Human's abilities to perceive light is quite limited. Essentially, we have three type of light sensors that each give a linear signal, and if two different light sources give the same signals they will look the same. This is convenient for monitor makers though, since then computer monitors do not to be able to produce any kind of light, just a representative sample.

What if our eyes where different though. What if when we saw light, we can see all the frequencies it contained (in the same visible light spectrum as we do now), and at what intensity.*

Is there a way that we could still make computer monitors look realistic (i.e. have the ability to produce any combination of light frequencies from each pixel)? Or would any possible computer monitor necessarily look "black and white (and red and blue and ...)" to us?

*I have a couple of ideas as to how this could work. The one way would be to associate each frequency with a slower frequency, (that the brain can handle), and to have the cone convert light waves into these slower neural waves. The other idea is to again have only one type of cone, but allow the brain to change what frequency it detects. The brain would subconsciously sweep the visual spectrum, the same way it moves eyes subconsciously to see more.

  • 6
    $\begingroup$ You are basically asking: what if we could see exactly like we do today only with a mechanism that is both impossible to exist because there is no way to do this biologically and neurologically and doesn't make sense because evolution (no expensive unnecessary stuff) and the spectrum is a continuum and heisenberg. The end result is exactly the same since you implied humans were otherwise the same, it's just impossible and not as elegant as current human sight. The brain can already do what you are asking and interprets the signal at the same time. All you add is some not needed data $\endgroup$ – Raditz_35 Jan 18 '18 at 8:57
  • 2
    $\begingroup$ Both your ideas are biochemically impossible. I think a question about full spectrum display is OK (after all multispectral imaging is a real thing), but if you want your aliens to have full spectrum sensors, just handwave it, because way you describe them is worse than not describing them at all. $\endgroup$ – Mołot Jan 18 '18 at 10:21
  • 3
    $\begingroup$ Our ears do that; we perceive a spectral analysis of the sounds, and not a blend. That's why when several piano keys are pressed at the same time we hear a chord, but when to subpixels are lit we perceive only one color. The auditory system does that by having the auditory nerve carry one fiber for each distinct frequency. The trade-off is that the spatial resolution of the sense of hearing is abysmal; by contrast, the spatial resolution of the sense of sight is exquisite, at the cost of not being able to perform spectral analysis. $\endgroup$ – AlexP Jan 18 '18 at 11:25
  • 1
    $\begingroup$ @JBH We presumably wouldn't have a concept of pink as an indivigual color. To us, there would be many different pinks, depending on what mixture its made up of. $\endgroup$ – PyRulez Jan 18 '18 at 19:30
  • 1
    $\begingroup$ Wait, are you talking about seeing, like, a fourier transform of light? $\endgroup$ – Pingcode Jan 19 '18 at 0:12

First thing you should consider is that brain would nevertheless have limited ability to distinguish colors, so it would be still possible to make finite number of light sources in every pixel displaying something perceived as "the full spectrum" (it would probably not be ideal as our current monitors are not).

Then you may look for ways to produce arbitrary color. You can make light of precisely tuned frequency using the quantum dots. Yet it would still be alike LED screens so you will need to put enormous number of them into every pixel.

Other way may be to build something like the cathode ray tube, but with light. Basically you will need to find a way to change direction of light ray with electricity (nanomechanism with mirror?) and then lead it over the screen like in the tube (i.e. in rows). Then instead of the electron emitter you put a light source which frequency you may change somehow. It may be just a bunch of quantum dot LEDs which you fire in desirable combination. This way you do not need all sorts of LED colors in every pixel and just single set of them somewhere in the back of your monitor.

| improve this answer | |
  • $\begingroup$ Insofar as we know, photons can only be diverted by gravity (unlike the electron streams used in CRTs). Bombarding phosphorous with an electron beam creates light, which is a light source, so why replace it? (we use LEDs today only because they're cheaper and produce sharper images). CRT "pixels" were the combination of three colors. What's the difference between three and three thousand? Pure yellow would still be pure yellow. $\endgroup$ – JBH Jan 19 '18 at 1:23
  • $\begingroup$ @JBH, with three thousand instead of three there will be 2997 extra color sources in addition to RGB. And photons can be directed via the same method that is used in tag scanners - simple rotating mirror. Unite photon source with beam modulator (Kerr cell or anything similar), and that'll be one step closer to what OP wants. So I disagree with downvoter(s), this one can work. $\endgroup$ – ZuOverture Jan 19 '18 at 3:26
  • $\begingroup$ @ZuOverture, A rotating mirror for millions of emission sources in a 2-sqft area? Meh. Besides, my primary complaint is that the OP thinks his aliens would see color differently when I doubt very much they would. Increasing color depth increases the number of colors in the image, but not the number of frequencies per-color. For any specific color, the light between the monitor and the eye is no different between his aliens and humans. So why should the monitors change? How is this answer providing more of a "full spectrum" of pink than any other monitor presenting the same color of pink? $\endgroup$ – JBH Jan 19 '18 at 3:51
  • 1
    $\begingroup$ Raman spectroscope does see colors differently. It's not producing a convolution of pairs of functions resulting in 3 or 4 numbers on the output, like most image sensors (and monitors) nowadays, it gives full scale spectrum for every point in the image. If there didn't exist numerous fitting functions for 3 or 4 points, then there would be no difference. But there is difference, and jaboja points to a method of producing those precise spectra. $\endgroup$ – ZuOverture Jan 19 '18 at 4:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.