What makes #9B51E0 purple?
April 20, 2018 7:22 PM   Subscribe

Color: From Hexcodes to Eyeballs "Why do we perceive background-color: #9B51E0 as this particular purple? This is one of those questions where I thought I’d known the answer for a long time, but as I inspected my understanding, I realized there were pretty significant gaps. Through an exploration of electromagnetic radiation, optical biology, colorimetry, and display hardware, I hope to start filling in some of these gaps."

Bonus: Tyrian Purple
"In Phoenician mythology, the discovery of purple was credited to the pet dog of Tyros, the mistress of Tyre’s patron god Melqart. One day, while walking along the beach the couple noticed that after biting on a washed up mollusc the dog’s mouth was stained purple. Tyros asked for a garment made of the same colour and so began the famous dyeing industry."

Extra bonus: Aniline Purple
"Whilst the origin of the modern industrial revolution is often traced to the Ironbridge Gorge northwest of Birmingham in the UK, less credit is given to one of the birthplaces of the modern organic chemical industry. William Henry Perkin (1838-1907) who at the age of 18 in the year 1856 had set out with idea of making quinine (C20H24N2O2) by oxidising allytoluidine (C10H12N), had accidentally produced instead the first ever synthetic dye aniline purple, better known as mauveine."
posted by aneel (17 comments total) 59 users marked this as a favorite
 
luckily, we're starting to get 4K tunable laser displays, so hopefully soon a large part of these concerns will vanish and we can begin addressing pixel coloring in nanometers.
posted by Xyanthilous P. Harrierstick at 7:53 PM on April 20, 2018 [1 favorite]


I'm surprised they didn't get into blackbody radiation when talking about the incandescent bulb, because that is exactly what that graph is there. Any body that you heat up will give off that continuous spectrum (no jaggies or waves) with that one peak corresponding to its temperature. Heat it up more and the whole graph gets taller and bigger and the peak moves to the right ...the area under the graph is the total energy in the system and the peak is it's strongest color. So, heat something up and it glows red-hot then orange, yellow, white-hot, and then blue-hot. These are the colors of stars. The jaggies and lines in a solar spectrum are inroduced by excitation and absorbtion in its atmosphere...the surface is a blackbody radiator, as are the filaments in a bulb...the color of its light comes from its temperature.
posted by sexyrobot at 8:22 PM on April 20, 2018 [5 favorites]


That's fascinating, I never realized how the gamut of a screen is actually a relatively small slice of what our eyes can actually perceive, and there's a big chunk of spectral blues and greens that cannot be recreated by a RGB display. It makes me wonder about film's ability to reproduce color versus digital displays, but I suppose film is also based on an RGB color space so it's probably not much better.

This also goes a long way to explain why the 5050 RGB light string that lights my porch doesn't make a satisfying rainbow when I set it to continually shift colors. Each LED is dutifully displaying every combination it has at its disposal, but there's a whole chunk of the rainbow it simply can't touch.
posted by Mr.Encyclopedia at 8:44 PM on April 20, 2018 [2 favorites]


God I love color science!

Thanks, aneel. Awesome article — downloaded and saved for reading more closely later.
posted by darkstar at 8:58 PM on April 20, 2018


Man, color is complicated.

I'm a bit confused, though. If screens can't reproduce a huge chunk of blues and greens, what is happening if I look at something in that space on a screen? How can he show the space on a diagram at all? He says the colors aren't accurate, but the diagram suggests that monitors can't display anything even close to the colors that are clearly being displayed. So how does that work?
posted by tau_ceti at 10:45 PM on April 20, 2018


The whole CIE diagram encompasses the entire range of human trichromatic vision plus some, so one can plot any color on it. Even the best gamuts currently in use only cover 70ish percent of normal vision at present. Most TVs only cover 70% of the Rec.2020 HDTV color space, though some of the newest Ultra HD HDR wide gamut displays get well up into the 90s. Of course, they only cover 60-70% of the UHD color gamut so far..

The other big issue with color rendering are that the color of the individual subpixels almost never match the standard values, throwing off the color reproduction. My first 1080p set could go well outside the standard gamut in blues, but the other colors were less saturated than standard, so even after calibration it was never quite "right," though it was fine. Your brain cares a lot less about this stuff than we'd like to think. It just fills in whatever is missing. That's why you only actually need two colors to make an apparent color photo if you choose your subject wisely. The brain will assume the missing color is there as long as it doesn't look like it should be highly saturated
posted by wierdo at 10:57 PM on April 20, 2018 [2 favorites]


Oh, the childhood fights with my sister that could have been avoided if I'd been aware of my mild blue/green color blindness years earlier!

"She took my green dress when I told her she could borrow the blue one!" "She said I didn't leave the phone number, but it's right there in the green notebook, like I told her! If she says there is no green notebook she's lying!"

It wasn't until I lost a team-building exercise at a work retreat because my clue was "Follow the person in the blue floral blouse" that I realized it wasn't just my sister being mean and I was actually seeing things differently.

The weird thing is, we had basic vision and hearing tests done every year in school. They included those tests for color blindness where there's a circle full of little colored dots, and if you were color-blind you wouldn't be able to see that some of the dots formed an Arabic numeral in a foreground color while the rest were a background color or colors. I always passed those with flying whatchamacallits.
posted by The Underpants Monster at 11:40 PM on April 20, 2018 [10 favorites]


If screens can't reproduce a huge chunk of blues and greens, what is happening if I look at something in that space on a screen? How can he show the space on a diagram at all?

The diagram is a map of all possible colours. The locations outside the RGB gamut are not filled in with the correct colour, but with the closest colour an RGB monitor can display. That's why the cyan at 490 on the curved boundary looks identical to the cyans in a horizontal line to the right of it. The one on the left is supposed to be even more saturated, but your monitor can't do it because it's already as saturated as it will go.
posted by a car full of lions at 1:13 AM on April 21, 2018 [1 favorite]


Yeah, the problem of using the apparatus under investigation to conduct the investigation itself has always been one of the most intellectually fiddly aspects of learning about human perception & cognition. Color is one of my go-to examples for just how incredibly deep and wonk-like it's possible to get about a concept that many finish their lifelong learning about in second grade.
posted by Fraxas at 2:24 AM on April 21, 2018 [7 favorites]


Great article! I learned somethings, and I study vision for a living!

Yep, colour is complicated... and he didn't even get to the really hard problem of how the human brain interprets colour. #thedress says it all: given the same physical stimulus (and therefore the same wavelength, same cone excitations, etc.), different people will perceive white/gold or blue/black.
posted by tickingclock at 2:56 AM on April 21, 2018 [2 favorites]


The question is "How do you know what "colour" a person perceives.
The problem is that when presented with a particular colour you have been taught that it is , say, green. Now I may have been presented with the same thing and also have been taught that it is green. So we both call it green, but do we actually see the same colour .
posted by Burn_IT at 7:07 AM on April 21, 2018


Think of all the designers and techies that had to deal with NTSC on top of all this.
posted by RobotVoodooPower at 7:46 AM on April 21, 2018 [2 favorites]


(Insert old joke about how NTSC stands for Never The Same Color.)
posted by introp at 7:52 AM on April 21, 2018 [1 favorite]


Fun fact: there are only 3 known color combinations that 'vibrate' when placed next to each other: red and blue (see bottom of 'NTSC' link above (though a slightly more cyan blue works even better), a deep(ish) green and (somewhat)reddish orange, and deep purple and yellow. That's in descending order of strength with red/blue being the strongest.
posted by sexyrobot at 9:36 AM on April 21, 2018 [3 favorites]


That's a really nice, concise summary, the history of the Wright - Guild experiments was particularly interesting.

Having picked purple as the example at the start, I was expecting it to lead into something about the line of purples being a sort of looping of the spectrum.
posted by lucidium at 2:22 PM on April 21, 2018 [1 favorite]


There are tricks that make it possible to see "impossible colors."
posted by sjswitzer at 2:28 PM on April 21, 2018 [3 favorites]


Additional links:
XKCD Color Survey Results
First, a few basic discoveries:
  • If you ask people to name colors long enough, they go totally crazy.
  • “Puke” and “vomit” are totally real colors.
  • Colorblind people are more likely than non-colorblind people to type “fuck this” (or some variant) and quit in frustration.
  • Indigo was totally just added to the rainbow so it would have 7 colors and make that “ROY G. BIV” acronym work, just like you always suspected. It should really be ROY GBP, with maybe a C or T thrown in there between G and B depending on how the spectrum was converted to RGB.
  • A couple dozen people embedded SQL ‘drop table’ statements in the color names. Nice try, kids.
  • Nobody can spell “fuchsia”.
Art world color war: "You are not Anish Kapoor, you are in no way affiliated to Anish Kapoor, you are not purchasing this item on behalf of Anish Kapoor or an associate of Anish Kapoor."

Tetrachromacy: "Most people have three cells, or receptors, in their retinas, but tetrachomats have a fourth receptor, which may be what allows for their heightened color perception. They are usually female, and it’s estimated that about 12 percent of women carry the gene for this fourth receptor."
posted by ErisLordFreedom at 12:55 AM on April 22, 2018 [2 favorites]


« Older The psychedelic renaissance or enclosing mind...   |   Accelerated evolution of super-cockroaches is... Newer »


This thread has been archived and is closed to new comments