Sunday, April 13, 2008

Colored Light and Form

When two lights of different colors illuminate a form, the lit areas interact with each other in unexpected ways.

In this oil sketch of a white head maquette, an amber-tinted light shines from below left, while a blue-tinted light comes from almost the opposite angle.

There’s almost no overlap between the regions lit by each of the lights. One light or the other covers almost every surface of the head. There are just a couple of small places untouched by either light: the dark area where the nose meets the eye socket, and the hollows above and below the ear.

This maquette is lit by two colored lights. One is yellow-green, and one magenta. The lights are placed closer together, so the illuminated areas overlap on the top of the head, the brow, and the cheekbone plane. In these shared areas, the colors mix to a pale yellowish white, brighter in tone than the lightest tones in each of two regions lit by one light only.

Here's another casting of the plaster head covered with silver aluminum powder. The surface now has a high level of “specularity” or reflectivity.

What happens with colored light on a reflective surface? Even with three different light sources (green, red-violet, and blue), the lights don’t really mix very much on the planes of the head. Instead, each light source accounts for a separate array of specular highlights.

Remarkably, our brains are able to construct an understanding of form based on these fragmentary bits of information.

But if we remove the color clues (and flop the image for a fresh look), the form is much harder to visualize. The highlights seem like random light spots on a dark head. This illustrates a key recent finding of visual perception researchers, namely that color plays a central role in the brain's active construction of form and depth, and is not—as is often supposed—a kind of extra frosting on perception.
Related GJ posts on
Character Maquettes, link
Studio Lighting Equipment, link.


stephen erik schirle said...

the last paragraph is super interesting, thanks!

Erik Bongers said...

I second that.
It got me thinking. Do I use colors to suggest form? Or only light and dark to model a shape?
I think the latter only.
This new knowledge might make me want to tune that a little bit.

And what about black and white drawings? I guess I would (unconsciously) use simpler lighting.

Hey, I guess this also explains why for B&W photography I prefer a simple graphic composition while for colour I like the complexity of things.
Hey, this may also explain the 'graphic' lighting of those old B&W films ! (remember The Third Man?)
I guess this finding means that a B&W image will always have lesser depth and that 'we artists' will (unconsciously) take that into account.

Yes, very interesting indeed.

Erik Bongers said...

Here I am again. Was still thinking about it.

Many (post-)impressionists sort of ignored light and dark and replaced it with 'light' colours like yellow, orange and even red while using 'dark' colours like purple and blue to suggest shadow.
However, most of them would use lighter yellow and darker blue and thus have a bit of actual light-dark working.
But some of those painters would drive it so far as to use pure colour and 'throw away them black and white paint tubes'.
And still the viewer would easily see the shape and depth.

Hehe...this may explain why there's no such thing as 'Black and White Impressonism'.

Dan Gurney said...

I find this post very interesting. It raises all sorts of questions I hadn't thought to ask before.

To begin with, I'm not sure I see the effect you're discussing. My eye/brain does not detect a noticeable difference in form and depth between the color and black and white images in the last two images in this post. (I believe myself to be very slightly colorblind.) The highlights on the B/W image look random, yes, but my perception of its form is pretty equivalent to that of the color image.

To test further, I went into my iPhoto collection and converted a few images to black and white to see if I would discern less form and depth with the color was removed. Maybe this was not a fair test since I had already seen the image in color, but the effect you're pointing to was not obvious to me. I felt like I was making an effort to see an effect that wasn't really there.

Does color play a big role in our ability to see depth, or is its contribution much more subtle, as it seems to me? Have studies been done on colorblind people so see if their colorblindness affects their spatial intelligence? Do animals who lack color vision bump into things more? What about bats and whales who don't rely on visual clues for depth?

How do proprioception, touch, and hearing contribute to our apprehension of form and depth visually? We stumble in darkness to the bathroom without seeing much. My hunch is that there is a good deal of interplay between the senses of sight, touch, and sound as we construct our perception of form and depth.

James Gurney said...

Dan, You raise a lot of great questions. I should explain that the old rule of thumb for artists was that you use light and dark tones (or value) to define form, and the color doesn't really matter with respect to the sense of form.

But the little silvery head with the colored lights is a slightly specialized case where the eye/mind sorts and groups all the highlights by color into sets. Each plane of the head with a given color highlight is parallel to the planes of all other highlights of the same color, due to the law of reflection. Therefore your mind unconsciously infers the form using these three sets of data points. Without the color information, your brain can't group the highlights, so they start looking more random, and the form suffers (at least it does for me).

Here's a quote from the vision researcher Kathy Mullen's website to address this topic from a scientist's perspective:

"For several decades color vision was considered very poor at seeing form and shape, and even called "form blind." We (McIlhagga & Mullen, 1997) dubbed this view of color vision the "coloring book model" because it describes a subordinate role for color vision in the extraction of shape and form: color vision simply "fills in" the contours and boundaries of objects that are primarily defined by luminance contrast (black & white). We have shown that this cannot be the way the brain uses color. Instead my results suggest that both color and luminance vision undertake the primary stages of spatial processing in a very similar manner, with little deficiency found for color vision."
Kathy Mullen Research

I'm sure all those additional senses you mention also come into play as well, not to mention stereoscopic vision.

Erik Bongers said...

Dan, I'm slightly colorblind too.
And I don't think it affects my ahum 'spacial intelligence', although I can never know that for sure since obviously I can't compare my sight to that of a errr... normal person.

And for the black and white picture not clearly being 'flatter' : I fear that we humans are so good at recognizing a face that we'll always be ably to 'imagine' a head even with very poor visual info.

I guess a random shape would show the difference more clearly. Especially if we were to see the B&W example first and try to imagine the 3D shape before seeing the color image.

Golden Plover said...

Thats a truly awesome stuff. I am really amused by this.