Look at the following image.
Look in particular at squares A and B. A appears to be dark grey, B appears to be white or whitish. But in fact, they are the same exact color.
Don' t believe it? Me neither! Or at least, I didn't until I went ahead and deleted most everything that is not A or B from this picture. When you do that, you get this:
This is from the Scientific American web site, where there are a number of illusions in a new piece on the subject of Neuroscience of Illusion. Here.
That's at least fifteen kinds of awesome.
As awesome as that is, I think it would be better if you put a title on the post.
This post has excess win
I didn't believe it till I tried altering it. It's neat to watch the illusion disappear. I think that although the checkerboard causes you to expect B to be light, and although the green cylinder gives a "reason" for that area to be in shadow, the only necessary part for the illusion is the area immediately adjacent to the squares. That is, when I edit the image in paint, the minimum disruption to make the color illusion "disappear" appears to be taking the 'eraser' and whiting out a border around both A and B. The rest of the picture can still be there, but you see the colors as equal.
(yeah, I should so be at the bench right now...)
Whoa. I just tried making borders in colors instead of white- the lighter the color, the closer it comes to making the illusion disappear. But black, for example, doesn't work.
Can anyone explain this?
(ok, I *really* need to stop playing on the internet...)
There is a title on the post. It is just an optical illusion so you can't see it!!!!!
Try this-- I masked the screen with overlapping crossed fingers to show just the two areas and they obviously were the same gray level.
Then I slowly spread my fingers and pulled them away from the screen, letting myself see more and more of the surrounding image. (Same thing you'd do with your hands while making complicated shadow pictures on a wall, basically)
The equal gray perception held on a _long_ while as more and more of the surrounding area was visible. Much longer than I expected it to do.
Similarly starting with seeing most of it and slowly closing down the 'iris' around each cell made them continue to appear different, a _long_ while. Definitely not a simple transition, not the same transition each way.
I strongly noticed that this transition each way felt (are we doing introspection here?) very familiar -- it feels like the transition looking at a black and white cratered moon or planet, where the depressions can invert into bumps as the perception of where the light's coming from switches.
Lovely. I hope someone extends this along the lines I noticed, if anyone else confirms that experience.
What's extremely interesting to me is that the "B" seems to contrast more than the "A" relative to their local background!
I didn't believe it either and I couldn't figure out how to hold my fingers to see just the A and B squares. So I ripped two little holes in a piece of paper so that only those squares could be seen. As I pulled the paper closer to my eyes, more of the picture was revealed and the apparent contrast changed. Brains are weird.
minusRusty, I cropped out the B and moved it over next to the A. The B is definitely darker, which adds greatly to the illusion.
Becca, if you want some fun (when you don't have to be at the bench, of course), paint in lines that are the color of those two squares. And my husband figured out why the black doesn't work. It's already as dark as it'll get, so you don't expect it to get any darker in the shadows.
Note to Photoshop Users:
Ever tried to find neutral gray by eye, to set the midpoint in the color picker on the curves screen? This illusion demonstrates quite succinctly why you can't.
Here's a trick to finding the all important neutral gray in your image.
Select the image layer in which you need to find a neutral gray. Create a layer directly above it, filled with neutral gray (50% each red, green, and blue). Set the blending mode for this layer to difference. Now create an threshold adjustment layer immediately above that. Click on the icon to bring up the slider which sets the threshold, and move the slider all the way to the left. The image will go white. Start moving the slider very slowly to the right (you can use the up arrow key for greater precision). The first black dots to appear will mark neutral gray in your image. To mark this precisely, zoom in down to the pixels, use the second eyedropper tool to put a numbered marker on the gray pixel(s).
There's your neutral gray.
Hide the fill layer and the adjustment layer, select your image layer, zoom in to your marker, call up the curves dialog, click the middle eyedropper, then click the marked pixels. You'll know you have succeeded if the foreground color changes to gray. Ta-daaaa.
To find the black points and white points is easier by eye, but you can use this same trick. Just delete or hide the gray fill layer, and use the slider in the threshold layer. Slide it all the way left, then slowly back right, and the first black pixels are the darkest pixels in your image. Slide it all the way right, the slowly back left, and the first white pixels mark the lightest parts of your image. You can mark these with that second eyedropper, then use the curves dialog to set your black and white points. (Hint: when selecting the white points, do not choose secular highlights--bright reflections such as glare. Choose the lightest white which is NOT such a highlight.)
Not a photoshop user? Sorry to bother you.
If you use the eyedropper tool from The Gimp (I'm not sure what this photoshop thing is you speak of) you can test the color value numerically. They are essentially the same.
You guys have wa-ay too much time on your hands!
There is a really cool interactive version of this at Shapiro Lab. Click on the top-left icon under "lightness illusions" and go crazy!
I'm finding that if I cover up either the half-shadowed light square that's below A or the half-shadowed dark square above B, I can easily see that A and B are the same color. Without either of those squares covered, A looks darker. I'm not sure how much that's influenced by already knowing the truth, though---after each "covering-up" experiment, I have to look away from the picture for a while before I can see A and B as different again.
Visual illusions are a wonderful demonstration of our ability to model the world we live in rather than see it as it is. Anything you care to do to cut away the cues that tell you this is a checkerboard with black and white squares will help destroy the illusion.
I took a screen capture of the picture and pasted it into PaintShop Pro. The color dropper confirms it. Both squares have RGB values of 121,121,121. Wow. I never would have believed it otherwise.