The Dress (color perception)

[Carol the Dabbler]

This is four or five years old, but I just now ran across it -- what are the colors is this dress:

[source]

The reason I'm asking is that some people see it as one color scheme, while other people see it as a very different scheme, and a few people see it as a third (or as other colors entirely).  I've just looked at the photo on two monitors that are set for significantly different color renderings, and I see it the same regardless, so I don't think it's the monitor that influences my perception.

I'm not gonna say yet how I see it, because I want to know how y'all see it.  I was thinking of making this a poll, but then I didn't want to influence your range of choices.  I urge you to form your opinion and word your description of the colors before you scroll down and read what others are saying.  Please note that there are no wrong answers, as long as you're telling us your honest perception.

Thank you!

[Hikari]

 Granted, I’m doing this on a small screen in the dark but I see brownish bands interspersed with thinner white  bands.  I guess I would call the brownish bands “bronze”. This is what I see.

[Carol the Dabbler]

Not sure what you mean by "small," but if you consider the smallness to be a problem, and you're on your phone, can you "stretch" that part of the screen?  Or if you're on a tablet or small laptop, can you increase your browser's magnification / zoom level?

 

[Arcadia]

I remember this one, but I don't remember what all the different color schemes were, so I'll play! First of all, what I see looks more like a lampshade than a dress, but hey.... at any rate, I see alternating bands of white and gold, or maybe it's yellow ochre.

[Carol the Dabbler]

33 minutes ago, Arcadia said:

I see alternating bands of white and gold, or maybe it's yellow ochre.

That's one of the more common ones.

What I see is very different (but similar to the other most common one), alternating bands of light somewhat-grayish blue (similar to Wedgewood blue) and dark grayish brown (charcoal brown?).

It amazes me that different people can see such different colors in the same photo (and apparently that's the case even when they're both looking at the same screen or the same print of this particular photo).

[besleybean]

I would be generous and say 'gold' and white.

[Arcadia]

3 hours ago, Carol the Dabbler said:

What I see is very different (but similar to the other most common one), alternating bands of light somewhat-grayish blue (similar to Wedgewood blue) and dark grayish brown (charcoal brown?).

Interesting. I can see that only if I squint. Although even then I see more of a light blue and charcoal gray.

I've "always" known that we all see color differently, but this is an extreme example. Very interesting.

[kimber8ada]

I see white and gold/brown.

[Van Buren Supernova]

Yes, I remember this when it came out. I saw both colors on different times of the day, try see it again in different lighting and see if you guys also see both.

I think it has something to do with color spectrum or something else.

And this is the newer version of object with similar problem my sil sent me.

The two colors seen are either

grey + turquoise green color or

pink + white.

[Carol the Dabbler]

VBS, I see the second colors on the shoe (though the colors are so pale that it could possibly be

beige rather than pink and/or gray instead of white

I don't see the other variant at all.

27 minutes ago, Van Buren Supernova said:

I saw both colors on different times of the day, try see it again in different lighting and see if you guys also see both.

I created this thread yesterday after dark in an artificially lighted room, and I'm looking at that dress photo now in a sunny room with no artificial lighting, and I still see the same colors, on both monitors (one on the "daytime" blue-ish white setting and the other on the "bedtime" setting which I've skewed even more toward pinkish white).  So I'm apparently one of the people who generally see the same colors on the dress, regardless.  And you're apparently one of the people who see it differently, depending.

Weird how something that I see as OBVIOUSLY one set of colors, some people see as JUST AS OBVIOUSLY a different set of colors -- and other people see as different at different times.  I can frequently see both sides of a philosophical question, and I can choose to see either the duck or the rabbit (in the Wikipedia page that I linked to in my initial post).  But I not only can't see either gold or white in that dress, I can't even imagine how anybody could possibly see either gold or white.

[J.P.]

As someone who grew up with photography and painting and stuff - I would say it's bronze/golden and white… something hanging in a shadow on a sunny day. But I could be wrong - I know I've seen this one, but don't remember the outcome.

[Carol the Dabbler]

1 hour ago, J.P. said:

I know I've seen this one, but don't remember the outcome.

The outcome is that some people see it one way and some people see it another way (and some people see it in other ways).  That's basically the point -- perfectly honest, competent, observant people can see things differently from each other.

But it's still really weird!!!

[J.P.]

Hmm… you might see things differently, but the fact is - this something had certain colours as it was photographed.

[besleybean]

Surely colours are just names people give to things...I mean red doesn't know it's red!

[Carol the Dabbler]

3 hours ago, J.P. said:

Hmm… you might see things differently, but the fact is - this something had certain colours as it was photographed.

I suppose one could do a factual analysis by measuring the wavelengths of light bouncing off the dress (or indeed, off the photograph).

But I have reasonably good color vision and I assume you do too, yet I see the dress as light blue and dark gray/brown, whereas you see it as white and gold (a far greater difference than a quibble over where orange leaves off and red begins).  A lot of people see it the way I do, and most other people see it the way you do.  I would love to see this explained scientifically.

[Caya]

Here you go, then:

https://www.inverse.com/article/30103-the-dress-black-blue-white-gold-illusion-neuroscience

[Carol the Dabbler]

I will freely admit to being an afternoon/evening person, and also to being no spring chicken, so I do fit their demographics of the typical blue-black person.

But what I really want to know is, how did they create that flipflop photo at the top?  (Is there an explanation somewhere that I missed?)  What do you white & gold people see in those two pictures?

To me, the darker of the two is just a slightly darker version of what I see in the "original" photo (dark brown & pale blue), whereas the paler version simply looks washed out or overlit to me (note that objects in the background are nearly invisible).  I will admit that the colors in the two versions do look different, but basically just light and dark versions of the same colors -- the "gold" stripes look like milk-chocolate to me, just a paler shade of the dark brown that I usually see, and the "white" is an extremely washed-out pale blue.

Of course the alternating photos are probably just an attempt to illustrate what different people see, rather than any sort of scientific demonstration, but I'd still be interested to hear how blue&black people see them vs how white&gold people see them.

[Arcadia]

I see gold and white in the light version and blue/black in the dark version. I can't make myself see one as the other. They both look like they're in shadow. My assumption, without further evidence, would be that they are both poor quality photos that have been photoshopped.

[Carol the Dabbler]

2 hours ago, Arcadia said:

My assumption, without further evidence, would be that they are both poor quality photos that have been photoshopped.

Yeah, I think they were probably just for the purpose of illustrating "what people see."

2 hours ago, Arcadia said:

I see gold and white in the light version and blue/black in the dark version. I can't make myself see one as the other.

Me neither.  And same with the original photo; I've tried seeing it as gold and white, and failed utterly -- it's not like the rabbit/duck optical illusion, where I can see either at will.

[Hikari]

On 9/4/2019 at 5:26 PM, Carol the Dabbler said:

The outcome is that some people see it one way and some people see it another way (and some people see it in other ways).  That's basically the point -- perfectly honest, competent, observant people can see things differently from each other.

But it's still really weird!!!

Are these supposed to be tests for color blindness?

I see a gold and white dress and a pink and white shoe.   Passed my color perception test at the eye doc which I take twice a year recently with good results.  The color quality on both photos isn't great, but unless the photos have been digitally altered on purpose to create eye teasers, either a shoe is pink and white, or it is blue and grey . . both of them can't be correct.

[Carol the Dabbler]

21 minutes ago, Hikari said:

Are these supposed to be tests for color blindness?

Nobody seems to be claiming that they are, so I assume not.

I'm by no means color blind.  I do tend to think of borderline colors in certain ways that may differ from some other people's views (e.g., I had a turquoise-colored skirt that I always referred to as my blue skirt, while my mother called it my green skirt -- which confused us both at first).  But The Dress isn't a simple matter of borderline colors -- you can't get much different than black vs gold!

Perhaps the key is that very few things are one pure, solid color.  If you look closely enough, you are likely to see that most things are dappled, and what we "see" is the combination of colors.  But that's only a guess.

18 hours ago, Caya said:

Here you go, then:

https://www.inverse.com/article/30103-the-dress-black-blue-white-gold-illusion-neuroscience

OK, another question for everyone:  In that alternating photo at the head of that article, does one of those versions look to you more-or-less like what you see when you look at the original photo, while the other one looks like what other people say they're seeing?  In other words, do the two versions constitute a fair representation of what the two factions see?

If the two versions were created using "fair" photoshopping techniques (meaning, no blatantly substituting one color for another), then perhaps the key is that the "black" is actually a dark brown (which is what I see).  In which case that article may well be correct about people making unconscious assumptions about the lighting.  If you lighten black you'll never get anything except gray -- but if you lighten dark brown sufficiently, you will get something like gold.  And of course a sufficiently washed-out blue will look white (especially to those of us who grew up in cultures where a very faint blue is thought to look whiter than actual pure white).

OK, I'm happy now.

[J.P.]

Not with a poor quality photo run through a PS filter. Actually, claiming that anyone could deduct that this washed out brownish is actually black in RL is BS. I could make the dress look red in one second btw, and ask the same question.

As for the shoe: I SEE light turquoise and greyish pink, but I also see the hand holding it  - and having processed tons and tons of such photos in my professional life, I'm quite sure that the shoe is pink and white in RL.

Lighting is the key. Human eye and brain can make corrections based on the ambient light colour, new cameras can do it also, but not always.

[J.P.]

You could also read

https://en.wikipedia.org/wiki/Color_balance
https://en.wikipedia.org/wiki/Colour_cast

This is a nice example from a photography blog https://foodphotographyblog.com/what-is-white-balance-color-temperature-and-kelvin/

https://photographylife.com/what-is-white-balance-in-photography     shows how colors in a photo can change by slight change in the sunlight.

[Carol the Dabbler]

1 hour ago, J.P. said:

claiming that anyone could deduct that this washed out brownish is actually black in RL is BS.

That's exactly what threw me at first, even though I've personally been seeing the darker color as dark brown.  Then it occurred to me just a little while ago that if it really is some sort of dark brown (not black at all), then a change in the lighting that would wash out the colors could very well make it look goldish (or as you say, washed out brownish), and the same change would of course make the light blue look whitish.

So I now think that's the whole key -- it's not black at all, it's a very dark brown.

[Caya]

Well, no:

https://www.businessinsider.com/the-roman-originals-black-and-blue-dress-2015-2