EnChroma: Color for the Color Blind | Hacker News Reader

EnChroma: Color for the Color Blind

enchroma.com

67 pointsGuiA25 comments

Quick summary of how it works: it's a passive filter, no electronic components. Nothing magical, but it is designed in a very clever way.

Our eyes have three types of cones, call them L, M, and S (you might think "RGB" but it's really not, but that's not a bad lie to believe). Between L and M, for example, there is a fair bit of overlap. So, these glasses filter out the colors in the middle, where L and M overlap the most. This means that the only colors which make it through are the colors which your eye can distinguish better, making the colors more vibrant.

This works because most colors are a mix of a bunch of different frequencies of light. Say, ROYGBIV. Maybe you have color blindness and a ROY mix looks the same as an OYG mix. These glasses will filter out OY, so you are left with R in one mix, and G in the other mix. These are easier to tell apart.

So, obviously not perfect, but it doesn't need to be. It just needs to be better than what nature gave you (if you're colorblind). It also doesn't work for people with, say, 100% deuteranopia, but most people with color blindness aren't that severe.

This means they're more like a hearing aid (shifting frequencies you're unable to hear into the spectrum your ear can process and simply adjusting loudness) than a "cure".

I'd want to try one out for the kick of it, but I can't imagine using it the way I would use regular vision-correcting glasses (or contact lenses). Possibly as an aid for driving (where colour vision can be critical) but not as a part of my everyday life -- much like Gunnar glasses, which I only use at work.

I guess I'll have to wait for gene therapy or bionic technologies to advance to the point where I can actually get something that lets me see colours properly.

Or, heck, tweak these so they shift IR or UV into the visual spectrum and they'd have my attention.

EDIT: Emulating "normal" colour vision is interesting but emulating mantis shrimp colour vision would make the glasses worth the price tag.

Explanatory note: Most humans have three types of color receptive cones; a few have four. Mantis shrimp have sixteen.

Fewer cones may be better for low-light vision. Apparently mammals grew out of a evolutionary niche where this was important, and they generally only have two cones while other vertebrates have four. Remember how dogs have poor color vision? That's true of almost all mammals. Primates are unusual among mammals in having 3 types of cones and better color vision. Meanwhile reptiles, birds, amphibians, insects have 4 types of cones and much better color vision than us.

The eye has a limited amount of real estate for light sensitive cells, and animals have a limited amount of energy to dedicate towards eye growth. More cones generally means fewer rods, which means poorer night vision. You can see this for yourself at night, because cones are concentrated at the center of the field of vision and rods are more common at the periphery. A common trick in astronomy is to look next to a star rather than directly at a star, it is not too hard to find stars which are invisible when you look at them directly but which become visible when you look to the side. Various nocturnal and deep sea creatures also feature a number of different adaptations, such as the tapetum lucidum (which reflects light to pass through the retina a second time) and special inverted rods which increase light sensitivity. These structures come with their own various tradeoffs. For example, the tapetum lucidum reduces image sharpness.

In spite of this, humans have very good eyes, with a balance of good low-light vision, good color vision, and good image sharpness. There are other animals better at each of these, but humans are still pretty amazing.

They're hearing aids in reverse. They filter out light instead of adding to it. Still, I'm colorblind, and when I tried them on in sunlight, it was an amazing experience. Maybe not a cure but damnit if I'm not happy these things now exist! Seeing that a carrot was orange instead of green for the first time was mind blowing and emotional. Enchroma will open a whole new market that might lead to a cure! Until these glasses started going viral, color blindness was an invisible disability that the real world mostly ignored despite its prevalence, because there was no market in fighting it.

So it's an optical notch filter?

It doesn't shift anything, just blocks the ambiguous "inbetween" colors.

Yes.

Just to clarify why calling it RGB isn't that wrong, for people who don't understand the LMS thing[1].

LMS stands for Long/Medium/Short (wavelength) cones, and I assume most of us know RGB is Red/Green/Blue.

Long cones are responsible for seeing long wavelengths ("λ") of the visible electromagnetic spectre ("EMS") are going towards Red and peak there before fading into infra-red.

Medium cones are responsible for seeing the medium λEMS peak at green, where the Short cones kick in for short λEMS, and they peak at blue.

Some colours are seen when two (or maybe even all three?) cones are simulated: Orange, Yellow, Cyan, Purple, etc.

--- [1]: I'm not an expert on the subject, bit I did some research into it because a friend of mine is partially colour blind.

The reason why I call RGB "wrong" is because the L and M cones both have peak sensitivities to greenish light. So the three cones would be something like "yellowish green" (564-580 nm), "bluish green" (534-545 nm), and "violet" (420-440 nm). It is a common misconception that long cones are mostly sensitive to red. People often say that, but it would be more accurate to say that long and medium cones are both mostly sensitive to green light.

At least two cones are usually stimulated, and often all three, even by monochromatic light. Color is perceived by the differences in stimulation between cone types.

RGB is a model which allows us to reproduce color, and it is useful because it can be mapped to LMS using a linear transformation. However, RGB is not how we perceive light.

I have a pair of Enchroma CX Receptors and blogged a while back about my experience.

http://madscientistlabs.blogspot.ca/2014/06/i-want-to-see-vi...

Basically, I knew I was colorblind from a very young age. I've refrained from describing things by color since then because sometimes I'd get it wrong and people would look at me like I was insane. Never bothered to learn the resistor color code despite being an electrical engineer. Bombed my geology labs because I couldn't distinuish rocks by their color (and because didn't know there was an exemption available for people like me).

The first time I tried the Enchomas on was a very emotional experience. Chickweed exploded from my lawn. Some red cars were a red I could never have imagined. I saw a distant dead brown tree in my yard that would have been almost invisible before. I suddenly understood why my wife liked a burgundy shirt I had never cared for before.

The reactions of some of the people in the videos linked here probably undersell the effect in some cases. Why? Several of the shots were taken indoors. The Enchromas work best in bright, full spectrum sunlight. I noticed in a few of the videos that people gravitate to the windows or head outside where the effect is more pronounced. Enchroma is very up front about this on their site.

The downside with the CX Receptors I have is their weight. The Receptors fit over my (rather strong) prescription glasses and aren't that comfortable when I'm bouncing around doing something physical. They are great for driving though. I just don't like when I have to take them off - the world dies a little when I do.

This is an amazing video:

https://www.youtube.com/watch?v=WCcxwieuDH0

A colorblind person tries these glasses on without knowing what they are. :-)

It's funny, because when he sees "purple" you have no idea if he sees the same purple people without color blindness do (and other commenters seem to confirm it isn't) but it's still enough of an upgrade that he's emotionally moved.

You can't beat that with a stick.

> and other commenters seem to confirm it isn't

Color isn't in the eyes, it's in the brain. Since these glasses help the brain receive the right balance of RGB, I see no reason to believe he isn't seeing purple. I'm colorblind and trying these on was a mind-blowing trip.

Would it be possible to apply the same filter to a sample video to make the colors pop for a color blind person? Maybe apply this filter http://i.imgur.com/EILB2lW.png via a VLC plugin?

If so, that would be a fantastic way to demo this product online. And also could be a great way for potential users to know if it would work for them or not.

Video only dictates the intensity of the 3 components (RGB), the spectral distribution of the components is dictated by the color filters in the TFT and its lightsource.

I.e. on the software side you don't have much influence over the spectrum that reaches the eye.

According to their test, I am a "strong pro-tan" and an unlikely to get much benefit from the glasses - I would have been curious to try the effect, but it would have been something rather special for me to consider parting with that much money for them! I can't say being colour blind has really affected my life (since I was a kid and used to colour the sky in purple anyway!) although I do sometimes worry when retouching images that I might get the colours way off. It would be interesting to see colour how other people do - as it is, it usually ends up coming in conversation, we have a good laugh at me trying to read the numbers on the test, and then I forget all about it again!

A good friend of mine is colorblind and I did a bunch of research on the topic of colorblindness a while back.

The problem(I'm sure it's great nevertheless) is that their technique doesn't add any new colors and doesn't work for full dichromats.

Today we've got the technology too supply different color information to each eye, it's even commonplace in the form of 3d television. I did a bit of testing on how the brain interprets this (fun party trick, use complementary colors for each frame and the resulting picture is grey without the glasses and can be psychedelic shenanigans with.)

How stable the resulting color is seems to be connected to object recognition, for a fruitstand there certainly were colors I didn't see before, for a fullscreen it oscillated between colors irregularly.

I would love to reverse engineer the interpolation function the brain is using. Since it's subjective it's difficult, even with a big sample size you can't match to control colors.

If anybody wants to get into this topic I don't have any papers since it was a while back, but topics such as MacAdam ellipse, color confusion lines and image daltonization should be good starting points.

I'm colorblind and thought theirs was one of the best online tests available, but it seems they've changed it. I took their old test a few times and it seemed quite accurate to me (based on the symptoms they list). With their current test I keep getting "inconclusive". Here are my results from the "good" test, I agree 100% with the symptoms but I don't agree with the simulation: http://enchroma.com/test/result/strong-protan/?completed=1&r...

What would be the equivalent design for people with normal color vision?

To see like a tetrachromat? There isn't one. These glasses only work for anomalous trichromacy. Those with just two working types of cone cells (dichromacy) will have to wait for gene therapy. Fortunately, anomalous trichromacy is the most common kind of color blindness.

While it wouldn't be as visceral as these glasses, you could mimic arbitrarily fine-grained color perception with a tunable bandpass filter. Set the band to 20nm and scan it through the visual range (400nm-750nm) quickly enough... you'd see some interesting artifacts. You'd also get a better feel for the emittance/reflectance curves of various dyes, plants, and lighting technologies. For example: OLED displays only emit light in a few narrow frequency ranges, so they would "flash" three times for every scan through the visual band. Blue flash, green flash, red flash. Repeat. Anything in sunlight or incandescent illumination would have much more continuous dimming and brightening.

While it sounds cool, I don't know of any device that can do this.

>To see like a tetrachromat?

Yes this is exactly what I was thinking.

Your narrow bandpass scanning idea sounds cool. Even if it was slow it would be interesting to look at something and watch it change over time - if you have three separated bandpasses then you should get something like colour vision, but where the colour of an object changed over time.

From the Enchroma website:

"Do these glasses work with normal color vision also?"

"The EnChroma Cx lens causes a very powerful “super-color enhancement” effect. The same “boost” that it gives to the vision of a person with color blindness can be appreciated with normal color vision as well. In addition, the lens also has a neutral color balance, which means that your sense of color accuracy is not compromised or disturbed.

We have received testimonials from some customers who have normal color vision that the glasses have beneficial effects on their mood, such as from seasonal affective disorder. While this is an interesting potential application, this particular use case has not been studied in detail yet and we do not have any official claims that can be made regarding this use.

The EnChroma Team all wears the Cx Sunglasses, even though most of us are not color deficient–and we love wearing them!"

Yup. Apparently they weren't even invented for colorblind people; they were popular among surgeons because the glasses would enable them to better differentiate what's going on in a pool of guts.

The story is that a colorblind friend of a surgeon thought they looked cool and just asked if he could try them on.

I was thinking if they could be used to try and get the equivalent of tetra-chromacy.

Reply on news.ycombinator.com