Parts of Mona Lisa's face have better spatial frequency or detail than other parts. Or, as a computer person would say, some parts of her face have more pixels and therefore are clearer than others. Higher spatial frequencies are seen better with central vision and lower spatial frequencies with peripheral vision. The way the painting is made, with a combination of spatial frequencies, makes the smile appear as it does. Take away the courser and more out-of-focus shadows that are seen with the peripheral vision and she looks completely different.

             A set of scientists from the Smith-Kettlewell Eye Research Institute in San Francisco, Christopher Tyler and Leonid Kontsevich, also decided to do some experiments on the Italian lady. Visual clutter is called "noise." If someone goes into a room that is completely filled with junk, it will be difficult to find an item, because the eye has difficulty focusing on everything. That is the theory behind this concept by the Eye Research Institute. The two scientists created 100+ different Mona Lisa portraits with different amounts of visual "noise," so each one looked fuzzy in a different way. Then they asked people to rate each of them a scale from happy to sad on a rating of one to four.

             Next, the scientists overlaid either the upper or lower half of what people thought were the "happiest" or "saddest" faces on the original Mona Lisa. They found that putting something over the mouth made the subjects see Mona smiling or frowning, but putting something over the eyes did not do anything to her expression. They say, in other words, that viewers get cues on a person's expression more from someone's mouth than eyes. .

             Dina Goldin, another researcher from Brown University in Rhode Island, has decided to get her two cents in too. The computer scientist explains that people have more muscles in their faces than any other parts of the body, so therefore is much more flexible.