Friday, May 2, 2008

Black is Light, White is Dark

Are these swatches labeled correctly?
Of course not. The swatch on the left isn’t black. It’s a mid-range gray, and so is the one on the right. The swatch in the middle isn’t white. It’s darker than the other two.

What if I told you that the swatch on the left is black acrylic paint, the sample on the right is a jet-black dress shirt, and the swatch in the middle is a white newspaper?

The x-factor is sunlight and shadow—and the pesky tricks that our visual systems play on us.

The samples were all lifted straight out of a single photo taken in my sunny front yard yesterday morning. I was sitting in front of a refrigerator carton painted black.

Even when the tones are adjacent (between 2 and 3), our minds tell us that the “white” is lighter. It's good to keep this in mind when we're painting.

The visual cortex uses context cues to override the luminance information from the retinas. Professor Edward Adelson of MIT developed the “checkershadow illusion” to show that the white square (B) in shadow is equal to the black square (A) in light (click to enlarge).

Here’s a rule to remember: In bright sunlight, a newspaper in shadow is darker than a black shirt in the light.

For more about the checkershadow illusion, check out Dr. Adelson's website and this interactive demo.
More on GJ about that painted backdrop, link.
Thanks to Professor Adelson.

Tomorrow: Spokewheeling


Erik Bongers said...

It's impressive how you keep putting so much effort in many of these post. I'm beginning to realize that your scientific approach to observation of your subject is a key factor for realistic art.
That makes for me the basic lesson learned from this blog (so far) twofold : good preparation of and understanding of what you are going to paint.

However, for this topic, rather than conclude with a rule, I think the lesson here is that there is no point for a specific rule. After all it's impossible to memorize the relation between all materials in all colors in all light and atmospheric conditions.

True, our eyes trick us in believing that even the shadow side of the newspaper is 'white'. Although this seems like quite a drawback, in fact it's a big advantage. We are intuitively able to 'know' that that dark patch is in fact pure white ! This 'gift' is vital for us to recognize objects in any given lighting situation.
On a canvas this skill will help us to see that a newspaper in the shadow appears too bright or too dark.

So as a rule for this topic, I feel that we should dare to trust our eyes and not our knowledge in the final evaluation of our canvas.

What appears to be proof that we shouldn't trust our eyes is in fact an indication that we should !

Another way to look at it: our eyes are amazing ! That shadow part of the newspaper is dark gray and yet our eyes are not tricked and recognize it as white!

I just realized that the basic lesson learned from this site for me now is now threefold : preparation and understanding the final evaluation, forget all the preparation and knowledge and trust your eyes.
Of course, that switch from rationale (we know it is white) to intuition (does it appear as white) is quite a leap to take.

Unknown said...

This is something I struggle with all the time. This rule is going to be scratched into my drawing board. Thanks !

Ginger*:) said...

I am not sure how I can ever thank my friend Jenn Morris enough for pointing me in your direction. Your blog is so full of wonderful advice, charming commentary and fascinating artwork, that I am embarrassed not to have found this place by myself.

Thank you for sharing,and thank you for this post in particular. I am not painting dinosaurs, but hippos and rhinos and crocodiles have great similarity.

Your blog will be a daily visit.

James Gurney said...

Welcome, Ginger, and thanks, Eric.

Erik, as always your comments get to the very heart of the matter, and they're probably more interesting and insightful than the original post.

Perhaps I should have started with the rule about the black shirt and the newspaper, and explained it. Howard Pyle gave the rule to his students to surprise them and to push them to increase the separation between light and shadow. I wanted to test the rule.

In truth, I discovered that you have to try to block some of the inevitable reflected light on the shadow side for the newspaper to come out equal or darker. As you say, there really is no hard and fast rule, and it all depends on circumstances.

And I agree with your central point: that our perceptual systems are a vital gift--the world would be chaos if we didn't organize it. One reason scientists like Adelson are doing this research is to train robots to see, and it ain't easy. Cameras alone don't see.

But more and more I'm realizing that it's not enough to tell art students who want to paint realistically just to "paint what you see." It does take knowledge and real effort to draw foreshortened forms correctly, to avoid overdefining detail and edges, and to judge values accurately.

craigstephens said...

Another fantastic post! This and Erik's insightful comments stirred up a lively discussion in my morning classes. Thank you both.

Anonymous said...

Hey James,

Great post, Very insightful.


Unknown said...


Michael Chesley Johnson, Artist / Writer said...

Excellent post, and something I've been looking into lately. ("Great minds," etc.)

I first heard this from Sergei Bongart, and I have heard it repeated by others:

"Nothing in the light is as dark as the shadow. Nothing in the shadow is as light as the light. In other words, you can have all the detail you want in the lights and all the detail you want in the shadows, but the lights should stay light, the shadows dark. The two should never mix."

Like others here, I've observed that this isn't necessarily true, especially if you are comparing different objects in a scene, such as a white house and a red barn. It works better if you're comparing parts of a single object, such as just the white house.