You have seen the videos. Someone puts on a pair of glasses in a garden, looks at a flower, and their face changes. Those clips are not staged in any meaningful sense, and the reaction is usually genuine. They are also a terrible basis for deciding whether to spend a few hundred on a pair, because the thing the glasses do is much narrower than the thing the videos imply.
Colour blind glasses do not add a missing colour. They cannot. What they do is remove light, and removing the right light in the right place turns out to be a surprisingly effective trick on some eyes and a near-useless one on others. Which group you fall into is decided by something you probably already know about your own vision, and it is worth working out before you buy.
What the filter is actually doing
Typical colour vision runs on three cone types. The long-wavelength cone, usually called red, peaks around 560 nanometres. The medium-wavelength one, green, peaks around 530. Those two peaks are close together and their sensitivity curves overlap across a wide slice of the spectrum. Your brain does not read either signal on its own. It reads the difference between them, and that difference is what red and green feel like.
In the most common form of colour blindness, the two peaks sit even closer together than usual. The cones are present and working, but one of them is shifted toward the other, so across a large range of light the two produce nearly identical responses. A difference of almost zero is almost no colour information, and reds, greens, browns, and oranges collapse into a narrow band of muddy sameness.
The filter attacks the overlap directly. It cuts a notch out of the spectrum right where the two cone responses are most similar, roughly the yellow-green region around 570 to 590 nanometres, and usually a second notch near 480. Removing that band means the light still reaching the eye is weighted toward wavelengths where the two cones do differ. The difference signal gets bigger. Colours that used to sit on top of each other pull apart.
It is the same idea as increasing contrast on a washed-out photo. You are not recovering information that was never captured. You are stretching what little separation exists until it becomes visible.
Why they help some people and not others
This is the part the marketing tends to skate over, and it decides everything.
- Anomalous trichromats have all three cone types, with one shifted. Deuteranomaly and protanomaly are in this group, and between them they account for the large majority of colour blindness. There are two overlapping signals here, so there is something for a filter to pull apart. These are the people the glasses can genuinely do something for.
- Dichromats are missing a cone type entirely. Deuteranopia and protanopia are in this group. There is no second signal to separate from the first, so widening the gap between two responses is not a meaningful operation. Filters mostly just make the world darker and shift its overall cast.
- Monochromats have one or no functioning cone type. Nothing a filter does will help, and light sensitivity is usually the bigger daily problem for this group anyway.
Most people who describe themselves as colour blind have never been told which of these applies to them, because a school screening tends to report a pass or a fail rather than a type. If you are considering glasses, finding out first is the single highest-value thing you can do. An optometrist can tell you in one appointment, and an Ishihara plate test will at least tell you whether there is something to investigate.
What the research says
The honest summary is that the glasses change how colours appear and do not restore normal colour vision, and that controlled studies have been much less impressed than the videos.
Studies that put the glasses in front of standard clinical tests, the Ishihara plates and the Farnsworth D-15 arrangement test, have generally found no reliable improvement in scores. That result is not as damning as it sounds, because those tests are designed to detect a deficiency rather than to measure everyday colour experience, and the filter shifts the whole scene in a way clinical plates are specifically built to resist. The manufacturers agree on this point: they state plainly that the glasses will not let you pass a colour vision test, which matters if you are buying them to get through an occupational screening. You will not.
Work looking at chromatic contrast rather than test scores has been more positive, with some evidence that extended wear over a couple of weeks increases the response to colours in the affected range and that some of that persists briefly after the glasses come off. The effect sizes are modest and the participant numbers are small.
Set against that is a large volume of consistent self-report: people say certain colours look more distinct, that they notice reds in foliage they had not seen, and that the effect is strongest outdoors in bright daylight. Self-report is weak evidence in general and unusually weak here, given how much anticipation surrounds putting the glasses on for the first time. It is not nothing either. A tool that reliably makes someone enjoy looking at a garden is doing something, even if it does not move a Farnsworth score.
The practical stuff nobody mentions
They need bright light. The filter works by throwing away part of the spectrum, so the image is darker. In full sun that is fine and the effect is at its strongest. Indoors under warm bulbs there is often not enough light left for the trick to work, which is why manufacturers sell separate indoor and outdoor versions with weaker and stronger filters.
There is an adaptation period. The first few minutes are the dramatic bit, and then your visual system starts re-normalising, as it does with any colour cast. Most wearers describe the effect settling into something subtler within an hour. That is normal and does not mean the glasses stopped working.
They distort as well as reveal. Removing a slice of the spectrum changes every colour, not just the confusing ones. Colours you could already distinguish perfectly well will look different, and some people find that more annoying than the original problem. This is a particular issue if you work with colour professionally.
Never use them for anything safety-critical. Not for checking whether a wire is red or brown, not for reading a warning light, not for driving on the assumption that you now see signals correctly. The colour you perceive through a notch filter is not the colour someone with typical vision perceives, and treating it as if it were is the one genuinely dangerous way to use them.
Try before you buy, or buy where you can return. The variation between individuals is large enough that no review will tell you what your experience will be. A generous returns window is worth more than any spec sheet.
The cheaper things to try first
A lot of what people actually want from the glasses is information rather than experience, and information is close to free.
- Phone camera apps that name the colour under the crosshair solve the is-this-shirt-green problem completely, instantly, and for nothing.
- System-level colour filters on iOS, Android, Windows, and macOS can be tuned for specific deficiency types and cost nothing to test.
- A single tinted contact lens in the non-dominant eye is an old trick that works on the same principle as the glasses, creating a difference between the two eyes that the brain can use. It is cheap, and an optometrist can talk you through whether it suits you.
- For work, labelling and layout changes do more than any optical fix. If you cannot tell two cables apart, the answer is a label.
So, do they work?
If you are an anomalous trichromat, you spend time outdoors, and you want certain colours to feel more distinct, then yes, within limits, and you should still buy from somewhere that takes returns. If you are a dichromat, the honest answer is that you will probably see a darker world with a colour cast and not much else. If you need to pass an occupational colour vision test, the answer is no, unambiguously, and the manufacturers say so themselves.
The framing that does the most damage is the one in the videos: that this is a cure, and that the moment of putting them on is the whole story. It is a contrast enhancer with real physics behind it and a narrow group it suits. Knowing which group you are in first turns a gamble into a reasonable purchase.
Not sure where you sit? The ten-plate colour blind test takes a minute and will tell you whether the wide-gap plates gave you any trouble, which is the signal worth taking to an optometrist. If you want the science on how many colours a typical eye can separate in the first place, we have written that up in how many colours can you see, and the flip side, people who may have a fourth cone rather than one too few, is in are you a tetrachromat.