Color and Motion Perception
We use what we know about perception to design our media in keeping with the contradictory imperatives to make media as good as possible within the constraints of technology and economics.
We can only detect a small part of the electromagnetic spectrum with our eyes.
The visual spectrum ranges from the shorter less than 450 nanometer blue wavelengths to the longer 650 plus nanometer red wavelengths. White light is composed of all wavelengths in the spectrum.
We can simulate almost any color by using the three basic colors Red, Blue, and Green, assuming we are talking about mixing light. Pigment is a different matter.
In the eye we have two types of receptor cells, rods and cones. The cones are differentiated into three types, each especially (or mostly but not exclusively) sensitive to a portion of the spectrum. There are red sensitive cones, green sensitive cones, and blue sensitive cones. The brain analyzes the varying amounts of each wavelength hitting the retina to determine the overall color of the illuminated object.
Cones are most sensitive to pure or saturated spectral colors (those colors that occur along the spectrum). Non spectral colors such as brown or chartreuse are seen as a result of the combined stimulation of the three kinds of cones.
We see 70% red, 30% blue, and no green and we say "reddish purple."
Cone cells (the least numerous - 7 million) are concentrated near the center of the retina. Good at color discrimination and daylight vision. Also for discerning detail.
Rod cells (the most numerous - 125 million) are on the edges of the retina. They are very light sensitive and most useful in darker situations. They are also very sensitive to motion. We catch movement "out of the corner of the eye." That means rods were stimulated.
Chromaticity Diagram gives us a way to visualize how non-spectral colors can be made from differing amounts of spectral light. Move around the periphery of the diagram and notice that you are moving along the frequency spectrum for visible light. That is you are changing hue. As you move inward toward the center of the diagram you are changing (decreasing) saturation. That is you are increasing the amount of white in the color. Draw a line between any two points on the periphery. The distances from each of the origin points define the colors along that line. Think of it as increasing a dimmer as you move toward a point and decreasing one as you move away.
See how blue and yellow could make white instead of green?
See how blue and red could make magenta?
See how green and red makes yellow?
Three or more colors just complicate the process slightly.
Red, Blue, and Green make white. Draw a line from Red to Green. Equal amounts of each would land you in Yellow. Draw a line from your Yellow point to Blue. Equal amounts of each would land you in white. Ergo . . .
Video (television) and computers operate within a limited part of the diagram. Color film does a little better, for now.
Human vision can distinguish about 70 million variations in color including hue and saturation.
The world is made up of particles too small for us to see. In fact things have to be relatively large for us to be able to distinguish them as individual items.
How close does one have to be to a brick building in order to distinguish individual bricks?
Consider Newspaper pictures. Examine one closely. It is made of dots. How close do you have to get to the page in order to see the dots? Consider TV. How close does one have to get in order to see the lines and dots?
Two dots one inch apart are indistinguishable at 2000".
We sit closer to small TV screens and farther back from large ones. We normally position ourselves at a distance equal to four times the screen height. How far apart could the lines of a TV screen be and we not recognize them as lines.
(Note the opportunity to design around being as good as possible while also being economically and technically feasible.)
Clearly, more lines would give a more accurate picture. But in all likelihood, more lines would require a more sophisticated machine (more expensive too).
Space between lines = height of screen/number of lines per screen.
H/N
Viewing distance (D) at which the lines could not be seen is 2000 times the space between the lines or 2000 x H/N
D = 2000H/N
But the average viewing distance (D) is four times the screen height (H).
D = 4H
4H = 2000H/N
N = 500
Add a few lines for additional information and for good measure. 525 lines, the current US standard.
Screen is 4:3 ratio (width to height). 4/3 x 525 = number of dots per line = 700
We actually use 630.
Sequentially blinking lights appear to move. Used in signs.
We know motion will look smooth if we sample the motion at least 24 times per second. For some people it is as low as 20 times/sec.
Strobe lights blinking on a dance floor at a rate slower than 20/sec make dancers look jerky. Above that the motion appears smooth.
This is how we determine the speed at which we flash individual frames of motion picture film (24 frames/second) and how often we refresh the image on a TV screen (30/sec).
One more problem. Even if the motion is smooth, we can still detect a flicker. A motion picture showing 24 frames per second would show smooth motion but the screen would appear to flicker. Hence the slang term for movies.
Flicker Fusion Frequency, the frequency above which a blinking light appears to glow steadily. 45 - 50 Hz.
Movies fix the problem by showing each frame twice so the shutter is actually opening and closing 48 times per second.
TV solves the problem by showing the picture by interlaced halves with each half shown every 1/60th of a second (60 refreshes per second is above the fff).
Here are some other interesting demonstrations of perception of motion
Disco strobes:
14/sec best
1 - 2/sec nausea
8 - 10/sec brainwave resonance, dangerous to some.
24/sec no fun. Smooth motion.
45 - 55/sec just another pretty light.