4 types of cones?

[Wayne Fox]

Not to stir things up considering the color of the dress thing on the internet, but this seems to be making the rounds as well, although as not to the extent.

After reading it I couldn't find any other sources that seemed to bear out this persons theory about some people having 4 types of cones instead of 3. The number of colors one can count is definitely affected by the display luminance, and wouldn't be surprised if there is some color management issues going in. 

Her credentials also seem rather dubious to be making this physiological claim of some humans. "Expert in Neuromarketing"

personally I can see 37 colors, only 3 of the bands blend with the band next to it. This is both on a mobile device and my color managed computers (macbook pro, and NEC).  I do think some people perception of color is more refined, but i find it hard to believe that it's because of the reason she offers.

[Czornyj]

This is just a shameless, outrageous BS. Simply ignore it.

P.S. - I can clearly see all 39 tones - as a child I was stung by a bee irradiated during Chernobyl disaster.

[Redcrown]

Fun to look at, but I'd agree that it's a questionable technique. Much like the Xrite color vision test. I score high on both tests. Got a 35 to 37 on this test. And I'm using old eyes, one damaged and repaired, and with custom prescription computer glasses. Seems unreasonable to me that I'd score high.

I've given the Xrite test to a few grandkids, aged 6 to 13. I've had them do it twice. First on my high end calibrated monitor with liminance set to 100 cd/m2 and then on an uncalibrated laptop whose default luminance is somewhere between 140 and 160 cd/m2. They score significantly higher on my calibrated monitor than on the laptop.

This "cone test" is an untagged png file, so all bets are off. When I test on-line (through my browser and monitor profile) I score 37. When I download the image and assign sRGB to it, I only score 35.

[ErikKaffehr]

Hi,

My understanding is that there are, or may be, humans with tetrachromic vision, but I guess that they are not very abundant.

Best regards
Erik

This is shameless, outrageous BS.

[Czornyj]

https://research.ncl.ac.uk/tetrachromacy/faqs/

Q: Is there an online test that I can do?

A: Unfortunately, computer screens do not provide enough colour information to be able to ‘tap into’ the extra dimension that tetrachromats may possess. It is therefore impossible for an online test to investigate tetrachromacy.

[Wayne Fox]

This "cone test" is an untagged png file, so all bets are off. When I test on-line (through my browser and monitor profile) I score 37. When I download the image and assign sRGB to it, I only score 35.

Some more research seems to support that some women may be tetrachromatic  although my question would be are they true tetrachromatic,(like some fish and birds) or do they just posses some cones which have a slightly different sensitive to blue than other cones and does that really offer them better perception of color?

Some of the references in the wikipedia article are interesting (although most are way over my head). Seems there is something to it ..

Although this stupid web thing certainly isn't the way to determine it.  So far I haven't found anyone that can only perceive 31 or fewer colors ...

I must be bored ...

[Simon Garrett]

It's rubbish.

I can't comment on tetrachromacy or necromancy or whatever it is, I mean it is rubbish to draw any inferences from that image.  

As already noted, it is untagged, so colours displayed are unpredictable.  More: for given RGB values, even if you know the colour space, the actual chromatic values displayed on a screen depend on the RGB primaries used in that particular monitor.  If you are trying to distinguish different spectral responses of different people, it is pretty meaningless to try to do it by means of displays on a monitor of unknown spectral output.  Even if the image were tagged, and even if displayed on a calibrated and profiled monitor, the spectral output is unpredictable as it depends on the chromatic values of the monitor primaries.  

For example, we all know that yellow can be created by red light plus green light.  On a narrow gamut monitor, a saturated sRGB yellow might be R=100%, G=100%, B=0%.  But a wide-gamut monitor, those RGB values would create a more intense yellow.  To get the same yellow on a wide-gamut monitor, you might need something like R=92%, G=98%, B=33% (I didn't make those figures up; those are the RGB values for fully-saturated sRGB yellow, but represented in a very wide-gamut space).

In other words: on different monitors, the same colour is made up of different spectral components: different wavelengths of light in different combinations and intensities.  

Even worse: that image is awful.  It's been heavily processed and compressed down to a jpeg quality of about nothing, so there are heavy compression artefacts on the image (especially strong ringing on the edges) and this will also have distorted the colour of each step.  

Apart from that, I'm sure that article is a genuine and inciteful addition to the gamut of human knowledge.

[ErikKaffehr]

Hi,

Any other side effects?

Best regards
Erik

This is just a shameless, outrageous BS. Simply ignore it.

P.S. - I can clearly see all 39 tones - as a child I was stung by a bee irradiated during Chernobyl disaster.

[Telecaster]

Some more research seems to support that some women may be tetrachromatic although my question would be are they true tetrachromatic (like some fish and birds) or do they just posses some rods which have a slightly different sensitive to blue than other rods and does that really offer them better perception of color?

I first learned of this in college, c. 1981. Still not aware of any thorough study of it, though. Given how close in peak wavelength sensitivity our "green" and "red" cones typically are, we just qualify as trichromats…so this fourth peak isn't likely to be very distinct. Are four peak wavelength sensitivities "better" than three? Depends on how/whether we put the extra one to use.

Although this stupid web thing certainly isn't the way to determine it.

I don't have a problem with the web thing…it's got people thinking a bit about how color perception works. Some of 'em even seem to get, rightly, that color is purely subjective and contextual.

-Dave-

[Tim Lookingbill]

I'm a bit puzzled but slightly in agreement about yellow being an irritating color mentioned in the article. I will not buy any clothes, furniture, wall paper or paint finish that is dominantly yellow but I'll eat food that is yellow without being irritated.

I'm also sensitive to noticing cyan to magenta hue tint shifts in yellow items caused by spiky spectrum artificial lighting which I now use as a quick test target when shopping for lights at my local hardware store. If the light renders cadmium yellow as lemon yellow I don't buy it.

And I counted 36 different colors but the jpeg compression of the test target really messes things up as Simon pointed out. I noticed more luminance differences over hue/sat which technically doesn't qualify as seeing a completely different color.

Also I didn't get a perfect score on that Xrite site color perception test that's been linked to in the past.

[GWGill]

And I counted 36 different colors but the jpeg compression of the test target really messes things up as Simon pointed out. I noticed more luminance differences over hue/sat which technically doesn't qualify as seeing a completely different color.

Anything displayed on a 3 channel additive color screen cannot possibly be used to test for Tetrachromacy - by definition you need at least 4 control channels and 4 different primary spectra to be able to independently trigger the responses of a Tetrachromat.

We already have 4 light sensitive cells in our retina, and they have 4 different spectral sensitivities, but we are not regarded as Tetrachromats because the responses are combined into just 3 channels of information going to our brains.

So (as I understand it), even if there is evidence that some people have 4 types of cone cells in their eyes, it has yet to be clearly established if any have 4 channels of information going to their brains. If the extra cones signals are just combined together into the usual 3 channels, then they aren't functional Tetrachromats, although they may be a little distant from the standard observer in their spectral sensitivities, and therefore "see color" a little differently from the average human.

[Simon Garrett]

Anything displayed on a 3 channel additive color screen cannot possibly be used to test for Tetrachromacy - by definition you need at least 4 control channels and 4 different primary spectra to be able to independently trigger the responses of a Tetrachromat.

We already have 4 light sensitive cells in our retina, and they have 4 different spectral sensitivities, but we are not regarded as Tetrachromats because the responses are combined into just 3 channels of information going to our brains.

So (as I understand it), even if there is evidence that some people have 4 types of cone cells in their eyes, it has yet to be clearly established if any have 4 channels of information going to their brains. If the extra cones signals are just combined together into the usual 3 channels, then they aren't functional Tetrachromats, although they may be a little distant from the standard observer in their spectral sensitivities, and therefore "see color" a little differently from the average human.

It must be a bit of a handicap having 4 types of cone, or having any perceptual system other than the "normal" 3 cone cells.  Virtually all colour reproduction systems, including digital photography, are designed to reproduce colour not by reproducing the overall spectral characteristics of the light of the scene, but by reproducing light that will result in the same sensory output from an idealised version of our three H, M and S cone cell responses. 

For anyone with different colour perception - different numbers of cones, or different spectral responses of any of the cones - photographic material won't look the same colour as the original scene.  Anything printed, television, cinema - any reproduced colour will potentially look the wrong colour. 

[Wayne Fox]

It must be a bit of a handicap having 4 types of cone, or having any perceptual system other than the "normal" 3 cone cells.  Virtually all colour reproduction systems, including digital photography, are designed to reproduce colour not by reproducing the overall spectral characteristics of the light of the scene, but by reproducing light that will result in the same sensory output from an idealised version of our three H, M and S cone cell responses. 

For anyone with different colour perception - different numbers of cones, or different spectral responses of any of the cones - photographic material won't look the same colour as the original scene.  Anything printed, television, cinema - any reproduced colour will potentially look the wrong colour. 

But to them that’s the “normal” color?  They don’t know what they don’t know?  Maybe it looks OK?

It appears most theories about it relate to 2 different cone response types in the X chromosome, which means a female (having two X chromosomes) can have both types of cones.  Seems to me either those two cones are very close to each other, or half the world is actually seeing things differently than the other half depending on which X chromosome they have.

For those actually interested (I think this leaves practical discussion and enters the realm of “curiosity”), here’s Wikipedia’s “expertise” on the subject. As they pointed out, even if the cones exist, it may not be possible for the optic nerve and brain to utilize the extra information.

"In humans, two cone cell pigment genes are on the sex X chromosome, the classical type 2 opsin genes OPN1MW and OPN1MW2. It has been suggested that as women have two different X chromosomes in their cells, some of them could be carrying some variant cone cell pigments, thereby possibly being born as full tetrachromats and having four simultaneously functioning kinds of cone cells, each type with a specific pattern of responsiveness to different wavelengths of light in the range of the visible spectrum.[16] One study suggested that 2–3% of the world's women might have the type of fourth cone whose sensitivity peak is between the standard red and green cones, giving, theoretically, a significant increase in color differentiation.[17] Another study suggests that as many as 50% of women and 8% of men may have four photopigments and corresponding increased chromatic discrimination compared to trichromats.[16] In June 2012, after 20 years of study of women with four types of cones (non-functional tetrachromats), neuroscientist Dr. Gabriele Jordan identified a woman (subject cDa29) who could detect a greater variety of colors than trichromats could, corresponding with a functional tetrachromat (or true tetrachromat).[18] [19]

Variation in cone pigment genes is widespread in most human populations, but the most prevalent and pronounced tetrachromacy would derive from female carriers of major red/green pigment anomalies, usually classed as forms of "color blindness" (protanomaly or deuteranomaly). The biological basis for this phenomenon is X-inactivation of heterozygotic alleles for retinal pigment genes, which is the same mechanism that gives the majority of female new-world monkeys trichromatic vision.[20]

In humans, preliminary visual processing occurs in the neurons of the retina. It is not known how these nerves would respond to a new color channel, that is, whether they could handle it separately or just combine it in with an existing channel. Visual information leaves the eye by way of the optic nerve; it is not known whether the optic nerve has the spare capacity to handle a new color channel. A variety of final image processing takes place in the brain; it is not known how the various areas of the brain would respond if presented with a new color channel.”

[Simon Garrett]

But to them that’s the “normal” color?  They don’t know what they don’t know?  Maybe it looks OK?

What they see other than reproduced images: yes, that would be their "normal". 

However, reproduced colour images: not necessarily. 

Here's why I think that (this is speculation, by the way!):  consider digital imaging systems.  They are designed to reproduce the tri-stimulus values for normal vision.  That is, normal vision is based on 3 cone cells, the so-called L, M and S receptors with responses such as here. 

The red, green and blue filters in camera sensors mimic those responses (approximately), and in a colour-managed system, the monitor will generate R, G and B signals that cause a stimulus in our L, M and S receptors that mimics the colour captured by the camera sensor.  It doesn't matter if the camera sensor response curves might not exactly mimic our cone receptors, nor that the R, G and B primaries in the monitor vary from one type of monitor to another, provided the appropriate transformations are done (by colour management) so that the monitor creates a combination of spectral colours that creates the right stimulus in our cone receptors.

But that won't necessarily work in someone with different cone cell responses - either different numbers of different cones, or different spectral responses of the cones. 

Hence, one could have an image on a monitor which looks the same colour as the original scene for anyone with "normal" trichromatic vision, but would look different to someone with 4 types of cone cell. 

[MarkM]

Hence, one could have an image on a monitor which looks the same colour as the original scene for anyone with "normal" trichromatic vision, but would look different to someone with 4 types of cone cell.  

I think this is probably correct in theory and a long way of saying that for tetrachromats there will be real-world color perceptions that you will not be able to reproduce using only three primaries (assuming tetrachromaticism is real). They might also perceive color differences where trichromats perceive metamers. Similarly, it's not hard to imagine a dichromatic reproduction system that works well for people with a certain type of color blindness that would only reproduce a subset of the color stimulus available to people with normal vision. In this system many colors that match for those with colorblindess would not match for those with three cone types.

[Fatih]

Wayne,

I found only 4 publications under D. Derval in PubMed, 1968, 1976, 1978 and 2014. May be there more D. Dervals in France. She was not the first or senior author on any of them so her academic title in dispute. I am not sure how she got such academic title, likely from that shady Chinese college.  she has no credible scientific evidence to support her theory.  Please ignore it.

Fatih Akisik

[Isaac]

Also -- "Physicists have pinned down precisely how pipe-shaped cells in our retina filter the incoming colours."