Is This an Optical Illusion?

July 9, 2006 11:27 pm

I am not really sure how to explain what happened, but I thought the occurrence was interesting enough to warrant an article. I had taken a photograph of two old doors. Below is what the the photograph looked like on the monitor.


If you look particularly in the areas of shadows, you will probably assume that even the areas of intense shadow are a shade of white or light gray. However, when I printed the image, the areas of shadow were a similar color to the color of the swatch below.


The printer I was using was an Epson Stylus PHOTO R2400. I tried different color profiles and different papers, but the color of the shadows varied only a small amount. After making two 13" x 19" pictures and a number of smaller ones, I decided that the printer must be malfunctioning. I then tried to print two other scenes. They printed perfectly. Now I really was wondering what to do.


The color of the swatch certainly did not resemble the color in the area designated by the black line. Next I scanned in the print on another machine and looked at the image. It appeared to have the same characteristics as the original.

Still puzzled as to what to do, I used the magic wand to separate the most densely shadowed area from the rest of the image and then did a copy and paste to form an new image.


The color of the shadows matched the swatch. When I looked at the main image after using the magic wand, I noticed the shadow area appeared darker when it was separated from the rest of the image by use of the magic wand.


When I put the original image next to the one (the image immediately above) on which I had used the magic wand to separate the shadows, I could see the difference.


When I tried this experiment by scanning in the printed image and working from it on a different machine and monitor, the results were exactly the same. That is why I call this piece my "Optical Illusion". I could also call it my expensive piece since I wasted a lot of paper and ink on it.

All supporting images are copyright, and cannot be
copied, printed, or reproduced in any manner without written permission from the artist.

  • The Paula Sanders Report is a regular Renderosity Front Page featured column, where Paula investigates and comments on graphic software, techniques, and other relevant material through her reviews, tutorials, and general articles.
July 10, 2006

SASE Jill for Genesis 8

DZ G8M UnderGear Set 1

SASE Blake for Genesis 8

Ruby for Genesis 3 & 8

Article Comments

sparrownightmare ( posted at 12:00AM Mon, 10 July 2006

Have you tried a different scanner? WHat color depth did you use, and at what resolution and DPI?

jan_scrapper ( posted at 12:00AM Mon, 10 July 2006

Great article. I was amazed how the dark color WAS really there!!! Indeed, an optical illusion. Your website is wonderful. Great artwork.

Becco_UK ( posted at 12:00AM Mon, 10 July 2006

There is a similar optical illusion around that uses a chequerboard with a cylinder casting a shadow. The shadow looks different shades of grey but is in fact the same shade. Do not always believe what your eyes see!

ggrace ( posted at 12:00AM Mon, 10 July 2006

I don't know if it is truely an optical illusion but I see what you mean. I have had similar "visions" In looking at things in heavy shadows. For example Ckeck out some shades of green in a dark forest where a bright ray of light sneaks thru a hole in the canopy. Your eye really can be fooled or is it your brain? Thanks for the fun ,,,this is what part of what art is about.

ysvry ( posted at 12:00AM Tue, 11 July 2006

a good lesson for you , most if not all of our senses work relatively. Maybe next time you can write about RGB to CYMK conversion were these colorshifts do take place. thx for the article.

Warlock279 ( posted at 12:00AM Tue, 11 July 2006

This is actually a very important phenomenon Ms. Sanders has brought up, and Becco_UK touched on it a little more. What your seeing is called "simultaneous contrast of value" [or something there abouts]. What happens is you more or less "see" all the colors in your field of view relative to one another. The easiest way to veiw this phenonmenon is take a medium gray value, and place it on a white, and a black square, and note how much darker it appears on the white one, and how much lighter it appears on the black one. This phenomenon is actually why a lot of people complain about their images looking too dark when they upload them here, whether realize what is actaully going on or not. Their image is the same as when it was, its just now displayed on a white background, making it appear a lot darker than it actually is. If you want to know more about it, you can find a lot of relevant links with Google. This illusion is noticable in color as well. Really interesting stuff, and something all artists should be aware. [Especially web page designers :p] Attached is a quick demeonstration of it. [Note the bigger you can make the image on your screen, the more more obvious the effect will become.] Anyway, cheers, and fun stuff, don't want to make this too long winded. :)

2ni ( posted at 12:00AM Tue, 11 July 2006

I think it's an optical illusion as those found on this site Especially, the one with the chess board and the green cylinder.

Paula Sanders ( posted at 12:00AM Tue, 11 July 2006

As I read the comments, I remebered that when I had a dark room and developed black and white prints, I always developed and fixed a white square and a black square to make sure that I had a little of both in each print. Sometimes the black turned out to be a dark brown and I had to reprint.

awrc ( posted at 12:00AM Tue, 11 July 2006

More along the same lines as the chessboard and cylinder example here: erception.html

davidoblad ( posted at 12:00AM Tue, 11 July 2006

Hi. I copied your door image and examined it with CorelPaint. Using the Magnifier tool one can see that just about door center, the pixels are grey scale. As ine proceeds upwards, the become a variety of sepia tones. If one uses a color sample tool, then most of the pixels near top of door are within a few shades of the swatch shown. Interesting how our eyes expect the color to be a shade of white. I just posted an equally interesting illusion that shows how your eyes fabricate mis-information. Caut and paste this URL to jump to my Gallery.. or search for 'davidoblad'.. Cheers!!!

Sixaxis ( posted at 12:00AM Tue, 11 July 2006

This is more than likely matrixing of color value and hue. WYSIWYG is only as good as what you see. You know as well as anybody that reflective and projected images will be interpereted differently bt the eye. I would throw a print from the original file into a lightbox against a print taken from a screenshot of the same print. That would at least end any questions about where the discrepencies are occouring. (Maybe) :)

deemarie ( posted at 12:00AM Tue, 11 July 2006

Thanks for the refresher art course Paula. It is mind-boggling how different the shaded area is perceived when it is near the white and when it is separated. Thanks so much for the article, it is something to keep in mind when both photographing as well as printing.

RaspK_FOG ( posted at 12:00AM Thu, 13 July 2006

It's just like what most others expressed; being fascinated with biology, chemistry, physics, and mathematics, I studied a little of all these a bit further. Research has shown that our perception is highly relative; part of it is reactive, while it can also be symptomatic (like when your neural connections become overloaded and you become temporarily near deafened or blinded). In your case, the dark areas on the the door panels make the ochre upper door parts seem a light beige. In fact, there is a simple way to overcome this: if you close your eyes so much you barely see, you'll notice there's almost no difference between the two. So, the illusion comes from you perceiving it as any different than the swatch (which, in fact, seems to be a cropping of the original picture). Another good example of this illusion is to draw a board full of black squares that only have a thin white grid separating them (think of it like plus white plus signs [+] connected to each other on a black backing); the grid will seem to be gray at the intersections.

T.Rex ( posted at 12:00AM Thu, 13 July 2006

I just want to point out the following about the eye- we have rods which register black/white, and cones which register R, G or B (which is why TVs and computer monitors work with RGB). In the retina, the intensity and hue of a color are interpreted in relation to the intensity and hue of other colors on the RGB and black/white scale. The info is then integrated in the retina to a multitude of hues and saturations in the ganglion layer before being transmitted to the brain. The problem comes in the integration when one has too contrasting a subject, like here, with black and white and shades of gray close to each other. It's based on the phenomenon called lateral inhibition. When the image reaches the brain, it tries to make sense of the input, which can further "distort"/misinterpret the color information. This is a well known phenomenon to reseachers in visual perception/physiology of visual systems. (I'm an anatomist, so I have a good insight in the workings here- summation of nerve impulses, lateral inhibition, etc.). I doubt there's anything wrong with your scanner, computer or printer. Good luck!

PaulCoddington ( posted at 12:00AM Thu, 13 July 2006

"When I put the original image next to the one (the image immediately above) on which I had used the magic wand to separate the shadows, I could see the difference." If this were a contrast illusion, it would not persist when you mask off parts of the screen with a sheet of paper. Besides, with the last image you provide, the Photoshop CS2 eyedropper confirms that the colours at the top of the door are different RGB values.

PaulCoddington ( posted at 12:00AM Thu, 13 July 2006

Another thought - monitor calibration may be out. On my monitor (Eizo L768), the shadows are definately not white/grey/black but in all samples provided here are clearly 'sepia'.

minieffects ( posted at 12:00AM Thu, 13 July 2006

This isnt an optical illusion. What you are seeing is the reflected light from the ceiling bouncing light onto the wall and creating a value difference. since the wall,doors, ceiling and floor are all white in color the light is bouncing from surface to surface.Also the grass reflects light because it is a lighter value and that color is being reflected onto the wall as well. Your seeing a hard shadow from the direct sunlight and soft shadows from reflected light within the doorway.This is an effect that many render engines try to simulate. Its refered to as Ray Tracing in the cg world. If you want to see how this works find a doorway with simular lighting and hold a white card such that the light reflecting off of the card bounces onto the wall. You are now affecting the value of the lighting on the wall. Make sense? Minieffects

Sardtok ( posted at 12:00AM Thu, 13 July 2006

Actually, there's nothing wrong here, but it's weird that the illusion doesn't work well on paper. Anyway, calling it an illusion isn't incorrect, this is basic color theory, how adjacent colors affect other colors. It's like the pink/magenta on 50% gray thing (anyone who's done painting should know this one), now the gray looks green(if I remember my color wheel correctly). And anyone who actually looked at the images while reading the article would see that this is what's going on(although it hasn't got to do with complimentary colors), as you can clearly see the brown in the color when it gets separated with a border. If I'm not entirely incorrect, the swatch is actually even a piece of the picture (notice the white line in it and some things that look like jpg artifacts). Anyway, never trust your eyes, they lie, well, actually your brain does, because of all the little things it does to make you see something that makes sense instead of just a bunch of colors. Minieffects: What you are describing is not ray tracing, but indirect lighting, sometimes called radiosity, some global illumination systems incorporates it and in some the two are not directly connected, but they use similar algorithms or something for their photon calculations. That might be the reason why it looks a greenish brown, but what was confusing here was that she didn't see it, because of the adjacent colors affecting it, as far as I can understand. Ah, I just read T. Rex' post, and he has a pretty nice and quick explanation of the physiology behind such color phenomena.

minieffects ( posted at 12:00AM Thu, 13 July 2006

Acutually radiosity was what I meant to say, thanks for the correction Sardtok. I get my cg terminolgy mixed up sometimes. I guess Im not really understanding what the problem is. Is it a monitor/printer calibration problem or that she doesnt believe what the photo shows. To me, what she sees as an optical illusion looks perfectly natural. When we are lighting sets for film we go to great lengths to simulate this indirect lighting to make an otherwise fake enviroment on a stage look real. All very interesting comments though. minieffects

Redmage777 ( posted at 12:00AM Fri, 14 July 2006

Things like this are invaluable to artist who use 256 or fewer colors in pixel art. Working this to your advantage could make the differace between garish low color images, or something passes that for "photo realistic"

Privacy Notice

This site uses cookies to deliver the best experience. Our own cookies make user accounts and other features possible. Third-party cookies are used to display relevant ads and to analyze how Renderosity is used. By using our site, you acknowledge that you have read and understood our Terms of Service, including our Cookie Policy and our Privacy Policy.