Famous optical illusion -- live

A beautiful live version of Adelson's famous visual illusion.

Here is the homepage of the original illusion:
http://web.mit.edu/persci/people/adelson/checkershadow_illusion.html
entr0pysays...

That was a well done video. Though, I've always thought that illusion is not actually an illusion, but just a trick. It always relies on ignoring the fact that one tile is in shadow and the other is in light when you go to compare them. If you physically did move the tile as animated above, it would suddenly appear much lighter when it moves into sunlight, because that is how light works. They must have gone to some work to render it in 3D, and then not have that one tile be effected by the scene lighting.

Samaelsmithsays...

>> ^entr0py:

That was a well done video. Though, I've always thought that illusion is not actually an illusion, but just a trick. It always relies on ignoring the fact that one tile is in shadow and the other is in light when you go to compare them. If you physically did move the tile as animated above, it would suddenly appear much lighter when it moves into sunlight, because that is how light works. They must have gone to some work to render it in 3D, and then not have that one tile be effected by the scene lighting.


I don't think they used any rendering. If you pause it at 0:58, you can see that her arm is equally lit where it should be in shadow and where it is in full light. I think it is done with the lighting that is coming in from offscreen in the bottom right corner to subtly illuminate the shadow of the cylinder.
Strangely, at 0:58 it's also more obvious that the two squares are the same shade.

draak13says...

Yeah, it's definitely not a trick. This is a famous illusion in still-life, and there is no bizarre rendering required. However, it is indeed a 'trick', and that trick is your assumption of what shades each of the tiles are. You look at the board, and immediately believe that this is a standard checkered board with exactly 2 different shades of tile. This assumption, and therefor your perception of the color of the tile, is false.

SamaelSmith had it right; there is a deception in how much shadow is actually being cast by the podium sitting in front of the metal floodlight. Consider the MASSIVE light shining above the stage, and consider the smaller floodlight in the back, and reconsider how much shadow you would actually expect there to be on the checkerboard. There would be only a faint shadow, not the dramatic shadow that they have cast across it. The tiles are colored to make it appear that there is a strong shadow, when there should only be a weak one.

Thus, it is your normally adaptive assumption that there is a strong shadow cast by the podium that causes you to believe that the dark tile looks white in contrast to the very dark tiles surrounding it.

>> ^entr0py:

That was a well done video. Though, I've always thought that illusion is not actually an illusion, but just a trick. It always relies on ignoring the fact that one tile is in shadow and the other is in light when you go to compare them. If you physically did move the tile as animated above, it would suddenly appear much lighter when it moves into sunlight, because that is how light works. They must have gone to some work to render it in 3D, and then not have that one tile be effected by the scene lighting.

draak13says...

Also, it's amazing how this illusion persists, despite the knowledge of exactly what's happening. Try covering up the surrounding tiles with a post-it note or your hand. As the information on the relative color disappears, the illusion loses its power...in particular, the white tiles above and below the tile in question. It's also amazing that the illusion resurfaces *immediately* after revealing the surrounding tiles =-P.

>> ^entr0py:

That was a well done video. Though, I've always thought that illusion is not actually an illusion, but just a trick. It always relies on ignoring the fact that one tile is in shadow and the other is in light when you go to compare them. If you physically did move the tile as animated above, it would suddenly appear much lighter when it moves into sunlight, because that is how light works. They must have gone to some work to render it in 3D, and then not have that one tile be effected by the scene lighting.

bmacs27says...

Ted is a great scientist. His life work as revolved around materials perception. He's also the guy behind this baddassery: http://www.gelsight.com/videos/

As many of you probably realize, almost every visual illusion was designed to illustrate some mechanism, or function of the visual system. This illusion is one of my favorites. It illustrates that the purpose of the visual system is not to directly "read out" luminance values from your retina, but instead to correctly interpret the reflectance properties of materials in the environment. Most often it's the properties of the materials themselves, not the properties of their illuminant which are important. However, the signals reaching your brain will depend drastically on both, and it is actually a relatively complicated process of 3d-scene reconstruction in order to parse out the reflectance properties you are interested in. Of course, the point of the illusion is to demonstrate just how automatic and subconscious this process is despite the obvious complexity.

offsetSammysays...

You make a good point here. This is actually quite different from the original illusion. The original demonstrated that our perception of color is based on context (e.g. the tile in the shadow is perceived as being lighter than the one not in shadow, even though their "absolute" color values are the same). If you were to create a physical model of the original by taking a normal checkerboard and shining a light on it, then moved the middle tile on top of a tile of the "opposite" color, as they do in this video, you would see that the tiles ARE in fact different colors. So really your brain is seeing things correctly from a contextual point of view, in that it's separating the surface color of the object from the color contributed by the light/shadow. If our brains didn't work this way we would have a lot of trouble navigating the world.

What I believe they've done in the video that's different from the spirit of the original illusion, is create a fake shadow. The "shadow" is actually painted onto the tiles (and also, importantly, the cylinder). If you were to take one of the darker tiles in the shadow and place it on top of the another dark tile, you'd find they're not the same shade, even though we assume they should be the same (the 'shadowed' tile would be darker by comparison). Notice that there is actually a rather bright light source pointing at the shadowed area, in which case the shadow only makes physical sense if the light source on the left is MUCH brighter, which is not the case. By the way, there is something of a shadow created by the light source on the left, but it is set up to be as subtle as possible and not have much real contribution to the image.

Your assertion that it's rendered in 3D could be correct, but this illusion should be possible to do (and much more cheaply!) without it.

>> ^entr0py:

That was a well done video. Though, I've always thought that illusion is not actually an illusion, but just a trick. It always relies on ignoring the fact that one tile is in shadow and the other is in light when you go to compare them. If you physically did move the tile as animated above, it would suddenly appear much lighter when it moves into sunlight, because that is how light works. They must have gone to some work to render it in 3D, and then not have that one tile be effected by the scene lighting.

Send this Article to a Friend



Separate multiple emails with a comma (,); limit 5 recipients






Your email has been sent successfully!

Manage this Video in Your Playlists




notify when someone comments
X

This website uses cookies.

This website uses cookies to improve user experience. By using this website you consent to all cookies in accordance with our Privacy Policy.

I agree
  
Learn More