Perception VS Reality. Perception Wins.
Thursday, June 25, 2009 at 11:11AM 
This one blew me away. See the image above? Unless you suffer from *color blindness, you should see a pink spirally box with a green spiral and a blue spiral vortex. Except they're not blue and green. The large colored spirals are the same green. Really.
Here is the original posting of this illusion on BuzzHunt. And some followup analysis on Discover.
*If you have some element of color blindness, this illusion may not work, according to Vischeck. This image to the left shows what some will see.
But! Even if you suffer color blindness (statistically more likely to be caucasian male), you will be able to appreciate the image to the right. In this amazing visual illusion by Edward H. Adelson, square A and square B are the exact same color. What's even more brilliant in this one is that our visual system is making cognitively complex adjustments to perceived color based on the concept of a shadow. So much of what we think is the raw data from the real world is having massive adjustments done to it before it's passed on to our consciousness. These 'illusions' remind me that interaction design is building around perceptions, not building around reality.
Optical illusion, Perception, Visual in cognitive science
Reader Comments (11)
OMFG! The checkerboard is truly amazing. If I stare at the B square for a few seconds, I can actually see the colour change in front of my eyes! It goes from white to grey as my eyes adjust! AMAZING.
Wow. I've seen the checkerboard one before, but this spiral one is amazing.
Glad you like them. I'd never seen the spiral one before, but they're both amazing aren't they! :-)
Wow! I can still hardly believe the blue and green are the same, I had to get the screen colour picker out to check.
On a similar topic, Mind Hacks posted about "Time illusions" yesterday. Think about this: if someone touches your nose and toe at the same time, then you feel them at the same time. But why? The signal from you toe has to travel much further, so you should feel it after your feel the nose touch! And it's all because your brain is making these adjustments, and you're actually living slightly in the past.
And there's an even cooler experiment...
http://www.mindhacks.com/blog/2009/06/pressed_for_time_per.html
The spiral illusion prompts me to think that such illusions might prove superior to the awful Ishihara test for detecting color deficiencies. According to Ishihara, I'm a deuteranope. I fail Ishihara every time. Yet I perceive the spiral illusion. Maybe instead of testing for correct color perception in contrived circumstances as Ishihara does, we should test for illusion; if I perceive the same illusions as you, then my brain is making the same adjustments as yours, even though our eye pigments might differ.
@Ross That's a really brilliant idea. The color blindness simulation that I posted was for a Deuteranope — so this points out one weakness of that simulation. I've noted in a http://www.uiandus.com/2009/01/28/uncategorized/insight-color-perception/" rel="nofollow">previous post on color perception that in fact there are two types of blue/green color receptors, and that people have one or the other. This is why people often can't agree on what color turqoise is. Perhaps this plays a role?
@Stuart Yes, I too had to pull out the color picker. The time illusion example you quote is a great one, I've never heard of that study before. Thanks!
I'd like to see one of those places. Trees normally grow up, and I guess if they all grew sideways it would be very disorienting.
http://typophile.com/node/60577" rel="nofollow">Here is an example of using carefully paired purple and yellow to make clearer "red on white" text in a tiny YouTube icon. It's not purely the same effect as the illusions, since sub-pixel rendering it relies on how pixels are laid out on an LCD display -- but it's neat and somewhat related.
That is cool. Sub-pixel rendering does some interesting stuff, using colors (even for black and white text) to get the optimal anti-aliasing.
[...] processing every other pixel differently had the effect of adding alternating light/dark stripes! We see contrast, not absolute color, so the numerically insignificant error was quite visible. Worse still, bands of 1 pixel stripes [...]
http://clientsfromhell.tumblr.com/post/316424686/client-we-like-the-design-but-could-you-make-the
(this didn't happen to me, I only found it.)