So wham! Rumor has it that apparently two years to the date, or at least the same week, we get a sequel to "The Dress." The not so red strawberries.
If you're not familiar with the dress you should probably be swept up by the Men in Black since it's clear you're not from this planet! But I'll explain while they arrive... "The Dress" was a simple photo post asking what color the dress in question was. It was not an amazing photo, nor was it posted from a celebrity account with gazillions of followers. Yet for some reason it took hold with every living being with internet access. Scratch that, I think even the tiny demographic of color fax machines received the photo with the question! Answers were split right down the middle: "Blue and Black" followed by "Are you crazy? It's White and Gold" or the reverse.
The question played out in an argument at the office, national and international news, and most definitely in the comments section everywhere. Of all random viral internet moments "The Dress" stands tall in the crowd. For those keeping track White and Gold actually was the majority winner in most cases. Winning Polls for "White and Gold" came in at 53.8% vs 39.8% on CNN with almost 300,000 votes, and 67% to 33% on Buzzfeed boasting an impressive 3.6 million votes! I actually was surprised by that since it's clearly Blue and Black...
Why All This Color Craziness?
Color constancy is an example of subjective constancy and a feature of the human color perception system which ensures that the perceived color of objects remains relatively constant under varying illumination conditions. We tend to think a color just is, and will be regardless of light. But in reality color is simple the way the light reflects off something. My son is very into Jonathon Bird, a successful underwater photographer and host of Blue World TV which I highly recommend! In one of his videos he points out that once you're past a certain depth in the water, a bowl or Skittles would look like different shades of grey. It's actually depth dependent and goes in the order of the spectrum. Red is the first to go at 15ft, orange is next at 25ft, yellow at 35-45ft, and finally green at 70-75ft. Luckily our brain uses good old color constancy to compensate and we still perceive our dear little reds.
So Onto the Strawberries in Question
This image caught some attention on Twitter after being created and shared by Akiyoshi Kitaoka, a Professor of Psychology at Ritsumeikan University in Kyoto, Japan. If you have not heard of the College, neither had I. So in the name of fact checking, I checked in and not only is it very reputable and prestigious looking, but after some random digging around I found it has CG and Game Club. What is CG and Game Club you say? CG&G Club is "creating computer graphics for games, from planning games, writing scenario, creating graphics, sounds and programming all by ourselves." Yeah um, Japan rocks! And Akiyoshi has some other really crazy illusions on his site (Warning - your eyes may not be ready for some of the movement without movement illusions!).
So I've read multiple articles about these babies, and all claim 100% no red pixels which is the part that really boggles the mind, since most of us clearly are seeing red strawberries on that plate. While I do agree with the true nature of the Red look of the fruit resulting from the phenomenon of Color Constancy, I do think there is some trickery at play since I was easily able to find reddish pixels in my photo editor of choice.
Could these subtle hints of red cause the whole Strawberry to seem it? I think our brain is doing most of the heavy lifting there, but it does probably help. Also I feel like the illusion might still work without the slightly reddish specks. But who knows, maybe as I hear they say with horseradish "a little bit goes a long way!" What do you think - are they Red?
I'm more impressed by this little cartoon girl, but slightly concerned about her smoking habit!
Shaun Maluga pointed out in the comments that this version has no red at all and the illusion actually works better.
The version you are playing with is different, and to me, has less of the effect. The strawberries look distinctly grey to me (minus compression artefacts) compared to the one you shared by Carson Mell in the article that demonstrates the effect better and has no red values (even from compression). Interestingly the original version Akiyoshi shared is the version you posted, the one he posted later has the better effect: https://twitter.com/AkiyoshiKitaoka/status/836743598469070849
I 100% agree that the 2nd example works better and has 0 red pixels... Seems like a switch occurred. Thanks for pointing that out. I was not expecting any trace of red in the original, then the article took a turn when I did.
Ha, very cool. The strawberries look reddish to me, but the Mac Digital Colormeter Utility says they are gray on both of my monitors (one color calibrated, the other not). I'm wondering how much the surround color plays into it, because the cyan is the direct complementary color on the color wheel.
If you really want to get into the question of color perception, check out this piece on IFL Science (which has the same strawberry picture, BTW) www.iflscience.com/brain/when-did-humans-start-see-color-blue/ Some researchers (yes, this is legitimate, not fringe stuff) hypothesize that as a culture, we didn't learn to see some colors until well into our history, much as we've had to learn to see depth in two dimensions via perspective.
Just a detail, the color absorption of water is actually distance dependent rather than depth dependent -- depth is part of that. Colors absorb following the ROYGBIV sequence, with red reduced quickly followed by orange, etc. The perceived color absorption is the same for an object right in front of you at 20 feet, for example, or one 10 feet away and 10 feet deep, all else being equal. The light travels through 20 feet of water either way. (I'm an underwater shooter, you may gather). You still see much of the color, and there's debate how much is our brain's ability to "replace" color versus our ability to "white balance" the trace amounts that are still there, or both. But, some experiences will stand out -- I cut my hand at 60 feet once (not seriously) and was fascinated to watch it bleed dark olive green.
Very cool! Thanks for sharing and that's crazy about the olive blood!
I don't really get the phenomenon. I mean I get that it is intriguing that our brains do cool things to make 'shorthand' of the visual world, but just like 'the dress' the photo's white balance is completely off. It's like going to the projector with some C41 and not adjusting the color balance, or shooting everything raw at 3000K and not setting the white balance. There's even enough data in the JPG in the title to do this in photoshop quite easily. The different between our brains and a digital camera is that our brains automatically attempt to figure it out so it still makes sense. For reference, here is the photo with the gray point set on the plate. It's not hard to see that the strawberries now look like strawberries, the plate looks like a plate (with even a hint of blue in the top left), and the table looks like you would expect a table to look. Maybe I'm just being annoying here?
Im pretty sure thats the point, that given incorrect color information (i.e. the grey strawberries in the blue tinted photo, a white balance that is off, etc) our brains unconsciously correct the mistake so that we "see" the true colors of the subject (strawberries) even when the colors we are presented with are off.
I think people assume everything they see is real, and when you can prove there's no red but your brain is showing you red it's interesting.
Two thoughts...the basic concept that Akiyoshi Kitaoka is illustrating overall with his examples are valid insomuch that our brains try to make sense out of what we see based on learned assumptions-thus the illusions...and we fight to believe what we see as "truth" and therefore "seeing is believing, " right? But...how many of of us have seen optical illusions before that bend our "truth" system? Even if it's just the simple ones on the place mat at the local diner.
This example is all about the strong presence of the color compliment of red-cyan, and the neutral tone of the berries and the white plate...in the process of trying to compensate the strong cyan, the brain is placing a complementary red value into the otherwise neutral grey where the berries are...and since we "know" strawberries are red-they must be red. For use on the internet, this image is converted over to an RGB space...but do I read Akiyoshi Kitaoka as having basically created it in a duotone space and converted? The theory is sound...just suspect that the manner in which this is shared is tossing some characteristic variables into the fray here as well. I can digitally "correct" this file...although due to the extreme artificial color imbalance, it will never look like the original shot.
This just about says everything I was going to say.
Actually in this case, it is a play with levels on different channels. Here is the corrected auto-leveled version of the picture: