3-D impossible structure : Omni Brain

The Homunculus

The Omnibrain is a psychology graduate student at an online university. He hopes that the three weeks and $29.95 that he is spending on his Ph.D. will get him a job at a Tier 1 research university. Do online universities have postdocs? Ok...just kidding, The Omnibrain is a real graduate student at a real school somewhere in the continental United States - or maybe Europe.

Search

Glial Cells

Access Omni Brain mobile here.

Access Omni Brain email here.

Axons

View blog authority

« Ann Coulter hates professors | Main | Online Visual Cognition Experiments »

3-D impossible structure

Category: Psychology
Posted on: March 5, 2007 9:10 AM, by The Omnibrain

via growabrain

Share this: Facebook Twitter Stumbleupon Reddit Email + More

TrackBacks

TrackBack URL for this entry: http://scienceblogs.com/mt/pings/34807

Comments

Excellent.

Posted by: Robert P. | March 5, 2007 9:47 AM

Presumably loooks VERY different from other angles... The illusion is caused by our extrapolating to other viewpoints...

Posted by: Jonathan Vos Post | March 5, 2007 11:12 AM

yup you're right jonathan. The beauty of accidental viewpoints.

Posted by: steve | March 5, 2007 11:21 AM

Heh, yep. I think I know exactly what it looks like from "non-ideal" viewpoints. The shadow tells everything ya need to know.

Posted by: G Barnett | March 5, 2007 11:43 AM

MC Escher made models of some of his impossible structures that looked real when viewed from the correct angle. He was exploring forced perspective and a variety of optical tricks at the same the great movie special effects pioneers were doing the same thing.

Posted by: John McKay | March 5, 2007 2:09 PM

The other thing that gives it away, apart from the telltail shadow, is the amount of lighting on the surfaces of the "impossible" bit. Pretty hard to fake without photoshop cheating. Must have been hell to photograph! Great shot.

Posted by: Jonathan Puckridge | March 5, 2007 6:27 PM

yup you're right jonathan. The beauty of accidental viewpoints.

Steve, don't mock the unsuspecting!
You know it's not a matter of "viewpoints".
You merged the entirely synthesized figure with some other "ordinary" objects (may be synthesized too, doesn't matter) and even bothered to paint fake shadows, how devious of you!
Long ago, as a teenager, I was puzzled by Escher staircase and decided to find out.
I modelled all stairs surfaces as a set of equations relative to their spatial positions and relative intersecting angles.
Interestingly this DOES have a (unique) solution : ALL angles are 0, that is all surfaces are coplanar, this is why it is possible to draw such figures.
This is the way your "3-D impossible structure" has been constructed, faking the shadows is really cheating, you should be ashamed ;-)

Posted by: Kevembuangga | March 6, 2007 12:26 PM

get a clue dude.... this isn't my creation nor am I trying to mock anyone.

Posted by: steve | March 6, 2007 12:45 PM

this isn't my creation nor am I trying to mock anyone.

So what are you trying to do "dude"?
Boosting your blog rank?

Posted by: Kevembuangga | March 7, 2007 2:57 AM

Steve, I hope this is the last time you post clearly doctored images on your blog! You should be ashamed. Not only is this an impossible image, but I've also never been to a science museum and so I've never seen the same thing with a special spot that says "stand here to see the impossible figure!"

Why would you have posted an image of an optical illusion on a blog that deals with vision research and psychology? Nothing makes sense today. I need a cookie.

Posted by: Brian | March 7, 2007 3:41 AM

"So what are you trying to do "dude"?
Boosting your blog rank?"

What the hell are you talking about? This is just an interesting example of visual perception!

Posted by: steve | March 7, 2007 8:56 AM

What the hell are you talking about? This is just an interesting example of visual perception!

Yes, but :
- The illusion is from 2-D not 3-D like the Ames room, i.e. you need ONLY a flat drawing to produce this.
- The image you use is a computed image not a photograph.
- Your comment to Jonathan was (deliberately?) misleading in that it pretends that it is a peculiar viewpoint of a REAL 3-D object which produces this "impossible structure" look.

Posted by: Kevembuangga | March 7, 2007 11:33 AM

"The image you use is a computed image not a photograph."
How do you know it's not a photograph?

Posted by: steve | March 7, 2007 12:01 PM

Hi, Kevembuangga.

Here's a link to lots of Escher 'impossible' figures that can be constructed in real life. I can't help it if you refuse to consider this as a possibility for the image above.

http://www.cs.technion.ac.il/~gershon/EscherForReal/

Your brain is making assumptions about the properties of an object based on incomplete information (i.e. information from a single viewpoint, which happens to be unique in its inability to convey the actual structure of the object). That's why it's interesting. Whether or not we can verify that this particular object in the above image is an example of this is essentially irrelevant: this is an easily reproduced phenomenon in the real world. A discussion of accidental viewpoints makes a lot of sense, then!

Posted by: Brian | March 7, 2007 12:02 PM

Kev, that is actually a photo. My comment about the shadows was not meant to imply that they were faked; rather, that they give essential information about the actual physical structure of the object.

To be excruciatingly specific, what you're looking at is a generally rectangular construct with two additions to it that are normal to the main body. One is at the lower right, and sticks out the same distance as the other extrusions in the plane of the figure. The other is at the upper left and sticks out farther than normal and is beveled at the end such that when viewed at the proper angle (that of the photo), it appears to be attached behind the right most vertical bar.

The shadows bear this out, as there's an anomalous shadow crossing the lower horizontal beam that shouldn't be there if this impossible figure had been actually created. It's cast from the upper left normal bar, crosses the lower horizontal bar, then terminates on the paper. The fact that the shadow terminates instead of joining with the shadow cast by the rightmost vertical bar shows that there's a discontinuous section in the model.

Posted by: G Barnett | March 7, 2007 4:29 PM

Go here for more information on the effect:
http://scienceblogs.com/omnibrain/2007/03/other_impossible_3d_structures.php

Posted by: steve | March 7, 2007 4:32 PM

G Barnett: Kev, that is actually a photo.
steve: Go here for more information on the effect:

Ok, but AGAIN, this is NOT a 3-D illusion, it only works because the SINGLE (monocular, photographic) viewpoint flattens the picture to allow for the fooling of our visual cortex 3-D "reconstruction".
In a real, live view, binocular vision will PREVENT the illusion to happen.
All those tricks have been known since at least 1966, see Richard L. Gregory's Eye and Brain: The Psychology of Seeing

Posted by: Kevembuangga | March 8, 2007 10:20 AM

actually.. you can still see the illusion in real life - if you are looking at it from the correct view point!

Posted by: steve | March 8, 2007 10:43 AM