3
The lens, the eye and perception
Introduction
A useful ability when framing up a shot is having the experience to predict how a particular subject or view will translate into a two-dimensional recorded image. A beginner without this skill may have to wander around the subject looking through the viewfinder at various set-ups, and with various choices of lens angle and camera distance, in order to see how these variables affect the shot. Cameramen and directors often have the developed visual ability to mentally predict the effect of mass, line and size relationships and how they will impact on the shot for any specific lens angle and camera position chosen. Viewpoint can be decided before a camera is taken out of its case. This is often called having a photographic eye and in one sense it is learning to see like a lens. How do we learn to see like a camera and why is it necessary?
Composing a shot involves the translation of a three-dimensional subject into a two-dimensional image. The eye and the lens are both used in this activity but the two imaging devices differ in their interpretation.
The lens of the eye focuses a two-dimensional image onto the retina of the eye and somehow the mind interprets the image. That ‘somehow’ is known as perception and has a significant influence on how an individual understands what he or she is looking at. There is always a subjective element in any individual's interpretation of their senses.
The lens of a camera focuses an image onto a recording medium but it is not an objective scientific instrument precisely translating the fieldof-view of the lens into an image. The conversion of the original subject into an image viewed on a screen is conditioned by what is chosen from a number of variables associated with the conversion at the moment of recording, and its later method of two-dimensional presentation. The variables include:
the lens – fno, focal length, camera height, camera distance, etc.;
the recording medium characteristics – film, tape and method of processing;
detail and resolution of lens and recording/transmission medium;
colour rendition;
lighting conditions;
filters;
the image size when viewed;
viewing conditions;
the subjective influence of context;
cultural influences.
These topics are discussed in more detail in later chapters.
The imprint of the lens
When a camera converts a three-dimensional scene into a film or TV picture, it leaves an imprint of lens height, camera tilt, distance from subject and lens angle. We can detect these decisions in any image by examining the position of the horizon line and where it cuts similar sized figures. This will reveal camera height and tilt. Lens height and tilt will be revealed by any parallel converging lines in the image such as the edges of buildings or roads. The size relationship between foreground and background objects, particularly the human figure, will gives clues to camera distance from objects and lens angle. Camera distance from subject will be revealed by the change in object size when moving towards or away from the lens.
For any specific lens angle and camera position there will be a unique set of the above parameters. The ‘perspective’ of the picture is created by the camera distance except, of course, where false perspective has been deliberately created.
Developing a photographic eye is learning how to manipulate these variables to achieve a particular image. These are the basic tools of visual design for the cameraman and they are discussed in more detail in the following chapters. The choices made can create style, mood or emphasize a significant element of the shot. Ignoring these lens/camera factors in an unthinking point-and-shoot technique will still control the appearance of the image because the lens/camera will be left to produce an image customized by whatever set of default characteristics are engaged.
The eye and a lens
There are similarities and differences between how we perceive the world and how the camera lens translates the world into images. What is often overlooked when making the comparison is the influence of the mind on the image produced by the eye.
Most people believe that seeing is a straightforward activity – just open your eyes and see what's there. But the component parts of seeing such as movement, depth, shape, colour and size, etc., are constructed in our heads and have to be pieced together by the brain. Visual information coming into our eyes is dismantled and then reassembled. Different aspects of seeing are dealt with by sub-sections of the brain; each area decoding one aspect of visual information. There is parallel processing by over 30 areas of the brain such as motion, colour, depth, etc. It is the brain that turns seeing into understanding. Like a film or TV frame, our eyes only capture static images: they are transmitted to the back of the brain where they are incorporated into seeing movement.
Figure 3.1 The contemporary theory about the brain and perception is that the visual brain has two systems:
There is part of the brain that generates images from the eyes (front projection); there is also a part of the brain that uses visual memories (back projection)
There is no adequate explanation of how these two parts of the brain are combined in perception. Perception is an active process by the brain that invents, ignores, distorts what is coming through the eyes
There are a number of facets of the brain's involvement in seeing that need to be taken into account when framing up a shot that may only be on the screen for a few seconds.
Size constancy
A shot, if taken with an appropriate lens angle, camera distance and viewed with a specified screen size and viewing distance (see next chapter), will replicate the retinal image, but this image will not necessarily be the same as how an observer views the same field of view.
With the above conditions, the camera will provide true geometrical perspective, but because we do not see the world as it is projected on the retina, the perspective of the shot may look wrong. For example, a holidaymaker takes a photograph of an impressive range of mountains. Looking at the print, the holidaymaker may be disappointed because the mountains look very small compared with their memory of the landscape (see Figure 3.2).
The mismatch between how we think we see a subject and how the camera records the same subject is due to a perceptual distortion called size constancy. This is the tendency for the perceptual system to compensate for changes in the retinal image with viewing distance. As a retinal image, a person walking away from a static observer halves in size as their distance from the observer doubles. The relationship between image size and distance from the observer is a constant, but in normal perception it is not seen as a constant. Perception adjusts the perceived size to match our knowledge of the size of the receding subject.
Figure 3.2 The mismatch between how we think we see a subject (a) and how the camera records the same subject (b) is due to a perceptual distortion called size constancy. This is the tendency for the perceptual system to compensate for changes in the retinal image with viewing distance
We habitually underestimate the change in size of a person walking towards or away from us and mentally picture them modified in size but only with a slight alteration to their ‘normal’ size. An audience will appear from the front to have similar size faces and yet, to an observer, the retinal image of the faces of the people in the back rows will probably be a tenth of the size of the faces of the people in the front row. We never recognize that the image of our face in the mirror is always much smaller than its actual size. These are all depth indicators we habitually ignore or make the necessary adjustment for, as in the phenomenon of the ‘upside down’ image that is focused on the retina of the eye. We ‘mentally’ correct this inversion of our field of view as we subconsciously ‘correct’ the change in size. Size constancy is what the brain does and the camera does not do and therefore when planning a shot we should not be misled by this habitual distortion.
A simple experiment demonstrating this phenomena is described by R.L. Gregory in Eye and Brain (1967):
Look at your two hands, one placed at arm's length the other at half the distance – they will look almost exactly the same size, and yet the image of the further hand will be only half the (linear) size of the nearer. If the nearer hand is brought to overlap the further, then they will look quite different in size.
To ‘see’ like a camera obviously requires overcoming this everyday mental adjustment in the perceptual process. Many artists have trained themselves to accurately draw their perceptual image whereas most of us are trapped, particularly when taking photographs, in the perceptual misconception of size constancy. In essence we see what we think is there, not what is actually there.
This characteristic of perception in habitually making adjustments to the size of a subjects at various distances from the viewer provoked a heated debate in the mid-nineteenth century when artists began basing painting on photographs. Always assuming that a lens/camera distance provided ‘normal’ perspective in a photograph (see Chapter 4, ‘The lens and perspective'), there was fierce criticism on what many people thought was the gross distortion in the painting depiction of size relationships. People were for the first time confronted with their adjustment of optical size as presented on the retina and the optical truth of a photograph transcribed into a painting (e.g., the holidaymaker disappointed with his photograph of a diminutive mountain range).
In 1858, Mrs Jane Carlyle complained about a Robert Tait painting of herself and husband in their drawing room, claiming that it was bad enough to be recorded for posterity with a frightful table cover, but what was worse was that their dog Nero, in the lower right foreground, was as big as a sheep. What was called the ‘false and ugly perspective of a "photographic" painting’ was in fact the true optical perspective showing size relationships as they were, not how we imagined they were.
Manipulating size relationships and the perspective depth of the shot is one of the principal compositional devices in film and TV productions. Seeing as a camera does requires not only retraining our habitual way of discounting the actual size of objects in a field of view, but also mentally conceiving the visual effect of choosing different lens angles and camera distances.
Figure 3.3 Perception is making sense of an image – searching for the best interpretation of the available data. The mind sees patterns and searches for the best interpretation. A perceived object is therefore a hypothesis to be tested instantaneously against previous experience. If it looks like a duck then it is a duck. That is, until we see it as a rabbit
How do we understand what we are looking at?
In setting up shots for film and TV it is important to remember that perception is not a simple common sense everyday activity that can be ignored. Understanding how the audience makes sense of the images presented to them, often in rapid succession, will ensure the required visual communication is effective. Unfortunately there are many competing theories about human perception.
The ‘perspectivist’ theory proposes that our understanding of a visual field is simply determined by the laws of geometric optics. There is no need to invoke mental processes. The retinal image, if it obeys the laws of linear perspective, correctly depicts the field of view. With this theory, a camera is an accurate substitute for an observer. The weakness with this theory, as has been demonstrated in discussing size constancy, is that there is often a mismatch between what we think we see and what the camera ‘sees'.
The Gestaltist theory suggests that mental operations play a much larger role in perception. There are a number of visual concepts the mind employs in the act of perception, which is much more complex than the simple mechanical process suggested by the perspectivist theory. A third theory, the constructivist, suggests that perception is an active process by the observer, who is constantly making assumptions and testing out visual phenomena until he is satisfied he has made the correct interpretation. These are not compatible ideas but there are elements of each that can be combined to suggest guidelines when attempting to compose a shot.
Figure 3.4 Searching for coherent shapes in a complex image, human perception will look for and, if necessary, create simple shapes. Straight lines will be continued by visual projection (cube shape from ‘Organisational determinants of subjective contours', Bradley and Perry, 1977)
Characteristics of perception
Most theorists agree that perception is instantaneous and not subject to extended judgement. It is an active exploration rather than a passive recording of the visual elements in the field of view and is selective and personal.
The mind makes sense of visual elements by grouping elements into patterns. Any stimulus pattern tends to be seen in such a way that the resulting structure is as simple as the given conditions. Making sense of visual stimuli involves testing by hypothesis. An unfamiliar or ambiguous image may be assigned a tentative definition until further information becomes available (Figure 3.3).
An American law lecturer once tested the accuracy of his students’ ability to witness an event by staging a fake crime in his lecture hall. A man ran into the hall disrupting his lecture and brandished a weapon of some kind and then left. The law teacher immediately asked his students to accurately describe what they had seen. Needless to say every student ‘witness’ had a different version and a different description of the bogus criminal. The simple point was made that most people are selective in their viewpoint. They see what they expect to see or what they can understand.
What a person perceives is dependent on personal factors as well as the visual elements in their field of view. Their understanding of an image reflects past experiences as well as their present state of mind Although it is probable that no two observers may observe a given scene in the same way and may disagree considerably as to its nature and contents, much of our perceptual experience shares common characteristics.
Figure 3.5 (a) (b) Sometimes you cannot see the tree for the wood. A shift in camera position may establish figure/ground priorities and allow the dominant subject to be emphasized
How the mind responds to visual information
Much of the theory of perceptual characteristics has been influenced by Gestalt psychologists. Gestalt is the German word for ‘form’ and these psychologists held the view that it is the overall form of an image that we respond to, not the isolated visual elements it contains. In general, we do not attempt to perceive accurately every detail of the shapes and objects perceived but select only as much as will enable us to identify what we see. This may depend on the probability of appearance of a particular type of object but the precision of our perception is sufficient only for our immediate need.
We may increase our visual concentration if we feel it is warranted but this enhanced visual attention may be of short duration. The tendency is for the perceptual system to group things into simple units.
The minimum amount of time needed to recognize an object (possibly 1/100 second) will depend on the familiarity and expectation of that specific image. An observer can perceive a large and complex picture that is seen everyday and is anticipated in a time that would be quite inadequate for the perception and understanding of a complex meaningless shape.
In searching for the best interpretation of the available visual data we utilize a number of perceptual ‘shorthand’ techniques that include organization of similar shapes and similar sizes. Shapes that are similar are grouped and form a pattern that creates eye motion. A ‘good’ form, one that is striking and easy to perceive, is simple, regular, symmetrical and may have continuity in time. A ‘bad’ form, without these qualities, is modified by the perceiver to conform to ‘good’ form qualities.
Perceptual steps
Perception is extraordinarily fast. This can be demonstrated by the deductions and judgements made when driving a car or, as a pedestrian, the perceptual calculations made when crossing a busy city street. Each element of the perceptual steps may operate instantaneously or occur in an order conditioned by the visual situation.
First, there is the need to separate figure from ground. Figure describes the shape that is immediately observable whilst ground defines that shape by giving it a context. A chess piece is a figure with the chess board as its ground. Identifying the shape of the figure – that it is a pawn – may provide complete recognition. Other subject recognition may involve colour, brightness, texture, movement or spatial position. Instantaneous classification and identification occur continuously but the perceptual process can be helped or hindered by the presentation of the subject (Figure 3.5(a) and (b)).
Figure 3.6 A simple shape such as a cube is easily seen in isolation but is camouflaged when swallowed up in a more complex figure. The centre of interest of a composition requires visual emphasis (Gollschadt diagram)
If the image is familiar, recognition may be instantaneous and therefore there is a redundancy of information. If the image is unrecognized then there may be a rapid search and match through memory to find similarities in mental images. When an unexpected image cannot be identified then either a guess is made or it is ignored.
People habitually overlook things they cannot understand. For example, a foreign news story in a TV news bulletin in which the political context or geographic situation is unknown to the viewer, ceases to be information and is ignored unless a connection can be established with an existing frame of reference. The reporter, who may have lived with the story for days, weeks or even months, may have an abundance of background knowledge to the specific two-minute item he files that day. This may cause him to overestimate the background information the viewer brings to the story. A similar extended preparation and filming of a narrative sequence may involve the production group investigating and discussing every nuance and significance of a 10-second shot. The first-time viewer of the shot has to extract all this days/weeks of considered deliberations during the 10-second running time of the shot. If a shot is viewed many times in editing, its visual impact can appear limited and easily understood. It is often tempting to shorten its screen time in order to inject pace and visual vitality into a sequence that has grown stale with repeated viewing in the edit suite. The audience, however, usually only see the shot once amidst a montage of other shots.
We predict what is likely to happen next from our experience of the past and rely on these assumptions to forecast the future. Shot structure and shot composition have to take into account this habit of searching for the cause of an effect.
Problems with perception
There is a basic distinction between ‘reading’ the space in a two-dimensional image, where a hypothesis of shape, depth, etc., has to be estimated by viewing from a fixed position; compared with the potential, in a three-dimensional situation, to move within the space to confirm a hypothesis. We cannot walk around a picture.
Our perceptual knowledge is gained from our experience of moving in a three-dimensional world. An essential element of testing and checking perceptual information is by moving through space. We use these acquired three-dimensional perceptual skills and frequently apply them to a very reduced image depicted in two-dimensions (e.g., a television screen) where we have no opportunity of testing out our depth ‘guesses’ by moving into the picture space.
Although the image created by a video or film camera may be similar to the image focused on the retina of the eye, there is the crucial difference of being unable to test out the depth indicators of a two-dimensional image by moving into its picture space. A moving camera can reproduce some of the image changes that occur when we move in space but not the visual depth checks achieved by binocular vision and head movement.
There is a considerable amount of visual information that is used in perception that is usually unacknowledged until an attempt is made to reproduce three-dimensions on a two-dimensional plane. If an untrained person attempts to draw a townscape they will soon realize that there are many aspects of visual representation of which they may never have been consciously aware. Although information about perspective of line and mass and vanishing points are present in the eye they are unexamined, even though they help us to determine distance.
As we move our viewpoint in space, so the appearance of objects alters. A plate may have the shape of a circle seen from above but viewed from any other angle its shape is never circular, but we persist with a mental image of a plate as a circle. We know that objects have an identity and a permanent form and ignore perceptual problems with the continuity of form produced by a changing viewpoint.
The two-dimensional representations of film and video provide image dimensions or size relationships of which we may not be aware. Sometimes the appearance of everyday objects are altered when lit from an unfamiliar angle or seen in extreme close-up. A low-angle, close shot of a golf ball against the sky in ‘Murder by Contract’ (1960), accompanied by the murmur of out-of-frame golfers, established an expectation of the normal object size, until the camera pulled back and disclosed that the normal two-inch golf ball was in fact a four-foot-high structure identifying the entrance to a golf course.
An unfamiliar setting or the absence of a field of reference to an object frequently creates difficulties in identifying what normally is instantly recognized. ‘It looks like an "x" but I must look at it longer to make quite sure’ can be caused by unexpected lighting or shot size. The meaning of a familiar image can be understood without conscious thought whereas intelligent interest is required to understand unfamiliar subjects. Recognition of images may be easily accomplished if the observer is favourably disposed towards them. That which tends to arouse the observer's hostility or antipathy may be either forgotten or ignored.
A tight shot of a woman feeding her baby with the woman's face looking down to the baby at her breast will engender, in most people, a feeling of warmth and human empathy. It is universal and timeless and will, in general, produce a feeling of uncritical endorsement. The same activity, if framed in a wider shot now showing mother and child in an exterior that includes shabby and broken coaches and caravans, dirty and half clothed children and a few mangy dogs roaming around a wood fire, will set up a completely different set of responses. Putting the original subject in a social context will provoke the viewer into bringing preconceived attitudes and social judgements to bear on the activity. It may even provoke anger that a woman should bring a child into the world in such an inhospitable and alien (to the viewer's frame of reference) world. What is included and excluded from the frame alters the way the central subject is understood.
Attention and perception
Perception is dependent on attention. If attention is concentrated on a small part of the field of view, little will be perceived of the rest of the scene. If attention is spread over a large area, no one part will be very clearly and accurately perceived. The total amount that can be attended to at any one moment is constant.
There is selective perception in everyday life, with people unable to attend to two different visual events; they either combine the two or their attention ends. It is not possible to continually attend to even one part of an image. After a short period, attention wanders, but by directing perception understanding improves.
Figure 3.7 Our attention is almost immediately captured by the ‘one’ that is different. The repetition of the brick shape provides an overall image unity whilst at the same time emphasizing the one exception
Attention can be split three ways even when watching a familiar TV event such as a weather forecast. The physical appearance of the forecaster, as well as their spoken commentary, will split the attention and to this is added a third part of the image that requires attention – the changing graphics of the weather chart. It is difficult to attend to all three elements even in this simple display without loss of attention to one part of the information presented.
Change blindness
Two psychologists at Harvard University devised an experiment to measure how attentive people are in an everyday situation. They were attempting to demonstrate how little we see of what comes through our eyes.
They set up a reception counter staffed by a young man to hand out forms. Individuals who were the subjects of the experiment approached the counter in turn. They were unaware at this stage that they were part of an experiment. When they handed the form back, the young man ducked down to get more information for them. While he was briefly out of their sight he was immediately replaced below the counter by another young man of different appearance including a different coloured shirt. The duplicate young man stood up and presented the material and gave directions to a room.
More than 75 per cent of the subjects, when debriefed, did not notice that the young man had changed between them receiving their first and second form. The 25 per cent who did notice may have been concentrating on a specific part of the young man and therefore noticed his appearance had changed when he re-emerged from behind the counter. Because our attention system allows us to actively select what to look at, we often miss large changes to our visual world that appear to be perfectly obvious to someone who knows what is going to change.
The brain fills in the gaps
The eyes are the slaves of our attention system. In film making this can be exploited and directed by subject emphasis in a shot, for example, a shadow behind a door or a face specially lit in a group of faces.
Viewing a horror film demonstrates the power of the viewer's imagination. It is not what is on the screen but what we think we see on the screen (e.g., the shadow behind the door). The less we see the more we imagine. There is nothing more vivid than the pictures we generate in our head.
Research suggests that our brains are constantly distorting what we see. We guess what is out there from past experience rather than having to build up images instantaneously. The brain fills in a vast amount of additional information. The brain just doesn't see – it invents much of it. Perception is a highly personal inner world.
Neuro scientists suggest that although the world appears to be visually high resolution and photographically sharp, a lot of this is filled in from memory. In recent years perceptual research has suggested the visual brain relies as much on information from our memory as from our eyes. We are using information from the past to imagine what is out there.
Vision is not a one-way transfer of information from eye to brain. There is also an exchange between the 30 known areas of the visual brain and the use of stored visual information. Our perception of the world is as much what we expect to see as it is about what we actually see. When composing images this aspect of perception can be exploited.
Summary
There is a mismatch between how we think we see a subject and how the camera records the same subject due to a perceptual distortion called size constancy.
The mind tends to group objects together into one single comprehensive image. The mind sees patterns and composition can enhance or facilitate this tendency or it can prevent it. A knowledge of how the mind groups visual elements is a valuable tool for good communication.
Test the strength of a composition by examining the individual visual elements it contains and check if they separately or collectively strengthen or weaken the overall form.
Although all these perceptual habits may seem obvious and unremarkable they often play a significant part in shot composition and cannot be overlooked when seeking to maximize visual communication.