I.

Introduction

Cinematography, art of making motion-picture films. Although Thomas Edison had patented the kinetoscope in 1891, incorporating many of the important features of the modern cine-camera, it was the launch in 1895 by the Lumière brothers in Paris of the Cinématographe, projecting films onto a screen for a large audience, that introduced the new mass art form of the cinema. It only needed the invention of practical, synchronous sound systems—Vitaphone in 1926 and Movietone in 1931—for the foundations of modern cinema to be complete (see Cinema, Early Development of).

The functioning of the cinema depends on two properties of the human eye and brain. The first is referred to as persistence of vision. If an image of a static light is focused onto the retina and is switched on and off slowly, we can see the flashing. If the frequency of flashing is increased there will come a point where the light appears to be steady. This is known as the critical fusion frequency. When the ambient lighting is at a low level and the retina is said to be dark adapted, this frequency is lowered.

A further effect of this phenomenon is that if the image of a continuous light moves on the retina, the persistence of the nerve impulses can cause a smearing of the image, which is most marked when the system is dark adapted. This is why a torch rotated in a circle in a darkened room appears to the eye as a continuous circle, since the source of light keeps returning repeatedly to the same position, before the retinal activity has appreciably faded.

Another property of the system of visual perception is referred to as apparent motion. If separate lights are switched alternately, at the correct frequency and spacing, we seem to perceive a light moving between the two positions. This property of brain neural networks is sometimes known as the phi phenomenon. It is probable that it also plays an important part in the illusion of smooth motion in the cinema, but cannot fully account for the illusion in static scenes.

Cinemas are darkened so that viewers' eyes are dark adapted and the critical fusion frequency lowered. In addition, the projector has multiple blades, so that each image is projected twice. This reduces the length of the dark part of the cycle and improves the subjective continuity. The rapid succession of still images (usually 24 per second) produces the impression of a continuous image, and movement seems to be smooth, even though actually presented in steps. The effect is further enhanced by the blurring of moving objects, caused by the relatively long exposures in a cine-camera, usually about of a second, only static or very slow-moving images being sharply “frozen”.

For reasons discussed below, most productions are still shot on film. The first large-scale cinema films to be shot using digital cameras have been released and digital technology is likely to take over from much of the present use of film cameras and cinema projectors eventually.

II.

The Cine-Camera

Although the fundamental purpose of a cine-camera, the taking of still photographs, is basically similar to that of other cameras (see Photographic Techniques), the need to take so many photographs per second means that the necessity for rapid transport of film through the camera dominates the design of the mechanism. Since, at 24 frames per second, one minute of filming uses over 27 m (90 ft) of 35-mm film, cameras are fitted with magazines that can hold 122 m (400 ft) or 305 m (1,000 ft) of film. In order to draw the film smoothly off such large rolls, and in order to wind it up again after exposure, the film has to run continuously in the camera. However, to take the photographs, the film must move in steps where the individual frames are exposed.

Continuous drive of the film is provided by one or more toothed drums or sprockets round which the film passes, held in contact with the teeth by guide rollers. The sprocket teeth engage in perforations along one or both edges of the film and drive the film smoothly.

The gate is a metal plate, with a rectangular gate aperture, against which the film is held flat, from behind, by the spring-loaded gate pressure pad. In front of the gate is the lens, which focuses inverted images of the subject being photographed onto the surface of the film, through the gate aperture, to form the frames on the film. In cine-cameras, the film usually runs vertically downwards and the height of each frame, plus the space between it and the next frame, is usually the length of four perforations, whereas, in a still camera, where the film lies horizontally, each frame takes up the length of eight perforations.

Between the lens and the film is a continuously rotating shutter, which alternately opens to expose the film in the gate and then closes while a fresh section of film is pulled down into the aperture. A typical shutter consists of a blade that is semicircular, so that the angle of the open section is 180°. The shutter is, therefore, open for half of the time and closed for half of the time. At 24 frames per second, this produces an exposure of of a second. On some cameras the exposure can be varied by altering the shutter angle.

The intermittent movement of the film through the gate is usually achieved by a claw mechanism that moves in a continuous cycle, engaging with the perforations, pulling the film downwards in a step, and then retracting to recover during the exposure. It then repeats the action. In order to guarantee that the film is accurately located and absolutely still during exposure, many cameras are fitted with register pins, which slide into the perforations and secure the film during the exposure and then retract for pull-down. In order to reconcile the continuous and intermittent drive of the film, the camera is loaded with small, free-standing loops of film above and below the gate.

To achieve accurate framing, most cameras have reflex viewfinders. The front surface of the shutter is set at 45° to the axis of the camera and is mirrored so that, during periods when the shutter is closed, the image is reflected at right angles onto a ground-glass screen, or fibre-optic screen, exactly the same dimensions as the frame of film in the gate. The camera operator has a magnifying viewer, through which he or she can view the image on the ground-glass screen, from the back of the screen. Many film cameras also have a miniature video camera that transmits the image on the ground-glass screen to a closed-circuit television system for monitoring and playback. This is referred to as a video assist.

The quality of the projected image depends on the area of each frame. For professional film-making, film 35 mm wide is the norm, although 16-mm film is widely used for television and documentary film-making. Frequently, a version of 16 mm, called Super 16, is used, where only one set of perforations is employed for film transport, freeing up extra space on the film. The aspect ratio of the resulting image can be used for television presentation, with part of the top and bottom of the frame masked in black. It is also well suited for showing on wide-screen televisions and high definition television systems. In addition the image can be enlarged to 35-mm wide-screen for theatrical distribution. Amateur use of 8-mm film has largely been replaced by the use of video camcorders. In some selected cinemas, enlargements of the image onto 70-mm film can be used, to produce the best quality of presentation.

III.

Digital Cinematography

Being entirely electronic, digital cameras have no moving parts. The various intensities of the image produced by the lens are focused onto a solid state chip called a charge-coupled device (CCD). The typical CCD has a much smaller area than a frame of film, being ’ of an inch (16.93 mm), but comprises over two million elements or pixels, laid out in a pattern 1,080 pixels high by 1,920 pixels wide. This ratio is called common image format (CIF). The intensity of light on each cell of the array is converted to 12-bit digital information for storage on high capacity memory. To record colour, the camera has a beam-splitting device that splits the picture into three identical images. The images are then passed through a primary colour separation filter (red, blue, and green respectively) to separate the relative intensities of each colour and the colours are recorded by three separate CCDs.

The data stream from a digital camera is recorded on a digital tape recorder. A lower resolution video version is later made for editing. (See Digital Editing, below.)

IV.

Laboratory Processing

The very large footages of film produced by cine-cameras are processed on big, continuous-processing machines, which develop the negative images on the film. These are then printed onto positive film on a printing machine. Most printing is done using continuous contact machines, where the developed negative and unexposed print films are run together, in contact, past a controllable source of light. The light passes through the negative, exposing the images directly onto the print film, which is then developed to produce the positive images. Other printing machines consist of a projector that uses a lens to project the negative image directly onto print film in the gate of a special type of camera. This type of step optical printer can be used for a wide variety of purposes such as enlarging (“blowing up”) from a 16-mm negative to a 35-mm print and for many special effects.

The light used for printing is produced by a special additive source, where the amounts of red, green, and blue light can be regulated independently to control both the exposure and the colour balance of the individual shots. By correcting small variations between shots, smooth photographic continuity is achieved. The technique of controlling the printer light settings is referred to as grading or timing the print.

Traditionally, the practice was for the first print to be sent back to the film-makers as soon as possible, usually the following day, so that they could check that the scenes had been filmed correctly. These prints are called rushes or dailies and, after viewing, were used by the editor to assemble the shots into the cutting copy or work print. They were joined or spliced in the right order and the right points at which to cut between the shots were determined. Recently, digital techniques have involved the editing decisions being made using virtual cutting copies in a computer. However, in fully budgeted films the first print is still made, as most film-makers wish to see their rushes on film to enable them to make accurate judgements about the photographic quality of the material as it will be seen on release in the cinema. If necessary, scenes may then be re-shot. If cost savings are critical, video versions of the rushes may have to be accepted as dailies, though high-quality telecine and video projection are necessary if the judgements are to be relied upon.

V.

Sound Recording

When precise synchronization between sound and picture is required, such as in dialogue scenes, where precise matching of the sound and lip movements is necessary, silent, or “blimped”, cameras are used and the sound recorded directly when filming. Sound recording is usually done on digital audio tape (DAT) recorders, normally in stereo. Synchronizing information is also recorded, usually in the form of time coding, so that sound can be matched to the corresponding images. Many film-makers still use clapper boards at the start or end of each take, as this speeds up the synchronizing of the picture and sound rushes for viewing, compared with lining up time codes on a special reader.

If these recordings are not of sufficient quality, for example, due to intrusive background noise on location shoots, they can be re-recorded later in dubbing studios, where the actor views short sections or loops of the picture, running continuously, with their original location recording on earphones. After repeating the lines several times in sync with the picture and earphone sound, they can then re-record the dialogue, not only accurately lip-synced, but also closely matching the tone of voice, inflexions, and emotional performance of the live acting.

VI.

Digital Editing

Nearly all films are now edited using techniques referred to as non-linear editing. The original negative is transferred on a telecine machine to high-quality video cassette, analogue or digital, and the images transferred to mass storage, in dedicated editing machines. The computer has several monitors and uses split screens to present the editor with all the shots for a given sequence, performance, and so on, and a graphic representation of the succession of shots and sound. Alternative cuts can be tried out and as little as one frame added or cut out until the editor is satisfied that the decisions have been made. The results can be reviewed frequently with the director and changes made extremely quickly. The coding system ensures that images and synchronous sound remain accurately matched and edited.

When the final form of the film has been decided, the computer produces a detailed list of the sections of film used, called the edit decision list (EDL). A specialist technician will then work through the list, and cut the picture negative to comply with the EDL. At this stage negatives of shots with special effects added are incorporated. These may include computer-generated images (CGI) produced by highly specialized equipment and technicians.

The computer also produces digital versions of the separately laid soundtracks. Any soundtracks in addition to the synchronous dialogue, such as sound effects, music, or commentary are recorded separately and their placing and combination decided upon and laid. The editor may carry out experimental sound mixes but it is usual for a formal sound mixing session, often called the dubbing session, to take place in a sound studio using the skill of a specialist “mixing” technician. In most cases the soundtrack is transferred to an optical negative, using a machine called a sound camera, to produce an optical soundtrack. The picture and sound negatives can then be printed onto the same film to produce a master “married” positive. This is printed again to a special type of positive film, from which duplicates of the negative are struck to produce the many hundreds of prints of a film required for a major release.

This strategy for the production of feature films is the virtual equivalent of the original scheme, only the form of execution of the editing decisions is digital. An alternative strategy carries the digital phase much further.

A.

On-Line Editing

In this case, the digital phase is carried right through to the penultimate stage of post-production. From the edit decision list, each section of original negative is separated out, with a few frames extra before and after the decision points, and has short lengths of spacing film or leaders and trailers, called handles, attached at both ends. In each case the leader has a start mark, from which the selected frames to be used are counted. The section is then scanned to a very high resolution by a laser scanning device, which uses blue, green, and red lasers to digitize the selected material. This pre-selection ensures that only the used sections of the negative are expensively scanned. Each section is coded with a consecutive identifying time code. Sometimes, a video copy is also made, with the numbers from the edge of the negative, called key numbers, inserted in the picture so that the assembly of the sections can be checked for accuracy.

The data of each section of negative is then imported into a non-linear computer and the segments are “stitched” together. At this stage, the colour and exposure grading of the production is carried out. On-line editors have taken over much of the function of the laboratory grader and are sometimes referred to as “colourists”. A number of effects can also be included, such as a fog filter. More complex effects (CGI) are produced by sending the selected digital data to separate facilities where the special effect, such as a light sabre, is added. The new digital data segment is then returned and assembled into the final version.

The graded and assembled digital data of the film is transferred to special ultra-fine grain film negative using a machine with red, green, and blue lasers at very high resolution. These machines record digital images to film at a rate from about 3.2 to 6.5 seconds per frame, so that one second of film action takes from 1 min 17 sec to 2 min 36 sec to transfer. One manufacturer claims to be able to record an entire feature film of 90 minutes duration in a week. From this negative, film prints for distribution to cinemas are made.

An alternative to the production of film prints is to carry digital techniques to the cinema screen directly using digital projectors. (See Digital Projection, below.)

VII.

Film Projection

A film projector has a gate, analogous to that in the camera, but in this case there are apertures both in front of and behind the film. Each frame is intensely illuminated from behind by a powerful light source, such as a sealed xenon arc. A lens focuses an image of each frame onto the screen in the auditorium. The film is pulled through the gate in steps by means of an intermittent sprocket, situated just below the gate. Unlike the camera shutter, the projector shutter has two blades and each frame is shown twice. This decreases the dark periods so that they are less noticeable to the dark-adapted eyes of the cinema audience.

An accurately focused beam of light, in the form of a narrow slit, is shone on the optical soundtrack running down the length of the film, to one side of the pictures. The track modulates the beam, varying the light that falls onto a photoelectric device, which converts the variations of the light into electronic signals. These are amplified and reproduced over the loudspeakers. The soundtracks may use a complex digital coding system to reproduce stereo and surround sound in the cinema, as well as to reduce the effects of electronic noise. Techniques for showing films with fully digital soundtracks are still not yet completely standardized and are likely to evolve.

Films are shown in cinemas on screens with different aspect ratios. Originally, the width and height of the screen were in a ratio of 1.33:1 (4:3). Most films are now shown with the top and bottom of the frame masked, so the ratio is 1.85:1—a system referred to as “wide-screen”. In a technique known as Super 35, it is masked down to a ratio of 1.35:1. Some films are photographed using anamorphic lenses, which “squeeze” the image horizontally. This makes them appear vertically elongated because their width is halved, relative to their height. It also means that a wider view is squeezed into each frame. These films are projected through corresponding lenses, which “stretch” the image horizontally to restore their proportions. The whole picture is therefore much wider and is shown on screens that are 2.35:1. Such films are usually referred to as “scope”, the earliest commercial version being CinemaScope^TM.

A few specially equipped cinemas can show 70-mm prints, running horizontally through the projector, using the very large images on the film for projection onto a giant screen. A special “rolling loop” system is employed to avoid the use of film claws and each frame is held flat, in an accurate position, by being pulled onto a glass plate by means of an air-suction system. Multiple soundtracks are used, in combination with the large pictures, to produce a spectacular show. Imax and Omnimax cinemas use these techniques.

Since the frame rate of standard cinema projection is 24 frames per second, motion will only be reproduced at a natural speed if the camera is run at the same speed. If the camera is run at higher speeds, the projector will slow the action down, producing the slow-motion effect. For example, if the camera is run at 48 frames per second, the projection will last twice as long and the action will be slowed down to half its natural speed.

Slower camera speeds will produce the opposite effect, speeding up the action. Time-lapse cinematography is achieved by using special motor and shutter systems on the camera, which can take frames one at a time. In this way a series of frames, taken over a very long time, is projected at a much higher rate, speeding up the action hundreds, or even thousands, of times. This technique is commonly used to reveal very slow movements, such as that of growing plants, or the emergence of a butterfly from its cocoon.

A.

Digital Projection

Except for the use of powerful lasers, the public use of which raises safety issues, the problems of illuminating a large screen with a digital image at sufficient intensity arises out of the difficulty of deflecting beams of light and controlling their intensity, which cannot be done using electrical or magnetic fields. A further problem is that the “look” of film projection arises mainly from the fact that all parts of the image of a frame of film are projected simultaneously and not scanned from side to side and top to bottom sequentially, as is the case with video.

However, the development of a new technology, known as digital light processing^TM (DLP), developed by the then Central Research Laboratories at Texas Instruments, has provided a method for the simultaneous projection of large numbers of pixels of light and the very high speed modulation of their intensity. This technique is based on the production of arrays of tiny mirrors, each of which can be deflected by a digital signal. The array, a digital micromirror device^TM (DMD), is constructed on the upper surface of a conventional CMOS static random access memory (SRAM) chip and consists of an array of 1,024 x 1,280, that is 1.3 million, aluminium micro-mirrors, each carried on a diagonal torsion hinge. Each mirror is 17 microns (17ìm) square, about a quarter of the diameter of human hair, and can tilt through an angle of +/- 10°, making a total movement of 20°. The close tiling of the square mirrors ensures that approximately 94 per cent of the area of the chip face consists of mirrors, making the total reflection very efficient and even. A system of electrostatic bias ensures that the mirrors are bi-stable, that is they can be “nudged” from one position to the other at high frequencies. The special version of the use of DMD devices for cinema presentation is referred to as DLP cinema^TM.

The digital data of a single frame used by the projector contains the information of red, green, and blue light to be projected to each point on the screen. The light from the xenon arc is split, by means of a dichroic colour splitting prism, into three primary colour beams, blue, green, and red respectively. Each beam shines on a separate DMD, which is used to modulate the amount of that colour in each pixel on the screen. After modulation, the three images are recombined for projection of the colour image to the cinema screen by the same dichroic prism that did the splitting.

The apparent intensity of light for each pixel is modulated by oscillating the relevant mirror. If it is held continuously in the “on” position the maximum intensity of light is reflected from the mirror through the lens to the screen. To reduce the brightness to the eye the mirror oscillates so that the apparent intensity is controlled by the ratio of the time during which it is in the “on” position and that during which it is in the “off” position. If it is held continuously in the “off” position the resulting brightness of the pixel is very low, producing a ratio of about 1,000:1 between the lightest and darkest parts of the image. This form of modulation is referred to as binary pulse width modulation and requires each mirror to oscillate up to thousands of times per second.

For each mirror, the intensity of the appropriate colour is controlled by a 15-bit “word” which is stored in the SRAM. When applied to the DMD each bit is loaded to all 1.3 million mirrors simultaneously and then the next bit and so on for 15 bits. The bits are loaded in bit weight order, that is the most significant bit (MSB) is loaded first and, if it is “on”, the mirror will be turned “on” for half of the available frame time (y x = sec). If the next most significant bit (MSB-1) is “on” the mirror will be “on” for the next one quarter of the frame time. If the MSB-2 is “on” the mirror will be “on” for the next one eighth of the frame time, and so on for 15 bits. Fifteen-bit resolution means that each mirror can produce 32,000 apparent intensity levels for each primary colour for each pixel on the screen. While the image is being projected the system is loading the data for the next frame, ready to download the next “word” sequentially, bit by bit, via the SRAM.

Unlike a film projector, the shutter of which cuts off the light from the screen while the next frame of film is being pulled down into the gate, the DMD reflects the image for virtually all the time and the substitution of the next image is almost instantaneous, making the picture virtually flicker-free. This also provides the most efficient use of the light, so that the projector can produce a picture that appears as bright as that from a film projector, even on large cinema screens. The simultaneous change of the entire projected image maintains the film “look” for cinema presentations.

The high-intensity of light from a powerful xenon arc is filtered by an infra-red mirror to divert heat from the DMDs. In addition, though much of the DMD consists of mirrors and is therefore not liable to overheat, it is water cooled as a safety precaution. The deflected light is reflected into a black “light-dump” and is converted into heat, which is dissipated by a cooling fan.

An alternative method of digital projection, used in prototype versions of the Kodak Digital Projector, developed by JVC, and known as digital direct drive image light amplifier (D-ILA) relies not on the tilting of mirrors but uses flat surfaces whose reflectance is electronically modulated by liquid crystal. In this case the array of pixels is 2,048 x 1,536, giving a total of over three million reflection cells.

Digital projection has many advantages. The cost of the production of hundreds of bulky film prints and their distribution to cinemas is very high. Distribution of digital data is much simpler and more efficient, offering large financial savings. Films may be distributed as specialized DVDs, as digital tapes, or on portable hard disc arrays. Alternatively, they could be transmitted to cinemas on high-speed optical fibre or satellite links. In multiplex cinemas the scheduling could be altered at short notice to change the number of screens showing the film.

The overwhelming advantage of digital projection is that every single show will be technically identical. The problems of damage to film prints caused by wear and tear and mishandling will be solved. The quality of release prints can vary and the dyes can change, causing variations in the appearance of film. With digital projection, the images shown at the local multiplex will be of the same high quality as those at the prestigious charity premieres in first-run cinemas.

However, these advantages inevitably raise issues for the film industry about data security and so encryption and other techniques to prevent “piracy” and theft are being developed.

At the beginning of the 21st century, digital projection was only available at a handful of first-run cinemas in the United Kingdom, but as the projectors enter large-scale production and the cost comes down, digital presentation is likely to expand rapidly. The relatively high cost of the projectors may be countered by extensive leasing of the equipment and will also be offset by the large reduction in the costs of handling film prints.