’t need cooling: it isn’t thermal IR, which is more like 9000nm.

But it’s also not the same class of image intensification like in a typical “night vision” goggle/scope:

You get some intensification by up converting IR to vis with high efficiency, while simultaneously passing vis (and the neat effect of combining both in one image without additional processing). You could get like 1-2x intensification from the wider spectrum (using a wider band version of the apparatus—the current one is fairly monochromatic), not 100x by amplifying photon counts.

Newatlas also fails to mention another interesting property: the amplified wavelength is tunable!

I'm tired of science reporting overselling technology. They are starting with the idealized outcome: "this film lets you turn glasses into night vision!" When the reality is far from that. New technologies in development should be framed in terms of the goals and applications, but then establish more grounded facts of where the state of the art is, the proof of concept that exists, and the state that the technology is currently in.

Who cares about night vision. Unless I missed a whole lot of development in physics, the ability to passively shift frequencies from outside to inside of visible range is the cool thing. I'd like to wear this material during the day, possibly with a visible light attenuator in front. The world may look different in beautiful or practically important ways when viewed at wavelengths outside our visual range, and we won't know how much until we start actually seeing them as we move about.

It seems as if every article from New Atlas has that problem, more so than other places. At least every article I notice here on HN.

I agree and I will no longer post New Atlas articles here.

It is still an interesting concept

Exactly; the article more or less obfuscates what the present proof of concept looks like. After opening with this dream of what sounds like an entirely passive system you could just select as a coating on your next Zenni Optical order, It makes no mention of current power requirements or how much space everything else takes up.

Would it fit in a Vision Pro? Is it currently 75 pounds of equipment spread over a test bench? The article is useless on these questions.

It's a selection/survivor bias going on too. The public doesn't want to read about some incremental gain in some subfield they have no expertise in. People interested in those are reading the journals and preprints and mailing lists and such.

Unfortunately, that means what typically makes it to widely shared mainstream science news is the clickbait sensationalist marketing hype. Sober analysis isn't good for TikTok...

I call it the "IFLScience" effect.

So if it is not in the thermal IR range, I suppose it is not particularly useful in seeing humans and other warm-blooded animals in the dark? As far as I know, there is not much ambient IR at night time.

What I would find most useful when driving at night is glasses that attenuate intense light rays, so that the increasingly-powerful oncoming headlights will not be blinding. I have a vague impression that channel-plate image intensifiers have this property, on account of individual channels saturating in their ability to release electrons.

That's what I was thinking as it was going on about driving. True "night vision" is going to have an even harder time with modern headlights coming at me than my eyes do.

It kinda reads like someone used chatGPT.

How would this level of "night vision" compare to nocturnal animals like cats or owls?

Interesting question. And where does the main commercial/military established night vision tech sit in comparison to the animal world too?

While interesting, this would almost be more interesting if it was discussing some of the other applications I a think of, such as photovoltaic cells. By converting IR light to visible light, they should be able to improve the efficiency of silicon-based PV cells.

Another application I can think of would be for greenhouses. By converting IR light to visible light, you can increase the amount of useful light for plants (and animals), increasing the photosynthesis (PAR) and plant growth. Mars comes to mind. Surely there would need to be a lot of refinement, but if it can increase useable light 2x, assuming the usable light increase occurs mostly 2 hours before dawn and 2 hours after dusk and dawn, and 2 hours after, for a 12 hour day, that's a 33% boost in efficiency. Half that to be conservative and ~15% is no joke, either

There is little IR light around, compared to visible light.

Plants could have evolved the ability to use IR light, but they didn't bother.

In fact, for plants it's more important being able to tightly regulate light absorption, which is why leaves are not black.

I’m not sure how I would measure this, but it doesn’t match my experience.

I have a Gen 3 WP PVS-14 that I use quite a bit. There is sufficient IR light outdoors pretty much all the time to see quite clearly with it, even with a new moon. The only time it’s difficult to use passively is if it’s extremely overcast. Even then, the tiny, weak IR LED on the device itself is enough to reach out to 50 yards or so.

Think about standing in front of a bright infrared source, like a space heater. You can't see the infrared, but you can feel it on your skin.

If you pay attention you can pick up more subtle infrared sources, for instance certain kind of light bulbs. But in general it feels pretty "dark."

this speaks more to the sensitivity of gen 3 white phosphorous than the abundance of infrared

as you say, it only takes a weak LED to work as a flood light

Only if it requires less input power than it outputs, which it probably doesn't.

A giant space film at the Lagrange point that upconverts (and down converts?) to only the photons that plants crave. Global warming and sunburns gone! Pollinators that see uv are done for, but we are going to kill them by accident anyway.

Edit: Also, put it in the windows of cars and houses and get rid of all the street lights so we can see stars again!

Stupid question, what stops us from doing visible-light binoculars with 1x zoom but an aperture larger than human eye pupil?

I know that regular magnifying binoculars can boost the magnitude for stargazing but is the magnification part necessary?

Edit: Apparently this law does: https://en.wikipedia.org/wiki/Etendue#Brightness_theorem

You can have light boosting binoculars with 1x zoom. Similarly, looking through the viewfinder of an SLR camera will show the light- gathering capability of the lens. The very best light gathering camera lenses (and binoculars) only collect 2 to 4 times more light for a given focal length compared to the human eye. Since our vision is so nonlinear in sensitivity, the 2-4x boost doesn't seem like much. The brightness theorem only says you can't "perceive an object to be brighter than it is at its surface" (my interpretation)

I get the nonlinear upconversion part, but what about the optical system? Are they saying that this preserves the direction of the incident photon so that it doesn’t require optics?

I sense this being export controlled technology.

nVG's are expensive, and if you want a system that combines the infrared heat spectrum overplayed with nvg's your going to pay an arm and a leg.

Where can I find an actual photo?

(Not just renderings)

It looks like the material produced is microscopic. The research article[1] is obscurely linked at the very end of the featured article. See Figure 3 for an SEM image of the material (not quite a photo, but close enough), and Figure 5 for photos of microscopic subjects through the filter.

[1]: https://onlinelibrary.wiley.com/doi/10.1002/adma.202402777

There isn’t one, the author is dishonesty presenting a phenomenon demonstrated by an optics lab as the technology they hope it could in theory eventually lead to.

Clickbait title. I'd bet the authors of the paper are fuming with the mis-representation of their results those journalists made.

I wonder why humans can't see infrared?

Seems like it would have been a good survivability enhancement that evolution could have taken care of?

Humans are diurnal. We foraged for fruit during the day and slept when the sun went down, so it's not very clear that being able to see infrared would help.

Plus evolution doesn't work by automatically adding the best body enhancements like a superhero. It might have happened that the genes for making red cones got duplicated (making one person a tetrachromat) and then that person had lots of children and those genes gradually evolved into infrared receptors and were successful, and the mutation was not too much of a burden in other ways. Or that's all very very unlikely and it simply didn't happen.

> Humans are diurnal. We foraged for fruit during the day and slept when the sun went down

Moonlight is plenty bright enough to see by, many people (eg: Sama-Bajau) catch non fruit food (such as squid) at night with a small oil | candle light and have done so for longer than European recorded history.

Aboriginal hunters in Australia catch particular animals that come out in the cool of the night using the moon to see by.

And do not forget the good old THC night vision enhancer.

https://www.sciencedirect.com/science/article/abs/pii/S03788...

But also see this

https://www.nature.com/articles/s41598-021-81070-5

YMMV

Animals who see infrared are cold blooded. My guess is that the hardware that evolution can easily produce to detect and process infrared gets flooded by the heat of warmblooded animal and is therefore not effective for them.

Most animals don't see into thermal IR ranges eitehr but the ones closer to the visible spectrum.

We could in theory avoid the problem if our eyes were at the top of antennas.

But yeah, genetic evolution does not explore states widely different from its origin. The more complex an organism, the more "conservative" the entire process is, and we are quite complex.

The amount of rod and cone cells in your eye fill all available space. Introducing an IR sensitive cell type would take away space from existing cells, limiting their usefulness. Having night vision would mean sacrificing something else. IR frequencies would also require larger sensor cells due to their larger wavelength.

As a side note, men generally have a higher ratio of rod cells, allowing them to see in the dark. While women have a higher ratio of cone cells, allowing them to better distinguish colors. Apparently, these tradeoffs were important enough to cause for specialization in our DNA.

> fill all available space

Couldn't we have evolved physically larger eyes to accommodate additional "visual apparatus" though?

Yes, but you will find a similar analogy here. Balancing tradeoffs is a central feature of evolution.

People with tetrachromacy have 4 cones Don’t know if that extra color perception, negatively influence other aspects of their vision

Also a lot of the IR spectrum is absorbed by water, so it wouldn't pass through the eyeball.

It would also be great if we had eyes in the back of our heads to see 360 degrees.

Oh, and we should be able to hear above 20,000hz in order to detect small prey more easily.

If we don’t have a particular adaptation there are a few basic explanations:

- it just wasn’t on the path: the RNG just simply didn’t produce this outcome

- it was on the path but it was sacrificed for something else (for humans this is typically our big, energy hungry brains)

- it was on the path but just didn’t offer a selection advantage

A scenario I just came up with: Half of all humans had IR vision, but when we got control of fire, they were outcompeted by those who weren’t “blinded” by the IR output of the flames.

Someone once told me that the visible frequency range matches the spectrum for which water is transparent.

This person also said that this is an indicator of life evolving in water. That seems unsatisfactory though since it has been a long while on land.

Evolution is the king of good enough. Things that work but have issues usually don't get 'revised'. For one of the funniest examples check out the recurrent laryngeal nerve which connects from your spine very close to your brain to your voice box... with a short little detour all the way down into your neck under your aortic arch back up to your voice box. [0] The main theory is it's a vestige of ancient fish anatomy and even shows up in GIRAFFES where it's over 4 meters long because it takes the same detour all the way down the neck.

[0] https://en.wikipedia.org/wiki/Recurrent_laryngeal_nerve#/med...

Wow! Thanks for sharing.

For the optic nerves in humans, a similar questiom comes -- Why does it connect to the rear side of the brain when the front is much closer! This perhaps relates animals having their eyes on the sides of the head and a relatively small brain positioned in a way that the eyes were closer to the rear side of the brain.

I'm far from an evolutionary biologist but AFAIK the majority of our brain is a late addition in the scheme of our evolution from sludge to human and vision is connected to one of the older parts and the new parts are up towards the front. Another case of the iterative nature of evolution, the frontal cortex where most emotion processing happens came about and enlarged a lot through the evolution of chimps that eventually branched off into humans.

Regardless of how long life has evolved on land, our eyes are filled with (mostly) water - so any frequencies that can't pass through water would be scattered and absorbed by the goop inside of your eyeball.

That makes sense. I should have thought of that myself. :-) Thanks!

The human visible frequencies also match the frequencies where the sun has most of its output.

Moreover, they also match the frequencies for which the atmosphere is transparent.

Can you imagine going through the design (well, not really design, but for the purpose of the joke…) of human eyes, like a code review?

- Nice resolution, really impressive stuff

- Can we fit in some more frequencies?

- Wait, are all the cells backwards?!?

And the massive blind spot that is handled by just having our brain be unable to notice it

"We'll patch it in firmware."

The image recognition algorithm is probably good enough to fill it in at the application layer anyway, so if the firmware patches don’t make it in we’ll see if anybody complains.

The firmware patch is wild. I carefully positioned a sign post so the sign was in my blind spot, and was amazed that I saw the full post clearly without the sign on it, just as if it had been photoshopped out. It is truly crazy how much our brains process the images we see, yet we assume we are looking directly at a representation of reality.

That applies to the minimal field of vision too.

But now advances in technology allow us to extend our sensory capabilities beyond natural limits

> let us see in the dark could change how we work and play after hours

More like change how we fight and surveil.

Safer than an Eyeshine but costlier than a pack of menthol Kools.

Did they show any demo images of what the output looks like?

Likely see military use it first once they declassified it

Naa they are just using sensitive CCD/CMOS without any IRcut filters. (well they have deployed 4rd gen photomultipliers as well, but thats slightly different)

Most human systems are still photomultiplier tubes because they're so much faster than digital night vision systems. It's tough to move/shoot/drive with the digital systems because of the lag from photon to photon.

Having the same thought.

It's about time for this. In the eighties I grew up watching the bionic man, night rider, MacGyver etc. The promise was broken. I'm old now and my knees are shot and I'm suffering with cateracts and a crap health care system that takes years to get it addressed. So hurry up science, I'm not impressed.

I was thinking Knight Visions from Snow Crash! Need to integrate some other techs for the full experience.

There is still plenty of time to make you a bionic man yet. As the Germans say, the rate of innovation is hitting breakneck fart.

The six million dollar man adjusted for inflation is a paltry $42M! Plus with advances in technology and medicine, it'll probably cost much less. I wouldn't count on it being covered by insurance though.

They will be glad to cover it. Anything to drive up premiums.

Wow, it could enhance nighttime activities

Reply on news.ycombinator.com