Author Topic: Tetrachromacy and how it relates to computer graphics  (Read 1997 times)

0 Members and 1 Guest are viewing this topic.

Online NiHaoMikeTopic starter

  • Super Contributor
  • ***
  • Posts: 9321
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Tetrachromacy and how it relates to computer graphics
« on: February 26, 2023, 03:50:18 am »
Computer graphics uses 3 primary colors based on the assumption that human eyes have 3 types of cone cells. For the most part, that's true. But a (relatively) recent finding discovered that up to 12% have 4 types.
https://www.extremetech.com/extreme/331169-human-tetrachromacy-is-real-heres-what-we-know
With decades of computer graphics being based on 3 primary colors, it's obvious that it's not likely to expand to 4 anytime soon. For those who have 4 types of cone cells, do they see pictures on a computer screen significantly less colorful than the real world? If computer displays had 4 primary colors, would the majority with 3 cone cells see any difference in the colors?
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15800
  • Country: fr
Re: Tetrachromacy and how it relates to computer graphics
« Reply #1 on: February 26, 2023, 04:26:22 am »
Dang, was octarine a real thing after all? :-DD
 
The following users thanked this post: newbrain

Offline I wanted a rude username

  • Frequent Contributor
  • **
  • Posts: 662
  • Country: au
  • ... but this username is also acceptable.
Re: Tetrachromacy and how it relates to computer graphics
« Reply #2 on: February 26, 2023, 04:59:37 am »
This claim of 6% of the population (0% of men, 12% of women ... it's sex-linked and requires heterozygosity) refers to what I'd call "weak" tetrachromacy, not the experience of the painter profiled in this article. Weak tetrachromats' vision is pretty much like normal trichromats'.

However, almost everyone could benefit from an extension of the colour space, as computer displays only render a subset of what people with normal vision can see. This is especially noticeable with deep violet hues, which I would guess even most dichromats would perceive. Note that this doesn't necessarily require a four-dimensional colour space.
 
The following users thanked this post: Someone, tooki

Offline Whales

  • Super Contributor
  • ***
  • Posts: 2122
  • Country: au
    • Halestrom
Re: Tetrachromacy and how it relates to computer graphics
« Reply #3 on: February 26, 2023, 12:04:01 pm »
This is especially noticeable with deep violet hues

My newest monitor does goofy stuff in the purples, see the attached photo vs screenshot.  sRGB preset chosen and annoying stuff turned off.  It's much more obvious in real life than the camera photo suggests.

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 9003
  • Country: us
  • Retired, now restoring antique test equipment
Re: Tetrachromacy and how it relates to computer graphics
« Reply #4 on: February 26, 2023, 02:51:00 pm »
Purple and violet are different colors that look similar.
 

Offline Tomorokoshi

  • Super Contributor
  • ***
  • Posts: 1212
  • Country: us
Re: Tetrachromacy and how it relates to computer graphics
« Reply #5 on: February 26, 2023, 03:08:32 pm »
Green is a difficult color. Also, most computers and displays are stuck with 8-bits per channel storage and display formats that were declared by consortia of "professionals". Their main purpose was to make something cheap, not something extensible. We are left with an inadequate infrastructure that will be very difficult to move beyond.

10-bits per channel would have started to make the situation a little better, while 12-bits per channel would be close to the limit of perception. To drive 12-bits would require DAC hardware closer to the 14-bit to 16-bit range.

With only 8-bits per channel there isn't enough color-space resolution to cover shadows in the blacks, highlights in the whites, and to avoid banding in the mid-range.
 

Offline Brianf

  • Regular Contributor
  • *
  • Posts: 73
  • Country: gb
Re: Tetrachromacy and how it relates to computer graphics
« Reply #6 on: February 26, 2023, 03:19:26 pm »
Talking about colours by name is pointless as many cultures and languages have very different concepts for what 'colour' something is.
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: Tetrachromacy and how it relates to computer graphics
« Reply #7 on: February 26, 2023, 05:14:12 pm »
Green is a difficult color.

Also there is this legacy of not having optimum CRT phosphor material for green. So the green phosphor used in CRTs produce yellowish kind of green, which does not perfectly map to the human vision system, and greatly limits the gamut; basically if pure grass green color is needed, one has to mix some blue, but then you have less saturated color because of the unwanted yellow tint mixing with the compensating blue, resulting in addition of white.

The sRGB color space is also compromised by this legacy limitation, even if no CRTs are used anymore. Even AdobeRGB is; look up the usual gamut picture and the green end corner is way too much on the right, and not in the middle.

Red is a tad too yellowish, too, but it's not as bad as the green, and red can be rationalized by efficiency at least. Deeper wavelength (say 660nm) would allow wider gamut, but look much dimmer because of human eye response is weaker there.

Blue phosphors were excellent from the wavelength perspective. Light output (efficiency) and lifetime sucked, though.

Now with LCD screens etc., which use dyes, this does not matter. LEDs have their own limitations, but high-intensity green LEDs do not have that yellowish tint problem; it's actually the opposite, they are a tad too much on the blue side, by having the peak at 530nm instead of 540, if my memory serves right.
« Last Edit: February 26, 2023, 05:16:06 pm by Siwastaja »
 

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 9003
  • Country: us
  • Retired, now restoring antique test equipment
Re: Tetrachromacy and how it relates to computer graphics
« Reply #8 on: February 26, 2023, 05:29:49 pm »
Talking about colours by name is pointless as many cultures and languages have very different concepts for what 'colour' something is.

It is not pointless to name colors.  Variations in color names across culture and language is an interesting subject, but within English-speaking engineering it is handy to name the colors according to standard English usage.
It is important to distinguish "color" from "wavelength", since the light reflected by (or transmitted through) real physical systems is polychromatic (range of wavelengths), as is the light emitted by phosphors.
One can say that a given wavelength (e.g., 510 nm from a narrow-band laser source) has a color (green), but "regular" green light has a finite width spectrum.
 
The following users thanked this post: SiliconWizard

Offline schmitt trigger

  • Super Contributor
  • ***
  • Posts: 2431
  • Country: mx
Re: Tetrachromacy and how it relates to computer graphics
« Reply #9 on: February 26, 2023, 05:38:06 pm »
Color spaces and its rendering are a fascinating subject which long predates electronics displays. During the late 1970s I was both an amateur photographer and also worked for Kodak. At our photography club we would discuss and compare color rendering and faithfulness.
Of course, we all recognized that color rendering is always a compromise: a film and light source which provides the best skin tones and thus suitable for portraits, would not be the best choice for landscape photography.
If I recall correctly, the zenith of color photography was achieved with a low speed Fuji Velvia slide film. But that was in the early 90s.
Those were the days…
« Last Edit: February 26, 2023, 05:39:37 pm by schmitt trigger »
 

Offline BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: Tetrachromacy and how it relates to computer graphics
« Reply #10 on: February 26, 2023, 05:53:42 pm »
This is especially noticeable with deep violet hues

My newest monitor does goofy stuff in the purples, see the attached photo vs screenshot.  sRGB preset chosen and annoying stuff turned off.  It's much more obvious in real life than the camera photo suggests.
I looked at your second 'true' screen shot on my studio grade CRT and 3-chuip DLP video projector and the violet color gradient is perfect.  (Note that the color you chose is tilted toward purple where a dark violet would have been a tint further to the left.)
Your camera shot of your LCD screen is crap.
« Last Edit: February 26, 2023, 05:56:23 pm by BrianHG »
 

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 9003
  • Country: us
  • Retired, now restoring antique test equipment
Re: Tetrachromacy and how it relates to computer graphics
« Reply #11 on: February 26, 2023, 06:54:21 pm »
Color spaces and its rendering are a fascinating subject which long predates electronics displays. During the late 1970s I was both an amateur photographer and also worked for Kodak. At our photography club we would discuss and compare color rendering and faithfulness.
Of course, we all recognized that color rendering is always a compromise: a film and light source which provides the best skin tones and thus suitable for portraits, would not be the best choice for landscape photography.
If I recall correctly, the zenith of color photography was achieved with a low speed Fuji Velvia slide film. But that was in the early 90s.
Those were the days…

In the glory days of color transparency (positive) films, comparison of different emulsions was a popular topic.
Many people thought Velvia 50 to be too saturated:  one photographer I knew attributed that choice to the normal cloudy weather in Japan, requiring extra saturation in the film to get vivid colors.
For normal color photography, my short-lived favorite (after the demise of Kodachrome) was Kodak Lumiere, but E100G Ektachrome also worked for me in broad daylight.
Unfortunately, my all-time favorite Koday EPY (ASA 64 tungsten balanced) went extinct along with the other tungsten-balanced transparency films.
 
The following users thanked this post: schmitt trigger

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: Tetrachromacy and how it relates to computer graphics
« Reply #12 on: February 26, 2023, 07:39:44 pm »
Velvia 50 was excellent IMHO and I did not find it too saturated; it also had surprisingly good dynamic range and clipped in a nice way around highlights. Contrast that to the "new" Velvia 100 of late 2000's, which had kind of digital look to it, with even more contrast and clipping highlights. Did not like that. Fuji's Sensia 100 which was basically just consumer-branded version of the (way more expensive) Astia 100 was my favorite slide film, though.

I do remember exposing Velvia at ISO25 and pull processing it (as I did my own E6 processing) and this reduced the contrast and color saturation a tad, it looked really good.

Speaking of which, the film dyes (as in, dyes in slide films, or motion picture print films) did better job getting that correct wavelength for green, which CRT monitors sucked at. You could see some beautiful deep pure (or cyanish) green colors on films.

Many CRT projectors used filter dyes mixed with the CRT cooling glycol to help shift the output peak wavelength a bit and improve the color rendition, but this came at expense of light output, and it still wasn't spot-on.
« Last Edit: February 26, 2023, 07:45:40 pm by Siwastaja »
 
The following users thanked this post: schmitt trigger

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15800
  • Country: fr
 

Offline Tomorokoshi

  • Super Contributor
  • ***
  • Posts: 1212
  • Country: us
 

Offline CatalinaWOW

  • Super Contributor
  • ***
  • Posts: 5571
  • Country: us
Re: Tetrachromacy and how it relates to computer graphics
« Reply #15 on: February 26, 2023, 09:56:18 pm »
In reply to the OPs question, no tetrachromatic people do not see our three color displays with less color than trichromats.

A close but imperfect analogy would be to ask:  Do people with 10 bit grey scale resolution see 8 bit images less vividly than people with 8 bit resolution?

Tetrachromats may find the three color displays less satisfying than trichromats, because they can see color differences in real life that can't be displayed on our three color screens and which trichromatic people don't complain about because they can't see the difference in real life.

Most of the discussion above is related to the differences between perfect representation of the real world and our displays.  I believe it is actually impossible to perfectly represent the real world for a variety of reasons, so there will always be arguments about how good "good-enough" is.  And that is without going into the lala land of audio-phoolery.  Video-phoolery?
 

Offline BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: Tetrachromacy and how it relates to computer graphics
« Reply #16 on: February 26, 2023, 10:27:45 pm »
 

Offline BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: Tetrachromacy and how it relates to computer graphics
« Reply #17 on: February 26, 2023, 10:29:10 pm »
I can't show you how pink this pink is...

 

Offline CatalinaWOW

  • Super Contributor
  • ***
  • Posts: 5571
  • Country: us
Re: Tetrachromacy and how it relates to computer graphics
« Reply #18 on: February 27, 2023, 12:41:55 am »
A chromaticity diagram really makes understanding of this easier.  There are many colors that simply can't be represented by three phosphors.  Something like half of the color space is not represented by standard phosphors.  LEDs can cover more of the color space but still leave a lot of room.  And because the shape of the boundary of the diagram is a complex truncated archlike shape, adding phosphors (or LEDs) just improves the polygonal approximation.  And given the satisfaction achieved with three for most people it seems really unlikely that we will ever have systems with eight or sixteen monochrome LEDs just to get high 90s coverage of the color space.  But it might be fun as a research project to use an array of the TI Digital Light Valves to build such a display and then test people to evaluate benefit vs completeness of color coverage.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 17428
  • Country: us
  • DavidH
Re: Tetrachromacy and how it relates to computer graphics
« Reply #19 on: February 27, 2023, 02:34:37 am »
For those who have 4 types of cone cells, do they see pictures on a computer screen significantly less colorful than the real world? If computer displays had 4 primary colors, would the majority with 3 cone cells see any difference in the colors?

The difference between tetrachromates and trichromates is the same as the difference between trichromates and dichromates.

Ask a dichromate sometime and they will say that there are colors that they cannot distinguish as they look like shades of the same color.  A tetrachromate can distinguish colors that look the same to a trichromate.

As pointed out above, one of the genes is sex linked, meaning that it is on the X chromosome, so only women can be tetrachromates, and that is why men, who only have one X chromosome, are more likely to be color blind; they lack a backup of the gene in question.

 

Online NiHaoMikeTopic starter

  • Super Contributor
  • ***
  • Posts: 9321
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: Tetrachromacy and how it relates to computer graphics
« Reply #20 on: February 27, 2023, 03:45:30 am »
This claim of 6% of the population (0% of men, 12% of women ... it's sex-linked and requires heterozygosity) refers to what I'd call "weak" tetrachromacy, not the experience of the painter profiled in this article. Weak tetrachromats' vision is pretty much like normal trichromats'.
Is there some other genetic factor that affects if the tetrachromacy is usable? Or do they just haven't learned to recognize the subtle differences in color?
Green is a difficult color. Also, most computers and displays are stuck with 8-bits per channel storage and display formats that were declared by consortia of "professionals". Their main purpose was to make something cheap, not something extensible. We are left with an inadequate infrastructure that will be very difficult to move beyond.

10-bits per channel would have started to make the situation a little better, while 12-bits per channel would be close to the limit of perception. To drive 12-bits would require DAC hardware closer to the 14-bit to 16-bit range.

With only 8-bits per channel there isn't enough color-space resolution to cover shadows in the blacks, highlights in the whites, and to avoid banding in the mid-range.
That's exactly what HDR is supposed to solve.
A chromaticity diagram really makes understanding of this easier.  There are many colors that simply can't be represented by three phosphors.  Something like half of the color space is not represented by standard phosphors.  LEDs can cover more of the color space but still leave a lot of room.  And because the shape of the boundary of the diagram is a complex truncated archlike shape, adding phosphors (or LEDs) just improves the polygonal approximation.  And given the satisfaction achieved with three for most people it seems really unlikely that we will ever have systems with eight or sixteen monochrome LEDs just to get high 90s coverage of the color space.  But it might be fun as a research project to use an array of the TI Digital Light Valves to build such a display and then test people to evaluate benefit vs completeness of color coverage.
I have noticed that the colors on OLED and IPS displays are very good.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline I wanted a rude username

  • Frequent Contributor
  • **
  • Posts: 662
  • Country: au
  • ... but this username is also acceptable.
Re: Tetrachromacy and how it relates to computer graphics
« Reply #21 on: February 27, 2023, 05:29:20 am »
Is there some other genetic factor that affects if the tetrachromacy is usable? Or do they just haven't learned to recognize the subtle differences in color?

It seems to have a genetic basis. For example, in this study of 7 carriers of protanomaly and 24 carriers of deuteranomaly:

> However, one participant cDa29 ... behaves as if she has access to an additional cone signal: she made no errors at any value ... and her response times not only are faster than those of other participants ... but are roughly even across the stimulus space
> Perhaps the most important result of our Rayleigh discrimination test is that most carriers exhibit a region of the Franceschetti space in which their discrimination is impaired in the same way as is found for color-normal controls. Thus, by this test, the great majority of carriers show no sign of tetrachromacy.
> cDa29 is the single exception.


As usual, further research is required.
 
The following users thanked this post: NiHaoMike


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf