Can my eyes be fooled that easily? I’ve seen optical illusions before, but not color optical illusions.
I’m a bit unnerved, I must admit. How do I know that the red I’m seeing is really red?
Yup, it’s really that easy. Our brains interpret color based on context, just like they do perspective, size, and angle. Fascinating, no?
It’s a bit frightening. You spend time in kindergarten identifying colors, you learn the Spanish words in junior high, then you come to find out what you think is yellow really isn’t yellow…
That’s interesting and a bit confounding. I understand the first two but the third one is a little disconcerting. I thought that we would always see a color as that color with variation in gradation. Wrong answer, Ruby.
One of the major factors is that our brains constantly “remove” shading to determine the “actual” colour of a lighted object.
For example, when we look at a parked car, we get a general sense that the vehicle’s paint is all of one colour, even though the specular highlights and shading mean that the light that meets our retinas from it covers an extreme range of shades.
Without thinking about it, we subtract the shading and highlighting to arrive at the object’s “true” colour.
A considerable part of the third illusion’s effect comes from the reorientation of the cross. The major factor, of course, is that we also compensate for the different colours of “ambient light” to determine the “true” colour, but the striking apparent difference in lightness/darkness is partly due to the “grey” cross being placed so that we mainly compensate for the highlight, while the “yellow” cross is placed so that we mainly compensate for the shadow.
Hokey Smokes! The second one did nothing for me, I figured they were all the same color, but 1 and 3 are pretty amazing!
That broke my brain.
:: has same reaction as Unintentionally Blank ::
Damn, I thought illusion 3 was a hoax.
I just took a picture of it with my Digital camera.
I copied it to my desktop, opened it in Paint.
I cropped the center pieces and brought to upper left corner and delete all else.
Cut out center of each center piece and dragged them to a white area. They do match.
Under 8x zoom they are grayish with some yellow gray and some blue gray but generally the same mix.
I couldn’t believe it. They really are basically the same color :smack:
Especially if you’re color blind…
Really surprising.
#3 didn’t work on me. Thank you, color blindness.
#1 was pretty good.
#2 was obvious, and I’ve seen it a million times. (They look like the same color to me.)
#3 blew my mind. 
AFAIK, I’m not color blind, yet #3 didn’t work for me very well at all. I could sort of see that they were the same color, but it was very nearly the same level of stretching as Tom Cruise saying he’s studied the history of psychiatry.
The first two worked. I shoot a lot of video and I wasn’t surprised at all by the first or the second.
But the last? I saw two totally different things. I saw 1) yellow WITH purple and 2) purple WITH yellow. They were definitely not the same. Maybe I’m colorblind.
Here’s a little help. This piccy is the two crosspieces isolated and reoriented. (You can still see the bleeding around the edges.)
The “base” colour mix on both is 128,128,128, with the darkest shades going to 79,79,79. It is objectively grey.
The overall picture provides enough information to simulate a “real” scene viewed with two differently-coloured lights, or through two differently-coloured translucencies. We automatically remove the “distortions” to intuit the “real” colour of the object on the fly, much like we automatically adjust for perspective distortions to “see” the real dimensions of an object. This gives us colour assumptions that can be messed with in the same way that forced-perspective illusions confound our assumptions about scale.
The “Yellow-tinted” scene screens out about half the blue, so that what we “know” is “white,” (which should be 255,255,255, using the conventional 256-shade levels that computers give us) is represented as 255,255,144 – light yellow. We automatically adjust and “know” that the scene really contains twice as much blue. So on that side of the image, we adjust the perfect grey of 128,128,128 and “see” something like 128,128,255 – a darkly-hued blue.
The “Blue-tinted” scene gives us full blue and screens out around half of the red and green, so that what we “know” is “white” is represented as 144,144,255 – light blue. Compensating, we automatically increase our assumption about the red and green levels, so that the perfect grey of 128,128,128 is adjusted in the head to something like 240,240,128 – a darkish yellow.
Yeah, what he said.
Uh, as an afterthought I guess it might be helpful for some folks who aren’t used to working with RGB values that in my post above the colour values listed are Red,Green,Blue. (ie; 0,0,255 is pure blue, etc.) Sorry.
The first two I get, but the last one is not that remarkable as some people have indicated. I see two gray cross pieces although the one on the left is a darker gray than the one on the right. Where’s this yellow cross people are seeing? Is it supposed to be the one on the right or left? Maybe I’ve just discovered some color vision difficiency on my part, or maybe my brain just refuses to be confused.
When I was a freshman in college, I started thinking about how we were taught in elementary school that colors corresponded to specific frequencies of visible light. But of course, that is very incomplete and highly misleading. (At least, it’s not just plain wrong, like teaching that the Coriolis effect determines the spin of water in toilet bowls.)
Color isn’t quite a physical property of light or a physical property of objects. It is a complex phenomenon that depends on the source of light, the properties of the object that determine what frequencies are absorbed and reflected, how receptors in the eyes respond to the light they receive, and how the brain interprets the signals from the eyes.