I say perceptions, you say illusions ...
Thu, Nov 8 2007 15:06
Andrew Coates has a picture of what he coins an "illusion". I prefer to think of it as a perception, not an illusion. In the picture a square is rendered as being in the shade. So the mind adjusts what the colour actually is in full light based on surrounding information. The reason for this is so as we can readily identify things in changing light conditions. So there's that perception. And the other perception is the colour is as rendered, and it's the effective colour that's important. That's the view of the pixel in isolation, which is how a computer screen, a graphics card, or any other non intelligent hardware would "see" it. But let's assume the picture is in fact a 3D rendering of a cylinder on a checkered surface, and there is in fact shadows. If you move the cylinder so as there is no shadow over the centre square, that square would have to be rendered a lighter colour. But if the only information the computer has is the original picture it can't do that unless it tries to interpret surrounding data, much like what our brain does. The computer either has to have more information (such as the original 3D model), or it has to become smart like us; the alternative being the inability to render it correctly and the potential false assumption that the colour of the square has changed when the shadow is removed.
To the mind, it's not important to match the shadowed colour with a colour that isn't shadowed, rather the interpretation of the true colour is more important, and hence that's what we perceive. Ah the joys of the memories of all those hours I spent in psychology lectures :D