Here's looking at you/me, kid

August 28, 2008 2:31 PM

More good stuff for people who like visual ("optical") illusions (previously): A nice Scientific American article, a particularly creepy illusion, and a link to the "Best visual illusion of the year" contest. Given that the eye/mind/brain is so easy to trick, a person might wonder what's really out there in the world.
posted by cogneuro (26 comments total)

"Creepy" is really great!
Thanks.
posted by Mister_A at 2:38 PM on August 28

I mentioned this in a previous thread: the winner for the best illusion totally blew me away. Afterimages are not supposed to work like that! It's not phenomenally trippy like a lot of other illusions, but it's very, very theoretically interesting.
posted by painquale at 2:47 PM on August 28

I bet there is actually no spoon.
posted by Dipsomaniac at 3:01 PM on August 28

It's not theoretically all that astounding but I enjoy the Castle Optical Afterimage Illusion
posted by bonobothegreat at 3:03 PM on August 28 [2 favorites]

Note that for the winner of the Best Visual Illusion, I found fixing on the obvious black dot - not the center of one of the images as the instructions state - much more effective. Thanks, painquale, I wouldn't have gone back to it if you hadn't said it was good! Great post cogneuro.
posted by alasdair at 3:03 PM on August 28

As for "creepy"...

*cough*

(self link)
posted by SmileyChewtrain at 3:07 PM on August 28

And thanks for the links - These illusions are great! I also like this blog, Illusion Sciences, which does some great interactive explanations of mind boggling illusions.
posted by SmileyChewtrain at 3:11 PM on August 28

That illusion of the doubled eyes and mouths *hurts* to look at.

Amazing.
posted by 5MeoCMP at 3:35 PM on August 28

Yeah, the doubled eyes and mouths started to make me feel nauseous (out of disorientation, not disgust) when I looked at them for too long.
posted by invitapriore at 4:03 PM on August 28

Ditto the nauseous feeling for the double eye one. While the others are interesting, that one is the most (literally) mind-bending.
posted by Solon and Thanks at 4:04 PM on August 28

appologies - i have nothing constructive to say other than that i love these illusions. any more?
posted by healey at 4:04 PM on August 28

As soon as I see optical illusion and creepy, I assume I'm going to see upside-down-eyes-and-mouth Lincoln.

(note: may look better in your intro psych textbook)
posted by Durn Bronzefist at 4:05 PM on August 28 [1 favorite]

I'm not actually seeing the "creepy" illusion (page 2). It always looks like the left one is looking over my right shoulder and the right one over my left shoulder. Also, I don't get the drawings of the girls with folded arms (page 6). They both look like they're looking to my right. Am I missing something.

(The headache-inducing double eyes and double mouths is surprisingly effective, however. It's like the brain thinks I'm going cross-eyed and is trying to compensate.)
posted by ErWenn at 4:43 PM on August 28

After posting that comment and then looking at the contest, the thumbnail version of the "creepy" illusion successfully gave the impression of them looking at each other. I think I just wasn't far enough away from the picture for it to work.
posted by ErWenn at 4:46 PM on August 28

ErWenn, I am having some serious-ass eye problems right now and got nada from creepy and the cross-arm girls. It's not just you! The double-feature people made my head explode a la Scanners.

Will have to revisit when the orbs are back in better shape.

However, I love me some eye-tracking maps!
posted by Lesser Shrew at 4:54 PM on August 28

I'm getting nothing from the doubled eyes/mouths one. To me, they just look like people with four eyes and two mouths each - no headache, no nausea.
posted by wanderingmind at 5:00 PM on August 28

No nausea or disorientation from the double-eyes/mouth pictures here, either. Maybe because I work in photography, and look at pictures of people frequently?
posted by seanmpuckett at 5:05 PM on August 28

If you have strabismus (as I do), illusions that require the brain to overlap images don't work; I've never been able to "get" 3-D movies as a result. Last time I tried at an amusement park, everyone jumped back when things flew out at them, except me.
posted by woodway at 5:22 PM on August 28

I've always liked this illusion, which is similar to the "creepy" one.
posted by Stylus Happenstance at 6:10 PM on August 28

I've always liked this illusion, which is similar to the "creepy" one.

Not just similar, but identical in every way that matters. And there is a nice Marilyn/Einstein version of it, too. I wonder why 'creepy' won a prize.
posted by jfrancis at 7:02 PM on August 28

jfrancis: I wonder why 'creepy' won a prize.

It didn't win for presenting a new illusion; it won because it used the illusion in a novel way. The pdf on the award site makes this clear:

The illusion sheds new light on the role of luminance cues in computing eye direction. Previous findings have led to the proposal that gaze direction judgments follow a simple heuristic that assigns the iris to the dark part of the eye. In the present figure, each eye is physically darker on the right side than on the left. Despite this, the eyes appear to be looking to the left when viewed at close range. This observation demonstrates that gaze estimation is not always dominated by gross luminance distribution across the eye.

It's fascinating to me how often a finding in psychology can spawn a new experimental method, or a whole new experimental paradigm, that can be used to explore seemingly unrelated cognitive features.
posted by painquale at 8:32 PM on August 28

The Illusion sciences blog is worth a look (or a subscription).
posted by TheophileEscargot at 11:27 PM on August 28

I'm getting nothing from the doubled eyes/mouths one. To me, they just look like people with four eyes and two mouths each - no headache, no nausea.

Yeah, I'm not getting any headache or nausea, but I can kind of see what 5MeoCMP and invitapriore are talking about. There is something about that image that causes my eyes to keep flicking between the two sets of features. When I focus on one, my peripheral vision sees the other and assumes that I'm looking at the wrong thing and tries to compensate.

It's kind of interesting in a I-don't-have-complete-control-of-my-body kind of way.
posted by quin at 10:17 AM on August 29

I've always wondered how they discover these illusions. ("OK, I'll stare at the flashing cow for 60 seconds and then you stab me in the ear with a pencil and we'll see what happens.") Are there instances where they observe some physical property and then realize it would manifest itself in a particular set of circumstances or is it mostly just "Whoa! -- what was that?"
posted by joaquim at 12:19 PM on August 29

joaquim- I'm working in a perception lab that uses visual illusions, so I might be able to shed some light on this. The short answer is that the stimuli evolve; the selection pressure on that evolutionary process is these illusions must have (at one time) been able to answer theoretical questions about how brains process visual input (or just to blow people's friggin' minds- that works too).

We already know some things about how the brain processes visual stimuli. Many cats and monkeys died to give us this information. But there's a lot we don't know. There are many possible ways our brain could go about its business. So vision researchers not interested in sticking probes into monkey brains sometimes cook up visual stimuli and measure how people respond to them in an attempt to distinguish how likely different possibilities are. The stimuli are created specifically to create responses in people that will help us say "by virtue of the measured response, that theory must be wrong; maybe this alternative theory is right".

So, for example, in the lab that I'm working in, we're interested in figuring out how the brain integrates continuous/ongoing stimuli with discrete/instantaneous stimuli. There are some stimuli already known to help compare the perceptual experience of continuous and discrete stimuli (e.g. the flash-lag illusion); we tweak those stimuli in various ways to see if our responses to the tweaked versions contradict any of the hypotheses out there about how continuous and discrete stimuli work. If we do find such a useful illusion, we publish a paper about it, saying "oh, these streams of visual processing can't work in that way; it's much more likely that they work in this way, because that's consistent with how we perceive this illusion".

Our lab just found a fucking killer illusion. So- we already know that we can generate paradoxes in how people interpret a scene; for example, we can present a moving object on a laptop screen that appears to be further to the right than a reference object that is itself on the rightmost edge of the laptop screen. These paradoxes of comparison aren't too new, and the farther apart the things compared are, the easier they are to generate. There are also well known illusions that rely on switching interpretations of the same patch of space, but people switch from one interpretation to another over time- they don't have both interpretations simultaneously. But through a series of incremental changes to a flash-lag style stimulus, we have produced an illusion that generates contradictory perceptions in the same space and time, right on top of each other. When we demoed it at the lab meeting, everyone's jaw hit the floor. That's all I can say; we haven't published yet.
posted by Jpfed at 1:09 PM on August 30 [2 favorites]

Jpfed, those illusions sound amazing! I really want to know more about the simultaneous interpretation illusion - that could plausibly blow some theories of illusion interpretation switching out of the water (I'm thinking of Pylyshyn's account in particular). I hope that when these illusions are published or made available online, you'll post a comment here to let us know! Or, if you can, drop me a MeMail when they're available. I'd love to see them.

(Actually, you make it sound like the rightmost-edge-of-the-screen paradox of comparison might already be publicly available. I'm not familiar with it. Can I get a site? Or a cite?)
posted by painquale at 2:21 PM on August 30