Can You Pass the Online Color Vision Test?

[collinsmark]

It might be useful to recall that a couple of decades ago, there was a lot of research and development that went into laser televisions. These were big screen TVs where lasers were used instead of the more conventional filtered lights, for rearscreen projection. Of course this was all canceled when plasma and LCD flatscreen TVs took off as they did.

But laser TVs illustrate the point. You can't get narrower bandwidth of the primary colors any more than using lasers. Lasers have extremely narrow bandwidths. And as you may know, you can't just simply change the peak wavelength of a given laser -- lasers operate at particular wavelengths.

In order to approximate the "near green" as shown in color B, the laser TV would shine the green laser together with a little bit of the blue laser, and that would fool the trichomatic human eye into thinking that color B was being shown, even though it wasn't really, in reality. It was formed by a very narrow peak at wavelength A plus a touch of the very narrow blue wavelength. And that's enough to fool the silly human into thinking that B was being shown. Most humans can't tell the difference.

 

[Algr]

What does it have to do with RGB monitors making poor test equipment for tetrachromats? (Which was the original issue.)

Monitors are RGB. Tetrachroma is RGBX by definition. Even if X is inadvertently included in one of the primaries, you would not be able to control it independently from whatever other channel it was lumped in with. A black and white monitor technically outputs RGB, but without independent control of the colors, you can't tell anything about a person's color vision using one.

 

[DaveC426913]

Your diagrams are incorrect, when applied to an RGB monitor. You have put two green wavelength peaks (A and B) in your diagrams. And RGB monitor is not capable of producing color peaks so close together. An RGB monitor produces one, and only one, peak of green wavelength.

You're misinterpreting the diagram. I debated how much detail to put in before it got confusing.

Those vertical lines are not narrow light frequencies, they are combined display colours. They are (for argument's sake) RGB 0,193,191 and 0, 191, 193. A monitor is certainly capable of displaying them.

 

[DaveC426913]

I don't know how useful this conversation is anymore. We are talking past each other. You are certain of your setup and results, and I find you keep missing the mark. Whereas you feel the same way about mine.

But I am in a stronger position because your claim (all X's cannot be done) is harder to back up. You must show all claims are invalid, including mine. (It is pointless for you to put forth a argument then turn around and demonstrate that it is false.) All I have to do is demonstrate that a single argument (Y could be done) can't be invalidated. I have the luxury of putting forth an argument that suits my premise; the onus is on you to falsify the argument of my choosing.

 

[collinsmark]

You're misinterpreting the diagram. I debated how much detail to put in before it got confusing.

Those vertical lines are not narrow light frequencies, they are combined display colours. They are (for argument's sake) RGB 0,193,191 and 0, 191, 193. A monitor is certainly capable of displaying them.

"Combined dispaly" colours? I'm not sure what that means. Perhaps we are talking past each other due to a misunderstanding of how an RGB monitor works.

Below is a more qualitative (not to scale) way of how an RGB monitor produces the colors RGB [0,193, 191] and [0, 191, 193]. Does it make sense now how this is not an adequate tool to use for a tetrachromacy test?

Notice that after combining (neglecting any slight overlap, which I think is a valid approach, given the narrower, primary color bandwidths produced by the monitor), the individual peaks do not change frequency (reciprocal of wavelength), at least not significantly. They only change their relative intensities.

 

[collinsmark]

Here's a link to a paper that gives a more quantitative approach on RGB monitors' relative intensities of color rendering.

http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CB8QFjAA&url=http://www.wseas.us/e-library/conferences/2009/tenerife/CSECS/CSECS-15.pdf&ei=iUFrVceJJcu2oQTTsoD4DA&usg=AFQjCNFac3dN0tdylrOgXBpzcTfc9QMXgQ&sig2=9f7jyevPl-hsCKbbX9hnqA&bvm=bv.94455598,d.cGU&cad=rja

Figure 1 shows the relative sensitivity of the cones (receptors in the eyes of a typical trichromat).

Figures 3 and 4 show the relative intensities of a CRT monitor and an LCD monitor. Note that the red, green and blue color bandwidths are much narrower in the monitors (which are transmitting the light) than they are in the cones (which are receiving the light).

Edit: Here's an interesting quote from the paper. It's in reference to the CRT monitor, but the same idea applies to the LCD monitor, albeit different peaks:

"Such spikes are not
commonly found in nature, and consequently the
CRT emissions almost never match the spectral
power distribution found in the original scene. The
color match can only be arranged basing on the
eye’s inability to distinguish between different
spectral power distributions (metamerism) [4,7]."​

 

[DaveC426913]

"Combined dispaly" colours? I'm not sure what that means.

I'm saying the diagram is not depicting two narrow frequencies of light; it is depicting two colours (i.e. 0,191,193 and 0,193,191).

Perhaps we are talking past each other due to a misunderstanding of how an RGB monitor works.

No. I know how they work. Being in both software/.hardware as well as photography, I have a better thna average understanding.

Below is a more qualitative (not to scale) way of how an RGB monitor produces the colors RGB [0,193, 191] and [0, 191, 193]. Does it make sense now how this is not an adequate tool to use for a tetrachromacy test?

Perhaps this is a place where we can meet. As I said before, it is certainly not ideal to use RGB monitors for study, I was simply making the point that a tetrachromat could distinguish colours better than a typical person, even when looking at a monitor.View attachment 84305

Notice that after combining (neglecting any slight overlap, which I think is a valid approach, given the narrower, primary color bandwidths produced by the monitor), the individual peaks do not change frequency (reciprocal of wavelength), at least not significantly. They only change their relative intensities.[/QUOTE]
Correct. And a tetrachromat will be able to perceive them better.

I'm annoyed, I should have just done the diagram accurately, and damn the information overload.

 

[Algr]

I was simply making the point that a tetrachromat could distinguish colours better than a typical person, even when looking at a monitor.

I don't think you can say that at all, even if your interpretation of how monitors work is correct. What if the addition of the fourth color introduces noise in the other three? Or the brain might have a fixed amount of attention to such details - more color channels means less detail in each channel. (Think of audio, where we have hundreds of "channels", but only two "pixels".) Or natural variation in color sensitivity might overwhelm any benefit a 4th channel might provide.

 

[DaveC426913]

What if the addition of the fourth color introduces noise in the other three?

There is no fourth colour.

Or the brain might have a fixed amount of attention to such details - more color channels means less detail in each channel.

No. There is no evidence that a tetrachromat has more receptors, simply that some of the green ones have a mutation that results in them having a peak stimulation frequency a little lower than others.

And they are not "channels". We all have zilliions of cones. If they are stimulated, a signal is sent to the brain. What stimulates them is outside the brain's purview. (If I manually stimulated your red cones with a tiny electric pulse, you would see red. Your brain does not know what stimulated the cones.)

Furthermore, you're talking as if we have to speculate what they might experience. We don't. There are tetrachromats, and their ability to distinguish colours better than typicals has been studied. The only question here is whether the narrowband frequencies of RGB monitors has a confounding effect.

 

[DaveC426913]

Below is a more qualitative (not to scale) way of how an RGB monitor produces the colors RGB [0,193, 191] and [0, 191, 193]. Notice that after combining (neglecting any slight overlap, which I think is a valid approach, given the narrower, primary color bandwidths produced by the monitor), the individual peaks do not change frequency (reciprocal of wavelength), at least not significantly. They only change their relative intensities.

I have been working on a 2nd diagram, using your example as a basis, if that one makes you happier. It demonstrates the same principle. A tetrachromat's combo of receptors will show a more pronounced difference between 0,193,191 and 0,191,193 than a typical person's receptors will.

I just have a limited amount of time in which to create these graphs.

 

[Andy Resnick]

And what's your third favorite?

Good question... 119.2, I think.

 

[Algr]

1) There is no fourth colour.
2) having a peak stimulation frequency a little lower than others.

As far as I can see, these two statements directly contradict each other.

 

[DaveC426913]

As far as I can see, these two statements directly contradict each other.

That's because you ambiguously applied the word: "colour".

You said there's "a fourth colour". There isn't. Colour is a property of a brain. A single colour can be comprised of a multitude of frequencies. But they are all part of the same spectrum we are used to, so no extra colours.

There is the visible spectrum, all of which we see, through stimulation of our cones that peak in their sensitivity at certain frequencies of light. Tetrachromats merely have a bunch of cones whose peak is more green-yellow than the usual green (ie. lower frequency).

 

[collinsmark]

That's because you ambiguously applied the word: "colour".

You said there's "a fourth colour". There isn't. Colour is a property of a brain.

For a typical human, being a trichromat, there are three primary colors, red green and blue. That's because in trichromats (typical humans) there are only three types of color cones with associated optic nerve connections and brain processing. The brain is able to process these three colors and combinations of them, into what we perceive as other colors (cyan, yellow and magenta are examples of combinations of these three primary colors).

Due to a limitation called metamerism, typical humans cannot distinguish between a single, narrow wavelength in between these primary color wavelengths, and a combination of two primary color wavelengths.

For example, there is significant overlap in the sensitivity of the green and red cones. Filtered light photons with a narrow wavelength around 570-580 nm have an approximately equal chance of exciting the green cones as they do the red cones. This is perceived as yellow. But typical humans can be fooled into perceiving the same color by viewing red light combined with green light.

For tetracrhomats there is a fourth primary color. For them, other colors are linear combinations of these four primary colors. For them, the difference might not be as pronounced as the difference between a trichromat and a colorblind person, but there is a difference.

The human tetrachromat can perceive an actual rainbow differently than a rainbow displayed on a standard computer monitor. Particularly, the difference is in the yellowish part of the spectrum. In a real rainbow, they perceive a color that is not possible to reproduce by a simple combination of red, green and blue light.

Tetrachomats also have a metamerism limitation, but it involves indistinguishable colors involving combinations of their four primary colors, not three as it is with most people.

A single colour can be comprised of a multitude of frequencies.

Yes, due to a limitation (some might call it a weakness) of metamerism.

But they are all part of the same spectrum we are used to, so no extra colours.

Well, there are indeed extra colors for a tetrachomat compared to a trichromat.

Although it might not be quite as pronounced, it's the same idea as saying a trichromatic human can see extra colors compared to a colorblind person.

There is the visible spectrum, all of which we see, through stimulation of our cones that peak in their sensitivity at certain frequencies of light. Tetrachromats merely have a bunch of cones whose peak is more green-yellow than the usual green (ie. lower frequency).

And the tetrachomats perceive that color distinctly -- differently than a trichomat who could be fooled into thinking that a combination of red and green light is that same color.

 

[waternohitter]

I scored 27, not bad

 

[Algr]

If I manually stimulated your red cones with a tiny electric pulse, you would see red. Your brain does not know what stimulated the cones.)

What color would a tetrachromat see if you stimulated her X and B cones, but not her R or G cones?

 

[DaveC426913]

What color would a tetrachromat see if you stimulated her X and B cones, but not her R or G cones?

Teal.
Almost, but not quite the same colour as they would see if we stimulated their G and B cones.

 

[Algr]

Unless we get an actual tetrachromat to post here it ought to be impossible to answer that question. My suspicion is that she would see an "impossible" color that was unlike anything she had ever seen before.

I wonder how you are evaluating the overlap between the green and X cones.

Imagine you were building a 3-chip video camera. You have red, green, and blue filters to place in front of each sensor chip to create color. But you discover that your filters are rather pale: each one let's in 100% of the desired frequency range, but 50% of the other wavelengths. You might think that this would result in a poor camera that can only produce pastels. Actually, once you calibrate your camera correctly, you'll find that it can produce excellent color - the full range of trichromatic vision. Compared to a camera with pure filters, your color channels would be noisier, but your luminance channel would be cleaner - possibly a desirable tradeoff. (I'm simplifying how cameras work here, let's not get off topic.)

For a tetrachromat, the overlap of green and X cones might affect the subtlety of the X channel, but would make no difference in the identity of that channel at all. She would simply learn as a baby that the X and G cones can only be so different and would regard the maximum difference as totally unlike colors.

Edit: I forgot to close the connection: With the pale filter camera there are certain RGB values that can't occur. 255,128,128 would be maximum red, and would be adjusted to 255,0,0 on the monitor. If a pixel yielded 255 despite the corresponding pixels being below 128, it would produce an "impossible" color such as in the first paragraph. A camera could recognize this as a broken pixel and reject the data. Until we can actually perform this experiment, we don't know what would happen with a person.

 

[DaveC426913]

Unless we get an actual tetrachromat to post here it ought to be impossible to answer that question. My suspicion is that she would see an "impossible" color that was unlike anything she had ever seen before.

No!

There really is a tetrachromat who has described her experiences! It's not some magical unicorny colour. Essentially, she can simply tell when a teal skirt and teal blouse don't match when others think they do.

I will see if I can dig up the article.

 

[Algr]

No!

Yes!

 

[zoobyshoe]

No!

There really is a tetrachromat who has described her experiences! It's not some magical unicorny colour. Essentially, she can simply tell when a teal skirt and teal blouse don't match when others think they do.

I will see if I can dig up the article.

Dave, did you see the extravagant claims made in the link posted in the opening post?
http://nymag.com/scienceofus/2015/02/what-like-see-a-hundred-million-colors.html

 

[Drakkith]

Got a 3 here.

 

[DaveC426913]

Dave, did you see the extravagant claims made in the link posted in the opening post?
http://nymag.com/scienceofus/2015/02/what-like-see-a-hundred-million-colors.html

No I didn't.

There was an article a few years ago that included some discussion with a tetrachromat, describing her experiences - straight from the horse's mouth.

 

[Algr]

What she says sounds more like synesthesia then tetrachromacy. When such a person says "I see green, purple, yellow, red" we have to be careful because we don't know what those words mean to her. One test would be to show her a single laser projecting patterns and textures on a wall. If she is still seeing all these "colors", then we have to reject her whole vocabulary and concentrate on test patterns.

 

[zoobyshoe]

There was an article a few years ago that included some discussion with a tetrachromat, describing her experiences - straight from the horse's mouth.

That's what that article is: an interview with a tetrachromat (alleged).

What she says sounds more like synesthesia then tetrachromacy.

I had the same reaction. She sounds like some weird sort of synesthete rather than someone who has one more color experience than everyone else.

 

[DaveC426913]

That's what that article is: an interview with a tetrachromat (alleged).

I had the same reaction. She sounds like some weird sort of synesthete rather than someone who has one more color experience than everyone else.

Then it's not the same articIe or the same woman.

 

[zoobyshoe]

Then it's not the same articIe or the same woman.

The linked article is an interview with Concetta Antico. She's also mentioned in the wiki article on Tetrachromacy as a proven tetrochromat. In fact, she seems to be the poster girl for the condition-there are many articles about her in various publications. So, rightly or wrongly, you guys have to deal with what she says about the condition.

 

[Algr]

It's particularly odd that she says she enjoys movies. With the fourth color unarticulated, she ought to find them odd and unrealistic, like we see two-strip color.

 

[zoobyshoe]

It's particularly odd that she says she enjoys movies. With the fourth color unarticulated, she ought to find them odd and unrealistic, like we see two-strip color.

I actually like the "after" version better. I dislike pale yellow green in plants. Also, the "after" version highlights the blue/orange complementary relationship. There was a whole movie filmed in that palette and it worked really well. (Can't remember the name of it, but it was about a bunch of National Guardsmen trapped in a southern swamp during the rainy season. The cinematography was, in fact, the best part of the movie.)
------------------------------------------
For my money the whole Concetta Antico thing sounds like an error or maybe even a hoax. She is, when all is said and done, hyping herself as "The World's First Tetrachromat Artist" :
http://concettaantico.com/

This doesn't sound like what you'd expect from the addition of a fourth color others don't have:

I see colors in other colors. For example, I’m looking at some light right now that’s peeking through the door in my house. Other people might just see white light, but I see orange and yellow and pink and green and some magenta and a little bit of blue. So white is not white; white is all varieties of white. You know when you look at a pantone and you see all the whites separated out? It’s like that for me, but they are more intense. I see all those whites in white but I resolve all these colors in the white, so it’s almost like a mosaic. They are all next to each other but connected. As I look at it, I can differentiate different colors. I could never say that’s just a white door, instead I see blue, white, yellow-blue, gray.

She seems to see perfectly normal colors, but in places others wouldn't see them. Rather than tetrachromacy, it's more suggestive of some kind of hyperactivity in the area of the brain where color qualia are added to experience. Provided she's reporting an authentic experience and not just spinning a tale to make her art unique in people's minds, then tetrachromacy as an explanation strikes me as barking up the wrong tree. Some kind of neurological anomaly seems more like it.

However, someone more conversant with how cones work might be able to explain her experience in terms of the 4th kind of cone. I don't know.

Her paintings are particularly exiting in terms of color, but I don't think she's doing anything that requires a novel mode of perception. The impressionists and expressionists did the same with, presumably, three normal color cones.

 

[collinsmark]

For my money the whole Concetta Antico thing sounds like an error or maybe even a hoax. She is, when all is said and done, hyping herself as "The World's First Tetrachromat Artist" :
http://concettaantico.com/

I agree in-so-far that we certainly can't rely on her own personal descriptions of colors as evidence of tetrachromacy.

Neglecting changes such as medical operations, injuries or aging (and maybe drugs), any particular person only has a single point of reference to how they see colors. So one's own personal description of colors doesn't get us anywhere. The interview goes on and on about her own descriptions of colors -- way to much in my opinion -- and that doesn't mean anything. Wasted words, if you ask me.

But scientific tests should be able do it. Perhaps paint swatches with two different, yet metemer pigments, both with equal texture and gloss and such -- pigments with different spectral properties, yet indistinguishable to most humans. The lighting needs to be controlled too. The ambient light fixtures must have proper spectral properties (perhaps direct sunlight might be a better option).

Maybe a better test is one that is more direct. Perhaps creating two color metemers with combinations different hot gasses, each gas bulb producing very narrow spectral lines. The test would then be looking at a matte, white surface reflecting the two different light combinations. Of course one of the light combinations would of course have peaks in between the standard RGB primary colors, and the other not so.

A simpler version might be like the gas bulb test, but instead simply using different combinations of narrowband light filters, filtering sunlight. (Such filters might be constructed for use in photography or astrophotography. [but probably not for color photography])

If she could reliably and repeatably distinguish the two pigments or two light combinations, when most other people cannot, that would be evidence of tetrachomacy.

[Edit: Or lasers! Anything involving lasers is cool. For example, use a yellow laser together with a micro-mirror device to produce an image. Then use red and green lasers together with the micro-mirror device to produce a metemer image (after adjusting the brightness, and red/green color balance).]

Concetta Antico was apparently tested with a sort of "minimum motion" test where different colors were flickered at a varied rate on a background with particular spectral properties (the minimum motion reference comes from how humans can perceive motion from still pictures displayed in quick succession, such as in a motion picture). I can only seem to find the preliminary results of the test (I wish I could find something which describes the test in more detail, but alas, my googling skills are sub-par as of late [Edit: found something. See next post]). Here's a link to the preliminary results (from her website): http://concettaantico.com/?smd_process_download=1&download_id=2054