The use of the lens in film-making is sometimes described as ‘painting with light’. The metaphor suggests that using a camera lens is like using a brush: that the result is under the total control of the user, the artist. That is not really the case. A lens acts on the light rays which pass through it according to the laws of physics. The photographer, still or cine, and the film-maker who directs the photographer’s efforts, is actually working with those laws and their consequences.
It is probably most useful to think of the camera lens as a viewer’s eye, borrowed for the occasion, and directed towards whatever the film-maker wants the viewer to see. Like a human eye, a camera lens is a converter of three dimensions to two, collecting light rays which stream from a three-dimensional scene and projecting them through a variable aperture—the iris—onto a flat surface: either onto a strip of film or onto the photosensitive screen of a video camera. The idea is to create, on that flat surface, an image which is, as far as possible, exactly what a viewer’s eye would see if the eye were in the same position as the lens.
Thus both the eye and the camera lens are concerned with projection onto a plane, the conversion of three dimensions into two. In this respect, both painting and film-making are concerned with a similar task and film-makers have something to learn from the history of painting. For in a sense, the art of the photographer is the fulfilment of the long search by painters for a technical solution to the problem of representing three dimension on a flat surface: a way of recording where the straight light rays emanating from a scene intersect with a flat sheet.
The artists of the European Renaissance worked hard on this problem, particularly in relation to the design of trompe-l’oeil scenery for the new dramatic art of opera. Leonardo da Vinci and Albrecht Dürer devised a number of methods and kinds of apparatus for ensuring that the two-dimensional image corresponded exactly to the painter’s-eye-view. Leonardo suggested:
‘Have a large sheet of glass about half the size of a sheet of good quality paper, and affix it firmly in front of you so that it is between you and the thing you want to portray; then stand back so that you are at a distance two thirds of an arm’s length from the glass, and use an instrument that will hold your head still so that you cannot move it. Then either cover or close one eye and with the brush or pencil trace on the glass what you see…’
Another way of obtaining a similarly accurate result was the camera obscura—a ‘dark room’. This was already well known during the Middle Ages. Light rays pass from an exterior scene into a blacked-out room through a pinhole which focuses them onto a whitewashed wall at the back. Such ‘rooms’ were built in many sizes. Some painters used the camera obscura and its close relative the portable camera lucida—the light room—to help compose landscape scenes. Today’s children still play with small pinhole cameras made of cardboard, with a translucent paper screen at the rear.
The great advantage of the pinhole camera is that it has no focus. That is to say: all objects, at whatever distance, are equally sharp. Each point of light in the picture cast on the back wall is a tiny image of the pinhole. The smaller the pinhole the smaller the dots of light making up the image and therefore the sharper the focus. When the pinhole is infinitely small, the overall focus becomes infinitely sharp. Unfortunately, an infinitely small pinhole lets through infinitely little light, so the picture is also infinitely dark. A pinhole camera must always find a compromise between sharp focus and a visible image.
The introduction of a lens solves this conundrum, though at the cost of introducing new limitations of its own. Unlike a pinhole, which brings all objects into focus on its screen, a lens at a given distance from a screen can bring into focus only objects at one particular distance in front of it. That distance is a function of the distance between the lens and the screen, the focal length of the lens. If an object is not at the appropriate distance its image will be out of focus. Anyone who has used a still camera will know that a decision has to be made about which objects in the scene are to be in focus and which not.
A camera obscura equipped with a glass lens was first demonstrated by Professor Daniel Barbaro in Padua in 1568: ‘Close all the shutters and doors until no light enters the room except through the lens, and opposite hold a sheet of paper, which you move forward and backward until the scene appears in sharp detail. There on the paper you will see the whole view as it really is.’1 But it was probably not until the seventeenth century in Holland that lenses were first regularly used in place of pinholes to project images for artists to paint from. The time was one of intense exploration of the science of optics. It was also—and perhaps not by coincidence—one of the high points in the history of Science 1995 of European painting. Some scholars believe that The Netherlands painters Jan Vermeer and Peter de Hoogh may have used a camera obscura fitted with a lens to produce a number of their domestic interior scenes. For in some of Vermeer’s pictures objects or characters in the foreground are painted slightly out of focus, seeming to throw the middle-ground and background into the distance. It is the first time such an effect is evident in the history of painting.
Whether or not the great Dutch master really did use a camera obscura, there are other reasons for photographers and documentarists to count him among the originators of their craft. For he certainly was among the first painters to confront the problem facing all television film-makers: how to suggest the illusion of everyday reality and develop realistic spatial depth within a small flat image intended to be viewed in normal light among everyday surroundings. Vermeer’s use of a lens may well have been the inspiration behind his manipulation of perspective and the use of different planes of focus in his compositions. But it was his ability to raise the naturalistic, the ordinary and the everyday to the highest plane of art which makes study of his work a valuable exercise for photographic image-makers today.
Early lenses were made of simple single pieces of glass. But such lenses suffer from aberrations and distortions, particularly at the edges of the picture. White light tends to be split into its constituent colours, straight lines may come out as curves. To overcome this difficulty, today’s lenses are made of multiple sections—they are compound lenses. Such lenses can almost get rid of aberrations, though not without introducing other problems.
Light suffers losses, by reflection, diffraction and scattering, when it passes from one medium into another. In a compound lens, light has to cross between a number of pieces of glass cemented together, losing something in the process at each crossing. The result is a spreading of what should be a point of light into a small disc, the ‘circle of confusion’—much the same as what happens in a pinhole camera when the hole is not small enough. This causes the picture to become slightly fuzzy or blurred. Additionally, as some light is lost altogether in crossing between so many pieces of glass, the brightness of the image is reduced.
Earlier film cameras used a number of lenses, often mounted on a revolving turret for convenience, rather than the single zoom lens which modern cameras favour. Today zoom lenses are almost universal. But, as the zoom facility is only made possible by stacking yet more pieces of glass together, there are still cine photographers working today who prefer to use—even insist on using—fixed lenses, even though modern optics suffer far less from the defects of compound lens design than did those of the past.
The two defining characteristics of a lens are its focal length and its aperture.
The focal length of a lens specifies the distance between the lens and the screen on which it focuses but it is not necessarily a direct measure of that distance. A very short focal length may not allow enough room for the camera shutter between the lens and the film. A long focal length lens may be too unwieldy and need placing rather less than the focal length in front of the electronic screen. Modern lenses deploy various optical techniques to fold the light rays and overcome these difficulties. The best one can say is that a lens with a given focal length produces an image on the screen which is the same size as the image which would be produced by a pinhole at the focal length distance from the screen. In other words, a 100mm lens produces an image the same size as would a pinhole 100mm away from the screen.
A change in the focal length appears to accomplish no more than a change in the magnification of the image but this result is much more important than it may at first seem. It makes focal length one of a lens’s key characteristics, responsible for controlling the perspective of the image. Though we speak of focal length, the distance between lens and screen is not what we are really interested in; much more relevant to us is the fact that the angle of view which a lens provides is directly related to its focal length.
The standard lens of a 35mm still camera usually has a focal length of some 50mm. This is intended to provide an angle of view equivalent to that of the naked eye. Film and video cameras are slightly different but their standard lenses are designed in the same way. Yet the view seen by the naked eye is not always what is required. Often a wider or a narrower view is called for. This is achieved by changing the focal length of the lens.
Given screens of the same size onto which to project, it is clear that the longer the focal length of the lens, the narrower the angle of its view (Fig. 6.1).
Fig. 6.1 The longer the focal length the narrower the angle of view.
Thus a longer focal length lens will appear to project a larger, closer image. A short focal length lens will project a wider view (Fig. 6.2).
Fig. 6.2 A shorter focal length lens will appear to project a wider image; a longer focal length lens will project a narrower view.
The focal length of the lens not only affects the apparent size of an object, it also determines the apparent spacing between different objects arrayed at different distances in front of a camera.
Consider a figure standing some distance in front of a building. If one keeps the size of the figure on the screen constant by moving the camera bodily backwards and forwards, the narrower the angle of view the smaller the amount of background that will be included in the image. The result is that a narrow angle lens will seem to compress distances and a wide angle lens will exaggerate them (Fig. 6.3).
This phenomenon is responsible for many familiar effects of perspective distortion. A narrow angle lens will make far-off mountains seem higher, distant buildings more imposing, a tyrant’s statue in the background more threatening. Shoot with a narrow angle lens along a street and pedestrians or cars or telegraph poles seem more crowded together. Shoot down the length of a train as the carriage doors open and they will seem to disgorge a tight press of humanity.
By contrast, a wide angle lens makes any space seem larger and deeper. Small rooms are stretched, submarines made to seem less claustrophobic, city boulevards are magnified, freeways seem extended to infinity. Studio drama is often shot through wide angle lenses to make cramped sets appear more spacious.
Fig. 6.3 A wide angle lens will seem to exaggerate distances and a narrow angle lens will compress
Any movement through space towards or away from the camera is affected too. Because of the foreshortening effect, shoot an athletics race through a narrow angle lens from some way off and because distance is compressed, the runners will appear to be expending huge amounts of energy yet covering hardly any ground at all. Conversely, shoot a slow chase with a wide angle lens and characters moving in line with the axis of the lens will seem effortlessly to move the distance in leaps and bounds.
Narrow angle lenses must be used with care. Because of their greater magnification, camera movements will seem to be exaggerated. Pans and tilts need to be slower than with a normal or wide angle lens to maintain the same speed of movement on screen. When shooting with a narrow angle lens, the camera needs to be set up on a stable tripod to avoid any unwanted movement. Operating hand-held is inadvisable.
Use of wide angle lenses often results in distortion of the image. This arises from a number of factors, some being lens defects, others being the consequence of the geometry of projection.
Perspective can become greatly exaggerated. The size of an object projected onto the screen is proportional to the distance from the lens. An object half the distance away from the lens is projected at twice the size. In a close-up of a figure with an outstretched hand in which the body is twice as far away as the hand, the hand will appear to be magnified to truly monstrous proportions.
This applies to all lenses, though it is most noticeable in wide angle lenses because they tend to work closer to the subject than narrow angle lenses. It is not the result of a defect. It comes from the fact that the proper perspective can only be presented to the viewer’s eye when that eye is placed at the geometrically correct distance—the focal length—from the television screen. More exactly, since the image is actually enlarged to fill the television screen, the correct viewing distance is equal to the focal length of the lens multiplied by the magnification used to recreate the image. For example, if a frame of 16mm film shot through a 10mm lens is magnified some 35 times to appear on a television screen, it should properly be viewed from a distance of 350mm (about one foot two inches) to make the perspective seem natural. In reality the viewing distance of a television screen is never anything like that close. As a result many images shot with wide angle, short focal length, lenses appear to be distorted. The distortions are less evident when using narrow angle lenses, as the distance from viewer to television screen is more commensurate with the focal length of the lens.
Other perspective effects may be equally disconcerting. Straight vertical and horizontal lines may become curved in wide angle views, especially towards the edges of the frame. This is often what the eye really sees, but while the brain automatically compensates for it in interpreting the image, the camera gives us the image raw and unadjusted.
There are also problems associated with the need to project the image onto a flat plane. Because the light rays are channelled through nearly a single point, their focus lies not on a flat plane, but on the inside of a sphere centred on the lens. The focal length is the radius of that sphere. The shorter the focal length (the wider the lens angle), the smaller the sphere and the more pronounced its curvature. The longer the focal length (the narrower the lens angle), the longer the distance from lens to screen and therefore the larger the diameter of the sphere and the flatter its curvature (Fig. 6.4). A section of the surface of a very large sphere is not very different from a flat surface. But a section of a small sphere is.
Fig. 6.4 Projection of a curved image onto a flat plane produces distortion away from the centre; the wider the lens angle, the more the distortion.
All projection of a curved image onto a flat plane produces distortion away from the centre. The greater the curvature the greater the distortion. The result is that distortion around the edges of pictures becomes quite noticeable in images shot through a wide-angle lens.
This is a particular feature of very wide angle lenses, sometimes known as fish-eye lenses, which may in addition sacrifice some optical accuracy in the interests of cramming in as wide a viewing angle (sometimes as much as 180 degrees) as possible.
The distortions may not be very noticeable in shots that are relatively static. But objects moving through the frame will seem to alter their proportions as they do so. Panning or tilting the camera will make the objects in view change shape and size as if they were oozing. Such an image is disturbing to the viewer, perhaps only acceptable in its extreme form if the aim is to suggest disorders of mind or of perception: psychoses, nightmares, drunken fantasies, panic attacks.
If the first defining characteristic of a lens is its focal length, the second is its aperture. This determines the amount of light which passes through on its way to the screen. The larger the aperture, the more light. The more light, the brighter the image. Thus aperture is a key factor in regulating the exposure of the image and therefore the overall tone of the picture.
On a still camera there are two ways of adjusting exposure: aperture and shutter speed. The faster the shutter, the shorter the exposure and therefore the darker the picture. The two adjustments: aperture and shutter speed, are both used together to provide for the particular circumstances in which the photograph is taken. But on both film and electronic cameras, the recording medium—film or tape—moves forward inexorably by 24 (or 25 or 30 depending on the technology) frames per second. The shutter speed is effectively fixed, and cannot be used to adjust for different light conditions. So a great load falls on aperture control. It also means that the lighting conditions under which moving images are filmed are much more constrained than with still photography. Hence film cameramen and women must be prepared to provide artificial lighting in circumstances in which a still camera would be able to operate by available light.
Measurement of the aperture of the lens of a still camera is given in ‘f’ numbers. These depend not only on the diameter of the iris, but also on the focal length of the lens. The shorter the focal length of the lens—the wider the viewing angle—the greater the amount of light passing through the lens. The ‘f’ number is mathematically derived by dividing the focal length of the lens by the diameter of the actual aperture (in other words: is the ratio between them). As both are defined in millimetres, the result is a non-dimensional number. Thus ‘f’ numbers grow larger as the effective aperture gets smaller—f22 lets through less light than f8. The brightness of the resulting image is inversely proportional to the square of this number. So as to make brightness increase twofold with each stop, the convention has arisen to make each ‘f’ number approximately the square root of a multiple of 2. Thus the usual ‘f’ numbers are f1 (square root of 1), f1.4 (square root of 2), f2 (square root of 4), f2.8 (square root of 8), f4 (square root of 16), f5.6 (square root of 32), f8 (square root of 64), f11 (just under the square root of 128), f16 (square root of 256) and so on.
These are entirely theoretical figures, however, resulting from the mathematical relationship between the size of the light passage and the focal length. Film and video cameras usually characterize aperture by empirical ‘t’ numbers, based on the ‘f stops but with the light transmission physically measured by the lens manufacturer.
Generally, the aperture of the lens is set by the photographer to capture the image with the best possible overall exposure. Here there is a considerable difference between film and video. The ability of film to record many levels of illumination is far greater than that of video. The latitude of film means that even though shadow detail may not be visible in a straight print, the information is actually still recorded on the film and can be brought out by special development. The same goes for detail in the highlights of a picture. Thus the difference between highest and lowest usable exposure is far greater on film than on video. It is possible, for example, to shoot a film interview against a bright sky with the interviewee’s face in shade and still get acceptable exposure of both. On video, the face would be almost in darkness and the sky burnt out to white. The video camera operator must be much more careful than the film camera operator to place the exposure right in the centre of the range.
If the aperture of the lens affected only the exposure of the picture, it would be a relatively simple matter to set correctly. But aperture also has an impact on focus, in particular on the range of distances in front of the lens between which objects will be in focus. This is known as the depth of field and is an important aspect of picture composition.
A pinhole projects a uniformly sharp image. A lens with an ideal pinhole-sized aperture would also create a picture which is focused all over. A wide open lens, by contrast, creates an image in which only objects at one particular distance are in focus. Real lenses are not ideal pinholes but neither do they entirely lack the pinhole effect. A real practical lens will project a picture in which a whole range of distances will be in focus. The smaller the aperture, the greater that range—in other words, the depth of field—will be.
If the lens is focused to infinity, the closest point to the camera that will be acceptably in focus is called the hyperfocal distance. This distance is proportional to the focal length divided by the ‘f’ stop (giving the actual physical size of the aperture). Given the same ‘f’ stop, the longer the focal length the further away the hyperfocal distance, therefore the closer to infinity, and therefore the narrower the depth of field. At the same time, the larger the ‘f’ number, the smaller the aperture, therefore the closer the hyperfocal distance to the camera, and therefore the greater the depth of field. The limiting condition is, of course, with a pinhole of zero aperture, when everything in the image from the lens to infinity will be in focus—but given zero exposure.
The depth of field in a shot will therefore depend on both the focal length of the lens and its aperture. Working close in with a wide angle lens—one with a short focal length—gives a cine-photographer great freedom, as the depth of field is large and the camera can follow most actions without much fear of losing focus. Working with a narrow angle, long focal length lens is much more difficult and demands far greater precision.
Since adjusting the aperture of the movie camera lens cannot be compensated by changing exposure, as with a still camera, it is important for a film-maker to recognize that working in different levels of illumination will have an impact on the depth of field. In dark environments, where it is necessary to work with the lens as wide open as possible, depth of field will be seriously reduced. Conversely, under very bright illumination, the lens will be stopped down and the depth of field correspondingly increased.
Note
1 The Faber Book of Science 1995.