Another bizarre illusion

September 8, 2014 • 2:23 pm

I just realized that Matthew must have a thing about fooling vision, for he not only sends us many cases of crypsis (e.g., “spot the nightjar”), but optical illusions as well.

Here’s a good one he sent as a tw**t from Kyle Hill:

Screen Shot 2014-09-08 at 4.18.40 PM

Yes, the lines are abstolutely straight; you can check that out with a ruler or a piece of paper. Now why do they look curved?


41 thoughts on “Another bizarre illusion

  1. it takes a half a second, but you can focus on the grey lines and ignore the blocks and they are perfectly straight.

    Optical illusions like this don’t seem to work well for me. Of course, my vision is already so screwed up, it’s almost useless anyway.

  2. Yes, our brain does a lot of image processing (or photoshopping). For example, if you close one eye, you see your nose. If you close the other one, you also see your nose. But when you open both eyes, your nose disappears, it is eliminated (photoshopped out) by your brain.

  3. Wow, that’s pretty cool. At first I couldn’t believe it. I had to hold a straight edge to the screen to check.

  4. Here’s my hypothesis: The gray lines separating adjacent rows are perceived by your peripheral vision as subtracting from the white squares, but adding to the black squares. So the boundary between rows seems to waver up and down.

    But when you look at it directly, your eye correctly perceives the gray as neither white nor black, so the line looks straight.

    1. I’m thinking it has to do with the patterns of white and black vertically and your eye/brain trying to interpret those as continuous lines. I think your brain is trying its best to make a 3D model and is interpreting the varying widths of the black and white as “closer” and “further.” The fact the verticals don’t line up then means it perceives an undulation to try to keep continuity with the verticals. It then tries to make sense of the gray by overlaying onto a topography that doesn’t exist.

    2. I like Greg’s hypothesis, as I started to think along the same lines. But I think it might be wrong. I downloaded the image and processed it a bit to test that idea. If I understand correctly, then that hypothesis suggests that inverting the image should invert the bulges, but it doesn’t (at least to my eye).
      I think that Vin might be right. Your brain tries to interpret the lines of bright and dark squares as low-resolution (pixelated) contour lines on a map and imagines a consistent 3D landscape.
      I tried converting it to different colors (still works but maybe less effective).
      The grey lines seem important; mapping them to black or white reduces the effect. Subtracting a local smoothed background does not eliminate the effect. Running an edge-detection filter definitely reduced/eliminated the effect. Making all the black squares white seems to remove the effect.

    3. I see it the other way round! – where white and black are adjacent for half the width of a square, it’s the white that seems to bulge at that end, not the black as your description would imply.

  5. It seems to me to be akin to a Moire pattern, except that it’s more irregular than the usual examples of that effect.

    Or as my spell checker would have it, a More pattern.

  6. So, I showed this to my 14 year old son and he says “that is called the cafe wall illusion” first documented on the wall of a cafe in Bristol, England. And then proceeds to show me a picture of it in a book about visual illusions. I’m gonna have to amp it up to impress him.

  7. I can imagine many scenarios during human evolution which would select for this strange phenomenon. Most of them involve herds of zebras and gently rolling hills of the African savanna. I haven’t yet worked out all the details.

  8. There is an interesting ‘glitch’ in the picture–between rows 6 and 7 the illusion doesn’t apply (for me, anyway), that is, the line between them looks straight. Does everyone else see it this way?

    In those two rows the squares all align, so the illusion must have something to do with the differently-sized squares above and below the grey line.

    1. I noticed it too. And there the squares touch corner to corner, as they are the same size. The top of the image looks straight too, as the squares seem to be the same size.

      On rows that appear to curve the black and white squares are different sizes causing them so shift somewhat from their adjacent rows.

      It’s maybe the amount of black or white that you see that makes the line seem to curve.

    2. Looking closer, the line appears to curve away from the row with a smaller white square toward the row with the larger one.

  9. I’ve been using optical illusions like this to show my dad how easy it is to fool our brains. I haven’t tied this to religion yet. I’m hoping he will make the connection on his own.

  10. The previous occupant of a house I lived in in Palo Alto had lined the kitchen drawers in Contact paper something like this. It was sometimes hard to grab just the silverware/cutlery you wanted.

Leave a Reply