Adventures in CyberSound

Persistence of Vision
by Stephen Herbert

"Virtually every ... account of the perception of movement in film texts [is] wrong. The impression of movement is not due to persistence of vision. The very persistence with which this 'explanation' has been recited says more about the hermetic and impressionistic world of some film scholarship than it does about the actual mechanisms involved." Bill Nichols and Susan J. Lederman

More than fifteen years ago, a number of writers pointed out, in detailed studies, that the long-quoted mechanism for motion perception in cinema, 'persistence of vision', an effect usually attributed to a 'defect' of the eye, ( or in some accounts the 'eye-brain combination' ), was an archaic concept long left behind by psychologists and physiologists specialising in perception. Most film writers were ( and still are ) unaware of, or unconvinced by, these writings, and continue to write of persistence of vision as being the mechanism by which we see motion pictures. They argue, when challenged, that the term is 'simple to understand', 'elegant' even 'poetic'. They are either unaware or unconcerned that it is incorrect.
The word 'persistence' has a meaning, and it seems that we do not perceive motion pictures through any kind of persistence, either in the eye or in the brain. So let's take a look at what those researchers wrote back around 1980, as their work is still valid ( though still largely ignored ) today.

"The notion of 'persistence of vision' seems to have been appropriated from psychology in the first decade of the century, the period during which cinema came into being. But while most film scholars accepted 'persistence of vision' as the perceptual basis of the medium and proceeded to theorise about the nature, meaning and functioning of cinema from that base, perceptual psychologists continued to question the mechanisms involved in motion perception; and they have achieved insights that necessitate the re-thinking of many conclusions reached by film scholars during the past 50 years." Joseph and Barbera Anderson

"To begin with, 'persistence of vision' is itself an imprecise term. We can only guess that film writers are referring to what psychologists call 'positive after images'. When a person stares at a light, he or she can still see it after the light has been turned off. Positive after-images retain the colour and brightness relations of the original stimulus. Common sense would suggest that the positive after-image is a plausible explanation of motion perception in film since it allows one image-frame to 'bleed' into another, despite the fact that the beam of light projecting the film-frame itself is intermittent. But this fusion occurs regardless of whether motion is perceived or not. The appearance of a continuously visible series of images, in other words, is a phenomenon distinct from the appearance of motion." Bill Nichols and Susan J. Lederman

It may help to understand the phenomenon of apparent motion in motion pictures therefore, if we think of it as two things:
Firstly, it is necessary to ensure an apparently continuous, reasonably flicker-free image on the screen. Early experiments showed that a minimum of about 10 separate frames must be projected every second to give the illusion of movement. But this is not enough; the image will flicker very badly if a projector with only a single blade ( to cover the period of film pulldown ) is used. A simple experiment will show that the flicker rate must be of the order of 50 per second for it not to be obvious. For much of the 'silent' period, films were shot at roughly 16 frames per second and shown on a projector with a three-bladed shutter. Each individual frame was shown three times, so around 48 screen images were projected every second, ( close enough to fifty to give a reasonably flicker-free result ). Taking and projecting speeds gradually increased so that by the time 'talkies' arrived a standard speed of 24fps was decided upon, which meant that the shutter could be reduced to two blades to achieve the same effect of 48 screen images (each frame shown twice) per second.
Secondly, if the screen image is composed of a series of slightly differing pictures, as in a cine film, the image will appear to move.
It is sometimes said that to ensure a flickerless, apparently continuous image, as explained in (1), persistence of vision is necessary, and that apparent motion (2) is a psychological effect of the brain.
Those proposing this explanation at least recognise that the important part of the phenomenon, that which causes a series of static images to apparently move, is a psychological effect that does not require 'persistence of vision' to work, and they suggest that 'persistence of vision' is only necessary to ensure that the image which we see on the screen appears continuous and does not flicker.
However, it has in fact long been determined that the so-called 'persistence of vision' is also probably irrelevant to the effect of (1), a continuous, flickerless image.
So where did this 'persistence' ( or after-image ) notion come from?
Roget, in describing the phenomenon of spokes of a wheel appearing to be curved when viewed through a series of vertical slits, also mentions the...

"illusion that occurs when a bright object is wheeled rapidly round in a circle, giving rise to the appearance of a line of light throughout the whole circumference; namely, that an impression made by a pencil of rays on the retina, if sufficiently vivid, will remain for a certain time after the cause has ceased."

As Joseph and Barbera Anderson have pointed out, there is no logic in ascribing how we see motion pictures to this effect:

"The illusion he describes is not an illusion of motion, nor does he claim that it is. Roget has described a case in which a series of moving points results in a static image, while in cinema a series of static images results in the illusion of motion."
Later in their article, they point out that:

"There is no motion on the screen, just a succession of still images. If there were persistence of these images in the eye of the viewer, figures on the screen would pile up, one on top of the other, resulting in a kind of chronophotographic display."

As previously mentioned, it has been suggested that, although perhaps not the mechanism by which successive still pictures are perceived as a moving image, 'persistence' ( i.e. the after-image ) is necessary in the cinema, in order to ensure that we are unaware of the blank ( dark ) screen which occurs between the projection of each image; and that if it were not for this persistence,the result would appear to flicker. Firstly, motion picture systems exist which do not have a blank screen between images; e.g. Steenbeck viewing machines, in which a prism block 'mixes' one image into the next. So even if 'persistence' were necessary for conventional projection with an intermittent shutter projector, the existence of systems that do not have blanking periods means that it can hardly be considered important to our understanding of how we perceive apparent motion in motion pictures.
Even the suggestion that 'persistence' ensures that we do not see flicker in a conventional projection system is in fact not the case, for there is another problem with this 'persistence' or after-image theory. What actually happens after a bright image is removed is that we see a succession of after-images, some positive and some negative. Obviously, the negative after-images are irrelevant; we do not perceive negative images when we look at motion pictures. More importantly, the first after-image, which is positive, appears some 50 milliseconds after the image ceases. In that period of time, not one but at least two screen images (at 48 flashes per second; that is, 24 frames per second with a two-bladed shutter) would have been viewed. Put simply, the first after-image of a film frame is not actually perceived until after at least one following frame has been projected.

"It is very unlikely that after-images contribute to the fusion of successive image presentations in film" Nichols & Lederman

So 'persistence of vision' is not even the mechanism by which we are made unaware of the blanks between images. So why don't we see those blanks? Quite simply because they are dark periods of nothing, and are sufficiently short that they do not register in our brain.
Some film-system engineers have long been aware of the problems with the 'persistence of vision' theory. In 1913 Kasimir Proszynski, who designed the popular Aeroscope 35mm cine camera and was a pioneer of 'flickerless' shutters for cine projectors, gave a lecture on the subject : "The theory received a rude shaking" reported the Amateur Photographer, "when Mr. Proszinski gave his very lucid lecture before the Royal Photographic Society last week. That there is a certain amount of persistence is undeniable....but it is very weak, and this purely physiological effect, in Mr. Proszinski's opinion, is altogether insignificant, and may be put out of the reckoning as far as the phenomenon of the moving picture is concerned. ..In his opinion the continuity of the cinematographic vision is due to no such physiological phenomenon as persistence, but is purely an illusion. It is psychological, and not physiological at all." ( Cinema Notes, Amateur Photographer, 3 March 1913 ).
A final word from Joseph and Barbera Anderson:

"Since most film scholars may feel unprepared to conduct experimental research, they may be tempted simply to acknowledge the inadequacy of the persistence of vision explanation and proceed with their work. The temptation should be resisted, for in any theoretical discussion of the cinema basic assumptions are embedded about how we see form in motion."

So having put to rest the suggestion that we see apparent motion in motion pictures by means of 'persistence of vision' - how then do we perceive the effect?
Well, it's a bit complicated.....and perhaps we'll look into that, some other time.
Stephen Herbert*

References
Joseph and Barbera Anderson, 'Motion Perception in Motion Pictures'
Bill Nichols and Susan J. Lederman, 'Flicker and Motion in Film', in The Cinematic Apparatus, New York, 1980
Michael Chanan, The Dream that Kicks, Routledge & Kegan Paul, 1980

* Stephen Herbert runs The Projection Box and publishes monographs about pre-cinema, early film and optical recreations

Source: http://www.grand-illusions.com/percept.htm

Back to the Top or return to the Essays Index