Color Motion Illusion

The explanation for this phenomenon combines both physics, neurology, and perceptual psychology. If you have a friend that is into theatrical lighting that will help a bunch. Also there may be a little difficulty translating from American into English.

If you get the materials together, we could set up a time where I guide you through the construction using skype.

Don't get too worried if this discussion gets a little complicated, but the setup is easier than the description. Also the artistic uses go well beyond this setup. This is just the easiest way for me to get you to the phenomenon.

Theatrical lighting gels are plastic filters for the big hot lights. They usually come in sheets about 18inches by 18inches or you can if you want fifty feet purchase them by the roll. A theater with a lighting person will have tons of this stuff lying around.

You need a large gel that has a wavelength cuttoff of about 550 nm in the blue. I think this is a Roscoe Filter 84 Blue. You need another gel that has a cuttoff of about 550nm into the yellow. I think this is a Roscoe 24 Amber.

Cut out a large white circle with the cardboard about 2 feet in diameter. Use the black tape to make thick stripes that radiate outward. Stick the circle on the end of your variable speed drill so that you can rotate it at a slow continuous rate (about one revolution every 10 seconds)

You will also need a piece of sheer white fabric like a white silk chiffon. and a strong light or a slide projector.

I will try to send the rest of the description in the form of a diagram


If you set it up like the diagram this is what you will see. Although the whole circle attached to the drill will be spinning at the same rate, you will see the half in white light moving normally and the half in orange light will seem to be stationary. It is absolutely startling. I have an explanation of why this happens which I won't burden you with until you build the illusion.



I thought I would write a little bit more on the color illusion to give you a sense of how it fits into the sciences. I studied this illusion when I was a post-doc at Harvard and as a professor at Tulane University. I can throw the technical papers at you later if you want. I hope you don't mind but I am going to do this in stages starting with a description of what an observer will see.

First this is such a startling illusion that if you have not seen it you will not believe my words. If you don't believe me you would be in good company as I have had a paper rejected by a reviewer who did not believe the illusion on which my research was based. The paper was published and the reviewer bypassed, because the Editor of the Journal had seen the effect.

Take a slowly moving striped pattern, the stripes made from two distinct colors. Let the stripes be moving at about 3 degrees visual angle per second. This would be like moving your finger across the computer screen in about 4 or 5 second. If you take the colors of the stripes and balance them for grey scale energy (ie if we were a black and white camera the stripes would disappear) you would have a moving image where all of the motion information is carried by color alone. Balancing the energy for any arbitrary two colors is difficult and generally must be done for the individual under computer control.

When you balance the image so that motion is carried by color alone, the motion will seem to slow down - on the order of 1 to 2 degrees per second. A motion of about 2 degrees per second can be stopped by color. This is a very peculiar sensation. You can follow each striped section with your eyes, know it is moving, and still swear it is motionless.

On the rotating circle demonstration, the part of the circle covered in amber light is using a trick of how we are neurally hooked up to make the motion information only visible to color sensitive parts of your brain alone - I will call this part "equiluminant". When the circle is rotating slowly, you will see the "Luminant" side moving and the equiluminant side moving significantly slower, maybe even stopped. This is true even though you know that the whole circle is moving at the same rate.

I will stop here for a moment.

Lee Z.


This note will focus on color perception and how images impinging on the retina are folded into the visual cortex.

First - There are two types of photo sensitive cells on the retina:rods and cones.

Rods are strictly used for night (or photopic) vision. Although the rods are color blind, they are so sensitive that almost any light in the visual field will saturate them meaning that they are not in play with this illusion.

Normal daylight (or scotopic)vision is determined by the cones on the retina. We have three types of cones determined by their spectral sensitivity. For this discussion, I will call them R for cells sensitive to long wavelengths (red sensing) G for cells sensitive to the medium wavelengths (green sensing) and B for cells sensitive to the short wavelengths (blue sensing). I can go into a much deeper discussion on metamers, how these cells can determine color, individual differences, and species differences, but this is not the place.

For our discussion, there are three cone types, R, G, B and they sense a particular part of the light spectrum.

Light sensed by these cones in a particular piece of the retina is combined in the following way before it is passed back to the visual cortex. This is from the work of Hering.

(R + G) = Luminance pathway, this is sort of the amount of black and whiteness we sense in the image. Most of us are most sensitive luminance with wavelenths around 550 nm

(R - G) = R/G Chromatic pathway. This is the amount of redness or greenness of the light. Notice if the spectral sensitivities of the red and green cones overlap we get red green color blindness. A situation found in about 20% of males and very rarely in females.

B - (R+G) = B/Y Chromatic Pathway. This is the amount of blueness or yellowness we perceive of the light. Notice that the luminance channel information is intimately involved in this color pathway.

The distribution of our perception of light along these pathways explain why we don't see greenish reds or blueish yellows. Also, we will perceive an intense yellow when presented with light that is around 550nm. The addition of a little bit of blue light at 470nm will change our percept of the light to bright white.

Next part details how the illusion works

Lee Z.



Dear Ian,

Again sorry about the long story. This note contains information on how the set up for the illusion works.

All of the filters, lighting, spinning wheels and strange set up are using a trick to how we are perceptual hooked up to isolate visual information along one of the perceiver's chromatic pathways.

The rotating disc provides motion information. All of this motion information is filtered through the blue gel so the motion information carried in the light is filtered about 550 nm into the blue. The amber light when reflected off the white chiffon, strongly signals the R and G cones on the retina. So when the system is all working, the  R and G cells are seeing only a flat bright light. this effectively saturates the Luminance pathway.  The motion of the rotating disc behind the blue filter is only seen by the B cells on the retina. This means that the motion is only being sensed along the B/Y chromatic pathway.

Luminance pathway - senses motion and space better, also is higher resolution

Chromatic pathways - senses motion going 1 to 2 degrees per second slower. Loses space and is willing to attach itself to the largest luminance motion signal in the area. You can sense movement along the chromatic faster than the luminant channel

Both pathways seem to be present in our scotopic vision at all times.

This will make more sense once the illusion is set up.

The next note will have comments.

Thank you,

Lee Zimmerman



Hi Ian,

First comment.

The illusion is big, not subtle. When you get it set up right it will blow you away and drive any physicist crazy (oops, I hope you aren't a physicist)

Also, although I have focused on the motion aspects of the color percept, the spatial implications are equally as intriguing. For example, luminant and equiluminant (color only) can be projected onto a white cloth and when I move the cloth the equiluminant dots will stick to it like they are painted on.

Second comment

From a science education experience it offers a possibly wonderful exhibit possibilities. As you can tell, understanding this illusion involves a story that encompasses physics of light, Young-Helmholz trichromacy theory, Metamers (which could show why birds have better color vision), Herings  opponent color theory, actively folding the two theories together to obtain the illusion, visual neurophysiology and what illusions can tell us about information processing in our own brains. Inviting the scientist to figure out what is happening in the illusion is inviting them to engage the scientific process. Plus color is cool

Third comment.

The neuropsychology of this illusion has several real world implications - these have not been completely explored and have long stories behind them if you want to know.

1) Some dyslexia responds to viewing the world through colored lenses. This was a lesson in scientific politics for me.

2) Robert Snowden, vision researcher at Cardiff University, found that subjects driving in a simulator, would perceive themselves as moving slower than they actually were under fog conditions and would speed up to compensate. (Driving 90 miles an hour when they thought they were driving 60).
He was roundly thrashed for these results. What may be happening is that fog creates the color motion illusion.

3) Professional ball playing athletes may be wired up in a way that their color pathways have less impact on their overall motion perception.

Thank you,

Lee Z.