Hacker News Reader

Comment by shutupalready | Hacker News Reader

> Jordan’s “acid test” involved coloured discs showing different mixtures of pigment, such as a green made of yellow and blue. The mixtures were too subtle for most people to notice: almost all people would see the same shade of olive green, but each combination should give out a subtly different spectrum of light that would be perceptible to someone with a fourth cone.

From what I understand, a modern monitor is capable of displaying the entire range of colors perceptible to the human eye. Am I mistaken about that?

So would it not be possible to create a web-based version of these "coloured discs", and then we can test ourselves.

View on HN

> From what I understand, a modern monitor is capable of displaying the entire range of colors perceptible to the human eye. Am I mistaken about that?

It kinda depends on what you think as a color. Computer monitors rely heavily on that (normal) human vision can not distinguish from eg. red-green and "true"[1] yellow, even though they are completely different from physical perspective. So a typical computer monitor does not bother to create "true" yellow and instead cheats by outputting red-green when asked for yellow.

If we had an hypothetical tetrachromat with additional cones sensitive for yellow light, her color vision would span the same range as normal (trichromat) color vision, ie cover the same piece of EM spectrum. But normal computer monitors would not be able to reproduce for her what she sees in real world.

Then there is the somewhat related discussion about violet and red-blue which is quite curious when you begin think about it.

[1] I'm using "true" color to refer light with specific wavelength

From a purely physical point of view, a computer monitor can only produce a spectrum of light (photon histogram) that has spikes on the wavelengths for red, green and blue. These spectra reproduce the vast majority of colors that are perceptible to humans with normal vision. However, the "colored discs" in the article exploit the fact that two colors that are perceived as identical, can in fact be produced by (at least) two different spectra of light. Tetracromats can differentiate between some of these spectra, whereas people with normal vision can not.

(A simple example: The color cyan can be produced by additively combining the spectra of blue and green, but can also be produced by creating photons with only the wavelength of cyan, 490-520nm according to Wikipedia. A person with normal color vision would not be able to differentiate between these two light sources, whereas a measurement device or perhaps a tetrachromat could. A computer monitor can only produce the first of these).

So the answer to your question is no, unfortunately.

Look at this image (http://upload.wikimedia.org/wikipedia/commons/thumb/6/60/CIE...) from Wikipedia. It represents all the colors (chromaticities) that humans can see. A device that reproduces images using primary colors can reproduce all the colors within the vertices of the primaries used. For example, an RGB monitor can reproduce all colors within the triangle formed by the particular R, G, and B subpixel primaries as plotted on that diagram.

What you'll notice though is that because the diagram has a curved edge (this is the spectrum of visible light actually), no polygon can encompass all colors.

I've always thought it would be cool if someone made a CRT using two prisms (splitting light from a blackbody), such that a narrowband of light from one prism is combined with a narrowband of light from the other prism in different intensities. This would be able to reproduce all colors a human can see, assuming the ability to tune to a specific portion of the spectrum is fast enough to scan over all pixels that constitute the image.

> no polygon can encompass all colors.

But I suppose that a polygon could encompass all colors if it was big enough to include also wavelengths of non-visible colors.

Ooh, a combined monitor/tanning lamp/heater! Great idea! /s

> Am I mistaken about that? Yes. See: https://en.wikipedia.org/wiki/Gamut

Google 'tetrachromacy test', first hit http://www.blogadilla.com/2008/06/08/are-you-a-tetrachromat/

that claim is quite dubious. I don't think it's possible to create metamers for people with normal 3-color vision on an RGB monitor. To do that you need an extra degree of freedom.

https://en.wikipedia.org/wiki/Metamerism_(color)

Hmm... I'm not sure about that. It's an interesting question. It's equivalent to the question: can we create two images on a computer monitor that appear the same to a person with normal color vision yet appear different to a colorblind person?

I think it would depend on the particular mapping from 3 to 2 dimensions... it might be possible.

> can we create two images on a computer monitor that appear the same to a person with normal color vision yet appear different to a colorblind person?

It surely depends on the type of colorblindness, there are many [1].

[1] http://en.wikipedia.org/wiki/Color_blindness

If you're monochromate or dichromate, I doubt that this is possible. Monochromates or dichromates are simply missing one or two color components out of three.

If you have anomalous trichromacy, you can by creating a monitor with pixels of different colors.

In any case, I don't see how it is possible to identify a tetrachromate with a trichromate (RGB) monitor. I tried to increase the contrast and change the hue of the color test to understand what was different between the apparently similar colors but I wasn't able to find any difference.

Using JPEG for a color test doesn't seem well thought out. Using GIMP (Colors, Map, Rotate Colors), one can make a letter appear in the center circle. I guess if the other circles had any letters in them, the compression ruined it.

Without using a monitor, a potential test would be to look at flowers like Rudbeckias(in real life of course) that have evolved to have patterns that tetrachromat insects can see. If you can see these patterns, you may be a tetrachromat.

Reply on news.ycombinator.com