Limits of the Visible Spectrum

Transcript + Footnotes

The cosmos is a bright disco of light of every scale… a huge spectrum of electromagnetic radiation, and our pathetic little human eyeballs… can see less than 1% of it.

Today on chromaphobe… Why can’t we see in the infrared or the ultraviolet? How come we can only see in the visible spectrum? 

Let’s get the lame answer out of the way… if we could see in the infrared… it would be red, not infrared. The definition of the visible range is simply what light humans can perceive, so if you change our perception, then you also change the definition.

For the interesting answer, we instead have to ask, why isn’t our visual spectrum wider. For that, let’s look at the entire electromagnetic spectrum.


At the very extreme here, we have x-rays and gamma rays. This is all very high-energy ionizing radiation with wavelengths shorter than 124nm. This “light” is the reason you have to put a lead blanket over your junk when you are getting an x-ray at the clinic, because this is the stuff that will rip your DNA apart and wreck your spermies or give you cancer. We aren’t equipped with x-ray vision because it wouldn’t take long for those x-rays to destroy our eyes.

On top of this, we aren’t evolved to see light that isn’t there, because for the most part, there is no ionizing radiation in our environment… and thank god, if there were, life would likely never have evolved at all, with or without vision. The only significant source of x-rays on earth is from lightning… and a vision system enabled only by lightning would be… stupid. You heard me superman…


Instead, the thing that enables our vision is the light coming from the sun because pretty much all light you see: a green tree, a white lapdog, a moonlit trashcan is just reflected sunlight… or at least that was true 100 years ago. Nowadays I wouldn’t be surprised if the majority of light that hits YOUR eyes, yes you specifically Kyle… is coming from the device you’re staring at right now…

Get outside… not NOW now… uh, wait til this video is over… oh, and while you’re at it, wait until after you subscribe to this channel.

Speaking of screens, do any of you ever wonder how our eyes are going to evolve after staring at RGB OLED screens for the next 1000 years?


Anyway, let’s pull our full electromagnetic spectrum back up and let’s compare that to the sun’s emission spectrum, or the range of wavelengths that are contained in sunlight. This spectrum excludes ionizing radiation, since the sun is simply not HOT enough to emit those high energy photons, so at the short side, sunlight only contains down to about 200nm. On the long side, the cutoff is a bit more ambiguous, but there are not a lot of photons past the 10,000nm mark. That’s a lot of the EM range we can get rid of now, so let’s zoom in on just the portion illuminated by sunlight. As we covered already, we can drop the ionizing radiation below 122nm, and drop the Far UV that the sun can’t emit.


Next, the cosmic sunlight has to travel through the atmosphere, which acts as a giant filter called the atmospheric window. The middle ultraviolet, 200-300nm gets absorbed by the OZONE in the atmosphere, so the sunlight that reaches the surface of the earth and is available for vision is limited to wavelengths longer than 300nm, so that’s another chunk gone, bye bye mid UV.

For some animals, this 300nm IS the limit of their UV vision, so… just spitballing… if we were to put that hole back in the ozone layer, just imagine what a service we would be doing to those animals, enhancing their color vision like that…. Anyway…


The next step is that light must travel through the optical media of your eye… the cornea, lens, aqueous and vitreous humors before it reaches the photoreceptors in your retina. These materials, mostly the lens, strongly absorb the near UV light between 300-400nm.

Why did we evolve this filter? Well, while the non-ionizing near-UV light is not going to literally rip your DNA apart, it does still have enough energy to catalyze harmful chemical reactions such as photooxidation that will ultimately, over time, degrade your retina and lead to diseases like macular degeneration. It’s likely that we evolved these filters to protect our retinas.

Footnote 1

Some people believe this harmful light doesn’t stop at 400nm and extends into the visible spectrum in a phenomenon called blue light hazard. These same people are happy to fear monger you into buying mostly useless blue-light blocking glasses. The truth is, your lens is already taking care of the worst of those wavelengths by blocking out the UV light, and if the blue light was really bad for our retina, we probably would have evolved a lens that blocks out the blue light too. I’ll definitely have another episode on blue-blocking lenses soon though.


Then how come so many other animals from bees, birds, fish and mice lack this UV filter, which allows them to see much deeper into the UV than we can? Doesn’t it also wreak havoc on their eye parts? Yeah, the difference is… we live a lot longer.

When was the last time you met an 80 year old mouse, besides Old Mr. Jingles from the Green Mile, adorbs. A mouse retina’s exposure to UV light probably does cause harmful photochemical reactions, but the mouse simply doesn’t live long enough for that damage to accumulate to any dangerous degree. Hell, it’s only now that humans routinely live well past our reproductive ages that problems like cataracts and macular degeneration, both accelerated by UV light, actually become a problem, but without this UV filter in our eye, we may all be going blind at 30, and THAT is why we can only see light down to 400nm… usually…


I mentioned cataracts earlier, these happen in old age when the lens turns cloudy or opaque due to a lifetime of absorbing UV to protect the retina… wow what a good friend. Cataract surgery removes the entire lens, thereby removing the UV filter and allowing near UV light to reach the retina. This filterless condition is called aphakia. While increased UV exposure associated with aphakia is dangerous for the retina, aphakia does indeed broaden the visible range and allow people to see UV.

So what does UV light look like?

You may know that each of the L, M and S cones of the retina are most sensitive to 560, 530 and 420nm light, respectively. The sensitivity curves are depicted by these nice Gaussian shapes that indicate the wavelengths of light that those cones are most sensitive to. If you’ve seen any other videos on this channel, then you know it’s the relative excitation of these cones that evokes color… but what you see here are the alpha-bands of each cone. What this image ignores, or rather, what the UV filter of the lens normally suppresses, is that each of those cones has a hidden secondary sensitivity peak called the beta-band in the UV.

While any single wavelength in the visible range will excite the cones to different levels and evoke color, the beta-bands of the three cone types all kinda overlap. This means all the cones are excited to the same level by UV, so UV light actually evokes mostly WHITE to subjects with aphakia, usually with a tinge of blue. To them, objects that reflect UV light simply look a little lighter… more washed out.

So not exactly the transcendent experience you may have imagined and actually, despite the world looking a bit different, subjects with aphakia aren’t able to perceive ANY new colors at all. In fact, if we take a rainbow and simulate how an aphakic would see it… kinda lame.


A lot of people find this unbelievable, but I think that’s just because there is so much conflation in popsci between wavelength and color. Wavelength does not actually define color, seriously… and seeing more spectrum does not equal seeing more colors.

Footnote 2

No, seriously! It is a convenient explanation to say that green is all light between 490-550nm, but almost none of the green you see consists of a single wavelength, and some colors cannot be reproduced by a single wavelength or continuous range of wavelengths. Light does not have a color until it interacts with your visual system, and depending on the parameters of your vision, the same light spectrum will evoke a different color. We only label 530nm as green, because it evokes green to the “standard observer”. Fish, birds, dogs and even myself have visual systems that differ from the standard observer, so 530nm does not evoke green for us. A better definition of color is to describe it in LMS color space, or the relative excitation of the three cone types, but even this is not perfect. Color models that seek to define color just keep getting more complex as we learn more about the human visual system.

If humans – as a species – evolved to lose the UV-filter in our eyes, and the sensitivity of our cones shifted along the spectrum to increase our visible range, we would continue to see all the same colors, but they’d just map onto different wavelengths. What was ultraviolet would now appear blue and what was blue would now appear green, etc.

Footnote 3

If we really space the cones out evenly along the visible spectrum, we may be able to see a few more hyper-greens (where the green cone is excited in absence of the other cones) that were not perceptible before, but we would lose color resolution in other parts of our gamut. Overall, we wouldn’t notice “new” colors, just some different colors.

It would still be really cool and maybe even useful to see UV light. After all, so many other animals have evolved UV vision with no UV filter; what advantage does it give them? Reindeer use it to better see plants against the snow. Insects use it to differentiate similar looking flowers. Eagles use it to see rodent urine… and all of these uses have one thing in common… they are hugely speculative. The truth is, we have very little idea why any of these animals need UV vision. And if I were to ask a vision scientist, “hey what do humans need blue cones for?” They probably wouldn’t be able to give me a coherent answer. Such is evolution… anyway…

Footnote 4

FOOTNOTE: The pressure that led primates to very recently evolve separate green and red opsins is very contentious. There is a whole wiki page on it. Why our ancestors 400 million years ago originally evolved the blue cone (SWS opsin), and why we have retained it, or even why primates SWS more recently shifted to shorter wavelengths is… really anybody’s guess.


There IS another way to see UV light without eye surgery, because there are plenty of reports of people seeing light when exposed to UV. However, these cases can all be attributed to AUTO-FLUORESCENCE.

When UV light is absorbed by the lens, it can re-emit light or fluoresce in the visible range.

Think of blacklights. They emit UV light, which you of course can’t see directly, but you know a blacklight is around by the visible light that fluoresces off of your white sneakers, or the hotel bathroom.

In your eye, it is the lens that is absorbing the UV and partially fluorescing it in the visible range. Any one of you can “detect” UV light through this mechanism, but there are some caveats.

  1. First, the direction of the fluoresced visible light is scrambled relative to the direction of the incoming UV light, so no image is preserved, just formless flashes.
  2. Second, to perceive this visible light, the UV either has to be very strong, or the visible light in the environment has to be weak or absent, situations which happen rarely in uncontrolled situations.

So it IS UV-vision, but its the kind you’d get if you asked a jackass genie to grant you UV vision.


So that is the short side of the spectrum all wrapped up, why we can’t see UV. Let’s pull our spectrum back up again so we can consider the infrared. Why does the visible spectrum stop here at about 700nm? Well, it doesn’t exactly stop, but the deeper you go in the IR, the more intense the light must be to be perceived. That’s why the boundary between the IR and visible spectra is reported very differently depending on the context, anywhere from 680nm, which is a practical limit to vision, to 850nm, which is only achievable in laboratory tests.

This graph shows that your vision becomes 10x less sensitive for every 35nm that we shift towards the infrared. By 850nm, your retina is so insensitive that you essentially need to shine a laser right into your eyeball to be able to see anything, so anything longer than that, definitely not the visible range.

…and for the 9th time on this channel, I will remind you… don’t… shine lasers in your eyeballs. It’s not only stupid, but it won’t even help you see the brand new color of “infrared” because infrared just appears uniformly… red…

How cool would it be if you could see microwaves? It would make your kitchen a much more exciting place.


Unlike the UV limit, the IR 700nm limit is not based on the sun’s emissions or an atmospheric filter or a physiological filter… its simply based on the sensitivity of the light detecting molecules in your retina: the opsins. We know that those opsins can easily be tuned to higher or lower wavelengths… that’s how we have these three opsin types that allow us to see color. Then why can’t the opsins be tuned to even longer wavelengths, allowing us to see deeper into the IR?

The thing is, it’s not just humans, it is very rare for any vertebrate visual system to have cones that are more sensitive to the infrared than our red cones, and therefore very rare for any animal to have true infrared vision. Biology indicates that there is some fundamental limit that prevents an opsin from tuning into the IR. Quantum physics explains why.

Footnote 5

When opsin proteins bind to normal retinal (11-cis-retinal; A1), the red cone can reach a maximum peak wavelength (lambda max) of about 570nm, a bit longer than the human max. However, some animals use the same opsins with a different form of retinal (11-cis-3,4-dehydroretinal; A2) to achieve a peak wavelength of up to 620nm. Opsins with this retinal need less energy to trigger the phototransduction cascade, and therefore the IR limit is extended, but the tradeoff is lower sensitivity and noisier vision (Corbo 2021). Certain fish and amphibians that use the A2 retinal (see porphyropsin) may be able to see a bit deeper into the infrared than the rest of us.

Footnote 6

Pit vipers and vampire bats have heat “vision” that is essentially IR perception. This works completely differently to normal vision and does not use the eyes or anything similar to opsin proteinss. Whether images are actually formed by this “vision” is contentious, so I’ve not included it here as IR vision, but maybe another good topic for a future video.

Okay, I made it sound complicated, it’s not that complicated. The core of your vision is the retina’s ability to convert light into an electric signal on your neurons. In your retina, you have cone cells, in your cone cells you have opsin proteins, and in each opsin protein you have a retinal molecule. When a photon hits the retinal, it excites one electron in that molecule to a higher energy state – a process called electronic transition. This process then triggers the retinal molecule to twist and blah blah blah signal reaches your brain.

However, a certain amount of energy is required for this electronic transition to occur, and lower energies will just jiggle the retinal instead of twisting it. As a photon’s wavelength increases, its energy decreases. Longer than 700nm and a photon is not packing enough juice to reliably achieve that electronic transition and our vision is desensitized to it…

So, take away those jiggly photons and we establish an IR limit. All that’s left of our electromagnetic spectrum is 400-700nm: the practical visible range.


But wait, you may be thinking… if one infrared photon doesn’t have enough energy to achieve the electronic transition, can’t we combine the energy from two photons? Absolutely. The problem is… two photons need to reach the opsin at exactly the same time… and I mean exactly with… *count* 12 zeroes, else the molecule will just get a double jiggle. Those photons have to arrive within picoseconds of each other to add their energy to a single electronic transition that will twist the molecule.

In normal environmental conditions this scenario is highly unlikely, but let some scientists fire a pulsed laser into your eye to concentrate all that infrared light into a few picoseconds and it’s much more likely for two photons to arrive at the same opsin at the exact same time.

Do so in pitch dark and you may be able to see the infrared laser light. You won’t see red though. Two 1060nm infrared photons have the same energy as one 530nm photon, so this 2-photon effect appears not as red light, but as green light.

So it IS infrared vision, but it’s the kind you’d get if you asked a jackass genie to grant you infrared vision.


So why isn’t the visible range wider? We don’t need it to be.  Look, human color vision is still some of the coolest shit on the planet, and if we want to know what something would look like with an altered visual system, we’ve got hyperspectral imaging and/or instagram filters to simulate it. For now, just be glad you’re not colorblind.

And as always… please don’t shine lasers in your eyes.

This is chromaphobe.






Leave a Reply

Your email address will not be published. Required fields are marked *