why do colors seem to appear in different planes on my LCD screen?

April 25, 2007 1:08 PM

When I look at an image (on my LCD computer screen) that is composed of a combination of solid colors separated by black borders, I perceive the colors as appearing on different planes of depth, as if the image were three-dimensional. If I close either eye, the illusion collapses. Am I the only one affected by this, or is it a consequence of the way LCDs display color?

To attempt to illustrate what I'm talking about, I created a GIF that, to me, strongly produces this effect (no visual trickery involved):



I perceive red, magenta, orange, and cyan to be closer than blue, yellow, and green. Red is always closest, blue is always furthest, and green seems to fall exactly in the middle.

Can anyone explain to me why this is, and if you can even see what I'm seeing?

posted by pmbuko to science & nature (15 comments total)

Image displayed in preview... grrr. Here's a direct link.
posted by pmbuko at 1:10 PM on April 25, 2007

Do you wear glasses or contacts? The effect you're seeing, I think, is due to the refraction of light through your lenses. At least, that's what I experience with my glasses.

It's the same thing as a prism.
posted by odinsdream at 1:27 PM on April 25, 2007

are you sure this effect only works on lcd monitors? In painting I often use color to enhance a sense of depth in an image. Due to the norms of atmospheric refraction warm colors often loom forward while cool colors recede. Just as saturated colors leap forward while grayer colors recede. Could this be similar to the effect you are perceiving?
posted by subtle_squid at 1:32 PM on April 25, 2007

This is something that happens in your eyes, not in your display. It's normal; it happens to everyone.
posted by Steven C. Den Beste at 1:37 PM on April 25, 2007

Based on my own experiments with LCD panels, the different pixels are horizontally offset a fraction of a pixel. There isn't any (visually apparent) depth difference. This little horizontal subpixel shift is what's used for ClearType and other subpixel font antialiasing techniques.
posted by chairface at 1:43 PM on April 25, 2007

Here's a screen shot that will show the subpixel shift in your monitor. View at 100% for best effect. The little crosses are all perfectly aligned in the bitmap. Any shifts you see is entirely due to your hardware.
posted by chairface at 1:48 PM on April 25, 2007

I agree with subtle_squid, it's probably your eyes "seeing" atmospheric/aerial perspective. The Wikipedia article has a few good examples (or you can just look around, it's everywhere).

One guess as to why you don't see it when you look at a CRT is that the CRT may be showing more glare, which distracts you ever so slightly and reminds your brain that it's really a flat screen.
posted by anaelith at 2:02 PM on April 25, 2007

Here (1.7MB image) is how the subpixels in my (dusty!) monitor are laid out. Yours is probably similar.
posted by aubilenon at 2:03 PM on April 25, 2007

I see this effect on LCD screens all the time... I liken it to those 'magic eye' pictures that give false depth perception. What appears to be a regular grid of pixels is really slightly shifted red, green, and blue pixels, so different colors actually inhabit different positions relative to where we think the borders of the pixels are. This slight positional offset confuses our brains into believing certain things are farther away than they really are.

If you think about how an autostereogram works (see here for an example), you can see how slight shifts in the placement of pixels create the illusion of depth. Then take a look at chairface's linked subpixel shift image and you can see how this occurs naturally due to the design of the hardware.
posted by foobario at 2:14 PM on April 25, 2007

I think what you are experiencing is an effect of the chromatic abberation of your cornea and lens:

In optics, chromatic aberration is caused by a lens having a different refractive index for different wavelengths of light (the dispersion of the lens)....Since the focal length f of a lens is dependent on the refractive index n, different wavelengths of light will be focused on different positions.

So when you are looking at something red, the sharpest image is farther from your lens (for a given configuration of your lens; the shape of the lens is changed by a special muscle, the ciliary muscle) than for green, which is farther than for blue, incidentally implying that the index of refraction of your eyes is higher for shorter wavelengths.

As you know, objects that are farther away focus on a plane which is closer to the lens than objects which are closer; the basis of the color illusion you see is that your visual system is using the cue of the depth of the focal plane inside your eye to give a sensation of distance in the absence of contradictory cues. That makes red seem nearest because its sharp focal plane is deepest in the eye.

The collapse of the illusion when you close one eye is interesting; I think that means your depth detecting system must 'know' it is vulnerable to being fooled when one eye is closed and blocks the illusion. The black border comes into play because the apparent continuity of surface which would exist without that border contradicts the illusion of depth and inhibits it. This is somewhat similar to the notorious illusion most people experience when they look at a face carved into a surface and see it as standing out from the surface in relief, despite the fact that their binocular vision starkly contradicts this perception.
posted by jamjam at 2:17 PM on April 25, 2007

Seconding SCDB. This used to happen to me with one color. I think it was yellow. Yellow always looked 'behind' the other colors. Now it doesnt happen anymore.
posted by damn dirty ape at 2:22 PM on April 25, 2007

Have you tried playing with different resolutions and refresh rates? Maybe switching the input types, say if your screen uses a VGA cable, try a DVI. Playing around with this stuff might make things appear a little different and stop the illusion.
posted by damn dirty ape at 2:23 PM on April 25, 2007

I'd never heard of chromatic abberation, it is an interesting theory. I wonder if it plays in to the effect I describe something I had assumed to be purely an effect of the brain.

Based on the photos any such effect with in the eye must be subtle as I've never noticed the effect with in my vision though it is often quite apparent in photography, as shown in evidence in your link.
posted by subtle_squid at 2:38 PM on April 25, 2007

There was actually a mediocre platformer for the SNES that came with a pair of paper glasses that had clear lenses in them that enhanced the chromatic aberration effect somehow, which the level designs exploited to produce a 3d effect. I wish I could remember the name of the game.

I remember wearing the glasses and trying to see if they worked on other things, like magazines and the like - they did.
posted by philscience at 9:52 PM on April 25, 2007

I wear prescription glasses to astigmatism. I tried taking them off just now and the effect is no longer present (even if I squint really hard to regain normal focus).

So, chromatic aberration it is.
posted by pmbuko at 6:50 AM on April 26, 2007