Welcome to the most eye-popping development in the history of vision!
April 5, 2016 6:39 PM Subscribe
Squant: the fourth primary color
Squant Story, by Crosley Bendix, Director Of Stylistic Premonitions, U.M.N.
Squant Photography, by Rolin Wandbagon. [Note: Since these pictures include the color squant, you'll need a special SquantView™ plug-in to see them properly. Plug-ins are available for both Mac and Win versions 3.1 through NT. Click here for troubleshooting.]
Squant Story, by Crosley Bendix, Director Of Stylistic Premonitions, U.M.N.
Squant Photography, by Rolin Wandbagon. [Note: Since these pictures include the color squant, you'll need a special SquantView™ plug-in to see them properly. Plug-ins are available for both Mac and Win versions 3.1 through NT. Click here for troubleshooting.]
I really don't feel like downloading software to install on my computer, so would anyone care to explain what the plugin does?
I know there was something a couple of years ago where you looked at an image for 30 seconds and then looked at a piece of white paper, and the afterimage would show you a color which can't be displayed on a modern display. (It was more saturated than a display permits.)
Is this something like that? Meanwhile, anyone care to give me a hint about what "squant" does look like? (i.e. "sort of like blue green" or something like that)
posted by Chocolate Pickle at 6:55 PM on April 5, 2016
I know there was something a couple of years ago where you looked at an image for 30 seconds and then looked at a piece of white paper, and the afterimage would show you a color which can't be displayed on a modern display. (It was more saturated than a display permits.)
Is this something like that? Meanwhile, anyone care to give me a hint about what "squant" does look like? (i.e. "sort of like blue green" or something like that)
posted by Chocolate Pickle at 6:55 PM on April 5, 2016
Those of us that can see octarine are saying, "That's cute", before going back to painstakingly (and very, very carefully) crafting our sapient pearwood staves.
posted by Halloween Jack at 6:56 PM on April 5, 2016 [53 favorites]
posted by Halloween Jack at 6:56 PM on April 5, 2016 [53 favorites]
It doesn't do squant.
posted by Confess, Fletch at 6:56 PM on April 5, 2016 [4 favorites]
posted by Confess, Fletch at 6:56 PM on April 5, 2016 [4 favorites]
I upgraded to system 8.1! It took care of things.
posted by unknowncommand at 7:00 PM on April 5, 2016 [5 favorites]
posted by unknowncommand at 7:00 PM on April 5, 2016 [5 favorites]
Crosley Bendix' mother is named Toy Celica (of the Coastal New Brunswick Celicas).
And that's all I really needed from this. Thanks!
posted by carsonb at 7:08 PM on April 5, 2016 [1 favorite]
And that's all I really needed from this. Thanks!
posted by carsonb at 7:08 PM on April 5, 2016 [1 favorite]
Good sir, wouldn't you say that's a little bit elaborate for a joke???
posted by carsonb at 7:12 PM on April 5, 2016 [4 favorites]
posted by carsonb at 7:12 PM on April 5, 2016 [4 favorites]
This is a hoax, isn't it?
posted by Chocolate Pickle at 7:12 PM on April 5, 2016
posted by Chocolate Pickle at 7:12 PM on April 5, 2016
Eponstyrical, amirite?
posted by theora55 at 7:13 PM on April 5, 2016 [10 favorites]
posted by theora55 at 7:13 PM on April 5, 2016 [10 favorites]
Oh finally! I got the plug-in installed after a lot of gnashing.
Squant is really quite tantalizing a color!
posted by carsonb at 7:16 PM on April 5, 2016 [2 favorites]
Squant is really quite tantalizing a color!
posted by carsonb at 7:16 PM on April 5, 2016 [2 favorites]
anyone care to give me a hint about what "squant" does look like? (i.e. "sort of like blue green" or something like that)
I mean… that's not how primary colors work. Can you explain what red looks like, using only blue and green?
posted by DoctorFedora at 7:17 PM on April 5, 2016 [6 favorites]
I mean… that's not how primary colors work. Can you explain what red looks like, using only blue and green?
posted by DoctorFedora at 7:17 PM on April 5, 2016 [6 favorites]
Meanwhile, anyone care to give me a hint about what "squant" does look like?You know what octarine looks like, right? Like that.
posted by L.P. Hatecraft at 7:20 PM on April 5, 2016 [6 favorites]
I mean… that's not how primary colors work. Can you explain what red looks like, using only blue and green?
Primary colors are monochromatic. So I can say that a wavelength of 650 nanometers will look red to me, and a wavelength of 530 nm will look green, and 470 nm will look blue.
So what wavelength is squant?
posted by Chocolate Pickle at 7:26 PM on April 5, 2016 [4 favorites]
Primary colors are monochromatic. So I can say that a wavelength of 650 nanometers will look red to me, and a wavelength of 530 nm will look green, and 470 nm will look blue.
So what wavelength is squant?
posted by Chocolate Pickle at 7:26 PM on April 5, 2016 [4 favorites]
So what wavelength is squant?
I'm checking the readouts and it says.... negativ?
posted by carsonb at 7:27 PM on April 5, 2016 [13 favorites]
I'm checking the readouts and it says.... negativ?
posted by carsonb at 7:27 PM on April 5, 2016 [13 favorites]
You know what octarine looks like, right? Like that.
I am neither wizard nor cat, so no, I don't know what octarine looks like.
Oh, screw it.
posted by Chocolate Pickle at 7:27 PM on April 5, 2016 [4 favorites]
I am neither wizard nor cat, so no, I don't know what octarine looks like.
Oh, screw it.
posted by Chocolate Pickle at 7:27 PM on April 5, 2016 [4 favorites]
New for tetrachromats?
The plugin opened as a text document full of ascii crap, which takes me back to the '90s. Coincidentally, Negativland takes me back to the '90s.
posted by klangklangston at 7:29 PM on April 5, 2016 [5 favorites]
The plugin opened as a text document full of ascii crap, which takes me back to the '90s. Coincidentally, Negativland takes me back to the '90s.
posted by klangklangston at 7:29 PM on April 5, 2016 [5 favorites]
I like how the plugin lets me smell squant on Netscape 3!
posted by SecretAgentSockpuppet at 7:31 PM on April 5, 2016 [8 favorites]
posted by SecretAgentSockpuppet at 7:31 PM on April 5, 2016 [8 favorites]
...took his degree in Social Esthetics from Montana State College...
My alma mater! Although it hasn't been named MSC since the 1960's. I've long advocated they reinstate their degree program in Social Esthetics. But then, the whole Cultural Affectations department has really gone downhill since those days.
posted by traveler_ at 7:33 PM on April 5, 2016 [3 favorites]
My alma mater! Although it hasn't been named MSC since the 1960's. I've long advocated they reinstate their degree program in Social Esthetics. But then, the whole Cultural Affectations department has really gone downhill since those days.
posted by traveler_ at 7:33 PM on April 5, 2016 [3 favorites]
There's squant evidence this is anything other than a very old prank.
posted by Mr.Encyclopedia at 7:39 PM on April 5, 2016 [2 favorites]
posted by Mr.Encyclopedia at 7:39 PM on April 5, 2016 [2 favorites]
Wikipedia:
Crosley Bendix is a radio personality (played by Joyce) for a series of commentaries. Bendix was touted as the Cultural Reviewer, and typically greets listeners with a cheery "Good hello!" Bendix introduced listeners to "squant", the fictitious "fourth primary color", which was also the only primary color to have its own unique scent.posted by Chocolate Pickle at 7:41 PM on April 5, 2016
Ah, Negativland. It's been a while.
posted by gusandrews at 7:42 PM on April 5, 2016 [4 favorites]
posted by gusandrews at 7:42 PM on April 5, 2016 [4 favorites]
"There's a moral to this story. Do everything with computers, and forget about all that analog garbage!"
They always were ahead of their time.
posted by gusandrews at 7:43 PM on April 5, 2016 [3 favorites]
They always were ahead of their time.
posted by gusandrews at 7:43 PM on April 5, 2016 [3 favorites]
Incidentally, "SquantView Plug-in" is the name of my new band. We do covers in tetrachromatic harmony.
posted by Greg_Ace at 7:54 PM on April 5, 2016 [4 favorites]
posted by Greg_Ace at 7:54 PM on April 5, 2016 [4 favorites]
Squant is on top of the refrigerator behind the Mr Coffee coffee maker.
posted by reedcourtneyj at 7:58 PM on April 5, 2016 [2 favorites]
posted by reedcourtneyj at 7:58 PM on April 5, 2016 [2 favorites]
The associated smell of squant is pot in a basement apartment after someone's done a Mark Leyner bit for an hour and is now trying to tell you about the Illuminatus books but you realize you haven't been listening to them for awhile or indeed ever.
posted by nom de poop at 8:01 PM on April 5, 2016 [10 favorites]
posted by nom de poop at 8:01 PM on April 5, 2016 [10 favorites]
DoctorFedora: "Can you explain what red looks like, using only blue and green?"
Easy. Red is the opposite of green.
posted by RobotHero at 8:28 PM on April 5, 2016 [2 favorites]
Easy. Red is the opposite of green.
posted by RobotHero at 8:28 PM on April 5, 2016 [2 favorites]
Can you explain what red looks like, using only blue and green?
it is feistier
posted by threeants at 8:36 PM on April 5, 2016 [21 favorites]
it is feistier
posted by threeants at 8:36 PM on April 5, 2016 [21 favorites]
This reminds me of the day I learned the word "umami" which until then I'd had to use a cribbed phrase "country goodness" to denote.
posted by janey47 at 8:50 PM on April 5, 2016 [13 favorites]
posted by janey47 at 8:50 PM on April 5, 2016 [13 favorites]
I do not like squant eggs and ham
posted by Kabanos at 8:56 PM on April 5, 2016 [7 favorites]
posted by Kabanos at 8:56 PM on April 5, 2016 [7 favorites]
The associated smell of squant is pot in a basement apartment after someone's done a Mark Leyner bit for an hour and is now trying to tell you about the Illuminatus books but you realize you haven't been listening to them for awhile or indeed ever.
If this all happened to me in real life, do I still have to download the plug-in?
posted by louche mustachio at 9:00 PM on April 5, 2016 [10 favorites]
If this all happened to me in real life, do I still have to download the plug-in?
posted by louche mustachio at 9:00 PM on April 5, 2016 [10 favorites]
This isn't that far off -- people have tried selling monitors with 4 primary colors. Really there's nothing special about red, green and blue except that everyone has agreed to use it.
posted by miyabo at 9:06 PM on April 5, 2016
posted by miyabo at 9:06 PM on April 5, 2016
Uuuuuummmm, if that RGBS diagram is to be believed, wouldn't yellow cease to exist, or at least always have a squanty tint? Squant = yellow. I appreciate the gag but...
posted by threecheesetrees at 9:29 PM on April 5, 2016
posted by threecheesetrees at 9:29 PM on April 5, 2016
The 4 color monitor thing is to help capture more of the color space than you can usually get out of an RGB monitor... I've seen them demo'd and they look pretty freaking sweeeeet, despite what Wikipedia says about it.
Obviously squant is parody, but there are of course some people who actually see 4 primary colors, you can read about them on Wikipedia for a start.
posted by MikeWarot at 9:30 PM on April 5, 2016 [4 favorites]
Obviously squant is parody, but there are of course some people who actually see 4 primary colors, you can read about them on Wikipedia for a start.
posted by MikeWarot at 9:30 PM on April 5, 2016 [4 favorites]
OK Go needs to update their song.
Four primary colors
They are all the colors you could want
There's so many colors
And they're made of red green blue and squant
posted by BiggerJ at 9:40 PM on April 5, 2016 [1 favorite]
Four primary colors
They are all the colors you could want
There's so many colors
And they're made of red green blue and squant
posted by BiggerJ at 9:40 PM on April 5, 2016 [1 favorite]
I want to hear more about the annual yearlong color fair.
posted by Lyme Drop at 9:41 PM on April 5, 2016 [1 favorite]
posted by Lyme Drop at 9:41 PM on April 5, 2016 [1 favorite]
miyabo,
No, yellow is really hard to hit, it's at an equal stimulation point of medium wave and long wave stimulation and, short of 590nm light, that's hard. So some monitors just emit 590nm.
By hard, I mean >600nm light is trivially red, 430nm to 490nm is blue, 510nm to 570nm is green. Those ranges are around 60nm wide. Yellow is really perceived between 570 and 600. That's only about 30nm, and really, the clearly yellow range might be as small as 15nm (at least for monochromatic sources).
My operating theory (by which I mean, I could be wrong, this is just what comes from working on DanKam) is that the "primary colors" are really the colors you can get wrong and still see roughly the same thing. You can actually make a CYM (Cyan Yellow Magenta) display and it works fine, but small shifts and everything breaks. That's actually what you want with ink though, because you want to limit how much ink you apply, both because materials can only take so much, and because ink can selectively only reduce reflectivity, it can't construct light. Shifting selective reflectivity a little, from the balance points, gives you desired colors. What's funny is you can generally combine red and green and get a yellow, it's just rather dark and muddy. That's because the red ink is blocking most of the green ink's allowed reflectivity, and vice versa. There's usually a bit of overlap though, between the low end of red and the high end of green. It ain't much, but it's there.
Your eye doesn't know whether a photon arrived from a constructive (RGB) or destructive (CYMK) source. It's just looking for relative levels of photon wavelengths.
As they say of course, color is in the brain, not the eye. Heck, most people don't even realize brown is dark orange/yellow. Black/White, RGB, and the skin tones are different systems.
posted by effugas at 9:42 PM on April 5, 2016 [16 favorites]
No, yellow is really hard to hit, it's at an equal stimulation point of medium wave and long wave stimulation and, short of 590nm light, that's hard. So some monitors just emit 590nm.
By hard, I mean >600nm light is trivially red, 430nm to 490nm is blue, 510nm to 570nm is green. Those ranges are around 60nm wide. Yellow is really perceived between 570 and 600. That's only about 30nm, and really, the clearly yellow range might be as small as 15nm (at least for monochromatic sources).
My operating theory (by which I mean, I could be wrong, this is just what comes from working on DanKam) is that the "primary colors" are really the colors you can get wrong and still see roughly the same thing. You can actually make a CYM (Cyan Yellow Magenta) display and it works fine, but small shifts and everything breaks. That's actually what you want with ink though, because you want to limit how much ink you apply, both because materials can only take so much, and because ink can selectively only reduce reflectivity, it can't construct light. Shifting selective reflectivity a little, from the balance points, gives you desired colors. What's funny is you can generally combine red and green and get a yellow, it's just rather dark and muddy. That's because the red ink is blocking most of the green ink's allowed reflectivity, and vice versa. There's usually a bit of overlap though, between the low end of red and the high end of green. It ain't much, but it's there.
Your eye doesn't know whether a photon arrived from a constructive (RGB) or destructive (CYMK) source. It's just looking for relative levels of photon wavelengths.
As they say of course, color is in the brain, not the eye. Heck, most people don't even realize brown is dark orange/yellow. Black/White, RGB, and the skin tones are different systems.
posted by effugas at 9:42 PM on April 5, 2016 [16 favorites]
threecheesetrees: To represent all the combinations, they'd need three-dimensional spheres with cutaway views. And, seriously, overlapping 3D spheres? Who can afford that? That's some Mind's Eye shit!
posted by BiggerJ at 9:50 PM on April 5, 2016 [2 favorites]
posted by BiggerJ at 9:50 PM on April 5, 2016 [2 favorites]
Well, it ain't squat.
posted by oneswellfoop at 10:02 PM on April 5, 2016
posted by oneswellfoop at 10:02 PM on April 5, 2016
...took his degree in Social Esthetics from Montana State College...
Chortle.... we scoffed at that program, from our lofty vantage at Harvard State.
posted by msalt at 10:10 PM on April 5, 2016 [3 favorites]
Chortle.... we scoffed at that program, from our lofty vantage at Harvard State.
posted by msalt at 10:10 PM on April 5, 2016 [3 favorites]
Squant is the color of my true love's hair
In the mornin', when we rise
In the mornin', when we rise
That's the time, that's the time
I love the best
posted by Atom Eyes at 10:32 PM on April 5, 2016 [1 favorite]
In the mornin', when we rise
In the mornin', when we rise
That's the time, that's the time
I love the best
posted by Atom Eyes at 10:32 PM on April 5, 2016 [1 favorite]
More suggestions for OK Go:
Squant and red make what?
Squant and red make CLURDEN!
Squant and green make what?
Squant and green make DREECH!
Squant and blue make what?
Squant and blue make TUNPUS!
This is something every school should teach
posted by BiggerJ at 10:41 PM on April 5, 2016 [2 favorites]
Squant and red make what?
Squant and red make CLURDEN!
Squant and green make what?
Squant and green make DREECH!
Squant and blue make what?
Squant and blue make TUNPUS!
This is something every school should teach
posted by BiggerJ at 10:41 PM on April 5, 2016 [2 favorites]
"This isn't that far off -- people have tried selling monitors with 4 primary colors. Really there's nothing special about red, green and blue except that everyone has agreed to use it."
"Everyone has agreed to use it" because it most closely approximates the natural color contrast of our trichromatic cones.
It's like saying there's nothing special about sweet, sour, bitter and salty, it's just that everyone has decided to cook with them.
posted by klangklangston at 11:24 PM on April 5, 2016 [1 favorite]
"Everyone has agreed to use it" because it most closely approximates the natural color contrast of our trichromatic cones.
It's like saying there's nothing special about sweet, sour, bitter and salty, it's just that everyone has decided to cook with them.
posted by klangklangston at 11:24 PM on April 5, 2016 [1 favorite]
what really blew my mind a little while ago was the discovery that red, green, and blue are the primary additive colors, and cyan, magenta, and yellow are the primary subtractive colors.
Cyan = green + blue
Yellow = red + green
Magenta = blue + red
posted by DoctorFedora at 11:30 PM on April 5, 2016 [3 favorites]
Cyan = green + blue
Yellow = red + green
Magenta = blue + red
posted by DoctorFedora at 11:30 PM on April 5, 2016 [3 favorites]
Cyan = white - red
Yellow = white - blue
Magenta = white - green
posted by Chocolate Pickle at 11:34 PM on April 5, 2016 [9 favorites]
Yellow = white - blue
Magenta = white - green
posted by Chocolate Pickle at 11:34 PM on April 5, 2016 [9 favorites]
I know there was something a couple of years ago
Might be chimerical colours. The colour space defined in the neuronal structure of our brains is slightly larger than the one set up by our eyes (or something like that); but you can experience the extra colours - dark yellow, hyperbolic orange and so on - by opponency.
posted by Segundus at 11:40 PM on April 5, 2016 [5 favorites]
Might be chimerical colours. The colour space defined in the neuronal structure of our brains is slightly larger than the one set up by our eyes (or something like that); but you can experience the extra colours - dark yellow, hyperbolic orange and so on - by opponency.
posted by Segundus at 11:40 PM on April 5, 2016 [5 favorites]
Tisquantum (Squanto) taught dull interlopers to plant colorful maize with glittering fish.
posted by pracowity at 11:40 PM on April 5, 2016 [1 favorite]
posted by pracowity at 11:40 PM on April 5, 2016 [1 favorite]
The browser plugin is a hoax, for those who were still wondering. Every OS and browser made in the last fifteen years or so supports squant more-or-less natively (32-bit color depth). Some browser plugins did exist back in the '90s, but since only fancy and expensive CRTs had squant capability they were just an early Web novelty like VRML.
Most Web content is still in 24-bit RGB color, but you'll sometimes see references to "RGBA" color. The "A" stands for "aucterin", which is French for "squant". (Blame the CIE.)
posted by neckro23 at 12:14 AM on April 6, 2016 [14 favorites]
Most Web content is still in 24-bit RGB color, but you'll sometimes see references to "RGBA" color. The "A" stands for "aucterin", which is French for "squant". (Blame the CIE.)
posted by neckro23 at 12:14 AM on April 6, 2016 [14 favorites]
If you imagine red, blue, and green as the 3 base points on the color triangle (or chroma triangle) then squant is the top point of the color pyramid (or simply put, the 4th point in the 3 dimensional solid chroma tetrahedron) so it's opposite red, green, and blue because, in a sense, they are all flat on the floor, and the squant point is poking into your bare foot in the dark.
Mindi, when?
posted by otherchaz at 1:36 AM on April 6, 2016 [1 favorite]
Mindi, when?
posted by otherchaz at 1:36 AM on April 6, 2016 [1 favorite]
"'Everyone has agreed to use it' because it most closely approximates the natural color contrast of our trichromatic cones."
I'm sure someone will come along with more energy to explain this in detail, or possibly you know better and your comment is oversimplified and exaggerated for convenience, but, really that's misleading. Because our eyes are trichromatic, a three color additive system is probably most convenient and efficient. But it certainly doesn't most closely reproduce the experience of viewing the real world, for many reasons. You're always going to improve the display quality and come closer to reality with more subdivisions of overlapping bell curves of wavelength ranges.
It's true that when you simplify human vision you get three (+ rods) signals of varying amplitude going into the brain's visual processing and therefore, in theory, you only need to replicate those exact inputs, but in practice that's difficult to do with three light sources. This is a flawed analogy, but think about perceptually-oriented lossy sound encodings like MP3s versus lossless high sampling rates and bit depth -- the former is built around a perceptual model that tries to reduce the signal to only the stuff that gets processed by our ears and brains and which produces a perception that is as close to the original as possible, while the latter tries to include as much of the data of the original signal and reproduce it as closely as possible. The closer to perfect is your model of perception and the more you're able to craft your signal to duplicate the perceptual response to the original signal, then the better this works and, yeah, that's all you really need. But in practice this is really hard to do and the fidelity is at best "good enough". Our visual perceptual color models are the various color spaces and the reason there are quite a few of them and they each have their advantages and disadvantages is precisely because modeling this perceptual space is very difficult and incomplete. All three-color systems are basically lossy perceptual systems that are good enough but very far from perfect. If we had the equivalent image recording and reproduction technology as we do audio, including the bandwidth, it really would just be simpler and better to just reproduce the original signal as closely as possible -- which would be reproducing the spectrum for every point in the visual field.
Don't let the trichromatic aspect of human vision fool you into believing that trichromatic sources are the ideal solution just because it's matching up 3 to 3. The devil is in the details and unfortunately the actual produced display curves and the actual nerve and perceptual response curves only match very loosely, mostly because the nerve and perceptual curves are really complex in a way that's so difficult to match on the display side such that, if fidelity is your goal, you might as well just throw more sampling and hardware at the problem in the form of just trying to duplicate the original signal and worrying less about predicting and evoking the desired perceptual result. Not that anyone really does this, because good enough is good enough and right now these three color reproductions are still right in the sweet spot. But that might not always be true. If full-spectrum sampling and reproduction at every point in an image field became technologically/economically feasible, that would open the door to an enormous variety of improvements in image recording and manipulation and reproduction and the three-color system will seem as quaint and make-do as the prior two-color systems did (which, for those reading along, it is possible to reproduce full-color response using only a two-color display device).
posted by Ivan Fyodorovich at 2:01 AM on April 6, 2016 [6 favorites]
I'm sure someone will come along with more energy to explain this in detail, or possibly you know better and your comment is oversimplified and exaggerated for convenience, but, really that's misleading. Because our eyes are trichromatic, a three color additive system is probably most convenient and efficient. But it certainly doesn't most closely reproduce the experience of viewing the real world, for many reasons. You're always going to improve the display quality and come closer to reality with more subdivisions of overlapping bell curves of wavelength ranges.
It's true that when you simplify human vision you get three (+ rods) signals of varying amplitude going into the brain's visual processing and therefore, in theory, you only need to replicate those exact inputs, but in practice that's difficult to do with three light sources. This is a flawed analogy, but think about perceptually-oriented lossy sound encodings like MP3s versus lossless high sampling rates and bit depth -- the former is built around a perceptual model that tries to reduce the signal to only the stuff that gets processed by our ears and brains and which produces a perception that is as close to the original as possible, while the latter tries to include as much of the data of the original signal and reproduce it as closely as possible. The closer to perfect is your model of perception and the more you're able to craft your signal to duplicate the perceptual response to the original signal, then the better this works and, yeah, that's all you really need. But in practice this is really hard to do and the fidelity is at best "good enough". Our visual perceptual color models are the various color spaces and the reason there are quite a few of them and they each have their advantages and disadvantages is precisely because modeling this perceptual space is very difficult and incomplete. All three-color systems are basically lossy perceptual systems that are good enough but very far from perfect. If we had the equivalent image recording and reproduction technology as we do audio, including the bandwidth, it really would just be simpler and better to just reproduce the original signal as closely as possible -- which would be reproducing the spectrum for every point in the visual field.
Don't let the trichromatic aspect of human vision fool you into believing that trichromatic sources are the ideal solution just because it's matching up 3 to 3. The devil is in the details and unfortunately the actual produced display curves and the actual nerve and perceptual response curves only match very loosely, mostly because the nerve and perceptual curves are really complex in a way that's so difficult to match on the display side such that, if fidelity is your goal, you might as well just throw more sampling and hardware at the problem in the form of just trying to duplicate the original signal and worrying less about predicting and evoking the desired perceptual result. Not that anyone really does this, because good enough is good enough and right now these three color reproductions are still right in the sweet spot. But that might not always be true. If full-spectrum sampling and reproduction at every point in an image field became technologically/economically feasible, that would open the door to an enormous variety of improvements in image recording and manipulation and reproduction and the three-color system will seem as quaint and make-do as the prior two-color systems did (which, for those reading along, it is possible to reproduce full-color response using only a two-color display device).
posted by Ivan Fyodorovich at 2:01 AM on April 6, 2016 [6 favorites]
It's really simple.
People think that there are three dimensions, but physicists know that you really need a six-dimensional matrix to hold all spatial information about a point - x,y,z,dx,dy,dz - so that both its instantaneous position and velocity can be treated.
In squantum biophysics, you need an n(6) matrix (*) for each colour; r,g,b, dr,dg,db, in order to properly describe perceptual squantum qualia of vision. The squant plug-in and display options map this vector space , tesseract-like, into a translated analogue that, while composed of normal light, allows the brain's visual system to synthesise the impression of squantal space This takes a little practice, but the effects are profound, and can be quite easily achieved by anyone with an iIQ of Y+5 (where Y = your IQ). In other words, if you can't see it, you're too dumb - and you are, you great lummox. Go back to your garish NTSC gamuts and leave the grown-ups in peace, you drooling eejit.
(*) This is the simplified theory: in fact, due to polarization, phase and diatromatic coupled spin properties, you need n(14pi/omega squared) to properly characterise vision in Sydjamesian space. However, the Sony 4K monitor capable of displaying this costs three times CERN's entire opex and capex budget for fiscal years 05-15, so nobody has yet been able to save up enough money to properly see this
posted by Devonian at 3:52 AM on April 6, 2016 [2 favorites]
People think that there are three dimensions, but physicists know that you really need a six-dimensional matrix to hold all spatial information about a point - x,y,z,dx,dy,dz - so that both its instantaneous position and velocity can be treated.
In squantum biophysics, you need an n(6) matrix (*) for each colour; r,g,b, dr,dg,db, in order to properly describe perceptual squantum qualia of vision. The squant plug-in and display options map this vector space , tesseract-like, into a translated analogue that, while composed of normal light, allows the brain's visual system to synthesise the impression of squantal space This takes a little practice, but the effects are profound, and can be quite easily achieved by anyone with an iIQ of Y+5 (where Y = your IQ). In other words, if you can't see it, you're too dumb - and you are, you great lummox. Go back to your garish NTSC gamuts and leave the grown-ups in peace, you drooling eejit.
(*) This is the simplified theory: in fact, due to polarization, phase and diatromatic coupled spin properties, you need n(14pi/omega squared) to properly characterise vision in Sydjamesian space. However, the Sony 4K monitor capable of displaying this costs three times CERN's entire opex and capex budget for fiscal years 05-15, so nobody has yet been able to save up enough money to properly see this
posted by Devonian at 3:52 AM on April 6, 2016 [2 favorites]
Recently discovered I'm a polyorchid. I've been waiting for just the right time to out myself. At last, Squant.
posted by lometogo at 4:03 AM on April 6, 2016
posted by lometogo at 4:03 AM on April 6, 2016
This thread really exemplifies that Metafilter sweet spot where some people are just fucking around and some other people are discussing really interesting stuff that I had previously never thought anything about.
posted by Literaryhero at 4:15 AM on April 6, 2016 [16 favorites]
posted by Literaryhero at 4:15 AM on April 6, 2016 [16 favorites]
I can't believe there are brain cells I have that were actually able to sort of recognize this site from when it came out a zillion years ago. The brain is a weird thing.
posted by fungible at 5:10 AM on April 6, 2016 [1 favorite]
posted by fungible at 5:10 AM on April 6, 2016 [1 favorite]
So, Hillary won the squant primary?
posted by Obscure Reference at 5:39 AM on April 6, 2016
posted by Obscure Reference at 5:39 AM on April 6, 2016
Guys I'm super hungover is squant real
posted by beerperson at 6:33 AM on April 6, 2016 [5 favorites]
posted by beerperson at 6:33 AM on April 6, 2016 [5 favorites]
Yeah, maybe it's real and maybe it's not. As opposed to Purple, which definitely doesn't exist.
posted by Mcable at 6:34 AM on April 6, 2016
posted by Mcable at 6:34 AM on April 6, 2016
Are Nike Hyper Elite Platinum basketball uniforms actually squant? (They just look unwashed) to me.
posted by Carol Anne at 6:43 AM on April 6, 2016
posted by Carol Anne at 6:43 AM on April 6, 2016
A few months back, there was a tetrachromaticity test image making the rounds, which a number of my friends reposted on Facebook. Failing to take into account that there is no accepted way to digitally describe a tetrachromatic color space, or that our monitors all use three colors. I rebuked one of my friends over this.
Now that I've found this plugin, I guess I need to apologize to her.
posted by adamrice at 7:29 AM on April 6, 2016 [1 favorite]
Now that I've found this plugin, I guess I need to apologize to her.
posted by adamrice at 7:29 AM on April 6, 2016 [1 favorite]
This thread really exemplifies that Metafilter sweet spot where some people are just fucking around and some other people are discussing really interesting stuff that I had previously never thought anything about.
and you are not sure which is which.
posted by srboisvert at 7:31 AM on April 6, 2016 [3 favorites]
and you are not sure which is which.
posted by srboisvert at 7:31 AM on April 6, 2016 [3 favorites]
But could Bronze Age ancient Greeks see squant? The squant-dark sea?
posted by branduno at 7:32 AM on April 6, 2016 [5 favorites]
posted by branduno at 7:32 AM on April 6, 2016 [5 favorites]
If there is another primary, 500nm green (which looks nearly gray to colorblind people like me) would be a great choice, making available a whole bunch of highly-saturated greens that are inaccessible now, at the peak receptive wavelength for the night-vision rods.
There has been occasional research toward a four-primary color matching system, but more recent research suggests that the visual system mostly treats rod input as desaturation of whatever the cones are reporting.
posted by enf at 8:15 AM on April 6, 2016 [1 favorite]
There has been occasional research toward a four-primary color matching system, but more recent research suggests that the visual system mostly treats rod input as desaturation of whatever the cones are reporting.
posted by enf at 8:15 AM on April 6, 2016 [1 favorite]
I can't believe there are brain cells I have that were actually able to sort of recognize this site from when it came out a zillion years ago. The brain is a weird thing.
Wasn't this a Slashdot April Fools gag in, like, 1999 or something?
posted by thecaddy at 8:16 AM on April 6, 2016
Wasn't this a Slashdot April Fools gag in, like, 1999 or something?
posted by thecaddy at 8:16 AM on April 6, 2016
See also the "Eclipse of Titan" illusion that allows you to briefly see shades of green and cyan that don't appear in nature. This is done through selectively overstimulating one or more types of cones in your eye.
posted by Robin Kestrel at 8:28 AM on April 6, 2016 [3 favorites]
posted by Robin Kestrel at 8:28 AM on April 6, 2016 [3 favorites]
The human eye's chromaticity diagram is roughly a rounded triangle, with red, green, and blue at the triangle's vertices. That's why we chose red, green, and blue as our additive primaries (and also why the RGBY TVs with the yellow LEDs didn't do much to extend the monitor's gamut).
posted by rocket88 at 8:33 AM on April 6, 2016
posted by rocket88 at 8:33 AM on April 6, 2016
The sRGB primaries are kind of arbitrary: they are a compromise between useful coverage of the chromaticity plane and reasonable manufacturability. The original NTSC specifications called for a larger color triangle that got scaled back over time to trade off saturation for brightness. Many displays today can't cover even the smaller sRGB triangle, but displays with larger gamuts that go beyond it are starting to come onto the market, for instance in Apple's new hardware.
sRGB red and blue are in about the right places, getting close in both cases to the line of purples. You don't want to push the wavelength of either of them much further out than it is or you'll start to lose brightness, especially for protanopes whose red cuts off at a lower wavelength than for other people.
Green is the one that requires more careful consideration of tradeoffs. If you move it to a lower wavelength, you lose the ability to have saturated yellows, and yellow is already the least perceptibly saturated hue to normal vision. If you move it to a higher wavelength, you've effectively got a red-yellow-blue color triangle, with gray on the yellow-blue line, and have lost green entirely.
A four-primary system with a lower-wavelength green and a new yellow primary ought to maximize coverage, but it's not going to do you any good to just feed normal video into it, since all the colors in your source have already been clipped to the normal sRGB triangle. You'd need new images that are encoded in terms of the larger range.
posted by enf at 11:37 AM on April 6, 2016
sRGB red and blue are in about the right places, getting close in both cases to the line of purples. You don't want to push the wavelength of either of them much further out than it is or you'll start to lose brightness, especially for protanopes whose red cuts off at a lower wavelength than for other people.
Green is the one that requires more careful consideration of tradeoffs. If you move it to a lower wavelength, you lose the ability to have saturated yellows, and yellow is already the least perceptibly saturated hue to normal vision. If you move it to a higher wavelength, you've effectively got a red-yellow-blue color triangle, with gray on the yellow-blue line, and have lost green entirely.
A four-primary system with a lower-wavelength green and a new yellow primary ought to maximize coverage, but it's not going to do you any good to just feed normal video into it, since all the colors in your source have already been clipped to the normal sRGB triangle. You'd need new images that are encoded in terms of the larger range.
posted by enf at 11:37 AM on April 6, 2016
"I'm sure someone will come along with more energy to explain this in detail, or possibly you know better and your comment is oversimplified and exaggerated for convenience, but, really that's misleading."
Sorry you feel that way, and it is oversimplified, but it's really pretty literally true.
This isn't because a three-color system gives the most accurate results, or even because it most closely matches our overall color sensitivity, but because those three colors have the highest degree of mutually distinguishable contrast. You could represent more colors by doing more complex modeling, but the more colors you have, the harder it is to distinguish them from each other — right down to the ultimate end of a continuous spectrum with no real boundaries at all. And by the time that you're really concerned with more accurately reproducing entire visual fields, you should already know that it's unnecessary, because you never see the actual light — your brain is doing all sorts of crazy color processing all the time, especially with color consistency mapping, to the extent that a classic experiment is to show that you can show someone the exact same frequency flux (with no reproduction) in two instances and have the brain perceive it as two different colors (the "Mondrian tests").
So, yeah, going beyond RGB allows you to expand your colorspace, and modeling that off of perceivable frequencies does incrementally improve your color-depth perception, but what you see isn't really what's displayed and past a certain point color perception rapidly becomes less about how many frequencies you can accurately display and more about how your brain works with the models that it's given.
posted by klangklangston at 11:42 AM on April 6, 2016
Sorry you feel that way, and it is oversimplified, but it's really pretty literally true.
This isn't because a three-color system gives the most accurate results, or even because it most closely matches our overall color sensitivity, but because those three colors have the highest degree of mutually distinguishable contrast. You could represent more colors by doing more complex modeling, but the more colors you have, the harder it is to distinguish them from each other — right down to the ultimate end of a continuous spectrum with no real boundaries at all. And by the time that you're really concerned with more accurately reproducing entire visual fields, you should already know that it's unnecessary, because you never see the actual light — your brain is doing all sorts of crazy color processing all the time, especially with color consistency mapping, to the extent that a classic experiment is to show that you can show someone the exact same frequency flux (with no reproduction) in two instances and have the brain perceive it as two different colors (the "Mondrian tests").
So, yeah, going beyond RGB allows you to expand your colorspace, and modeling that off of perceivable frequencies does incrementally improve your color-depth perception, but what you see isn't really what's displayed and past a certain point color perception rapidly becomes less about how many frequencies you can accurately display and more about how your brain works with the models that it's given.
posted by klangklangston at 11:42 AM on April 6, 2016
Website uses HTML5. No HTML5 browsers work on Windows NT. Some help here?
posted by clawsoon at 1:33 PM on April 6, 2016
posted by clawsoon at 1:33 PM on April 6, 2016
This is going to lead to a whole new field of study: Squantum Physics.
posted by Burn_IT at 3:28 PM on April 6, 2016 [1 favorite]
posted by Burn_IT at 3:28 PM on April 6, 2016 [1 favorite]
"Setup was unsuccessful." :-(
Time to dig out my Windows 98 box...
posted by numaner at 7:16 PM on April 6, 2016
Time to dig out my Windows 98 box...
posted by numaner at 7:16 PM on April 6, 2016
Dammit, I forgot about the new tertiary and quaternary colors:
Squant, red and green make what?
Squant, red and green make FRADLOW!
Squant, red and blue make what?
Squant, red and blue make SCLUSS!
Squant, green and blue make what?
Squant, green and blue make POEDRALT!
Also four primaries make White Plus (TM)
posted by BiggerJ at 10:39 PM on April 6, 2016
Squant, red and green make what?
Squant, red and green make FRADLOW!
Squant, red and blue make what?
Squant, red and blue make SCLUSS!
Squant, green and blue make what?
Squant, green and blue make POEDRALT!
Also four primaries make White Plus (TM)
posted by BiggerJ at 10:39 PM on April 6, 2016
« Older Rule 34 - If something exists, there is porn of it... | Djrum: aiming for the dancefloor, but ending up... Newer »
This thread has been archived and is closed to new comments
posted by benito.strauss at 6:42 PM on April 5, 2016 [32 favorites]