Is my brain playing tricks on me?

February 4, 2010 9:33 PM

How does this floating man optical illusion work?
posted by ieatwords to Grab Bag (13 answers total)

Looks like it works through Photoshop to me.
posted by Meatbomb at 9:34 PM on February 4, 2010

Man stands on road near oil-stain?
posted by deCadmus at 9:34 PM on February 4, 2010

The wet spot looks like his shadow?
posted by wongcorgi at 9:35 PM on February 4, 2010

The man is simply standing next to a patch of ground that has been darkened, perhaps with water or spraypaint. It's actually a cloudy day (notice the lack of shadows from other objects in the scene), so without direct sunlight the man doesn't cast a real shadow. The darkened ground is where is shadow would be if it were a sunny day and the man were hovering.

It may help to cover the 'shadow' with your hand and look at the man's feet. He's quite plainly standing on the pavement.
posted by jedicus at 9:36 PM on February 4, 2010

Thanks jedicus. Amazing how the entire illusion vanishes simply by covering the dark spot.
posted by DieHipsterDie at 9:40 PM on February 4, 2010

Seconding everyone, also the supposed shadow doesn't really make sense for him hovering above the ground. If the sun was bright enough to make that dark of a shadow the shape would be a lot more irregular and wider along the shoulder to shoulder access.
posted by nanojath at 9:41 PM on February 4, 2010

Axis, that is.
posted by nanojath at 9:42 PM on February 4, 2010

So it turns out that shadows are a strong cue to the human visual system on some level. This picture is just exploiting that fact. Jedicus' description is what your visual system is "deducing" from the shadow. However your visual system doesn't sit there and think about stuff, really.

I've never seen a better demonstration of this kind of illusion than Daniel Kersten's "ball in a box" demo:

http://vision.psych.umn.edu/users/kersten/kersten-lab/images/ball-in-a-box.mov

I strongly recommend checking out his pages on the topic here:

http://vision.psych.umn.edu/users/kersten/kersten-lab/shadows.html -- the 57mb QT file on that page has a lot of related effects. There are links to scientific articles and other investigations into related effects. I've had beers with Kersten and he's a very smart, very good vision scientist.
posted by fake at 9:43 PM on February 4, 2010

I think the framing of the picture helps with the illusion too. Notice how the distance from the top of the guy's head to the top edge of the picture is roughly the same as the distance from the bottom of the stain/supposed shadow to the bottom edge of the picture, putting the guy slightly above center in the picture, which reinforces the perception that he's hovering.

I'd never seen that one before. Neat!
posted by Balonious Assault at 9:44 PM on February 4, 2010

Just to give a weird, outlying data point, I don't think I even see the illusion. It took me about thirty seconds of puzzling this out before I even noticed the grease spot. I see how that could be an effective illusion, but I think I over-thunk it already.
posted by The Potate at 9:52 PM on February 4, 2010

I didn't get the illusion effect either.

As above, "Man stands on road near oil-stain?".

Interesting.
posted by knapah at 2:45 AM on February 5, 2010

If you don't see the illusion, try standing back from your monitor a ways. You should see it when the perspective/distance cues from the ground disappear.
posted by fake at 6:23 AM on February 5, 2010

Agreed- the only reference your eyes have as to where the guy is is the grease stain/shadow. He's too far from the background to judge his position, and the asphalt is uniform enough that the positional cue from the shadow overrides the desire to map him right onto the pavement.

(If there is any photoshoppery, I would guess that they might have taken two photos, one with the guy, and one without, and wiped any real shadows out. If you look at the folks in the back, they do have small shadows under them.)
posted by gjc at 7:30 AM on February 5, 2010