Feb 27 2015

What Color Is This Dress? It’s An Optical Illusion

This is pretty amazing – almost as much for how quickly this has gone viral as for the effect itself. There is now an intense debate going on in the intertubes over whether this dress is black and blue or white and gold. Take a look and decide for yourself. Buzzfeed has a poll which currently puts it at 72% white and gold, and 28% black and blue. Right now there are about 2 million votes, so that is probably statistically significant.

I see black and blue, no matter what screen or version of that picture I look at. It does not seem to be an issue with the monitor or viewing conditions.

The reason, in my opinion, this has gone so viral so quickly is that people are legitimately freaked out by the realization that how they see the world is ultimately a subjective construction of our brains. Taylor Swift tweeted about the debate:

“I don’t understand this odd dress debate and feel like it’s a trick somehow. I’m confused and scared. PS It’s OBVIOUSLY BLUE AND BLACK.”

That about sums it up. She thinks it must be a trick (it is – a trick of the brain), and is scared and confused. At the same time she is caps-lock-certain that her perception of the dress’s color is the objective truth.

This is clearly an optical illusion. The type of illusion is called color constancy. Our brains evolved to favor consistency over accuracy, in both memory and perception. If we see a tiger running through a sun-dappled forest, it’s important that we perceive a constant entity, not a morphing and changing image.

The actual color that falls upon our retina will change dramatically in different lighting conditions. This might trick a perceptive system into thinking that one item is actual multiple items, divided along lines of shade and light. In order to perceive the item as the single continuous thing that it is, our brains evolved color and shading correction algorithms White, for example, will appear blue in dark light, but our brains still see white – it corrects the blue perception into white.

Here is a black and white version of this illusion – the checkerboard illusion. The shade of squares A and B are identical, but our brains see them as light and dark. It makes assumptions about shading, and then corrects for the shadow effect, so that we correctly perceive the light squares as light, even when they are in shadow.

Below is a really intense color illusion. The blue and green stripes are actually the exact same color. Our brains perceive them differently because of the surrounding colors, which force our brains to make different assumptions about shading, and therefore they correct the color in opposite directions.

The dress is a similar color constancy illusion, but is also an ambiguous stimuli illusion. Ambiguous optical illusions are ones in which our brains are given conflicting information, or there are different ways to resolve the image that are equally valid. Remember the spinning girl illusion? This remains one of my most popular posts, for the same reason this dress controversy has gone viral. Our brains can make different assumptions to “see” the girl spinning clockwise or counterclockwise. There are lots of this type of illusion – is it a young girl or old woman, which way are the cubes facing, do you see a wine glass or two faces, etc.

The photo of the dress just happens to hit the sweet spot of ambiguity in terms of lighting and shading. Different people’s brains will therefore make different assumptions and correct for either apparent overexposure or underexposure. Do you have to correct for the glare of bright lights, or the dulling of colors because of shade? Remember, white appear blue when it is shadowed, and our brains correct the blue to white.  Our brains can correct the reflective part of the dress darker to be black, or the dark parts of the dress lighter to appear gold.

Conclusion

This is a fun viral phenomenon, and one that is a useful teaching moment. The dress color debate is the result of an optical illusion. Don’t be “scared and confused,” this is just how our brains work.

I want to emphasize that this is not just a isolated weird case. This is how our brains work all the time. What we perceive is a constructed illusion, based upon algorithms that make reasonable assumptions about distance, shading, size, movement, and color – but they are assumptions, none-the-less, and sometimes they can be wrong or misleading.

By the way, it appears that the dress is objectively black and blue (see the photos here), which means that 72% of people are correcting in the wrong direction.

151 responses so far

151 Responses to “What Color Is This Dress? It’s An Optical Illusion”

  1. carbonUniton 27 Feb 2015 at 9:35 am

    (really having some problems getting connections to theness.com this morning…)

    I don’t see what the factor is with this dress pattern which is causing confusion.

    The Today Show did a segment on it this morning.
    http://www.today.com/popculture/white-or-gold-blue-or-black-dress-sets-internet-ablaze-2D80518434

    There is a picture there which does show the dress in a manner in which the black areas have a golden tint to them and the blue seems lighter. But the effect looks like something caused by the lighting shimmering off the material to me.

  2. dkilmeron 27 Feb 2015 at 10:02 am

    Poor Steven’s blog….

    One thing. You say, “Different people’s brains will therefore make different assumptions and correct for either apparent overexposure or underexposure”. I don’t think it’s an exposure correction. I think it’s a white-balance correction. Our brains are a lot better at that than cameras are.

  3. tmac57on 27 Feb 2015 at 11:25 am

    I love this kind of example that starkly exposes the differences in individual’s perception rather than just talking or writing about it in abstract terms.
    I ‘clearly’ see the dress as white and gold, so everybody else who sees it as black and blue are bonkers!

    🙂

    Sending out prayers for your blog to heal soon!

  4. chopchopchopson 27 Feb 2015 at 11:35 am

    I’m pretty sure it’s due to how you interpret the lighting. Is it in a dark, bluish shadow, or is it in the same bright yellow light as the rest of the scene? It’s illustrated really well here: http://i.imgur.com/9N4KNLn.jpg
    and it seems like a real-life example of this classic optical illusion:
    http://www.scifun.ed.ac.uk/pages/about_us/shows_cube-illusion.html#

  5. RCon 27 Feb 2015 at 11:44 am

    All sorts of technical problems with the site today.

    I see blue and black – almost royal blue – if I look at just the top right corner I can kind of make the black go gold, but no matter how hard I try I can’t get any of the blue to turn whiteish. (There’s also a black/white cow patterned dress bottom left, which is clearly white – which isn’t helping that). Like Steven, I’ve tried different screens, etc, and I just can’t get it to turn.

    My wife saw white and gold at first and is seeing blue/black now. It’s definitely strange. The manufacturer says the dress is blue/black for those wondering.

  6. mumadaddon 27 Feb 2015 at 12:25 pm

    If I had to choose between white & gold and blue & black I’d pick the former, but actually I see neither; I see (light) blue (but definitely blue) and browny-gold. Does that mean I have a brain tumour?

  7. mumadaddon 27 Feb 2015 at 12:32 pm

    Larger picture here:

    http://www.bbc.co.uk/news/blogs-trending-31659395

  8. YtterbiJumon 27 Feb 2015 at 12:34 pm

    That blue/green spiral doesn’t work when you post the image scaled down like that with compression artifacts (notice all the dots in the gray area on the right side of the image, and the weird blending effects at the center of the spiral). So the colors actually *aren’t* the same in the blue and green regions.

    For anyone curious, here’s the full-scale GIF version which looks a lot better:

    http://theness.com/neurologicablog/wp-content/uploads/2015/02/color-constancy1.gif

  9. RickKon 27 Feb 2015 at 12:52 pm

    theness.com is having technical difficulties, and a very long-lived Vulcan died. Don’t tell me it’s coincidence.

    LLAP

  10. BBBlueon 27 Feb 2015 at 12:58 pm

    RGB colors in photo are 107-126-169 and 90-75-46, which are more like a gray-blue and orange-brown. Amazing to me that some brains automatically do the color correction to get blue and black out of that.

    http://wrd.cm/1ABgxO0

  11. tmac57on 27 Feb 2015 at 2:12 pm

    RC- I can also see the white as a light blue, but I cannot for the life of me see how the gold could be seen as black. I only interpret the white because of the background brightness and think of the dress as being in shadow (which could cast it as blue). When I look at the actual dress where it is clearly blue and black in other examples, I see no resemblance at all except for the design.

  12. Kestrelon 27 Feb 2015 at 2:31 pm

    I don’t see the dress as blue or black. It looks plainly white and gold, and I ran it through Photoshop to check the colors objectively, and the result was the same. This doesn’t strike me as an optical illusion at all. It seems much more like a lighting issue messing up the camera’s exposure.

  13. Danamron 27 Feb 2015 at 2:34 pm

    This is not a optical illusion.
    It’s bad Photoshop. I have seen several versions of the image now, most commonly the one included on this page, and I would like to make a couple of points.
    Firstly this is not the dress, it’s a picture of a dress. And that picture has been messed with. There are a ton of tools in Photoshop that could have made that dress any color.
    If you look at the un- “corrected” image, The dress is blue and black. The background is also NOT blown out.
    The image on this page has been brought into a editing program, and had the exposure slider slid 2-3 stops in the plus range. This results in the background blowing out and the loss of a true black point. As information is lost in the highlight areas the blue of the dress shifts more to a white. You will also notice that there are no reference colors available (skin tones for example) that would show the color shift.
    I am appalled at this whole mess.
    While it is correct to say that perception of color is very personal, this is not a example of that.

  14. Bronze Dogon 27 Feb 2015 at 2:45 pm

    My brain keeps telling me it’s a white and bronze dress in the shade. The picture with the red background from that Today Show link has me baffled as to how the black parts ended up appearing gold to me.

  15. PBenzon 27 Feb 2015 at 2:54 pm

    Using the color picker in OS X (10.10.2), the color is coming back as gold when I hover the medicine dropper tool over the photo. Same for the “loupe” tool in XScope. Yet when I used these tools to hover over the squares in the checkerboard illusion image, both the A & B squares are identical. So how can the dress be blue? I don’t get it.

  16. BillyJoe7on 27 Feb 2015 at 4:14 pm

    Steven,

    I don’t think you have the correct explanation for the checkerboard illusion.

    It’s more to do with colour contrast than shadow.
    The purpose of the “shadow” is to make the colour of B the same as the colour of A.
    But you see A as darker than it actually is because it’s surrounded by light areas. And you see B as lighter than it actually is because it is surrounded by dark areas.
    It’s essentially a colour contrast illusion.

    The same effect can be achieved by comparing a light disc on a dark background with a dark disc on a light background. Both discs can be the same colour but look different because of the surrounding colours.

  17. cosmicaugon 27 Feb 2015 at 6:18 pm

    http://lizclimo.tumblr.com/image/112252924494

  18. tmac57on 27 Feb 2015 at 6:51 pm

    Bronze Dog- “The picture with the red background from that Today Show link has me baffled as to how the black parts ended up appearing gold to me.”

    The thing is, that the image Steve has ( the one most people are reacting to), is very different from the ones where it is obvious that the dress is black and blue. Several sites have the images side by side, and they are radically different. The one in question is way more ambiguous and open to interpretation.
    I don’t think anyone who falls in to the black and blue camp would say that the obvious one and the ambiguous one are even close to being the same, even with the back grounds removed.

    cosmicaug -The Liz Climo comic was a good touch. I saw that on my Facebook feed too, as I am a fan of her work.

  19. grabulaon 27 Feb 2015 at 7:40 pm

    Is this whole thing an an April fools joke? I see a gold and white dress in indirect lighting, making the white look slightly greyish-blue, as white sometimes does

  20. mumadaddon 27 Feb 2015 at 8:08 pm

    Grabula,

    “Is this whole thing an an April fools joke?”

    I was wondering the same thing, but thinking more ‘meta’, or social experiment. Everyone I showed it to gave the same answer, but then it was only three people so you’d expect that within the stats described. The fact that the RGB values (at least according to BBBlue) back up my original assessment only reinforces my opinion:

    RGB colors in photo are 107-126-169 and 90-75-46, which are more like a gray-blue and orange-brown. Amazing to me that some brains automatically do the color correction to get blue and black out of that.

    http://wrd.cm/1ABgxO0

    Ugh, I just found myself identifying with Taylor Swift, who- or what-ever that is.

    (By the way I’d heard of that person, but didn’t know it was a girl; thanks, Steven Novella, for teaching me something I definitely didn’t want to learn)

  21. catticus finchon 27 Feb 2015 at 8:51 pm

    I’ve often wondered if our brains do this with sounds and music too. I’m a singer, and while I don’t have “perfect pitch”, I have accurate pitch. When I listen to certain voices of singers who aren’t pitch perfect, I still enjoy it and find it interesting. I always have wondered if my brain corrects the notes they are singing. Let me know if you have any insight on this. The song “Perfect Day” by Lou Reed is a good example of this: https://www.youtube.com/watch?v=QYEC4TZsy-Y

  22. The Other John Mcon 27 Feb 2015 at 8:53 pm

    BJ you are more correct but still this discussion is super confusing (no matter what site or TV channel) because of the complexity of color perception. First is terminology. Wavelengths of light and RGB triplets are NOT “colors”… Colors or hues are the perception that is occurring in someone’s head, which is related to but not identical to wavelengths and RGBs. Second, viewing the image is a giant uncontrolled experiment in color perception, with so many confounds. It’s hard to know where to begin: in calibrated displays, image formats, ambient lighting, display size and resolution and color space, viewer’s color vision, viewing angle and distance, etc. it’s not too shocking that people are reporting conflicting experiences…

  23. steve12on 27 Feb 2015 at 9:05 pm

    Obviously we know that several systems between retina process everything in a relative rather than absolute fashion, and the systems want consistency as a means of perceptual hypothesis testing.

    But why the hell do a sizable minority of people see this differently? Are there people who can switch back and forth a la necker cube?

    I ask this because for a brief moment when reloading the page, the dress turned black and blue for me. Could that be expectancy of some sort? I don’t know…

    I’d love to have some information about the white&gold folks vs. the black and blue folks..

  24. steve12on 27 Feb 2015 at 9:11 pm

    Hey OJM!

    THink this’ll be a topic in a few months at VSS?

    Do you think this is a display issue, and the 25% that see it one way just have, e.g., their display set differently?

  25. Willyon 27 Feb 2015 at 9:36 pm

    Haven’t read your post yet, but to me it’s clearly black and blue and to my wife it’s obviously gold and white. I am dumbfounded.

    steve12–My wife and I have different opinions when viewing the same screen. I continue to be dumbfounded.

  26. The Other John Mcon 27 Feb 2015 at 9:40 pm

    Hey Steve! I hope no one writes a book or makes a movie about this, but yes I’m anticipating some papers :-). Could be a couple diff images floating around I think I heard, plus displays are all over the board in terms of their color space. RGBs probably just happen to be right around the thresholds between white/blue? Would be my guess.

  27. The Other John Mcon 27 Feb 2015 at 10:11 pm

    Think if I showed all of you a really light grey, kinda white/blue/grey but just barely. Then I ask you to name the color. This will be variable across people. Then the color contrast assumptions that Dr N mentioned kick in, making the opposing color something different, too. We know color opponency happens in pairs (red/green, blue/yellow) plus we know color constancy is a real effect.

    But again, the psychology of color perception is bizarre counterintuitive and complex…so interesting!

  28. Mr Qwertyon 27 Feb 2015 at 10:51 pm

    For a semi-objective measurement, one could average out the colors (for example, 3-4 pixel wide gaussian blur in photoshop) and read the value from the center of the stripe. I got approx:

    “lighter” color: red:125 green:139 and blue:180 – or in web colors, #7d8bb4, which is some kind of light blue (actually this web page http://gauth.fr/2011/09/get-a-color-name-from-any-rgb-combination/ says “Cool grey”).
    It has most in the blue component, but it is also very close to grey which can be simply white based on white balance.

    “darker” color: R:100 G:87 B:75 or #64574b, or “Umber” (never heard that name before 🙂 ). It has more red than green than blue, which is consistent with a yellowish, brown or gold, but it is dark enough so again it depends on the context.

    So, as everyone already noted, the actual colors are ambiguous (I personally saw bluish white and light brown or dark gold).

    I think if there was just a bit of skin (or some other kind of reference) visible, most of the ambiguity would go away!

    The older I am, the more I learn how much color perception depends on balance/contrast, “priming” (what you expect it to be from the context), individual, etc. My parents work in publishing business and I’ve seen/heard them (jokingly) argue about colors many times, and still have those arguments with them (and for the record Mom, it’s blue, not a turquoise shirt).

    To go on a bit of a tangent now, one of the things I’ve learned recently is that our colour receptors (cones) aren’t actually most responsive (they don’t peak) at “red”, “green” and “blue” monochromatic wavelengths. It’s actually shocking (to me) how far away the peaks are from “red”, “green”, “blue”: http://www.color-blindness.com/wp-content/images/cone-sensitivity.gif

    The sensitivities also overlap a lot and it’s not possible to “tickle” green without activating red or blue receptors.

    This inability to activate one without activating others practically means that, if a camera/film doesn’t record the whole spectrum and the display doesn’t show the whole spectrum (which is not practical), it’s not possible to reproduce exactly the same effect in the eye, for each possible original color, by just recording and displaying a subset of the spectrum (like the 3 monochromatic RGB values)!

    However, due to color metamerism (http://en.wikipedia.org/wiki/Metamerism_%28color%29) and other effects it is possible to get to a “good enough” result with storing/converting/reproducing just a few (3) monochromatic wavelengths and I had no idea how much research people have done in order to achieve what we take for granted today, with our displays & cameras –
    http://www.cis.rit.edu/mcsl/research/broadbent/CIE1931_RGB.pdf / http://en.wikipedia.org/wiki/CIE_1931_color_space 🙂

  29. Vijayon 27 Feb 2015 at 11:02 pm

    This can help you see the white and gold: http://xkcd.com/1492

  30. BillyJoe7on 27 Feb 2015 at 11:50 pm

    Apparently the actual dress is blue and black.

    I see white and gold.
    So does my wife and two sons.

    But I’m being a bit generous to both of those colours. The white has a slight but definite bluish tinge and the gold a definite brownish tinge.

    My older daughter sees blue and black.

    And she doesn’t mean bluish white or whitish blue. It’s a solid blue colour.
    And the other colour is solid black, not a brownish black or blackish brown.
    There is no way that I could call the gold that I see black, or the white that I see blue.

    My younger daughter sees solid blue and a bronzy greyish black.

    I thought this illusion was pretty uniteresting until I made the comparison with my daughters.

  31. Pete Aon 28 Feb 2015 at 12:32 am

    The people who see the dress as white and gold are suffering from a colour assessment error, however, those who see it as blue and black are suffering from a different type of colour assessment error.

    The photographer caused the camera to make two very large errors: both the auto-exposure and the auto white balance algorithms used mainly the dress rather than the background — hence the drastic overexposure and the strong yellow colour cast.

    A light area of the blue part of the dress has RGB 180 194 230; HSB 223 22 90, which is pale blue.

    A light area of the black part of the dress has RGB 129 113 77; HSB 42 44 50, which is murky gold, it is definitely not black or dark grey.

    Therefore, as depicted, the dress is pale blue and murky gold — it is neither white & gold nor blue & black! The material appears to be especially shiny rather than matt, but most of this effect is caused by the drastic overexposure of the photo rather than being a characteristic of the material.

    After properly adjusting the image for brightness and white balance it becomes obvious that the dress is in fact deep blue and black.

    Summary
    This is not an example of an optical illusion, it is a sobering reminder that the advanced algorithms used in digital cameras cannot begin to compensate for photographers who are completely clueless. The very first step in photography is to obtain the correct colour balance — with both film and digital — for the intended rendering of the captured scene; otherwise, all viewers will misinterpret the image in a variety of ways, as has been clearly demonstrated by this photograph.

  32. BillyJoe7on 28 Feb 2015 at 5:26 am

    Pete,

    “A light area of the blue part of the dress has RGB 180 194 230; HSB 223 22 90, which is pale blue.
    A light area of the black part of the dress has RGB 129 113 77; HSB 42 44 50, which is murky gold”

    Therefore, if you’re not seeing a pale blue and murky gold dress, then it is an optical illusion by definition.
    Of course, in a sense, everything we see is an optical illusion.

    “The people who see the dress as white and gold are suffering from a colour assessment error, however, those who see it as blue and black are suffering from a different type of colour assessment error”

    Brains may not see it that way. Brains that evolved successfully are still here. If their colour assessment is part of that success, as seems likely, then “colour assessment error” is at least a relative term.

  33. mindmeon 28 Feb 2015 at 6:46 am

    Oddly when I looked at it in the morning it was more white and gold. By night time, I looked at it again and it was more blue and black.

  34. Feboon 28 Feb 2015 at 6:58 am

    It seems to me that all this talk about optical illusions, while interesting, does not really have anything to do with the dress in the photo. The color ambiguity in the photo is primarily a product of the exposure and temperature settings of the camera that took it.

  35. Ian Wardellon 28 Feb 2015 at 7:02 am

    Optical “illusions” are interesting. The implications about what it means for perception are mind blowing. But I still don’t think that people fully really grok these implications. They tell us that the external world is to a certain degree a creation built from visual cues.

    Also I disagree that the “world is ultimately a subjective construction of our brains”. I think that it is our minds which are responsible, not our brains. And it is incorrect to think that what we are seeing is therefore not depicting reality accurately.

    A blog entry of mine might be of interest:

    http://ian-wardell.blogspot.co.uk/2011/03/are-perceptual-illusions-always.html

  36. BillyJoe7on 28 Feb 2015 at 7:26 am

    Your blog entry is wrong from beginning to end.

    From the beginning…
    You say that it’s the shadow cast over B that makes us think A and B are different colours. But how can that be? B actually looks lighter than A. How can a shadow make something look lighter than it actually is?

  37. Ian Wardellon 28 Feb 2015 at 7:27 am

    My above post wasn’t very clear. It is the optical illusions which tell us that the external world is to a certain degree a creation built from visual cues.

  38. BillyJoe7on 28 Feb 2015 at 7:37 am

    “They tell us that the external world is to a certain degree a creation built from visual cues”

    No, the external world is real. Our brains simply misperceive it at times. Natural selection does not select primarily for accuracy, but for survival. It is advantageous for survival to accentuate contrasts, so accuracy simply has to suffer.

    “Also I disagree that the “world is ultimately a subjective construction of our brains”. I think that it is our minds which are responsible, not our brains”

    Minds are what brains do.

    “And it is incorrect to think that what we are seeing is therefore not depicting reality accurately”

    Yeah, A and B are actually really and truely different just as we perceive them to be. Colour sensors be damned.

  39. BillyJoe7on 28 Feb 2015 at 7:45 am

    Febo,

    If colour sensors tell us that the actual colours in the photograph are pale blue and murky gold and you see it as blue and black or white and gold, then it is indeed an example of an optical illusion.

  40. Ian Wardellon 28 Feb 2015 at 7:46 am

    @Billyjoe.

    Billyjoe, you have said on numerous occasions that my blog entry here is wrong from beginning to end. Yet you never actually specify any errors.

    You say:
    “You say that it’s the shadow cast over B that makes us think A and B are different colours. But how can that be? B actually looks lighter than A”.

    I say how it can be in my blog entry.

    A shadow is cast over “B”. So the mind has to account for that by adjusting our visual qualia so that what we see is consistent with our prior experience of the world — in this case of seeing chess boards and the like. So the mind adjusts our colour quale and makes B lighter.

    I have no idea what you don’t understand…

  41. Ian Wardellon 28 Feb 2015 at 7:49 am

    I didn’t say the external world isn’t real. It’s real but it’s partially a construction by the mind. If you’re talking about a mind-independent reality, I don’t think anything can be said about such a reality.

  42. BillyJoe7on 28 Feb 2015 at 8:06 am

    The external world is real. Our brains construct an inaccurate representation of it.

  43. BillyJoe7on 28 Feb 2015 at 8:14 am

    You do not explain the illusion, you go on to imagine what may be the case with a 3D version of the illusion, but you never return to explain why B is perceived to be lighter than A in the 2D version.

    Btw here is a video of a 3D version, which demonstrates that what you said about the 3D version is also incorrect. Watch while the woman moves the squares…

    https://m.youtube.com/watch?v=z9Sen1HTu5o

  44. Pete Aon 28 Feb 2015 at 8:17 am

    BillyJoe7,

    The evolution and mechanisms of human colour perception are well explained in the book Vision and Art: The Biology of Seeing by Margaret Livingstone and David Hubel.

    Trichromacy in human vision provides colour differentiation, e.g. detecting the stages of ripening fruit[1], rather than colour accuracy. Interestingly, a small percentage of humans are innately good at colour accuracy and some others can learn to become quite good at the task by working with colour control processes.

    1. If you look at the CIE 1931 xy chromaticity space you will observe that yellow and cyan occupy very narrow regions. Particularly note the colour temperature scale: the camera white balance must be set to the illuminant colour temperature otherwise there will be a colour shift along the blue/amber axis.
    http://en.wikipedia.org/wiki/Chromaticity

    The photo under discussion is strongly shifted in the amber direction, this shifts blue towards cyan and dark grey towards bronze / muddy gold. Human vision is extremely sensitive to colour shifts on this axis — especially so in the above photo because the only reference point of known colour and brightness is the white page surrounding the photo.

    Dr Novella wrote: “The dress color debate is the result of an optical illusion. Don’t be ‘scared and confused,’ this is just how our brains work.” Naming it an optical illusion, and attempting to explain it as such, is perhaps likely to cause people to be “scared and confused”. Properly explaining in terms of the appaling photography enables people to realize that the photographer caused the illusion rather than there being something inherently faulty/illusory with their perception.

    You wrote “Of course, in a sense, everything we see is an optical illusion.”, which must be true because consciousness, the self, and free will are all illusions 🙂

  45. Ian Wardellon 28 Feb 2015 at 9:48 am

    @Billyjoe

    I explain everything fully in my blog entry. I then give further explanations on here on countless occasions. And you still don’t understand.

    About the video. Obviously we are being conned here! Get a chess board and pieces. Light up the king and cut out the light square which has a shadow cast across it and see if it is the same colour as a dark square on the edge of the board.

    But don’t trust a video where we don’t know if what we’re seeing hasn’t been set up so that everything isn’t quite as it seems. I think shadows have been painted on or something . .I don’t know. I hate conjuring tricks. And in this instance it’s obviously confusing people!

  46. Ian Wardellon 28 Feb 2015 at 9:52 am

    I mean take a photo of the chess board and cut out the square.

  47. tmac57on 28 Feb 2015 at 10:00 am

    On a tangential but related story, remember those ‘3d’ pictures that were so popular years ago (the ones that you had to stare at cross eyed to perceive) caused quite a stir? Some people just never could see them, while others saw them almost instantly. The thing that I noticed when I was finally able to readily see them, was that I always saw the image as inverted from what everyone (and I mean everyone) else saw them.
    For example: If it was an image of a dolphin, I would see it not projecting out from the picture, but as a depression, as though someone pressed a dolphin down into the sand and then removed it, or like a footprint in the sand. No matter how hard I have tried, I always see them that way, unlike the regular 2d images that can be viewed one way, and then the other (I can switch back and forth on those). Not sure if this means anything about my visual cortex, but it has always puzzled me. Hey, at least I can see them, some people drive themselves crazy trying, and even think that others are pulling their leg, like some have thought about this dress color thing. 🙂

  48. Ian Wardellon 28 Feb 2015 at 10:03 am

    Anyway in this video:

    https://www.youtube.com/watch?v=z9Sen1HTu5o&app=desktop

    I want that cylinder removed. If this is not trickery then all shadows should disappear and the 2 squares should look the same colour. I don’t think so!

  49. Mr Qwertyon 28 Feb 2015 at 10:05 am

    Just noticed that my comment from yesterday this time is in “Your comment is awaiting moderation.” state – is this a bug with the site or..? 🙁

  50. Ian Wardellon 28 Feb 2015 at 10:10 am

    I notice someone in the comments says:

    “Of course the “shaded” bit was never white, who would think that?”

    https://www.youtube.com/watch?v=z9Sen1HTu5o&app=desktop

    So Billyjoe, if this guy is correct then we are simply being conned . .well . .you are anyway!

  51. mumadaddon 28 Feb 2015 at 10:30 am

    Argument from random YouTube comment? Man alive, Ian!

  52. Ian Wardellon 28 Feb 2015 at 10:34 am

    Er mumadad, Billyjoe thinks that video destroys my argument here:

    http://ian-wardell.blogspot.co.uk/2011/03/are-perceptual-illusions-always.html

    But in this video of a 3D version of the board we are clearly being conned.

    https://www.youtube.com/watch?v=z9Sen1HTu5o&app=desktop

    Necessarily the “shadow” is no such thing and must be painted on!

  53. mumadaddon 28 Feb 2015 at 10:37 am

    Ian,

    Do you generalise your opinion about optical illusions or do you just think this one is dodgy?

  54. mumadaddon 28 Feb 2015 at 10:40 am

    What I mean is, there are plenty of optical illusions demonstrating effects similar to this and many other kinds of effect. What do you make of optical illusions in general? Or audio illusions such as the McGurk effect?

    https://www.youtube.com/watch?v=G-lN8vWm3m0

  55. mumadaddon 28 Feb 2015 at 10:45 am

    Ian,

    Let me see if I understand what your positions is (I haven’t got time to read your blog entry, but I think I read it before anyway) – you believe that reality is exists exactly as the mind perceives it?

  56. Bronze Dogon 28 Feb 2015 at 10:46 am

    Recently posted in my YouTube subscriptions: Captain Disillusion’s friend, Mr. Flare shows what happens if you go into Adobe Photoshop with the image.

  57. RickKon 28 Feb 2015 at 10:47 am

    Ian Wardell

    In the original 3D checkerboard video:

    https://m.youtube.com/watch?v=z9Sen1HTu5o

    If you stop the video at exactly 0:45 just as she starts to drag the square, then cover the whole screen except for a thin strip exposing only the original square, the target square and the square she’s dragging, you will clearly see they are all medium gray.

    The shadows, and the juxtaposition with the other squares, fool our eyes into thinking the medium gray is lighter in the shadowed area, and darker in the unshadowed area. Removing the cylinder would not be enough to fully expose that the squares are the same color – you would also have to move them away from the adjacent squares.

    Here is an even simpler example that shows how adjacent colors affect our perception.

    http://www.123opticalillusions.com/pages/gradient_illusion.jpg

    The gray bar across the middle is one consistent color. Only the background is a gradient.

    No trickery. We’re not being conned by anything other than by the wiring in our brains – you know, that organ from which the illusion of “mind” emanates.

  58. Ian Wardellon 28 Feb 2015 at 10:52 am

    @mumadadd

    That video is not a 3D version of the 2D checker board. In the 3D version *the shadow is painted on*!

    Billyjoe needs to make his own 3D version. If he does that he’ll see that when the square is moved it emphatically *will not* be the same colour.

    However! Take a photo of Billjoe’s checker board. Cut out the square with the shadow over it, place next to the other square. Then VOILA! They’ll be the same colour.

    Hence this fake 3D version does absolutely nothing to negate my arguments in that blog entry!

  59. steve12on 28 Feb 2015 at 10:56 am

    Willy:

    “steve12–My wife and I have different opinions when viewing the same screen. I continue to be dumbfounded.”

    Funny you should say, Willy, because my wife came home shortly after I posted that, and she immediately said black and blue!

  60. The Other John Mcon 28 Feb 2015 at 10:57 am

    Not really an “optical” illusion either, it’s a perceptual illusion not related to optics

  61. RickKon 28 Feb 2015 at 11:05 am

    Oh I see…. re-reading your comments, you seem to think that it is a “con” that the shadowed square wasn’t white to begin with. Bizarre, but ok.

    No, the shadowed square looks whiter because of how our visual perception interprets the “shadow”. That’s the illusion. All 3 squares are medium gray, but the fake shadow and the coloring of the adjacent squares fool our perception.

    But no matter how many people view the illusion and interpret the squares to be different colors, it doesn’t change the actual frequency of light reflected off the squares. Our perception does not influence the objective reality of the shapes and light.

    Regarding your blog post: If you define “color” as the impression a particular has of a given wavelength of light in a given circumstance, then the two squares are indeed different “colors”. So, if someone online asks you the color of your skin, your answer would be: “I’m blacker than a native of Chad in very low light to pale caucasian in full sun”. Is that how you describe the color of your skin?

    However, if color is defined as a particular RGB or CMYK mixture or a particular wavelength of reflected photons, then the color is completely unaffected by perception. Measurable objective reality exists outside of our perception, and much of science is concerned with recognizing and correcting for misperception.

    And because a reality exists regardless of how we perceive it, “misperception” is possible, and illusions are real.

  62. RickKon 28 Feb 2015 at 11:13 am

    That last line wasn’t clear. It should be:

    “And because a reality exists regardless of how we perceive it, “misperception” is possible, and optical illusions – differences between reality and how we perceive it – exist.”

  63. steve12on 28 Feb 2015 at 11:27 am

    Ian,

    Don’t get angry. Re-read RickK’s post – it shows you unequivocally that it is not a con. Your blog post is just wrong. No worries – you have no training in visual perception. I’m a vision scientist and I can’t tell you for sure why people see the dress so differently.

    IF you stop pretending you know everything, you might be able to learn something.

    To wit – you have actually brought up the very same blog post previously, and BJ7 showed you that you were wrong with the very same video:

    http://theness.com/neurologicablog/index.php/neurosurgeon-thinks-the-brain-doesnt-store-memories/

    Yo could have TESTED you conclusions then and readjusted them with new evidence. Instead, you close-mindedly trudged on with the same nonsense. You can do better.

  64. Ian Wardellon 28 Feb 2015 at 11:45 am

    RickKon 28 Feb 2015 at 11:05 am

    “Oh I see…. re-reading your comments, you seem to think that it is a “con” that the shadowed square wasn’t white to begin with. Bizarre, but ok.

    No, the shadowed square looks whiter because of how our visual perception interprets the “shadow”. That’s the illusion. All 3 squares are medium gray, but the fake shadow and the coloring of the adjacent squares fool our perception”.

    Rick, what exactly are you disagreeing with me about?? You agree the shadow is fake, so you agree with me and disagree with BillyJoe. This 3D version is not in fact a 3D version of the checker shadow illusion.

    {sighs} How often must I repeat myself . .

  65. Ian Wardellon 28 Feb 2015 at 11:48 am

    RickK
    “However, if color is defined as a particular RGB or CMYK mixture or a particular wavelength of reflected photons”.

    That is *NOT* how colour is defined. This is the mechanistic philosophers *redefining* colour.

    Read my essay. Scroll to part 3.

    http://ian-wardell.blogspot.co.uk/2015/01/science-afterlife-and-intelligentsia.html

  66. Ian Wardellon 28 Feb 2015 at 11:54 am

    steve12

    “Your blog post is just wrong”.

    *Where* are there any errors in my reasoning?

    I’ve already pointed out in many posts above that the supposed 3D version in that video is nothing of the sort.

  67. Ian Wardellon 28 Feb 2015 at 11:58 am

    “Don’t get angry. Re-read RickK’s post – it shows you unequivocally that it is not a con. Your blog post is just wrong. No worries – you have no training in visual perception. I’m a vision scientist and I can’t tell you for sure why people see the dress so differently”.

    A vision scientist?? {sighs} Can you wrap your head around the fact that this is a philosophical issue??

    Rick agrees it’s not a real shadow! If it’s not a real shadow then how on earth does it have any implications for the original 2D “illusion”?

    Does anyone on this blog understand *anything* . . .

    God’s sake . .

  68. Ian Wardellon 28 Feb 2015 at 12:03 pm

    Colour is defined by the characteristics of the qualia we experience. I’m not interested in scientists redefining words.

    We might have slightly different colour experiences in different light, but the mind adjusts for this to a large extent. Hence when an object is covered by a shadow, the object still looks more or less then same colour.

  69. mumadaddon 28 Feb 2015 at 12:20 pm

    So what’s the answer to my question, ian?

  70. RickKon 28 Feb 2015 at 12:24 pm

    Ian – yes, the shadow is faked to fool our perception. BillyJoe doesn’t say otherwise. But the faked shadow DOESN’T MATTER. You haven’t won a point by realizing the shadow is faked – nobody but you is focused on the nature of the shadow.

    The “shadow” is just there to help twist your perception away from the REALITY of the color of the squares.

    The point is this: regardless of what we perceive, the two squares are the same color. That’s the reality. By any measure they are the same.

    In the gray bar picture I linked above, tell me Ian – is the bar in the middle a sing solid color by your definition? Because it objectively IS an unchanging, solid color, however we perceive it.

    You can dismiss color measurement as “mechanistic”. You can dismiss any attempt by humans to objectively measure reality. While we happily use such measurements to communicate clearly to each other in reliable, repeatable ways free of subjective perceptions and differences in context, you can close your eyes, cover your ears and insist objective reality doesn’t exist.

    But your philosophy is no more real than the cylinder’s shadow. It’s just an illusion with which you’ve fooled yourself.

  71. steve12on 28 Feb 2015 at 12:25 pm

    Oh, Ian. Getting angry is not the answer.

    This would be a philisophical question – if it was 1543. It isn’t. If you’re that into philosophy, go read some Dennett and get with the times of your discipline.

    “*Where* are there any errors in my reasoning?”

    The post is so vague that it doesn’t really explain anything. Goes over a few trite things that are in fact true I suppose. But you do make a prediction re: the illusion not working in 3D, and it’s been empirically falsified.

    Sorry, you lose.

    Again, if you would simply let go of the idea that you already know everything, you could at least go back to the drawing board. But even when you’re so obviously proven wrong, you persist with your bullshit. It’s astonishing, really.

  72. steve12on 28 Feb 2015 at 12:26 pm

    “The point is this: regardless of what we perceive, the two squares are the same color. That’s the reality. By any measure they are the same.”

    Right. This is what makes the nature of the shadow immaterial.

  73. steve12on 28 Feb 2015 at 12:28 pm

    But I suppose Ian can “philosophize” (i.e., BS) this central truth away with a syllogism or bizarre “thought experiment”.

  74. Ian Wardellon 28 Feb 2015 at 12:30 pm

    Rick, in order to defeat my argument we need a 3D version of the checker board *without* any fake shadows or other trickery. If you move the square then then it *won’t* be the same colour.

    So the fact that people admit that trickery is involved means that it isn’t a 3D version of the 2D “illusion”. So that video does not remotely negate what I say here:
    http://ian-wardell.blogspot.co.uk/2011/03/are-perceptual-illusions-always.html

    And at this point I feel like a broken record.

  75. Paul Danger Kileon 28 Feb 2015 at 12:34 pm

    Steve,

    THIS IS NOT A COLOR CONSISTENCY OPTICAL ILLUSION.

    My daughter asked about this. I told her this: “To see the [color-consistency optical illusion], you would need two things to compare. We don’t have that here. We have one dress, viewed on non-color-calibrated monitors. We don’t know if the image was white-balanced, and neither does anyone else that’s arguing about this.”

    I then explained what white balanced is. After that, we used Photoshop to analyze the image. I will re-look-up those values again, and get back to you with them. After I do that, you will agree that Taylor Swift was correct, because: science.

    As for the blue: it’s also gray, but it does have more blue than red or green. It’s close enough to white, and blue, that either answer is correct.

    Be back soon,
    Paul

  76. steve12on 28 Feb 2015 at 12:36 pm

    “Rick, in order to defeat my argument we need a 3D version of the checker board *without* any fake shadows or other trickery. ”

    Trickery! Sorcery!

    Ian: percetpual illusions ARE trickery. That’s the whole point.

  77. steve12on 28 Feb 2015 at 12:38 pm

    Anyway…

    Has anyone seen any polls that look at the demographics of G/W vs. B/B perceivers? If the colors switched for everyone it wouldn’t be nearly as interesting.

  78. steve12on 28 Feb 2015 at 12:47 pm

    Buzzfeed alone has two polls saying opposite things:

    http://www.buzzfeed.com/jimwaterson/age-makes-you-black-and-blue

    http://www.buzzfeed.com/claudiakoerner/this-might-explain-why-that-dress-looks-blue-and-black-and-w#.mgka4zMp1

    Obviously we’re not dealing with very reliable data here!

  79. RickKon 28 Feb 2015 at 1:56 pm

    Ian,

    Your argument in your blog is this: “If this were a real 3D object and we were to approach it and view it from various angles, then we would see that squares A and B are very different colours. Indeed their intrinsic colours would be precisely as we perceive them in the illusion above.”

    Great – Ok, I get your point. In other words, if there was no illusion, then there would be no illusion. Brilliant.

    But, of course, you’ve completely missed the point. In the words of Andy Dufresne, “Is it deliberate?” You’ve actually made no argument, so there is nothing to argue against.

    The point is that in the 2D “checkerboard with cylinder” illusion, ON THE PAGE, the two squares look to be vastly different shades. But in reality, they are not. In trying to help us interpret the scene, our brains influence our perception to the point that we don’t perceive the reality – that squares A and B are the same color.

    Are the squares the same color? Yes. Do we see them as the same color? No. Therefore, optical illusion.

    In the video it is even more dramatic. The square she moves seems to actually change color. But if we limit our field of view to just the 2 squares, it clearly doesn’t change color.

    Does the square appear to change color? Yes. Does it actually change color? No. Therefore, optical illusion.

    The fact that some of our brains interpret the woman’s dress as one set of colors, and other brains interpret it as another set of colors, doesn’t mean that the dress actually flashes back and forth from one color to another depending on who is looking at it. The apple is still red, even with the light off – just as it is still present, even when we can’t see it.

    Reality exists and is what it is, regardless of how we perceive it. Remove all humans and all consciousness from the universe, and everything from quarks to galaxy clusters continue unchanged.

  80. skylanceron 28 Feb 2015 at 2:00 pm

    If I understand him correctly, I think that Ian is calling the 3-D version of the blocks a “con” because in a real-world scenario (i.e., where the shadow was not painted on in an attempt to create an illusion) that created the same visual perception, the two squares would not be the same color once placed under the same lighting.

    It is interesting that these illusions are often used as lessons in how we should not trust our perception, when (in this one at least) our perception is actually correct when determining the underlying conditions (shade of paint used on each square), at least when we’re not being deliberately manipulated!

  81. RickKon 28 Feb 2015 at 2:04 pm

    Oh, and Ian, what’s the reality of this image? Is the bar in the center of this image a uniform shade or does it have a gradient, by your non-mechanistic, mind-brain-dualist, mystically-undefinable-qualia definition of color?

    http://www.123opticalillusions.com/pages/gradient_illusion.jpg

  82. BillyJoe7on 28 Feb 2015 at 2:19 pm

    Ian,

    The Bottom Line:

    Your claims that the 3D version is “faked” is what actually makes it the 3D version of the 2D version.
    I explained this to you in a previous thread but you ignored it.
    Here it is:

    In the 2D version, the shadow is painted on – obviously, since there is no 3D block of wood.
    And, in the 3D version, the shadow is also painted on!
    That’s what makes it a 3D version of the 2D checkerboard illusion!

    So, if the 3D version is “faked”, then the 2D version is also “faked”.
    But “faked” is an odd word to use when you’re setting up a visual illusion.
    The whole point is to fool you into perceiving something that is not real!

    In both versions, the setup is manipulated (better word) so that A and B are the same colour but are perceived to be different colours.
    To repeat- that is the whole point of the exercise!

    The Real Explanation:

    In both versions, A looks darker than it really is because it is surrounded by light areas.
    And B looks lighter than it really is because it is surrounded by dark areas.
    The reason the shadow is painted on is to make A and B the same colour.

    Do you understand this?…
    The shadow is painted on to make A and B the same colour.
    Otherwise we don’t have the illusion we’re trying to create!
    Please tell me you understand this.

    However, the shadow is also painted on the darker areas surrounding B.
    Hence there is still a colour contrast between B and the areas surrounding B.
    As a result B appears lighter than it actually is – and therefore lighter than A, with which it is, in actual fact, identical.

    To repeat, that is the purpose of both versions.
    The whole purpose of this visual illusion is to demonstrate that two actually identical colours can be perceived to be different colours

    Of course this destroys your theory about reality and that really sucks doesn’t it?

  83. BillyJoe7on 28 Feb 2015 at 2:41 pm

    Rick,

    Yes, your link is a very clear example of the colour contrast illusion.
    The checkerboard illusion is just a visually more appealing and more complex example.

    Of course they faked the central bar by deliberately painting it the same colour along it’s whole length.

  84. BillyJoe7on 28 Feb 2015 at 2:50 pm

    And nice take down of Ian in your previous post. (:

  85. RickKon 28 Feb 2015 at 3:15 pm

    “Of course they faked the central bar by deliberately painting it the same colour along it’s whole length.”

    Exactly! The nefarious materialist!

  86. mumadaddon 28 Feb 2015 at 3:32 pm

    Ian has a load of things he wants to believe, which is affecting what he does believe. Start with a conclusion, work your way back, have blatant errors pointed out, double down. Reality conforms to my wishes!

  87. Paul Danger Kileon 28 Feb 2015 at 3:50 pm

    I hope that you will agree, after reading my explanation (http://dangerismymiddlename.com/archives/10129), that the main issue here is that the image was poorly toned and not white balanced.

    The original RGB values do put a gold tint to the black, and lighten the blue to make it look whitish. We can actually see that’s not just perceptual: it’s in the actual RGB values.

  88. tmac57on 28 Feb 2015 at 4:04 pm

    I went to the Penn and Teller show in Vegas, and I gotta tell you, those guys were totally faking magic and trying to trick the audience. Shameful!

  89. John Lechagoon 28 Feb 2015 at 4:06 pm

    No, not really how our brains work. You first have to get past what the camera is automatically doing before it gets to our brains. The camera is on auto white balance. With so much blue, it wants to balance it to white by adding an orange filter, and therefore makes the black seem gold. It is not really an optical illusion of the human eye in these photographs because the camera auto white balance is contributing so much to the illusion. It seems white and gold because the camera’s software is trying to make it white!

  90. BillyJoe7on 28 Feb 2015 at 9:56 pm

    tmac,

    “On a tangential but related story, remember those ’3d’ pictures that were so popular years ago (the ones that you had to stare at cross eyed to perceive) caused quite a stir? Some people just never could see them, while others saw them almost instantly. The thing that I noticed when I was finally able to readily see them, was that I always saw the image as inverted from what everyone (and I mean everyone) else saw them. For example: If it was an image of a dolphin, I would see it not projecting out from the picture, but as a depression”

    You’re not supposed to look cross-eyed at these pics, you’re supposed to focus on a point behind the pic.

    Here is an example:

    http://www.vision3d.com/sghidden/3ring.html

    If you focus behind the pic, three normal interlocking rings can be seen.
    If you focus in front of the pic, you get a sort of negative of that.
    I actually never had trouble with these and even now after all these years can bring up both these images almost instantaneously and alternate between the two.

    (On the other hand, my wife simply cannot do it. She ends up feeling nauseous and has to stop. If I tried to show her one now she would immedately feel sick and wave me away. Some optical illusions also freak her out. There is one where a dot disappears if you manage to get its image to fall on the optic nerve. Completely freaks her out. Doesn’t even want to discus what’s happening)

  91. BillyJoe7on 28 Feb 2015 at 10:28 pm

    Pete,

    Thanks for the information about accuracy and errors in taking photographs.

    Okay, so it’s overexposure and incorrect white balance that resulted in this photograph of a dress looking different from the real actual dress.
    But you contradict yourself about whether or not it is an optical illusion…

    “This is not an example of an optical illusion”
    “the photographer caused the illusion”

    I don’t think an optical illusion has to be deliberate to be an optical illusion.

    As you said, the actual measured colours in the photograph are pale blue and murky gold. So, if some people see it as white and gold and the rest see it as blue and black, then we have a genuine, if inadvertant, optical illusion.

    Anyway, minor point in the scheme of things.

  92. Ian Wardellon 01 Mar 2015 at 6:53 am

    “In the 2D version, the shadow is painted on – obviously, since there is no 3D block of wood”.

    Painted on?? At best it might be exaggerated in order to bring out the “illusion” more effectively! We can see the cylinder is lit from the other side. If the shadow is painted on are you saying there is no shadow at all?

    All this is a red herring of course and doesn’t address remotely the substance of my blog entry. You’re such a incorrigible clueless nincompoop.

  93. Steven Novellaon 01 Mar 2015 at 7:09 am

    To those saying this is not an optical illusion but essentially a technical error (white balance, monitor balance, photoshop, whatever), I think you are missing the point.

    You are simply explaining why the color is ambiguous and not accurate. That is besides the point, however. It is still an optical illusion in how the ambiguous colors are interpreted by different people.

    Sometimes differences in perception may be due to the monitor – but not always. I have looked at the image on a monitor directly with other people – same image, same monitor – and some see what and gold while others see black and blue.

    That difference is a color consistency illusion.

    Regarding the black and white squares – I agree this is not a color consistency illusion, and I apologize if I was not clear. It is a contrast illusion. The point of the example was to show that our brains adjust our perception in order to make real-world sense. In this sense it is a very similar illusion, but the exact correction our brains our making is different.

  94. Ian Wardellon 01 Mar 2015 at 7:12 am

    The point is we can’t say the 2D checkerboard is an illusion, otherwise *everything* we ever see is an illusion. In which case the word “illusion” loses all its meaning. In particular we cannot distinguish the illusion created by trickery in the 3D version from say a 3D version where the shadow is real.

    What we are immediately presented with is a 2D world. The 3D world is a creation from our minds. It’s an “illusion” in your terminology BillyJoe. Of course this is ridiculous labelling it an illusion.

    I wish people would read that blog entry a tad more carefully!

  95. Steven Novellaon 01 Mar 2015 at 7:25 am

    Ian – everything we see is an illusion. That is the point. Our visual perception is a multifaceted constructed illusion – always.

    However, most of the time this illusion is a sufficiently useful representation of external reality that we don’t notice. The world behaves as we think it should based upon what we see. Other people seem to see the same things we do.

    Sometimes, in every day life, the assumptions and heuristics that our brains use to construct this illusion don’t work, and we are temporarily fooled by what we think we see. We usually laugh these off as a temporary “trick of the light.” Unless, of course, the illusion has emotional meaning, then we think we saw Jesus, a ghost, or a UFO.

    Optical illusions are often constructed specifically to cause a breakdown in the visual heuristics our brains use. Often these are created by neuroscientists specifically to elucidate those heuristics or some aspect of visual processing. Sometimes they happen by accident, as is the case with the dress.

  96. BillyJoe7on 01 Mar 2015 at 7:59 am

    Ian,

    Well, it’s Sunday morning over there I guess, so you’ve probably had a bad night at the pub and are paying for it this morning. In any case, you’ve become so incoherent that it’s probably not worth responding.

    However, just so you get this when you sober up – a picture might have a representation of a light source, block of wood, and shadow, but it does not have those actual objects in them. Okay? And, yes, everything we “see” is an illusion, it’s just that some particular aspects of that illusion – like the checkerboard illusion and, yes, that dress – are good illustrations of that very fact.

    It’s also a good illustration of the fact that your philosophy is €r@p

  97. Ian Wardellon 01 Mar 2015 at 8:21 am

    Steven Novella
    “Ian – everything we see is an illusion. That is the point. Our visual perception is a multifaceted constructed illusion – always”.

    I agree with this. Well it’s just a fact isn’t it which it would not be possible to disagree with! We never perceive reality *as it is in itself*. Everything is an illusion.

    But there’s illusions and then there’s other illusions.

    BillyJoe

    “However, just so you get this when you sober up – a picture might have a representation of a light source, block of wood, and shadow, but it does not have those actual objects in them. Okay?”

    Forget the fact it is a picture. If we were seeing an actual checkerboard we likewise would only immediately see a 2D pic. But our minds construct a 3D image.

    To say the shadow in the picture is painted on is to spectacularly miss the point.

  98. Steven Novellaon 01 Mar 2015 at 9:00 am

    Regarding the 3D version of the checkerboard illusion, as I understand it this is a contrived demonstration. The problem is, as the paper were moved from the shadow to the light its apparent shade should also change, but it doesn’t. The demonstration, in other words, should not work.

    The 2D version works precisely because it is not caused by actual light and shadow, but is a static image of the representation of light and shadow. The darker area tricks our brains into interpreting it as if it were shadow, and so our brains makes the necessary adjustments as if it were a real 3D image. That is the illusion.

    In the real world 3D setup, it is possible that the shade of the shadowed light square is the same as the lighted dark square, if you set everything up precisely (I imagine that would be hard to do). But moving the paper from light to shadow should not work because it’s apparent shade would not remain consistent.

    None of this, BTW, has any implications for the deeper discussion about illusion.

  99. Pete Aon 01 Mar 2015 at 9:19 am

    BillyJoe,

    How we categorise things often depends on our experience/expertise. If something belongs to two categories, A and B, A causes B, and I have expertise in A, then I (perhaps wrongly) tend to categorise it and talk about it in terms of A rather than B. This is because A has far greater explanatory power than B.

    If we categorise the dress colours anomaly in terms of B (an optical illusion) we will fall foul of the illusion of explanatory depth[1]. Some of the commentators on this website fall foul of this illusion each time they try to explain real world phenomena using the category philosophy rather than the category science. I think that they make this error because they have much greater experience in outdated philosophy than they have in modern science (especially cognitive science).

    You wrote “I don’t think an optical illusion has to be deliberate to be an optical illusion.”, to which I totally agree. However, suppose I set my camera on auto exposure then take two pictures: the proverbial black cat in a coal cellar; a white cat in snow. Both images would render as a grey cat on a grey background. All half-decent photographers who viewed my images would immediately spot the exposure error and describe the images in terms of photography (category A); they would not describe the images in terms of the apparently grey cats being optical illusions (category B). Some/many non-photographers who viewed my images might come to the conclusion that I’d photographed the same grey cat against two different grey backgrounds — technically these would be optical illusions (category B), but it would not be particularly useful to describe them in terms of category B because doing so has a very shallow explanatory depth.

    I try hard to avoid making category errors aka category mistakes[2], but I still get in a muddle when something belongs to two or more categories. I enjoy only the scientific category aspects of this viral dress colour phenomena, but it’s easy to forget that the reason it’s gone viral is not because the general public enjoys rational explanations, it’s because most people love gossiping about controversies.

    1. Rozenblit, L., & Keil, F. (2002). The misunderstood limits of folk science: An illusion of explanatory depth. Cognitive Science, 26, 521-562.
    2. http://en.wikipedia.org/wiki/Category_mistake

  100. RickKon 01 Mar 2015 at 9:30 am

    “The point is we can’t say the 2D checkerboard is an illusion, otherwise *everything* we ever see is an illusion. ”

    *sigh*

    In this thread, the term “optical illusion” is clearly referring to cases where our perceptions are fooled into seeing something that is very different from a clear, measurable reality – so different that when the reality is pointed out to us, we’re surprised.

    The source or reality of the “shadow” in either the 2D or 3D image DOESN’T MATTER. The point of the illusion is not whether the images are fair representations of a real scene. All that matters is this:

    Are the squares the same color? Yes.
    Do we see them as the same color? No.
    Therefore, it is an optical illusion.

    Ditto for the video.

    There’s no reason to get all philosophical or mystical (or name-calling mad). This discussion is entirely about how our perceptual mechanisms can be fooled.

    You keep referring back to your blog post where you conclude: “But in that case what justifies us in labelling it as an illusion? If this were a real object that we are seeing, then squares A and B are very different colours. Our senses are not deceiving us. Indeed if someone claimed to see the squares as being precisely the same colour, then it is doubtful that he could proficiently visually apprehend his environment.”

    In other words, if this was a real scene, there would be no illusion. You’re right – if there was no illusion, there would be no illusion. Again, all you’re fixated on the reality of the scene as portrayed by the 2D image. That doesn’t matter! All that matters is:

    Are the squares the same color? Yes.
    Do we see them as the same color? No.
    Therefore, it is an optical illusion.

    Our senses ARE deceived – the two squares look to be different colors on the page (whether the scene is real or not), but they are not different. Period. Senses fooled!

    Oh, and as for BillyJoe7’s comment about the shadow being painted on – let me explain the joke to you. In the video image some of the shadow is faked, painted on, and some of the squares are not colored as your mind perceives them to be. This is done to make the illusion work. As Steve said: it is contrived – the demonstration shouldn’t work, but it does – hence the illusion. BillyJoe7’s point was that in the 2D image, of course the shadow is “painted on” – the whole scene is “painted on” by definition. It is a drawing! Get it?

    Please don’t think for a minute that BillyJoe7 is the one here who is missing the point.

  101. RickKon 01 Mar 2015 at 9:32 am

    Oops, the above is directed to Ian. Incomplete copy/paste error. Apologies.

  102. Steven Novellaon 01 Mar 2015 at 9:32 am

    Pete – Your example suggests you are making the same error as others before you, essentially missing what I am trying to explain. I am not trying to explain the apparent colors of the dress (which I agree is a technical photographic issue) but why different people will interpret the same picture (including on the same monitor) differently. That specific aspect of the dress color phenomenon is indeed an optical illusion, no matter what photographic phenomenon caused it.

  103. Ian Wardellon 01 Mar 2015 at 10:13 am

    RickK

    “The source or reality of the “shadow” in either the 2D or 3D image DOESN’T MATTER. The point of the illusion is not whether the images are fair representations of a real scene”.

    Rickk I explain everything in my blog entry. Neither you or BillyJoe have addressed my argument. Instead you both have gone off on a wholly irrelevant tangent.

    At his point there’s nothing more I can say. If you guys don’t understand, then so be it. I have tried my best to explain, but at some point one simply has to throw up ones hands and give up.

  104. Ian Wardellon 01 Mar 2015 at 10:51 am

    OK just one very last comment. Read what it says under Equivocation. What the guy says is basically correct.

    http://en.wikipedia.org/wiki/Talk%3AChecker_shadow_illusion

    He also says:

    “The equivocation is more easily fallen into if you have reductionist inclinations”.

    Well you guys are reductionists. Is this why you can’t understand?? :-O

  105. BillyJoe7on 01 Mar 2015 at 11:10 am

    Ian,

    Now you are simply being disingenuous. Your entire post on your blog has been drawn and quartered. We are now beating a dead horse. What you need to do to revive it is to go through Rick’s last post line by line and explain where he goes wrong. That would be impossible because every line is irrefutably true. You know that of course which is why you are backing out of the discussion.

  106. Ian Wardellon 01 Mar 2015 at 11:13 am

    Rick said:

    “Are the squares the same color? Yes.
    Do we see them as the same color? No.
    Therefore, it is an optical illusion”.

    The first assertion is incorrect. What we *see* is the colour, not what we measure.

  107. BillyJoe7on 01 Mar 2015 at 11:22 am

    Steven,

    ” The darker area tricks our brains into interpreting it as if it were shadow, and so our brains makes the necessary adjustments as if it were a real 3D image. That is the illusion.”

    I still think you have the wrong explanation for the checkerboard illusion.
    It is simply a colour contrast illusion.
    It would work equally well it we simply had two identically coloured discs with one surrounded by a lighter colour and the other surrounded by a darker colour (colour = shade of grey).
    Or see the “gradient illusion” illustrated in the link provided by Rick:

    http://www.123opticalillusions.com/pages/gradient_illusion.jpg

  108. steve12on 01 Mar 2015 at 11:25 am

    “The first assertion is incorrect. What we *see* is the colour, not what we measure.”

    What does that mean?

  109. BillyJoe7on 01 Mar 2015 at 11:29 am

    Ian,

    “The first assertion is incorrect. What we *see* is the colour, not what we measure.”

    (:

    Andy Dufresne: How can you be so obtuse?
    Warden Samuel Norton: What? What did you call me?
    Andy Dufresne: Obtuse. Is it deliberate?

    Oh dear.

  110. Pete Aon 01 Mar 2015 at 11:31 am

    Dr Novella,

    You are probably correct in claiming that those who do not see the dress colours as they are rendered in the image are experiencing something that belongs to the very broad category of what is commonly referred to as “optical illusions”.

    Unfortunately, you haven’t (yet) provided a scientific explanation of why there are at least three very different visual perceptions of the image: the false colours pale blue and bronze resulting from the appalling photography; blue and black; white and gold. You have instead provided examples of very different types of optical illusions, which in themselves are intriguing, but they do nothing to explain this particular instance of an optical illusion.

    Attempting to explain the various perceptions of the image in terms of optical illusions is going to fail at some point, if it hasn’t already. Cognitive science alone can provide only a shallow explanation. A much deeper explanation of how we perceive and misperceive colour is provided in the book that I’ve already mentioned — Vision and Art: The Biology of Seeing by Margaret Livingstone and David Hubel. This book not only explains the biology of seeing and optical illusions, it goes on to explain in scientific terms some famously powerful works of art.

    Art is not science, but art produced without a basic understanding of the relevant science is, at best, haphazard. Likewise, attempting to explain an instance of an optical illusion without having a solid foundation in art is, at best, haphazard 🙂

  111. Ian Wardellon 01 Mar 2015 at 11:37 am

    That Andrew van der Merwe guy who explains under the “Equivocation” heading (http://en.wikipedia.org/wiki/Talk%3AChecker_shadow_illusion) is getting the same complete irrelevant responses as myself!

    Personally what I find more interesting than “illusions” is the fact that some people are utterly unable to comprehend relatively simple stuff!

  112. steve12on 01 Mar 2015 at 11:47 am

    Re: the nature of the gradient illusion…

    I think you’re right BJ7, there is a much more basic process here.

    I think the gradient illusion emanates form the premium given to detecting edges right from the earliest processing. Even before leaving the eye, lateral inhibition will adjust the gain of adjacent portions of the visual field, allowing accentuation of differences on either side of an edge (and for other reasons).

    This local accentuation takes place on both ends of the bar. But because the portion of the bar where neighbors are inhibiting is of different level of darkness, the lateral neural gating is not equal on the left and right ends of the bar, regardless of the bar’s uniformity. So, this local processing gives us robust edges at the cost of the accuracy of the overall contrast. There is a global vs. local processing conflict, and the illusion is because those local processes “win”.

    There are surely other factors at work here, but very early lateral inhibition can explain much of the phenomena.

  113. RickKon 01 Mar 2015 at 11:54 am

    Ian said: “The first assertion is incorrect. What we *see* is the colour, not what we measure.”

    Baloney

    Draw a wide stripe of the same color as square A and square B connecting the two squares, and we SEE the paradox revealed – we SEE that they look different but ARE the same at the same time. BOTH views – (1) that the squares are the same color and (2) that they are different – are perspectives we can “see”.

    Just as with the simple gradient illusion. Why don’t you address that one?

    The “Equivocation” link you posted makes exactly the same distracting and silly argument you make. Like you, he immediately starts suggesting ways to manipulate the shadow in the image – in other words, changing the image will reveal “voila!” that our eyes were right all along. He misses the point just as you do. And as one commenter said in the discussion: this is about the picture, not about the objects represented by the picture – “there are no objects here”.

    So, in the gradient “optical illusion” – tell us, using your perception-is-reality definition of color – is the center bar a uniform color or does it change color from one end to the other?

  114. tmac57on 01 Mar 2015 at 12:00 pm

    BJ7- Thanks for the tip. That explains why I see them differently. I was interested to read that some of the stereograms were specifically designed for cross-eyed viewing. I haven’t gotten the parallel method to work for me yet, but I’ll keep at it. Hard to not fall back into the cross-eyed method though.

  115. The Other John Mcon 01 Mar 2015 at 12:02 pm

    Tmac – some people can’t cross their eyes, some can’t diverge them, some can do neither, others both! Don’t drive yourself too crazy and please don’t hurt your eyes trying too hard 🙂

  116. tmac57on 01 Mar 2015 at 12:05 pm

    I guess Ian would describe this as a giant foot about to crush some tiny people then, instead of an example of forced perspective:

    http://media02.hongkiat.com/force-perspective-photos/crushed_foot.jpg

  117. tmac57on 01 Mar 2015 at 12:07 pm

    John Mc- I know! They could get stuck that way…at least that’s what my mom told me 😉

  118. The Other John Mcon 01 Mar 2015 at 12:09 pm

    Lemme modify Rick’s basically correct statements…

    Are the squares physically the same color as measured by wavelength spectrum and/or RGB triplets? Yes.

    Are the squares mentally perceived as the same hue/color? No.

    This is therefore a perceptual illusion. Perception is incongruent with physical reality.

  119. steve12on 01 Mar 2015 at 12:13 pm

    I just read the Equivocation link…

    “At no point in your rambling, incoherent response were you even close to anything that could be considered a rational thought. Everyone in this room is now dumber for having listened to it. I award you no points, and may God have mercy on your soul.”

    I do like the anti-science rant. Nice touch.

    Another person that is unwilling to do any of the WORK involved in understanding science, but wants to authoritatively discuss science nonetheless.

    It doesn’t work that way.

  120. RickKon 01 Mar 2015 at 1:18 pm

    Other John,

    I already tried that. Ian rejects RGB or CMYK measurement as reductionist. What you see is reality, not what you measure. We’re just too immersed in our materialist world view to realize that the universe is only what we perceive, not anything objectively measurable.

  121. ccbowerson 01 Mar 2015 at 2:32 pm

    When discussing how people interpret the dress, bringing up color measurements with software does not help in evaluating people’s response to the question of “what color is the dress?” We have to distinguish the question of what we actually see in the image, with what we interpret the image to mean for the color of the actual dress itself.

    When I say that I see a white and gold dress, I do not mean that in the actual image of the dress is white and the actual stripes are gold. I mean that it looks like a white and gold dress was used to create the image I see. The actual colors (in the image itself) look like a blueish light gray (interpreted as white) and a brownish tan (gold). Because it appears to be in the shade to some extent (due to the overexposed background), I assume some darkness of the image of the dress results in those shifts in color in the actual image.

    I’m not sure how black and blue come from that image… perhaps those people are correcting the image darker due to the brightness of the background (i.e., that the black and blue dress is lightened in the image by the brightness perhaps)

  122. BillyJoe7on 01 Mar 2015 at 3:52 pm

    Ian,

    “Personally what I find more interesting than “illusions” is the fact that some people are utterly unable to comprehend relatively simple stuff!”

    Yeah, like that a 2D picture is not actually a 3D object but a representation of a 3D object.
    That guy wants to remove the cylinder and hence the shadow from the 2D picture. The idiot doesn’t understand that there is no actual cylinder and no actual shadow in that 2D picture. And he doesn’t understand that if you change the picture so as to get rid of the illusion, then of course there’s not going to be an illusion anymore. I mean, how stupid can you get?

    Anyway, the checkerboard illusion is proving much too complex and difficult for you to understand, Ian, and nothing we can say now is going to improve the situation for you. You are going to remain ignorant about it and there’s nothing we or you are going to able to do at this point to change that fact. So, perhaps you should just keep to the simple stuff like that gradient bar where there are no representations of cylinders and shadows to confuse you.

    Anyway, kudos for finally getting up the courage to discuss this instead of ignoring it as on all our previous occasions when it has come up. You took a big risk. You lost. Sorry.

  123. The Other John Mcon 01 Mar 2015 at 3:59 pm

    Rick — I know it’s hopeless with Ian, I just liked what you wrote and wanted to make some minor tweaks for (possible) clarity…at least in my head

  124. Robneyon 01 Mar 2015 at 8:12 pm

    Ian,

    when I cross my eyes am I creating two separate realities?

  125. RCon 02 Mar 2015 at 10:52 am

    It’s amazing to me how many people don’t seem to get that this isn’t just a white balance issue (or who aren’t willing to read the article or talk to actual people).

    Of course the white balance is screwy – but that’s not interesting or important. What’s interesting is that two people can look at an image on the same monitor, at the same time, from basically the same viewing image, and one can see a royal blue dress with black stripes, and the other sees a bluish-white dress with gold stripes.

    There’s a significant difference in image processing going on here – it’s not monitors or white balance or different gifs – none of those are important or interesting here.

  126. Pete Aon 03 Mar 2015 at 8:53 am

    RC — I hope you find the following explanation slightly useful…

    If someone were to write an ambiguous statement then, by definition, we would expect different people to interpret it differently. We wouldn’t be surprised by the differences and we wouldn’t try to explain the differences in terms of an illusion — it’s simply an ambiguity that cannot be resolved without gaining further information about what the author was trying to convey to the readers.

    The dress photo is highly ambiguous in terms of both the lighting and the dress colours. The ambiguity just happens to be very close to a decision threshold used by our visual processing, at a subconscious level, that attempts to match what we see to our prior knowledge of the real world:

    1. If our visual processing decides that the dress is illuminated by, say, the light from only the sky on a clear sunny day (cold light) then we will see the dress as white & gold (or similar).

    2. If our visual processing decides that the dress is illuminated by, say, an overhead tungsten lamp (warm light) then we will see the dress as being (shades of) blue & black.

    There simply isn’t enough information in the photo itself to have caused the vast majority of people to see it as either scenario 1 or scenario 2. In order to understand why this has happened we need to delve into some of the underlying objective and subjective technicalities of human vision.

    The relevant part of our visual processing that attempts to match what we see to our prior knowledge of the real world is called chromatic adaptation or colour constancy. Its required equivalent mechanism in photography is called white balance; auto white balance is the name given to algorithms that emulate chromatic adaptation / colour constancy. NB: I’m talking about the white balance used to create the digital image, not about the white balance of the monitor used to view the image — these are very different things.

    The purpose of colour constancy (in both human vision and photography) is to compensate for scene illumination that is shifted away from the circa 5600 K spectrum of overhead sunlight with clear skies. The two axes in which colour shifts occur are blue/amber and green/magenta — particularly note that each axis has two complementary colours.

    The dress photo contains no obvious visual clues, such as skin tones, to guide our chromatic adaptation in the right direction. The white balance of the photo is strongly shifted towards amber plus it is overexposed: this causes the royal blue to render as light blue/cyan and the black to render as a dark amber (bronze, muddy gold, etc.). If the white balance of the photo was a little further towards amber then more people would see the dress as white & gold. Conversely, if the white balance was further towards neutral then more people would see the dress as blue & black. Similarly, if there was more overexposure in the photo then more people would see white & gold; with less overexposure, more people would see blue & black.

    The monitor screen used and its viewing environment will have an effect on each individual’s visual processing decision, but once each of us has made our initial decision it becomes permanently biased due to memory. E.g. if we initially saw the dress as white & gold it becomes almost impossible to view the image again and see it as blue & black because humans know that most objects in the real world have consistent colours and textures, even though the lighting changes dramatically throughout the day and the times of year. Dr Novella has perhaps well illustrated the memory effect in his statement “I see black and blue, no matter what screen or version of that picture I look at. It does not seem to be an issue with the monitor or viewing conditions.” Indeed, this is precisely why we use double-blind randomized controlled trials to accumulate robust scientific evidence rather than relying on a plethora of testimonials. (This is in no way intended to be an insult to Dr Novella).

    The reason why this photo has caused so much debate is because this is a rare instance of having just the right combination of abysmal photographic errors to cause maximal ambiguity in the perceived colours of an object: framing, white balance, and over exposure. It really isn’t an optical illusion, let’s all be totally honest: it’s a wonderful reminder that it’s far better to delete abjectly crap photos than to publish them online to the whole world!

  127. tmac57on 03 Mar 2015 at 9:23 am

    Pete A- I like the technical breakdown that you gave, but I think that you are being pedantic regarding the idea of an optical illusion. Indeed, the basic definition given by Wikipedia seems to support the idea that this is one of the classic examples of an optical illusion:

    An optical illusion (also called a visual illusion) is characterized by visually perceived images that differ from objective reality. The information gathered by the eye is processed in the brain to give a perception that does not tally with a physical measurement of the stimulus source. There are three main types: literal optical illusions that create images that are different from the objects that make them, physiological illusions that are the effects of excessive stimulation of a specific type (brightness, colour, size, position, tilt, movement), and cognitive illusions, the result of unconscious inferences. Pathological visual illusions arise from a pathological exaggeration in physiological visual perception mechanisms causing the aforementioned types of illusions.

  128. Pete Aon 03 Mar 2015 at 12:59 pm

    tmac57 — I’ve written all of my comments in full awareness of the excellent Wikipedia pages and many other superb definitions and examples of optical illusions. I detest myself, and everyone else, for being overly pedantic, but regarding the title of this article “What Color Is This Dress? It’s An Optical Illusion” I feel compelled to rebut what has been stated using the very few areas of art and applied science in which I have at least a reasonable level of proficiency.

    I have a deeply personal interest in this viral controversy because I specialise in painstakingly creating two types of art: that which is as unambiguous as possible in order to most accurately convey the subjective experience of actually ‘being there’; and just for fun sometimes, producing the quizzically bizarre!

    Well, I have to concede that the photographer who published the image that has resulted in this world wide debate has certainly produced “the quizzically bizarre”, but I think it’s fair to say that the credit goes to the imprecise algorithms in their camera rather than to the artistic skills of its user 🙂

    I try my very best to write my occasional comments on this website using evidence and/or science in order to fully support Dr Novella and the commentators who promote science and critical thinking. I sincerely thank everyone who has replied to my comments because I learn nothing from being right, almost everything of real value that I’ve learnt during my life has been from being wrong.

  129. Steven Novellaon 03 Mar 2015 at 1:48 pm

    Pete – what you described is the very definition of a type of optical illusion called an ambiguous image. In these optical illusions the brain may shift back and forth between different interpretations of the image based upon different processing assumptions. This is simply an ambiguous color constancy illusion.

  130. steve12on 03 Mar 2015 at 2:50 pm

    Pete A:

    What if I decided to use all those same camera settings because I WANTED to make a color constancy illusion? Would it be an illusion then?

    That’s essentially what vision scientists are doing when they make an illusion. They’re not trying to find a naturalistic situation that just happens to reveal something about our visual system. They’re purposefully stacking the deck. This guy/gal just happened to achieve this unwittingly.

    And a lot of ambiguous images you can switch – it’s not true that the way you see it once is how it always is. I have seen the dress as black and blue for a few fleeting moments now.

    http://en.wikipedia.org/wiki/Ambiguous_image

  131. tmac57on 03 Mar 2015 at 4:24 pm

    I have a question for those who can readily see the dress both ways, since I have not experienced this for myself:
    When it switches, is there a noticeable difference like when the Spinning Girl illusion clearly and unambiguously switches direction? Or is it some subtle change, where you go “Oh yeah, I can kinda see that as black and blue…sort of”?
    I ask, because some people have given me the impression that they see the black and blue as though it were as obvious as the photos where it is clearly a dark royal blue and black, and I just can’t wrap my brain around that.

  132. steve12on 03 Mar 2015 at 4:32 pm

    tmac-

    I dramatically saw it as black and blue (I almost always see it white and gold) for a fleeting moments three times. All three it was when I first looked at it, and for all three it reverted back immediately.

    I stared trying to switch it to no avail.

  133. Steven Novellaon 03 Mar 2015 at 4:33 pm

    I always see it as clearly black and blue. It never switches for me either.

    I guess some people’s brains are more solidly on one side or the other of the divide, with some are so close to the middle they can go either way.

    I was at a lecture on optical processing and the lecturer noted that for one illusion the equilibrium point is slightly different for everyone (this was a cool illusion, adjusting the shades of red and green dots until they became identical, at which point the 3D image collapsed to a 2D image because the 3D effect was dependent on black and white vision and the colored dots were the same shade in black and white). So she adjusted the shading until it hit the equilibrium point for each person. The same could theoretically be done for this illusion.

  134. tmac57on 03 Mar 2015 at 5:12 pm

    Well, that is really fascinating! I guess that is what has made this thing really blow up. The differences must be so stark that each side can’t imagine how the other could see it so differently.

  135. Pete Aon 03 Mar 2015 at 5:30 pm

    Dr Novella and steve12 — thanks for your replies. Yes indeed, instances where the brain may/does switch back and forth between two different interpretations is a profoundly fascinating example of [what I call] an optical illusion. My favourite is the contrived ambiguous silhouette of the rotating dancer. I’ve seen it many times and it still irritates me immensely that I haven’t yet mastered switching my perceived direction under my conscious volition! This is why I love studying cognitive neuroscience.

    Dr Novella, with the greatest respect to you, I’m fully aware that the brain switching back and forth is correctly classified as an optical illusion, but you stated “I see black and blue, no matter what screen or version of that picture I look at.” therefore I have to logically conclude that you and the many others who’s interpretation is constant cannot claim to be witnessing an optical illusion in this particular instance of the dress colour anomaly. You can claim only that your interpretation of the ambiguous information in the image agrees with many and disagrees with many others. I’m still failing to understand why different, yet largely consistent, interpretations of highly ambiguous information needs to be called an illusion experienced by the recipient of the information rather than simply being called out for the abject garbage that is being presented to the recipients.

    By all means use science to show me where I’m mistaken in what I’ve written thus far [which I’ve tried to back using reasonable sources of evidence]. Sorry, but quoting science articles that I’ve previously read, understood, and agree with, isn’t helping me to see my errors.

    @tmac57 — Since editing the image to mostly correct its huge errors, I can no longer see it as I initially perceived it. However, I’m still totally unable to see the original as royal blue & black. I’m not yet convinced that those who claim to see it as blue & black actually see it as royal blue & jet black unless they’ve seen an image of it portrayed as such.

  136. RickKon 03 Mar 2015 at 5:41 pm

    Pete A,

    The RGB values of the image are off-white with amber/brown stripes.

    Is the dress image blue with black stripes? No
    Do many people see the dress as blue and black? Yes

    Therefore, it is an optical illusion.

  137. Pete Aon 03 Mar 2015 at 6:05 pm

    RickK, if you are correct (and you likely are correct) then please explain the optical illusion to me in scientific terms that I can understand rather than using what is illogical to me. E.g.:

    Has RickK properly explained the optical illusion? No.
    Do many people think that the image is an optical illusion? Yes.
    Therefore, it is an optical illusion.

  138. RickKon 03 Mar 2015 at 9:21 pm

    Pete,

    The objective reality of the image is unambiguously a dress of bluish white and amber brown, as measured by color sampling.

    http://www.wired.com/wp-content/uploads/2015/02/Unknown.png

    Yet due to features of the photograph that have been discussed ad nauseum, many see it as royal blue and black. For those who see it as blue/black (I wish I could), it is an optical illusion – no different than the shaded squares illusion.

    It has nothing to do with the camera or the settings or the exposure or the skill of the photographer or the artistic merit of the photograph. How the illusion came to be doesn’t matter. All that matters is its ability to fool our perception and make us see what isn’t so. It is an optical illusion and there’s no basis (or purpose) for claiming it isn’t.

  139. BillyJoe7on 04 Mar 2015 at 5:17 am

    Rick,

    That’s about the fifth offer.
    I don’t think he’s buying.

    It seems so simple doesn’t it:

    Objectively measured colour is X.
    Observed colour is Y.
    Therefore optical illusion.

    But the photography lesson was interesting.

  140. Pete Aon 04 Mar 2015 at 6:38 am

    Ok, if objectively measured colour is X and observed colour is Y then it’s an optical illusion. The problem I have with this definition is the highly significant number of people who have colour vision deficiency must therefore be experiencing frequent or continuous optical illusions. We don’t categorise these perceived colour differences as optical illusions, we categorise them as colour vision deficiency aka colour blindness. Why? Because the latter category has far greater explanatory power.
    http://en.wikipedia.org/wiki/Color_blindness

    In very low light we use only our rods, we don’t see colours. The light crossover threshold varies greatly between individuals: at certain light levels some will see colours others will see only monochrome. Again, this isn’t categorised as an optical illusion.

    Obviously, the dress image induces optical illusions; as do countless other images that have huge errors in exposure and white balance. Carry on calling them optical illusions, but I prefer to call them crap photos. A photographer can’t learn to improve by being told that they’ve produced an optical illusion 🙂

  141. BillyJoe7on 04 Mar 2015 at 6:59 am

    Flogging the dead horse (:

    I would have thought that a reasonable assumption is normal vision. It never entered my mind that I would have to spell that out. If you are blind you won’t see any optical illusions. That doesn’t mean they don’t exist.

  142. Steven Novellaon 04 Mar 2015 at 9:22 am

    Pete – this has been covered, but since you addressed me:

    This is an ambiguous stimuli type of optical illusion, specifically dealing with color constancy processing.

    Some people always see black and blue. Some always see gold and white. Some shift back and forth. This seems to depend on individual “calibration” of vision processing. My calibration is apparently shifted to the blue-black result. Others are calibrated close to the breaking point, and their brains will shift from one set of assumptions to another, causing the apparent colors to change.

    Partly you are simply making a category mistake. You are explaining why the colors are not being represented accurately based upon the underlying photographer. That is valid, if that is your question. I am explaining why different people perceive those apparent colors differently. That is also valid in the context of perception.

    One other point that might help – you seem to be struggling with the demarcation line of what constitutes an “optical illusion.” A better way to understand it is this – all perception is a constructed illusion of the brain. There is no demarcation line. The only variable is how aware we are of the fact of this constructed perception.

    What we call optical illusions occur when there is a demonstrable difference between some objective measure of reality and our constructed perception of that reality. There is nothing fundamentally different about what is happening in the brain. The only difference is that the phenomenon can be made apparent.

    If everyone’s brains were calibrated in the same direction for this image, no one would have noticed or cared that the actual color of the dress differed from what people perceived. That difference would best be explained through both photography and our visual system, but this story would not have gone viral. What freaked everyone out was this – looking at the picture and seeing one pair of colors, then showing the picture to a friend who claimed to see a different pair of colors. They could not reconcile their different perceptions, and everyone freaked.

  143. Pete Aon 04 Mar 2015 at 10:50 am

    Steven — Thank you very much for taking the time to write your very clear explanation, it makes perfect sense to me. I sometimes completely fail to understand something because I can’t figure out what questions I need to ask.

    You were quite right about me struggling with the demarcation line of what constitutes an “optical illusion”, the more I thought about it the more confusing it seemed to be. I knew that all perception is a constructed illusion of the brain, but I was missing an essential piece of information, which you provided: “The only variable is how aware we are of the fact of this constructed perception.” That profound statement has explained several things that have been puzzling me for a very long time.

    The other essential thing I’d failed to properly appreciate was: “They could not reconcile their different perceptions, and everyone freaked.” I’m so familiar with the fact that people have vastly different perceptions of many things, especially sound and vision, that I naively assumed most people were familiar with it.

    I apologise for being a pain with all my comments and misunderstandings. Many thanks again for your time and patience.

  144. tmac57on 04 Mar 2015 at 11:38 am

    I’m wondering, can we understand this as a form of bias, in a similar way that people have other cognitive biases? Or would it be more of a metaphorical comparison? If there is any crossover in the two phenomena, that might open another door of investigation.

  145. RCon 09 Mar 2015 at 11:08 am

    @Rick

    “The objective reality of the image is unambiguously a dress of bluish white and amber brown, as measured by color sampling.
    http://www.wired.com/wp-content/uploads/2015/02/Unknown.png

    This link as a reference is interesting to me, as those RGB numbers most certainly aren’t bluish white and amber brown.

    Look at the colors without looking at the dress itself.

    The three colors that are pointed to as bluish white are:
    #838CC3 – Slightly desaturated blue -http://www.colorhexa.com/838cc3

    #8794BE -Slightly desaturated blue – http://www.colorhexa.com/8794be

    #656581 – Dark grayish blue. – http://www.colorhexa.com/656581

    I do agree that when looking at the RGB the black runs to brown a bit, but that’s pretty typical in overexposed images, or images in the wrong white balance – but there’s nothing even approaching white on that dress within the RGB numbers.

  146. Pete Aon 09 Mar 2015 at 12:04 pm

    @RC — I hope that you and some other readers will find Steven Pinker’s take on this phenomenon illuminating:
    http://www.forbes.com/sites/matthewherper/2015/02/28/psychologist-and-author-stephen-pinker-explains-thedress/

  147. RCon 09 Mar 2015 at 2:24 pm

    @Pete A

    I understand the issue, and what’s going on. I understand the issue of color constancy here. I don’t need to be illuminated.

    What I take umbrage with is RickK insisting that the RGB numbered colors are white and gold, when they clearly aren’t. Look at the image he posted, and the colorhexa diagrams of the colors present – they’re not anywhere near RGB white (#ffffff) – They’re slightly under-saturated blue.

    Slightly under-saturated blue is not “unequivocably” white. Greyish blue? Sure.

    Here’s one of the colors he describe as “Amber-brown”

    http://www.colorhexa.com/47352b
    #47352b color description : Very dark desaturated orange – in a CMYK color space, it is composed of 0% cyan, 25.4% magenta, 39.4% yellow and 72.2% black.

    There’s nothing ‘unequivocably’ white, or “amber-brown” in this image.

  148. Pete Aon 10 Mar 2015 at 9:19 am

    @RC — Thanks, I’ve just used ColorHexa on my measures values stated in my first comment…

    A light area of the blue part of the dress has RGB 180 194 230
    http://www.colorhexa.com/b4c2e6
    Color description: Very soft blue.
    CMYK color space: 21.7% cyan, 15.7% magenta, 0% yellow, 9.8% black.

    A light area of the black part of the dress has RGB 129 113 77
    http://www.colorhexa.com/81714d
    Color description: Mostly desaturated dark orange.
    CMYK color space: 0% cyan, 12.4% magenta, 40.3% yellow, 49.4% black.

    Even the “very soft blue” areas are a long way off being white.

  149. RickKon 10 Mar 2015 at 6:10 pm

    RC and Pete A

    Thanks. You’re right of course – I should not have said off-white even though that’s how my processing interprets the color of the jacket in the picture. So, those who see it as off-white are just as susceptible to the optical illusion as those who see royal blue and black.

    Thanks for even more confirmation of my point that this photograph, intentionally or unintentionally, creates an optical illusion for many if not most viewers. Much appreciated.

  150. Pete Aon 10 Mar 2015 at 8:18 pm

    RickK — You are more than welcome to challenge the objectivity of the dress colours. However, if you wish to argue about the subjectivity of the perceived colours then it is only you who owns the burden of proof for every argument that you present. Otherwise, you become indistinguishable from anti-science commentators such as Ian Wardell.

  151. Pete Aon 20 May 2015 at 9:02 am

    The Wikipedia page on this topic has been frequently updated. Some readers may be interested to know that:

    “The Journal of Vision, a scientific journal about vision research, announced on 12 March 2015 that a special issue about ‘The Dress’ will be published with the title ‘A Dress Rehearsal for Vision Science’. The submission deadline is 1 July 2016, and scientific work has been ongoing.[33]

    A 2015 study by Bevil Conway et al. where 1,400 people were asked how they saw the dress, found that 57% said they saw blue and black, 30% said white and gold and about 10% said they saw the dress as blue and brown. About 10% reported they had seen the dress switch colour. Relatively more women and older people saw the dress as white and gold. The scientist further found that if the dress was shown in artificial lighting almost all respondents saw the dress as white and black, while they saw it as white and gold if the background had blue lighting.[34]”. Retrieved 2015-05-20 from
    http://en.wikipedia.org/wiki/The_dress_(viral_phenomenon)

    The Conway et al. study, and two other studies, are mentioned here:
    http://blogs.scientificamerican.com/illusion-chasers/the-current-biology-of-the-dress/

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.