Exploring Color Mixing: Red & Green Chequer Pattern Spin Test

[cmb]

If I paint a ball in red and green chequer pattern and with 50% coverage of each, then spin the ball then it looks white.

If I make the pattern with smaller and smaller chequers, does it stay white and how small can the pattern be before it turns a different colour?

Once the pattern is a molecular size, and there are 50% red molecules and 50% green, then I guess it'll be a brown ball when I spin it!

 

[sophiecentaur]

Actually, it all depends on the actual colours you start with but it seems that you are describe basic Additive Colour Mixing, in which case you would expect to see Yellow and not white. There would need to be some Blue in there if you expect to see White.
The details are also important when you are asking what happens as the sources get smaller and smaller but as long as the sources do not overlap each other, the resulting perceived colour should not look any different from the result with larger chequers and a de-focussed observation (snake eyes or a frosted glass, for instance).

At molecular size things could be different if the electron states in the two types of pigment start to affect each other. Aamof, pigments tend to consist of mixtures of quite big molecules. Consider the dots on an Inkjet Printer; they are pretty much invisible at arms length but sometimes the technology assumes a certain amount of 'bleed' from dot to dot and that has to be catered for in the drivers.

 

[cmb]

I am not convinced by the comment 'electron states.. affect each other'. Of course I understand that rationale and it would be the case in some scenarios, but I don't think that is a good argument for why we see brown when mixing red and green.

Let me ask this in a different way then.

If I had a quantity fine red powder and a fine green powder in the order of a few 10's of micrometres in particle size, what colour would I see with the naked eye if I mixed them? If I then looked at that mixture under a bright white light and x200 optical microscope, what colours would I see then?

I'd like to know more about how our eye works when seeing mixed colours than the thing we're looking at. Is there a difference when the colours are alternated across our view, compared with mixed, i.e. if the colour cones are picking up alternately red then green then they put one interpretation on that, but if a cone is exposed to both red and green at the same time then then brain interprets that in a different way to alternating colour?

 

[hutchphd]

f I had a quantity fine red powder and a fine green powder in the order of a few 10's of micrometres in particle size, what colour would I see with the naked eye if I mixed them? If I then looked at that mixture under a bright white light and x200 optical microscope, what colours would I see then?

If there is enough light, the fundamental spatial acuity is determined by the "pixel density" of the cones. Apparently there are three types cones which respond roughly to red green and blue predominantly. Their deposition on the retina is not uniform but peaks strongly at the fovea and the distribution for each color is different. At the fovea there are ~500 per mm so that determines the expected maximum best angular color acuity. Seems this would make a fun experiment to try.

I'd like to know more about how our eye works when seeing mixed colours than the thing we're looking at. Is there a difference when the colours are alternated across our view, compared with mixed, i.e. if the colour cones are picking up alternately red then green then they put one interpretation on that, but if a cone is exposed to both red and green at the same time then then brain interprets that in a different way to alternating colour?

Apparently the color "flicker fusion rate" is about 15 Hz (as opposed to ~35Hz for white).
My apologies but I have no references for these facts pulled from my brain. Human vision is of course extraordinarily complicated and there are very few absolutes! I recommend some of the work of Edwin Land about color perception as an example

 

[Cutter Ketch]

but I don't think that is a good argument for why we see brown when mixing red and green.

I believe SophieCentaur was just saying things might get weird at the molecular level, not claiming that this is the proper explanation for a phenomenon you seem to believe in. I think for the spirit of your question we should probably stay away from the molecular level. It should be sufficient to talk about scales down to less than the smallest the eye can resolve. Then we can safely assume that the actual color characteristics of the individual dots do not change and we can stick to talking about perception.

Perception of color (and lots other things about visual perception) is extremely complicated. The brain is doing all kinds of interpretations and corrections. That is why there are so many neat optical illusions. I’m sure circumstances could be arranged where all kinds of strange things happen.

However, in the spirit of your question, I believe your premise is flawed. I believe sophiecentaur is correct. As long as the color of the individual patches does not change, your perception of the color will not change substantially as the patches get smaller. The combination of red and green will make a shade of brown, not white. When you mix colors it is possible for things to happen at the molecular level, but for most pigments (larger color particles that don’t dissolve in the medium) they don’t. When you mix paint together you get microscopic particles each reflecting its original color and the result is effectively the same color as you would see on your spinning ball. This is how Home Depot makes whatever color you ask for. There are lots of examples of this on YouTube. Search for spinning color wheel.

 

[sophiecentaur]

When you mix paint together you get microscopic particles each reflecting its original color and the result is effectively the same color as you would see on your spinning ball.

I would totally disagree with this. On a spinning ball / disc, the coloured areas all reflect just a filtered version of the white illuminant and any particular area will reflect a mix of the three colours Additively. The result (depending on the details of the pigments) will be a pale Grey (not a bright white but a pale grey because the contributions of the three colours will be only 1/3 of the area). When you mix pigments, otoh, there are layers of molecules which, individually, are partly transmissive and partly reflective. The perceived colour is not only due to the very surface (hence the frequent need for several coats) The lower layers do not get exposed to white but whatever wavelengths are left by the time light gets to them. Ideal Red in Ideal Green light will be Black. That is Subtractive Mixing. If the pigments are 'good' and pass / reflect only their intended wavelengths than the lower layers will have none of 'their colour' light. They will appear black. The only molecules reflecting any significant light will be those on the very surface. Paint is harder to describe than coloured dyes within a filter because it's a bit indeterminate what actually goes on in the first few layers. But you're going to get a more or less Black or very dark grey.

 

[Jeff Root]

Colors on a TV or computer monitor screen are additive: Adding
different colors of light together produces a brighter color. The
primary colors of light which most closely match human color
receptors are red, green, and blue. Mixing equal amounts of
red and green light together additively gives brown or yellow,
depending on the total brightness of the red and green light.

Opaque pigments in paint work by selectively reflecting the light
that strikes them, which is a subtractive process. A thick layer of
paint or many thin layers make it more likely that incident light
will strike a pigment particle and be reflected back out (when the
color of the light matches the color of the pigment). Subtracting
different colors of light from the incident light produces a color
that is roughly the average brightness of the different pigments.
The primary colors of paint pigments are red, yellow, and blue,
which can be mixed together to give the widest gamut of colors
for human color receptors. Mixing equal amounts of yellow and
blue paint pigments together subtractively gives green, with a
brightness roughly equal to the average brightness of the yellow
and blue pigments, since the pigment particles are essentially
side-by-side. Yellow is inherently lighter than blue in human
vision, so the more yellow and less blue in the mixture, the
lighter the resulting green. Mixing equal amounts of red and
green together subtractively gives a brown which is an average
of the included red, yellow, and blue pigment brightnesses.

Colored inks, dyes, watercolor paints, or wax crayons on a white
background work mainly by selective transmission, which is also
subtractive, but is different from how opaque paint pigments
work, so that I call the process "transparent" color mixing rather
than "subtractive" color mixing. Different colors of light travel
through the transparent pigments, are reflected by the white
background, and travel back through the pigments again to reach
the viewer. The thicker the layer or layers of ink or dye, the darker
the color tends to be, since less light is transmitted. Mixing
different transparent colors gives a resulting color darker than
any of the component colors of the mixture. The primary colors
of transparent pigments are yellow, cyan (or "process blue"), and
magenta (or "process red"), which, like the primary subtractive
colors of opaque pigments, are the three colors which give the
widest gamut of colors for human color receptors. And like the
subtractive colors of opaque pigments, yellow plus process blue
gives green, and process red plus green gives brown. The brown
will be darker than any of the individual transparent pigments.

The attached images depict these three methods of color mixing:
RGB is for light, RYB is for opaque pigments, and YCM is for
transparent pigments.

-- Jeff, in Minneapolis

 

[sophiecentaur]

I call the process "transparent" color mixing rather
than "subtractive" color mixing.

Very much a half way house and very dependent on the nature of the particles of the three primaries. I guess that's in the magic spells that ink jet technology uses. There must be a lot of suck it and see about coefficients used in the drivers. Top secret I imagine.

 

[hutchphd]

Is there a difference when the colours are alternated across our view, compared with mixed, i.e. if the colour cones are picking up alternately red then green then they put one interpretation on that, but if a cone is exposed to both red and green at the same time then then brain interprets that in a different way to alternating colour?

I suggest that you look at the history of color television in the early 1950s. There were two competing systems one used time and the other used space. The CBS system projected 144 fields a second through a rotating tricolor filter wheel ( 2 interlaced fields per frame, 3 colors so 24 color frames per second ). The color was said to be very good. The RCA system used dedicated triads of pixels simultaneously excited by three electron beams and is the system that became standard. Also it was directly compatible with the existing Black and White signal.

So both ways work visually.

 

[sophiecentaur]

I think that sequential systems are not as bright as the current system in which you see each colored pixel all the time.

 

[hutchphd]

At that time I believe the RCA brightness was limited by the stability of the red phosphor (I think it was the red) and so the the strength of emission was not that great. The shadow mask required to ensure beam aiming color separation also reduced the strength of the electron beam. So I don't know which was brighter. I have never seen the "color wheel" in action!

 

[sophiecentaur]

The shadow mask required to ensure beam aiming color separation also reduced the strength of the electron beam

Absolutely true. The holes in a shadow mask system (even the more modern versions), along with convergence problems make that sort of system very inefficient as most of the beam current ended up colliding with the mask. It was a brilliant system for its day but very lacking in many ways. Modern LED /LCD displays are much brighter and convergence is no longer a problem. (Replaced by the problems inherent with very large scale solid state circuit manufacture - TNSTAAFL)
But, despite history, it seems that sequential colour synthesis systems are just not as good as spatial separation of primary coloured sources. I can't think of a recent sequential system but that could be to do with movement portrayal as much as anything.

 

[Ibix]

I can't think of a recent sequential system but that could be to do with movement portrayal as much as anything.

I saw a projector that used sequential colour about 20 years ago. As far as I recall it had a single mirror array that ran at something like 60 frames a second and red/green/blue lights that flashed in time, so you got 20 frames of full colour per second. Worked ok for PowerPoint type stuff, and hardware that could capture and display video didn't come in cereal boxes like it does today, so that was fine. You could still do some funny stuff, though. If you looked at one side of the image and then flicked your gaze to the other side, the movement of your eye was rapid enough to decompose the image. Sharp white/black edges got a rainbow effect a bit like the stars at warp speed in Star Trek, which was quite cool (for nerdy values of cool) to be able to produce at will.

 

[hutchphd]

I believe the Apollo color cameras for lunar surface use were color- wheel based. I think they may have had a fairly slow frame rate. Also I remember seeing a "frame sequential" camera recently that used an undifferentiated LCD screen to produce the sequential color filters. I can't recall the context for use of camera...

 

[sophiecentaur]

I believe the Apollo color cameras for lunar surface use were color- wheel based. I think they may have had a fairly slow frame rate. Also I remember seeing a "frame sequential" camera recently that used an undifferentiated LCD screen to produce the sequential color filters. I can't recall the context for use of camera...

I think you're right about the those cameras. But camera technology is a different issue. Advanced astrophotography always involves multiple exposures with a range of filters to optimise image quality.
Having made my statement about brightness and sequential displays, I now recollect the Eidophor display which was a massive beast and which produced large projected images which were something like 80 times brighter than the CRT projectors of the time. One colour version used a colour sequential system but the most impressive system used three separate projectors. Line-up procedure was very long winded. However, fidelity was way down the list of priorities when presenting TV to vast crowds.