You probably saw the photo of the dress that caused an uproar on the Internet a few weeks ago (1). You probably also formed a firm opinion about whether the dress was blue and black or white and gold. Some of you may even have worried that you were colorblind when you saw a different color dress than your peers. As it turns out, however, the dress is simply a type of optical illusion—a difference in how our brains process visual information, not a problem with anyone’s eyes. Colorblindness, on the other hand, is caused by differences in our eyes, not our brains. Colorblind people are unable to distinguish certain shades of color, depending on the type of colorblindness. Thus, although they might be able to tell you what color the dress is, other colors may be impossible for them to distinguish. Fortunately, new technology is allowing people with a common type of colorblindness to see the world in full color for the first time.
What’s so great about color vision?
Vision is the most important sense for humans, providing 90% of the information we process. Among our other senses, hearing provides just 5% and our other senses provide even less (2). Color vision is particularly important in detecting the borders of objects (Figure. 1) (3), picking up on non-verbal social signals, and noticing details about the environment (2). Without color vision, humans lose a significant amount of critical information.
Figure 1 ~ Color vision allows humans to more easily detect borders of objects (A). People with protanomaly (B) and deuteranomaly (C) — two types of color blindness—may have difficulty noticing details in their surroundings.
How does color vision work?
Knowing how light behaves helps us understand the concept of color and how we see it. Light is often described as a wave. Similar to the way sound waves of different wavelengths give higher- or lower-pitched sounds, different wavelengths of light are perceived as different colors (4). For instance, the color red has a relatively long wavelength – about 700 nanometers (nm) – whereas purple has a shorter wavelength – roughly 400 nm (4). White light – the light that comes from a light bulb or the sun – is composed of light of all wavelengths (5). When white light hits an object, waves of some wavelengths are absorbed, and others are reflected back, depending on the physical properties of that object. The waves that are reflected back are the ones that reach our eyes.
When light waves reach our eyes, they are detected by special light-sensitive neurons called rods and cones (6). Rods are important for seeing in dim light, and cones are responsible for color vision (4). There are three types of cone cells that sense light waves from different parts of the color spectrum (6). Red cones detect light with wavelengths from 500-700 nanometers (nm) long, green cones 450-630 nm, and blue cones 400-500 nm (7). While the sensitivity of each cone type is maximized at different wavelengths, there is some overlap in which wavelengths the cones detect. When light hits the leaf in the image below (Figure 2), most wavelengths are absorbed except for those around 520 nm, which are reflected off the image. When a person with normal vision looks at the picture, the 520 nm wavelength light stimulates the green cones in our eyes to about 90% activity and the red cones to about 55%. Neurons connected to the cones send signals to our brain, which uses this information to tell us we are looking at the color green (Figure 2) (7). In this way, the brain interprets the signal intensities from the three types of cone cells and decodes the colors we are seeing.
Figure 2 ~ When light from the sun hits the leaf in this image, most wavelengths of light are absorbed. Only wavelengths around 520 nm long are reflected off the image. When the light reaches the eye, it activates green and red (but not blue) cone cells to different degrees. The brain translates this information and tells us we are seeing the color green.
What is color blindness?
So what happens in people who are colorblind? Actually, the term “colorblind” is a bit misleading. Most colorblind people see many colors, so a more accurate term is Color Vision Deficiency (CVD). This is because people with CVD usually have two functioning types of cone cells, but the third is missing or mutated. They have difficulty distinguishing between certain colors but not others. For instance, if a man doesn’t have any red cones, his brain won’t ever get the proper combination of signals it needs to tell him he is seeing red, orange, or purple. His brain receives the signals from only the green and blue cones. To him, colors containing red hues look more like various shades of yellow, brown, green, blue, and gray (8). Meanwhile, his perception of other colors may be unaffected.
In the most common type of CVD, a person has all three cones but there is a mutation in the red cone (protanomaly) or the green cone (deuteranomaly). These two types of CVD, which affect about 5-8% of men and 0.5% of women (8), are often collectively referred to as red-green colorblindness. [Men are much more likely to have red-green CVD because the genes leading to this type of colorblindness are located on the X-chromosome. Men only have one X-chromosome, while women have two. Thus, even if a woman has one defective copy of the gene, the X-chromosome inherited from the other parent will usually provide a normal copy of the gene (10).]
In protanomaly, red cones respond to light closer to the green part of the spectrum (450-630 nm) than to red (500-700nm). Deuteranomaly is caused by a mutation in green cones, causing them to detect light closer to the red region of the spectrum (Figure 3) (8). When the cones of a person with protanomaly pick up light waves at 520nm, the green cones are activated the same as someone with normal vision, but the red cones are activated to about 75% instead of 55%. Thus, although the individual with CVD is looking at something green, the brain tells the person that the object is closer to a brown or amber color (Figure 3B). Conversely, in deuteranomaly, a mutation in the green cones causes them to respond to light waves closer to the red range. Instead of 90% activation by 520nm wavelengths, the mutated green cones are activated only to around 60%. This means that the observer will see a color closer to yellow or olive.
Figure 3 ~ Cone cells in the eyes of people with normal vision detect a range of wavelengths. When wavelengths of 520 nm reach the eye of someone with normal vision, red cone cells are excited to about 55% and green cones are around 90% activated. In protanomaly and deuteranomaly, the spectrum of the red or green cone is moved slightly closer to the other, which changes the level of activation of the mutated cone.
How can colorblindness be corrected?
There is no cure for CVD. In the past, researchers and engineers have attempted to improve the vision of the colorblind by using tinted lenses that allow borders to be more easily detected (2). These lenses don’t allow the user to see more colors; they simply increase the user’s ability to distinguish differences in colors. However, a company called EnChroma has recently developed sunglasses that allow those with CVD to actually see more colors. The glasses were originally created as laser safety glasses for surgeons, but when a colorblind friend of the creator borrowed them, he found he was able to see many more colors (11). From this discovery, EnChroma was created and has since improved the technology so that people with more serious degrees of CVD can also benefit (8). The lenses are made of up to 100 very thin layers of a special material. These layers allow some light waves to pass through, while blocking other light waves. To correct for color blindness, the lenses filter out wavelengths that overlap between red and green cones in people with CVD so that the appropriate cones are activated when light waves pass through the lens (Figure 4) (9). The glasses block a lot of light, so they’re only meant for outdoor use. Nevertheless, colorblind users notice a dramatic increase in the number of colors they can see (11).
Figure 4 ~ The many layers in the lenses of EnChroma sunglasses filter out specific wavelengths of light such that the proper combination of cones is activated. This allows people with CVD to see shades of color they have never seen before.
So what about the dress? We can’t agree on the color of the dress because the light waves that reach our eyes are not always exactly what our brains tell us we are seeing. In fact, our brains do a little photo editing for us to give us the most accurate picture possible. The trouble is, brains sometimes differ in how much they edit an image (1). In other words, when you look at an object, your eye tells your brain its color and also other details like what type of light is illuminating the object (e.g. fluorescent lights vs. morning sunlight). In order to determine the true color of the object, your brain makes a split-second adjustment to the image based on the quality of light (1). This phenomenon is demonstrated in the optical illusion below. In the image, squares A and B are the exact same shade of gray. However, your brain sees that square B is in a shadow, so it makes an adjustment because it assumes the square must be lighter than it actually appears. Thus, it tells us that square B is lighter than square A (Figure 5). The photo of the dress confuses us similarly because one person’s brain assumes that the dress is lit by regular white light, so his brain tells him he’s looking at a blue dress. The brain of another person assumes that the dress is being illuminated by a bluish light or is in a blue shadow. In this case, his brain corrects for the lighting by editing out this bluish light so that the dress appears white to him (1). This is not a case of colorblindness but a situation in which the brain interprets information from the eyes differently from person to person. The reality? That dress is definitely blue and black (1).
Figure 5 ~ Squares A and B are the same shade of gray. However, your brain sees that square B is in a shadow. Thus, your brain edits the image and tells you that square B is lighter than it appears to be. (©1995, Edward H. Adelson. Used with permission [12].) This is similar to what happens for some people with the blue dress.
Hannah Foster is a PhD candidate in the Molecules, Cells, and Organisms program at Harvard University.
References
1) Rogers, A. (2015) The Science of Why No One Agrees on the Color of this Dress. <http://www.wired.com/2015/02/science-one-agrees-color-dress/> [retrieved March 4, 2015]
2) Ozorlite. Color vision deficiency. <http://colourvision.info/> [retrieved February 12, 2015]
3) Gouras, P. Color Vision. <http://webvision.med.utah.edu/book/part-vii-color-vision/color-vision/> [retrieved February 12, 2015]
4) Pappas, S. (2010) How Do We See Color? <http://www.livescience.com/32559-why-do-we-see-in-color.html> [retrieved February 12, 2015]
5) The Physics Classroom. Wavelike Behaviors of Light. <http://www.physicsclassroom.com/class/light/Lesson-1/Wavelike-Behaviors-of-Light> [retrieved February 12, 2015]
6) Web MD. Color Blindness—Topic Overview. <http://www.webmd.com/eye-health/tc/color-blindness-topic-overview> [retrieved February 12, 2015]
7) Cyberphysics. (2013) The EYE: Photosensitive cells. <http://www.cyberphysics.co.uk/topics/medical/Eye/photosensitiveCells.html> [retrieved February 12, 2015]
8) Color Blind Awareness. Types of Colorblindness. <http://www.colourblindawareness.org/colour-blindness/types-of-colour-blindness/> [retrieved February 12, 2015]
9) EnChroma. How it works. <http://enchroma.com/technology/how-it-works/#technology> [retrieved February 12, 2015]
10) Brownsword, J. (2010) Genetics Explain Why More Men are Colorblind than Women. <http://www.prurgent.com/2010-09-30/pressrelease122256.htm> [retrieved March 1, 2015]
11) Yam, K. (2015) New Glasses Transform the Way Colorblind People See the World. <http://www.huffingtonpost.com/2015/01/13/color-blindness-correcting-glasses_n_6446094.html> [retrieved March 4, 2015]
12) Adelson, E. (1995). <http://persci.mit.edu/gallery/checkershadow> [retrieved March 14, 2015]
To find out more about EnChroma and how the glasses were invented, take a look at this news report on ABC:
http://news.yahoo.com/video/sunglasses-fix-color-blindness-053406308.html
For a better understanding of what it’s like to be colorblind, check out this video:
http://www.dailymotion.com/video/x2cq5hi_what-it-s-like-to-be-color-blind_lifestyle
I do a project at my school on the Enchroma glasses and I was wondering If you knew the material that the lenses are made.
I searched it, but didn’t found it. It would be very helpful.
And by the way, I really like your article.
Thank you
*please respond
Some back and fourth from Facebook:
Jonathan Toomim: This article’s numbers and figures are off. Normal pigments for medium and long wavelength (“red” and “green” cones) are much closer together than this article makes it seem. While S2N says 90%, 55% for 520 nm for a normal trichromat, it’s actually more like 80%, 65%.
Also, calling them “red” and “green” cones is a common but deceiving labeling system. They’re best described as long (L) and medium (M) wavelength cones. The L cone is maximally sensitive to about 568 nm in normal humans, which is either a yellowish green or a greenish yellow. The M cone is maximally sensitive to about 530 nm, which is pretty much just green. Red comes from the difference between the M and L cone activations strongly favoring the L cone, but reds only activate the L cone weakly.
The page below is much more accurate, though a bit drier:
http://psych.fullerton.edu/eriko/research/ColorVision.html
As for the glasses (which are cool!), I think this page is a bit more accurate and specific about how they work:
http://enchroma.com/technology/how-it-works/
Hi Jonathan,
Thank you so much for your response. I appreciate you taking the time to point out the inaccuracies of the article.
You are absolutely correct that the exact numbers are not entirely accurate. I was attempting to illustrate the concept of colorblindness with the figures–not convey precise numbers. It is easier to show the spectral shift when the peaks are further apart, and it would be confusing if the numbers I wrote in the article and the figures did not match. However, I should have noted that the numbers and figures are for illustrative purposes and are not precise, and I should have provided a reference for anyone interested in knowing the exact absorbances of each cone.
As for “green” and “red” cones, you are also correct that this is, perhaps, deceptive terminology. I chose to stick with these terms because they are most familiar to the general public. Changing the terms did not seem important enough for the description of the concept to warrant the extra sentences. However, perhaps the article would have been more clear had I chosen to clarify the commonly-used, yet somewhat deceptive labeling system.
I also appreciate your references. For anyone interested in knowing precise numbers or looking for more information, these would be great sites to visit.
Thanks again!
Hannah