How many frames per second are appropriate for a motion picture? The answer depends on the intended use. Figure 6.17 shows a table of significant frame rates from to . Stroboscopic apparent motion begins at FPS. Imagine watching a security video at this rate. It is easy to distinguish individual frames, but the motion of a person would also be perceived. Once FPS is reached, the motion is obviously more smooth and we start to lose the ability to distinguish individual frames. Early silent films ranged from to FPS. The frame rates were often fluctuating and were played at a faster speed than they were filmed. Once sound was added to film, incorrect speeds and fluctuations in the speed were no longer tolerated because both sound and video needed to be synchronized. This motivated playback at the fixed rate of FPS, which is still used today by the movie industry. Personal video cameras remained at or FPS into the 1970s. The famous Zapruder film of the Kennedy assassination in 1963 was taken at FPS. Although FPS may be enough to perceive motions smoothly, a large part of cinematography is devoted to ensuring that motions are not so fast that jumps are visible due to the low frame rate.
Such low frame rates unfortunately lead to perceptible flicker as the images rapidly flash on the screen with black in between. This motivated several workarounds. In the case of movie projectors, two-blade and three-blade shutters were invented so that they would show each frame two or three times, respectively. This enabled movies to be shown at FPS and FPS, thereby reducing discomfort from flickering. Analog television broadcasts in the 20th century were at (PAL standard) or FPS (NTSC standard), depending on the country. To double the frame rate and reduce perceived flicker, they used interlacing to draw half the image in one frame time, and then half in the other. Every other horizontal line is drawn in the first half, and the remaining lines are drawn in the second. This increased the frames rates on television screens to and FPS. The game industry has used FPS standard target for smooth game play.
As people started sitting close to giant CRT monitors in the early 1990s, the flicker problem became problematic again because sensitivity to flicker is stronger at the periphery. Furthermore, even when flicker cannot be directly perceived, it may still contribute to fatigue or headaches. Therefore, frame rates were increased to even higher levels. A minimum acceptable ergonomic standard for large CRT monitors was FPS, with to FPS being widely considered as sufficiently high to eliminate most flicker problems. The problem has been carefully studied by psychologists under the heading of flicker fusion threshold; the precise rates at which flicker is perceptible or causes fatigue depends on many factors in addition to FPS, such as position on retina, age, color, and light intensity. Thus, the actual limit depends on the kind of display, its size, specifications, how it is used, and who is using it. Modern LCD and LED displays, used as televisions, computer screens, and smartphone screens, have , , and even FPS.
The story does not end there. If you connect an LED to a pulse generator (put a resistor in series), then flicker can be perceived at much higher rates. Set the pulse generator to produce a square wave at several hundred Hz. Go to a dark room and hold the LED in your hand. If you wave it around so fast that your eyes cannot track it, then the flicker becomes perceptible as a zipper pattern. Let this be called the zipper effect. This happens because each time the LED pulses on, it is imaged in a different place on the retina. Without image stabilization, it appears as an array of lights. The faster the motion, the further apart the images will appear. The higher the pulse rate (or FPS), the closer together the images will appear. Therefore, to see the zipper effect at very high speeds, you need to move the LED very quickly. It is possible to see the effect for a few thousand FPS.
Steven M LaValle 2016-12-31