Science: Light
Extra hue between indigo and violet   (+4, -5)  [vote for, against]
Address rotational value symmetry in HSV colour space

Yellow is the brightest colour, indigo the darkest. Red and violet are respectively darker and brighter than yellow and indigo. Green and blue are intermediate in brightness. This means that in terms of subjective impressions of how bright a spectral colour is, ignoring the extremes of the visual spectrum, value is rotationally symmetrical about some axis in the green or blue area. However, concerning colour terms, this is never the case. Implicational linguistic universals predict the distribution of basic colour terms in different languages, but there never seems to be a colour term for the part of the spectrum between indigo and violet.
Why should there be a colour term there? Because there's one for orange. In terms of the symmetry concerned, violet currently covers the "opposite" of both orange and red, yet people with optimal colour vision whose first language has a basic colour term for orange never have one for indigo-violet.

I propose that one be invented, to redress the asymmetry. Indigo-violet is a distinct colour because it's brighter than indigo but dimmer than violet, and just as orange is "bright brown", indigo-violet is "dark white" in a sense (or something). It has a distinct character like all the others but cannot currently be referred to. It's a real example of Hume's "missing shade of blue" (except it's not blue).
-- nineteenthly, Dec 03 2008

here's something interesting... http://laser.physic...04/reports/jasleen/
[xandram, Dec 03 2008]

Visual Snow http://www.visualsnow.com/
One source is called Hallucinogen Persisting Perception Disorder, from the use of LSD and other drugs [ldischler, Dec 04 2008]

Visual Snow forum http://thosewithvis...ral-Discussion.html
[ldischler, Dec 04 2008]

Four indigos http://en.wikipedia...ur_shades_of_indigo
[ldischler, Dec 04 2008]

eye floaters http://www.eye-floa...whatarefloaters.php
[Spacecoyote, Dec 04 2008]

Crappy picture of a CD http://i122.photobu...y/falsespectrum.png
In various ways - hues are wrong, blurry, JPEGification probably spoilt it too. [nineteenthly, Dec 16 2008]

Spectral monitor Spectral_20monitor
Doesn´t solve the pigment problem. [nineteenthly, Dec 22 2008]

Full visual spectrum Monitor Full_20visual_20spectrum_20Monitor
how to show violet properly: new hardware required. [Loris, Jun 25 2009]

intriguing, but a link to an example colour wheel would be helpful.
-- conskeptical, Dec 03 2008


OK.
-- nineteenthly, Dec 03 2008


I thought this would be an attempt to add letters to "ROY G BIV" in order to spell a real word.
-- phundug, Dec 03 2008


That would also be good, maybe giving a clue as to the name, which perhaps needs to be discovered rather than invented.
I have a problem. I've spent quite some time trying to get a realistic indigo-violet colour, but i'm hampered by the differences between HSV and RGB. The colour wheel looks OK until it gets to blue, then gets steadily less accurate until it bears no relation. What the hell is with RGB?! It's absolutely nothing like real colour at that point! Presumably this is because of not being able to go beyond blue.
-- nineteenthly, Dec 03 2008


It's "Roll Over You Great Big Innocent [ ... ] Virgin".
-- neelandan, Dec 03 2008


I've always thought it was that bloke in Wizzard.
-- nineteenthly, Dec 03 2008


Call the new color "hugh".
-- phoenix, Dec 03 2008


"Roll Over You Great Big Innocent Huge Virgin"?
-- nineteenthly, Dec 03 2008


Since you are rolling off the top of the visual sensitivity index (note how little contrast the eye perceives in that range) we aren't able to make color distinctions. If we were tuned to include that range we would need a whole new color wheel because the colors required to add up to white and black (compliments) would be different. We inaccurately presume that the rules of our very limited visual system are universal rules (we are gods after all not just fruit loving upright apes). Birds and dogs would have completely different color wheels and might well find our TVs, and art, completely worthless for not respecting the rules of their visual system, or "color wheel". There is no strong halftone between I and V in our retina.
-- WcW, Dec 03 2008


OK...

Two questions then, which really would help me:

Firstly, does a rainbow in a photograph, film, video, monitor or other visual representation look similar to a real rainbow, or is there a glaring difference, and if so what?

Secondly, does a clear sky look uniformly blue to you or can you see something else of a different colour in it? If so, what colour and what does it look like?

I'm not kidding. I'm beginning to suspect my eyesight might be a bit strange.
-- nineteenthly, Dec 03 2008


Can't speak for anyone else but for me when I stare at a clear blue sky it looks sort of like a moving pointalist painting or blue TV static with tiny dots of every color popping into and out of existance very rapidly.
I used to try and explain this to people but no one else seemed to know what I was talking about so I stopped.

Until now.
-- 2 fries shy of a happy meal, Dec 04 2008


Nineteenthly- as to your first question the answer is yes, modern media are finely tuned to fool the eye into thinking that a full spectrum of colors are present. The real thing is unbeatable but as far as colors go the whole idea is that it is as close as possible. The human brain however does lovely and wonderful things with the very limited information it receives and everyone's visual system developed in a novel and completely unique way. Your vision cannot be translated into my vision. Secondly, to me the sky is a gradient of color from one horizon to the other but i have experienced remarkable visual experiences during migraines that allow me to believe that a wide range of visual experiences are probable. Because the brain is struggling to make distinctions in the upper range the chromatic response will vary considerably as your eyes and brain saturate, adapt, and it will be heavily biased by nearby colors.
-- WcW, Dec 04 2008


//when I stare at a clear blue sky it looks sort of like a moving pointalist painting or blue TV static with tiny dots of every color popping into and out of existance very rapidly.//

Very likely this is the permanent aftereffect of LSD use.
-- ldischler, Dec 04 2008


//when I stare at a clear blue sky it looks sort of like a moving pointalist painting or blue TV static with tiny dots of every color popping into and out of existance very rapidly.//

That sounds like a combination of "floaters" and auto "phosphenes" - sometimes described as "visual snow".
-- zen_tom, Dec 04 2008


Thanks for the links. All this time I thought it was just me.
The information is both interesting and more than a little disturbing.

There *are* floaters in my vision and I used to tell my parents that I could see molecules when I was old enough to know what a molecule was.
The thing is, all bright solid colors look snowy to me and always have. It doesn't seem to impair my ability to discern between hues.
The fact that I tried both LSD and shrooms when I was a teenager would seem to support the hallucinogen after effect theory, if it weren't for the fact that I saw this way even as a child.
The dissociative bit has happened to me the few times I've been backed into a corner and forced to fight, (the whole time dilation, lack of pain thing is helpful in such situations but the effect scares the shit out of me so I go out of my way to avoid physical confrontation), and I've always had tinnitus.

Daaang.
I'm even more messed up than I thought.
-- 2 fries shy of a happy meal, Dec 04 2008


//I'm even more messed up than I thought// you're not a BIT messed up - stop it!
-- po, Dec 04 2008


If I let my eyes lose focus I see floaters, too. Apparently, they have nothing to do with LSD (which I've never used) and are caused by cells that naturally move around in the vitreous solution in the eye. This worsens with age. [link] Most people just don't notice it.
-- Spacecoyote, Dec 04 2008


[2 fries] I called them molecules when I was little, too. I didn't know what a molecule was, just that they were tiny and vaguely mysterious... so of course that's what those things floating through the sky at such a stately pace were! gotta love kid logic.
-- ryokan, Dec 05 2008


No logic like it.
-- 2 fries shy of a happy meal, Dec 16 2008


//What the hell is with RGB?!// RGB is a deeply flawed and misleading concept. The centres of sensitivity of the 3 most common types of cone cells are Orange, Green, Blue. The wavelengths needed to simulate colours by additive mixing are Red, Green, Violet. It is not possible to produce the sensation of violet by any mixture of RGB.

Take a digital photo of a spectrum (eg a CD) under sunlight or halogen light, and then compare the photo with the real thing. Violet is conspicuously absent in the photo. (Assumes good colour vision).
-- spidermother, Dec 16 2008


//16bit Pro Photo RGB// My assertion is that it is impossible to create the sensation of violet, even in theory, with any RGB system, no matter what the gamut or bit depth. Pure blue light affects both the "blue" and the "green" cones. To affect only the "blue" cones (and thus create the sensation of violet) pure light of a wavelength shorter than blue (ie violet light) is needed.

<Arm waving> As to the idea, I thought that only a minority of people could distinguish between indigo and violet, so still fewer would be able to distinguish an intermediate colour. Is that why the word for such a colour is missing? Also, naming colours is a cultural thing. I've heard that some cultures do not make, for example, the same distinction between blue and green as westerners do. Also, the word for the colour orange is relatively recent, being derived from the word for the fruit </Arm waving>.
-- spidermother, Dec 16 2008


Agreed. And for (most) humans, you would need a red-green-violet system to get good colour representation.
-- spidermother, Dec 16 2008


I should mention that since conventional HSV colour space is a transformation of RGB, my comments apply there equally. While I respect your desire to make a small addition in order to improve the elegance and symmetry of a system, the fundamental problem remains that indigo and violet are not represented at all; those names, when applied to any parts of an RGB or HSV colour space, are incorrect, and the same would apply to an intermediate colour.
-- spidermother, Dec 16 2008


Do you know Berlin and Kay´s "Basic Color Terms"? It established universal linguistic structures for colour terms, such as the fact that languages with different words for violet and red never lack a term for green. Homer used four colour terms, for yellow/green, red/purple, black and white There´s an order for the terms used. Russian and Hungarian are suggested as examples of languages with extra basic colour terms. Russian has sinij and goluboj for different blues. Latin also has two: caeruleus and azureus - sky blue and, well, "frog" blue i suppose.
What i´m saying is this. I don´t think the problem is primarily perception, but the stability and durability of the concept. If there´s an extra term, people will see the extra colour, within limits, and if perception of differences is poor, it could actually do with some help by inventing an extra term, but it might fail because it´s not "natural" to have it.
My solution to this, as you may have read, was to introduce a musical-style twelve-hue scale for the spectrum, whereof some (the "white" colours) would be more obvious and others (the "black" colours) less so.
Call it deacon blue.
-- nineteenthly, Dec 16 2008


//musical-style twelve-hue scale for the spectrum// Oddly, I thought of mentioning the 12 pitch equal temperament as another overly simplified and incorrect description, in this case of what musical intervals are. A friend told me that she always thought the notes on her piano sounded slightly wrong compared to where she thought they should be, and wondered if her ear was wrong. When I explained that the piano's intervals are out of tune (tempered) and that her ear was probably correct she was annoyed that her music teachers had never been able to explain it. Likewise, you are right in suspecting that a real rainbow does not look the same as its visual representation, but conventional teaching does not explain that the RGB theory is a fudge.

(Later) I can't convincingly see Indigo and Violet as distinct colours in a CD spectrum. I might have to invest in a decent prism, and learn to tell them apart, then see if I can graduate to indigo-violet.
-- spidermother, Dec 16 2008


[UB], you have real sun where you live, whereas we just have a watery grey thing. That might make a difference to us.
[Spidermother], the light in this room has an unusually high colour temperature, and if i fetch a CD i get a huge range of hues. The photo i´m about to post a link to is way off (and also crap), but it shows you a pretend spectrum. I wonder if your artificial light is a funny colour rather than your colour vision being off.

Later: That picture is shockingly awful! Is it the camera? I mean hue-wise.
-- nineteenthly, Dec 16 2008


My immediate reaction to that is, "you jammy bastard", since it´s getting dark at three pm here, but i expect the sun shines brighter on the other side.
-- nineteenthly, Dec 16 2008


//To affect only the "blue" cones (and thus create the sensation of violet) pure light of a wavelength shorter than blue (ie violet light) is needed.//

There is no wavelength of light which affects the blue cones without affecting either the red or green ones to at least some degree. The result of that is that while there are only three types of hue receptors, the inability to have color components with negative magnitudes means that no set of three primary colors will suffice to achieve all of the optimally-saturated hues a person could achieve.

On the other hand, there's a reverse observation that's also interesting: magenta is a hue which is clearly perceptible and recognizable, but which does not appear in the spectrum because the 'magenta' sensation cannot be created using only a single wavelength.

//On a related note, many people complain about compact flourescent bulbs purely because of the colour temperature the cheaper bulbs have. Try shopping around for ones with a rating of 5000K.//

More important than color temperature, IMHO, is spectral uniformity. I wish bulb makers would release spectral plots the way loudspeaker manufacturers used to. It's possible for a bulb to produce light which looks 'white', and yet for the colors of objects illuminated by that bulb to be distorted in weird ways.
-- supercat, Dec 17 2008


[supercat] Your explanation is probably better than mine; but the standard "the primary colours of light are red, green, and blue since they can be mixed to produce any colour" is complete bollocks, and the inability to produce violet is an easy way to demonstrate this. I'm making a point similar to [nineteenthly]'s in that red and violet are the last colours to be seen as wavelengths get longer and shorter respectively, so RGV has some kind of symmetry and sense, whereas RGB does not.

[UnaBubba] Perhaps your wife is a tetrachromat (rare, and found only in women) or (less exciting) you have a form of colourblindness. Or she might just be more practiced in distinguishing colours.

[supercat] I want the same from paint manufacturers. Isn't the colour rendering index supposed to do what you ask, better than the colour temperature?

[bigsleep] White LEDs have a much more continuous spectum than flourescents, as a quick CD test will reveal. They still have a peak in the blue. I use a couple of naughty incandescents for anything where colour is important and daylight isn't available.
-- spidermother, Dec 17 2008


//I'm making a point similar to [nineteenthly]'s in that red and violet are the last colours to be seen as wavelengths get longer and shorter respectively, so RGV has some kind of symmetry and sense, whereas RGB does not.//

RGB can display any hue, though there are some hues (violet among them) that it cannot display at maximum saturation. RGV would be unable to display blue at any degree of saturation.
-- supercat, Dec 17 2008


//I could try to simulate a balanced spectrum, but the human eye perception could mess with any plans of equally spacing wavelengths.//

The need for spectral uniformity comes not from human eye perception, but rather the spectral behavior of various objects one might try to illuminate. Some objects and materials transmit or reflect some wavelengths much better than other perceptually-similar ones. If such an object is illuminated by a light source with a non-uniform spectrum, it may appear strange and unnatural. Even a pan-chromatic black and white film could capture the fact that while two light sources may be indistinguishable when used to view one object, they may produce totally different results when illuminating another.

BTW, fun little science experiment: shine a yellow LED through a red piece of plastic. Primary-color theory would suggest that since yellow is red+green, the result should be red. If the "yellow" LED is in fact red+green, that will indeed be the result. Many yellow LEDs, however, emit a relatively narrow-wavelength yellow light, which will be significantly attenuated by the red plastic, but will still appear yellow.
-- supercat, Dec 17 2008


//RGB can display any hue// I would still beg to differ. RGB cannot display violet _at all_. If you look at digital images of a spectrum, the end labelled 'violet' is mostly blue, with some red mixed in, and is therefore purple, not violet. Violet light stimulates 'green' cones less than does blue light, rather than stimulating 'red' cones more.

I confess that I'm not entirely sure that RGV would be better than RGB, or would be able to display blue well. I suspect that RGB is used because people aren't used to distinguishing violet, and don't really miss it, whereas poor rendering of the more popular colours would be noticed.
-- spidermother, Dec 18 2008


//I confess that I'm not entirely sure that RGV would be better than RGB, or would be able to display blue well. I suspect that RGB is used because people aren't used to distinguishing violet, and don't really miss it, whereas poor rendering of the more popular colours would be noticed.//

Ideally, one would have primary colors chosen so that each primary color would stimulate one type of cone while leaving the other types completely unstimulated. Unfortunately, that isn't possible, since all wavelengths that stimulate what are commonly called the "blue" receptors also stimulate the red or green receptors to some degree.

Typically, the color "blue" is chosen to as to minimize the total amount of stimulation to both the red and green receptors. One can produce any hue of blue-cyan, but saturation will be imperfect because of a little stray red; one can produce any hue of red-magenta, but saturation will be imperfect because of a little stray green.

Adding cyan (shortest wavelength with essentially no red) and magenta (longest wavelength with essentially no green) as additional primary colors would allow better saturation of colors in the blue-cyan and blue-magenta regions, though one would still likely achieve optimal saturation only at the particular primary colors chosen. A mix of blue and magenta, for example, would stimulate the green receptors more than would a pure wavelength yielding the same apparent hue.
-- supercat, Dec 19 2008


OK then, how about magenta, yellow, cyan and violet? Is that completely off the wall? I think i broke my brain.
-- nineteenthly, Dec 19 2008


Red, green, and blue can work pretty well at producing other hues. No other hues can mix well to produce red, green, or blue. Adding other colors to red, green, and blue can improve things, but losing red, green, or blue would make things worse.
-- supercat, Dec 22 2008


Well, i did the spectral monitor thing. See link.
-- nineteenthly, Dec 22 2008


//supercat// I realised later that I was defining violet as a spectral colour but neglected to say so, and therefore my disagreement with you was purely semantic. Sorry.

I still suspect that violet light would be better at stimulting blue cones in isolation than blue light. It would be less energy efficient, of course, since the peak sensitivity is at blue, but exactly the same criticism can be made of the use of red rather than orange at the other end.
-- spidermother, Dec 23 2008


//I still suspect that violet light would be better at stimulting blue cones in isolation than blue light.//

The "red" cones are sensitive to both long and short wavelengths of light. That's the reason that the perceptual hue resulting from such wavelengths is like that produced by a mixture of red and blue light. If one were trying to minimize the amount of green-code stimulation for a given amount of blue-cone stimulation, monochromatic violet would be better than blue. On the other hand, monochromatic violet would stimulate the red cones a lot more than would blue. Blue is chosen as a pretty good compromise between cyan and violet.

BTW, the fact that the spectral response of a particular camera does not match that of the human eye does mean that RGB cannot accurately reproduce all hues. It simply means that the spectral response of the red/green/blue filters on the camera doesn't match that of the human eye.
-- supercat, Dec 23 2008


Frustratingly, sources differ on the response of 'red' (L) cones to short wavelengths. As far as I can tell, there is a second, short wavelength, peak in absorbance, but not in responsivity.
-- spidermother, Dec 23 2008


It seems to me that the violet response from "red" cones is the result of violet light being around twice the frequency of red. Is that at all possible? If so, doesn´t that make ultraviolet bluish-yellow?
-- nineteenthly, Dec 23 2008


STOP with all this colour rubbish, everything is black. All else is but a pigment of your imagination.
-- eight_nine_tortoise, Dec 23 2008


//It seems to me that the violet response from "red" cones is the result of violet light being around twice the frequency of red. Is that at all possible? If so, doesn´t that make ultraviolet bluish-yellow?//

I should ask my retired physics prof. Since he had cataract surgery, he became able to see a couple of extra lines in the mercury spectrum; I should ask what color those lines were.

For people with biological lenses in their eyes, too much UV is going to get absorbed in the lenses to produce a visible image in any color.
-- supercat, Dec 23 2008


Hmm, maybe i should avoid treating some of my patients´ eyesight properly, then ask them to describe what they see. The trouble is, most of them have glaucoma and macular degeneration rather than cataracts.
-- nineteenthly, Dec 23 2008


//[UnaBubba] Perhaps your wife is a tetrachromat (rare, and found only in women)//

Another study suggests that as many as 50% of women and 8% of men may have four photopigments.

//I'm not kidding. I'm beginning to suspect my eyesight might be a bit strange.//

[nineteenthly] I did not mean to side track that question with vs, what do You see when you look into a bright blue sky?
-- 2 fries shy of a happy meal, Jun 25 2009


anybody else look at the Wikipedia link to the 4 colours of indigo and see blue/purple, pink/purple, purple and dark green/blue(wtf) ?
-- FlyingToaster, Jun 25 2009


[Fries], no problem. I see large, stationary purple blobs which look a little like cumulus clouds. Since they don't move when i move my eyes, i don't think they're hallucinations, but they could conceivably be an optical illusion involving contrasting colours. It's true that my eyesight is strange, since i have vertical blind "stripes" and other scotomas which move around a lot. Someone suggested that for some reason i can see thermals, which apparently some glider pilots and hang-gliders can see, so i've heard.
[FlyingToaster], monitor problem? I see three colours which look nothing at all like indigo to me and a final colour which looks pretty like my idea of indigo.
-- nineteenthly, Jun 25 2009


Fascinating.

I am disappointed at the amount of research that has been done about these visual phenomena, <he says as though he has the right to gripe>, you’d think that something acknowledged by the ancient Greeks would have a bit more historical data to chew on.
Given your line of work and the fact that you experience visual quirks I am betting that you have dug up many tid-bits of knowledge unknown to the mainstream on this and related subjects.
If there's a chance that you ever felt like sharing this information with someone like, oh, say... myself for instance, then I think the chances of you finding a more motivated pupil would be slim.

Just say'n.
-- 2 fries shy of a happy meal, Jun 25 2009


// a more motivated pupil //

Ha hah, Hahahaha !

That almost counts as humor .... or a pun ......

Ohhhhhh ....

<listens for drum/cymbal sting>
-- 8th of 7, Jun 25 2009


//monitor problems?//
I don't think so, but referring to "Four Indigos" link, Indigo dye is actually blue-green go figger... I never thought about the differences between various purples though I guess "Persian Indigo" ("Shades of Violet" at the bottom of the page) is what I think of as Indigo. The sun burns too much to look at the sky these days but last time I ventured out sans shades it was pretty well just "sky blue". Eyes are useless outside the "visual spectrum" but I can feel X-Rays if that gets me membership in the club :)
-- FlyingToaster, Jun 26 2009


[Fries], i don't know why i get the purple blobs thing, but i have hypotheses about the rest. It doesn't make sense that i would be able to see ultraviolet, because it doesn't look like that (it's more like blue-white, apparently, because it stimulates all the cone cells), and in any case tetrachromats have their extra cone cells in the "lime to mustard" portion of the spectrum.
The stripy thing is probably connected to my other thing: specific objects trigger off hallucinations. Brick walls from a long way off, small diagonal tiles and in particular zip teeth set off migraine-type flickering patches which look like animated chessboards, a definite black and white colour, not the swirly things most people see when they look at a finely divided visual field, and occasionally complex visual hallucinations like images of thorns.
I think there's something strange about the way my brain processes some higher-order visual information. I also get double vision in one eye sometimes, which i would have thought impossible, but it happens.
[FlyingToaster], the thing is, because i'm a herbalist probably, when i think of indigo, i tend to think of the dye plant indigo and its dye rather than the colour, and that's very close to the colour indicated. Concerning the thermals, i think it might just be damp air.
-- nineteenthly, Jun 26 2009


//concerning the thermals...//
The brain can do a marvellous job of interpreting and re-presenting data not percieved consciously, though yours (the sky thing) would probably technically be considered very mild synesthesia if there really is no 1:1 cause/effect. Or of course from testing the medicinal mushrooms a bit much ;) Are you actually seeing double in one eye or is it just a feeling, like "persistance of hatness".
-- FlyingToaster, Jun 26 2009


Actually, on the subject of mushrooms, many years ago i did try them (they were semi-legal at the time, don't know about now) and in fact ate several times the safe dose along with some friends. They experienced a marked temporary mental change and nothing at all happened to me.
I don't mean afterimages, floaters or phosphenes. These are solid-looking stereoscopic visual impressions, which is why i think they're happening somewhere beyond the visual tracts and visual cortex. Synaesthesia is maybe what's going on there but i don't know. Not with the phosphene-like experience though. The blobs definitely behave visually like they're external to my head.
[FT], i know what you mean about your marvellously-named "persistence of hatness", but that isn't it. I close one eye, look at an object with or without glasses, and occasionally i see two of them, next to each other. It only works if it's a few minutes of arc across. I'm astigmatic, so maybe it's a lens thing.
-- nineteenthly, Jun 26 2009



random, halfbakery