Visual illusion may explain the allure of pointillist paintings

ResearchBlogging.orgWhat is so mesmerizing about pointillist paintings like Seurat's Sunday Afternoon at La Grande Jatte? At first, we're impressed by the technical virtuosity of the work. It's an immense painting that Greta and I visited many times when we were in college in Chicago (and now, whenever we return for a visit):

i-430d212eb0ceb232194fa91bc64ca991-monnier1.jpg

As you can see even in this reduced image, the painting is composed of tiny dots. But what you may not notice is that the dots in a given region of the painting aren't all the same color. Take a look at this detail:

i-5f7c6bc4cf59a6abb3d959f40281919b-monnier2.jpg

The leaves in the trees range from red to yellow to green to blue, and the water includes blues, purples, greens, and yellows.

When you visit the painting in person, you might get the impression that the colors actually change as you approach or move away from the work. Of course, they can't really be changing. But vision scientist Pascal Mamassian believes that what you experience as you view a pointillist work might be similar to an effect described by Patrick Monnier and Steven K. Shevell. Take a look at this image:

i-57c44c8ac41db3d106e20024fed060a5-monnier3-1.jpg

The orange/pink rings in the center of each large ring appear to all be different colors. But actually they are the same. Click on the image to see an animated figure that removes the other colors. What's particularly remarkable about this illusion is that in rings a and c, the color immediately next to the central ring is the same. The only difference between the figures is the frequency at which the rings change color. The same effect occurs between rings b and d.

The central rings in c and d appear to be much closer to each other in color than the central rings in a and b. Monnier and Shevell had several volunteers look at figures like this, comparing the multi-colored rings to single rings on white background. They adjusted the figures on white until they appeared to match the rings in the figure. The responses were consistent: everyone with normal color vision appears to experience the illusion.

Monnier and Shevell believe this illusion is due to an "antagonism" that occurs in the brain in response to just some the cones used in the eye to detect color: S-cones, which respond to the shortest "blue" wavelengths of light. That's because the differences between what was perceived and the actual color seen can be completely explained by the S-cone response: no differences in M-cone (green) or L-cone (red) responses were observed.

The antagonism only occurs at certain frequencies -- and this could lead to colors in the Seurat painting appearing to change when our distance to the painting makes the dots appear about the same width as the rings in Monnier and Shevell's study: about 0.15 degrees of visual angle.

P MAMASSIAN (2008). Ambiguities and conventions in the perception of visual art Vision Research, 48 (20), 2143-2153 DOI: 10.1016/j.visres.2008.06.010

Patrick Monnier, Steven K Shevell (2003). Large shifts in color appearance from patterned chromatic backgrounds Nature Neuroscience, 6 (8), 801-802 DOI: 10.1038/nn1099

More like this

New parents can come up with a seemingly endless array of vexing questions about their infants, from the best brand of stroller to the ideal song to sing them to sleep. The questions begin well before the child is born: what type of clothing should you purchase? What kind of crib? One question…
This story is some kind of awesome: For those who don't want to watch the whole thing, the observation in brief is that color perception is affected by color language. The investigators compare Westerners with our familiar language categories for color (red, blue, green, yellow, etc.) to the…
On Developing Intelligence, Chris Chatham shares a new study which demonstrates that performing new tasks actually reverses the accustomed workflow between different parts of the brain. Chris writes "Cole et al demonstrate that the causal influence is from [the dorsolateral prefrontal cortex] to…
Take a look at this short video clip. Can you tell which dot is blue and which is yellow? Click here to play Unless you have a rare vision impairment, this task should be easy for you. But read on, and we'll show you how you can become blind to this difference in as little as 40 minutes. The human…

My initial impression was that the rings in b and d were the same color (more orange) and that a and c were the same (more pink). Does that mean there is something wrong with my cones?!?

I don't think this would be news to an artist or photographer used to working with color. Artists may work with pigments or mixtures of colors, but they are much less concerned with the actual color of the component materials as the effect achieved by putting the component materials together into a visual design. Some Photoshop manipulations actually work to separate clusters of pixels of similar colored pixels in order to increase the perception of sharpness, contrast, and texture.

Also, it was noted a long time ago in studies of the visual system of cats that certain clusters or arrays of cones and rods fire in specific kinds of coordination. For instance, as part of their hypersensitivity to motion cat eyes are wired to extract information regarding edges. Any two cones or rods arrayed in a line of five cones or rods will fire a signal that indicates motion into or out of that visual line. In a sense, the cats brain jumps to conclusions based on only a partial firing of all the cones or rods in that line of five. It isn't unexpected that similar but antagonistic arrangements in the coordinated firing of cones or rods might accentuate sensitivity to colors or shapes. They have called this mechanism a feature extractor.

I have long thought that guys are wired to perceive female curves anywhere on the horizon. :-)

There's no lateral inhibition going on, like with movement illusions?

That is a very cool affect. So let me get this straight - all the central rings are the same colour?

I must admit I can't see the changing colour effect either and I tried using various colour settings for my monitor. Maybe the circles are placed to close together?

I do share your love for Seurat and pointillism however. Would be interesting to see a post on Bridget Riley.

Thanks

It's amazing how strong these illusions can be.Even after knowing the rings are the same colour I still see b and c as orange,and a and d as pink.

By A. Willow (not verified) on 26 Feb 2009 #permalink

It isn't a color illusion, but the effect is 10x stronger and more unbelievable. It is perhaps my favorite of all time:

http://www.ebaumsworld.com/pictures/view/3216/

It is a checkerboard with an object casting a shadow on the checkerboard. The two tiles marked "A" and "B" have identical shading, but look very different.

(one problem with the color rings above is that, due to antialiasing and the narrowness of the ring, the ring has slightly different shading at each pixel)

By Jim Battle (not verified) on 27 Feb 2009 #permalink

I have been using this effect in my photographic art for 25 years. My art form uses high definition photographs of magnified grains of color slide film (usually Kodachrome)using complex three dimensional lighting to elicit color juxtapositions which result in a pointillist photograph. In fact, I call what I do pointillist photography and it can be seen currently in the Palm Springs Art Museum and after March 19 at the Kingstad Gallery in Portland and of course here www.rosenblumstudio.com

CyberLizard: Your eyes are fine. I can't say the same about your reading comprehension ;)

As I understand it, A + C, B + D, and C + D should look the same. The effect is between A + B.