3
Since the end of the nineteenth century, people have been able to watch moving pictures through the medium of photographic film. It’s a medium that has changed surprisingly little during the past century, particularly compared to the progress of video and similar technologies, but film is still widely considered to produce the highest-quality images in the motion picture industry.
3.1 Film Image Formation
A length of photographic film consists of light-sensitive crystals bound together in a transparent material. There are other aspects to most common film formats, such as “sprockets,” “dye layers,” and “key numbers” (each of which will be covered later in this chapter), but the crystals and binding agents are the only requirements for image formation.
Film consists of silver halide crystals that are sensitive to light. When light falls on the crystals, they become chemically activated. At this point, there is no visible change to the crystals. However, the chemical changes mean that it is possible to use further chemical processes to separate the crystals that have been exposed to light from those that have not.
With regards to imaging, this process works when the light reflected from a scene is focused on a piece of film using an optical system (i.e., a combination of lenses, mirrors, and so on). There will be more light in brighter parts of the scene, and so the equivalent bright areas imaged on the film will activate the crystals at those points quicker than those that are less bright. If the combination of the brightness of the scene and the length of time the light from the scene is in contact with the film is balanced correctly (or, to use photography terminology, if the film is correctly “exposed”), a “latent” image is formed on the film. This latent image consists of groups of film crystals that have been chemically activated and is therefore invisible. If you were to take the piece of exposed film and examine it carefully, even with the aid of a microscope, you’d be unable to find any evidence of any of the features in the scene that was photographed. But the paradox is that you couldn’t put exposed film under a microscope without exposing it further. Even after the latent image has formed, any additional light that hits it, even a barely perceptible amount, will still affect the film. This means that the film must be kept in complete darkness before and after the desired image is exposed. To make the latent image visible (not to mention permanent), the film must undergo a development process.
Development involves putting the film through a number of different chemical baths, which contain the developing agents necessary to convert the activated crystals to metallic silver and to remove them from the film. Because the crystals are so small, a detailed image with a large tonal range remains as the negative image of the original scene.
This “negative” can be used as the basis for creating a print, by acting as a filter between a light source and another sensitive material. To create an image on paper, for example, an enlarger is used to project the negative image onto photographic paper, which in turn undergoes a chemical development process, resulting in an image that is the negative of the negative. Alternatively, the image can be projected onto another piece of film, which is developed to form a positive image that can be projected onto a screen.
Figure 3–1 Photographic image formation involves many stages before the image is even visible
The film itself usually consists of individual “reels” or lengths of celluloid several hundred feet long, which have “sprockets” running along one or both of the edges. The sprockets (or perforations) are holes in the film that are used to hold the film in place in the camera (or other equipment) and advance it a frame at a time.
The most significant aspect of all film material is that it is highly standardized, in terms of the size and positioning of the various elements (i.e., the sprockets and the actual frame locations), and these standards have changed very little during the entire history of the motion picture industry. Film created for theatrical release decades ago can be easily viewed even in today’s modern cinemas.
Figure 3–2 A positive film image is produced by developing a print of a negative
Figure 3–3 A reel of film (left) and a closeup of an undeveloped negative, with sprockets running down the sides
Many of the properties of a film image are determined by the composition of the film material, or “stock.” The film stock affects the level of detail in each image, the way that colors are recorded, as well as how sensitive the film is to light. Different stocks are created for different situations. For example, certain film stocks may be optimized for shooting subjects lit by tungsten light (i.e, they are “tungsten balanced”) and others for daylight scenes (“daylight balanced”). Also, different film manufacturers create film materials using proprietary formulas, and so the various stocks produced by each manufacturer may be considered different from each other.
The differences between stocks are due to differences in a number of properties of the film material, in particular differences in the size, shape, and distribution of the silver halide crystals, which in turn determine many aspects of the material.
3.2.1 Tonal Range
The tonal range of the image is determined by the build-up (or density) of opaque crystals in the film during development. Areas of higher density block out more light, and the difference between the opacity of the areas with highest density possible (or “D-max”) and the completely transparent (or “D-min”) indicates the tonal range of the image.1
Photographic film has a nonlinear response to light (rather like human vision), meaning that doubling the amount of light in a scene will not necessarily result in an image that is twice as bright. In fact, most types of film have a “characteristic curve,” a graph that can be plotted to show the film’s response to an increase of light. The curve follows an S-shape, with the exact shape of the curve determining in part how the image looks. Much of the somewhat ethereal quality of film images is due to the characteristic curve of the film.
The nature of the curve means that the mid-tones of the image are compressed (i.e., less detailed in terms of luminance differences), while the shadow and highlight regions have a much higher level of detail than linear formats. The nature of the curve also means that film tends to have a much greater “exposure latitude” (which is the overall luminance range that can be recorded for a scene) than other formats. This extended latitude grants a lot more flexibility during color grading, because scenes can be adjusted to increase or decrease the effective exposure within the recorded image.
Figure 3–4 Graph showing a linear response to light
Figure 3–5 Graph showing the characteristic curve of a typical film material’s response to light
Figure 3–6 The characteristic curve of film determines in part how the resulting image will look. © 2005 Andrew Francis
Figure 3–7 With a different characteristic curve, a different image is produced
Different stocks will have different maximum and minimum density levels and different characteristic curves, reflecting differences in the way that tones are reproduced.
3.2.2 Film Speed
Different film stocks are given a speed rating. This is a numerical value that can be used in conjunction with a number of other factors to determine the optimum exposure of the film material. In broad terms, the speed of a film is a measure of its sensitivity to light. A “faster” film (i.e., one with a higher speed value) requires less light (either in terms of the intensity of the light, or the length of time the film is exposed to the light) to provide a satisfactory image and is typically comprised of larger crystals. A number of external factors can also affect the speed of a piece of film, such as the type of light in the scene and the development process.
Measuring the speed of a material in a useful way is a complex issue, and several different methods attempt to do this. The most widely accepted speed rating system is the ISO system. Each film stock is assigned an ISO speed rating, and this value can be used in conjunction with a measurement of the amount of light in the scene (typically determined by using a handheld light meter) and the camera settings (notably the aperture and shutter angle) to achieve a correctly exposed image. Even so, the reported speed rating of a film stock is not applicable to every lighting condition, and so the cinematographer often has to adapt (or ignore) it for each situation.2
3.2.3 Graininess and Granularity
Although each silver halide crystal is microscopic in size, film images tend to have visible patterns of “graininess,” made up of randomly sized and shaped grains. This grain pattern is caused by individual silver halide crystals forming “clumps” (or by appearing to do so because of the way they are distributed within the material). The perceived size of the clumps (or the level of graininess of the image) is completely subjective, determined by the film stock and the visual contents of the image. Sharper images tend to be less grainy than blurred images (or those with fewer small details). It is also worth noting that enlarging images tends to make the grain more pronounced.
Similar to this is granularity, which is a more objective measure of the crystal structure in a photographic image. The granularity is determined by measuring the fluctuations of density in a developed image of a uniform tone (such as a gray card). The amount of granularity depends upon the film stock, the development process, and the amount of light exposure.
Figure 3–8 Close up, the grain patterns can be seen, particularly in regions of solid color
Figure 3–9 Images with lower granularity show less variance in density
Figure 3–10 Images with higher granularity show a greater degree of variance in density
3.2.4 Color Reproduction
As with many imaging technologies, film can record images in color by combining several monochromatic images. Modern film achieves this by combining multiple “dye layers.” Each layer is sensitive to a different part of the spectrum and only records light of a particular color. Typical color film has a red-sensitive layer, a green-sensitive layer, and a blue-sensitive layer, so that white light will affect all layers. During development, the layers are dyed so that each layer filters light through in the appropriate color. For negative images, this means the dyes are cyan, magenta, and yellow, and for prints, they are red, green, and blue. Thus, projecting white, unfiltered light through a print creates an image that is an accurate, full-color reconstruction of the original scene.
Different film stocks may reproduce colors in slightly different ways and may be aimed toward capturing color accurately or more aesthetically.
3.3 Film Formats
Film, and in particular, motion picture film, is rigorously standardized. This means that every aspect of the film image is strictly defined in terms of its size and position. Other, nonimage elements are also standardized, such as the positioning and size of the sprockets. This standardization is irrespective of the stock type or manufacturer, which is why footage can be shot on any particular film stock and still be projected.
There are several different formats though, each existing for different applications, each with their own standards. The formats differ in two main ways: the gauge of the film material and the aspect ratio of the image.
3.3.1 Gauge
A larger piece of film can record a larger image. This in turn determines the maximum level of detail that can be recorded within an image. All things equal, a larger piece of film (i.e., a larger gauge) implies the ability to record an image of higher quality. It does not imply, however, the capability to record more of the scene but simply to better resolve small features and details. When the images are projected onto the same screen, the images shot on larger gauge film will appear sharper and the apparent graininess will be much lower.
Gauge is measured as the width of the film material. The most common gauges are 16mm, 35mm (which is also a commonly used format for still photography), and 70mm. Of these, 35mm is the most popular for motion picture production, striking a suitable compromise between manageability and quality when projected onto the average cinema screen. Films with 70mm formats are used for productions intended for much larger screens, while 16mm film tends to be used by productions on a lower budget (or those that are not destined for cinema projection).
3.3.2 Aspect Ratio and Framing
The shape of an image on a piece of film is determined by its “aspect ratio,” the ratio of the width of the image to its height. Different film formats have different aspect ratios, and when certain aspect ratios are combined with certain gauges, these characteristics determine the exact size of the image. For example, the “full aperture” 35mm film format has an aspect ratio of 4:3 (or 1.33). It spans the usable width of the film (with some allowance given to the sprockets), of approximately 25mm. Therefore, the height can be calculated as roughly 19mm.3
In practice, the entire width of the gauge is rarely used. This is because part of the film area is used for optical sound tracks and because cinematographers (and audiences) prefer images that have rather wide aspect ratios (such as 1.85), probably because such images give a greater sense of psychological immersion when viewed.
This issue is compounded somewhat by the mechanics of the projector and the film camera. Rather than maximize the available film area when recording images so that each frame is placed directly next to another, every frame must occupy a fixed number of sprockets, depending upon the gauge. For example, 35mm film formats almost always use four sprockets (and are referred to as “4-perf” formats). With the 4-perf format, much of the film area is wasted, particularly with wide images. However, there are very good reasons for using this format. Most importantly, it allows a single projector to be used for projecting films of different aspect ratio; the only difference is that a mask (or “gate”) has to be used to block out light that is outside of the image area. The same is also true for film cameras because each frame is advanced by a number of sprockets. This ensures that the different 4-perf formats, for example, are compatible to some degree.
All of this means that the different film formats depend on the gauge of the material but also on the shape and positioning of the images on the film. Fortunately, because the parameters for each format are rigidly defined and maintained throughout the motion picture industry, problems with projection are rarely encountered.
Figure 3–11 A full aperture framing
Figure 3–12 An academy aperture framing
Figure 3–13 Super 35 with center framing
Special consideration should also be given to anamorphic (or CinemaScope) formats, which use special lenses to compress very wide images (typically with ratios greater than 2:1) onto much narrower film areas. In practice, anamorphically squeezed film formats may be handled in the same way as formats that are not squeezed, although the squeezed formats will appear distorted when viewed under normal projection (or on such devices as lightboxes). To view the images correctly, they must be unsqueezed, using projectors equipped with suitable lenses. Note that most cinema projectors are able to rotate between two or more lenses for this purpose, without requiring additional screening rooms.4
Figure 3–14 A cinemascope image as it appears on film, and as it is displayed when unsqueezed
The quality of a film image is difficult to quantify, especially compared to digital and video formats. Certainly the definition (or resolution) of a particular stock can be determined, such as by photographing a high-contrast edge and then measuring the recorded thickness of the edge. The same can be said for color reproduction, where test patterns can be filmed and then measured from the developed material later.
In more pragmatic terms, film quality is largely a subjective issue. Film images may appear sharper than they actually are because edges, particularly high-contrast edges, are exaggerated somewhat. In some cases, color reproduction may not be accurate, but many audience members (and a great number of cinematographers) prefer the look of film images to those that are more realistic. Similarly, graininess might be one of the more important factors in determining quality, but it is a completely subjective one. Granularity can be measured, but it tends to be inconsistent, with different amounts of granularity seen at different levels of brightness. Grain size—and to some degree, the resolution of film—cannot be directly correlated to video or digital images because the grains are not uniformly distributed or shaped like video and digital images. This becomes even more significant when you consider that film is a moving image. Because the grain position and size changes every frame, it’s likely that even in watching footage of a still scene, the grain differences are averaged out perceptually, and so a moving image sequence may be subjectively higher quality than a single still image taken from the sequence. Only one thing is certain: each successive generation (i.e., copy) of a film image is of lower quality than the preceding one. (Chapter 5 examines methods of analyzing film images.) Most likely, the most significant issues of film quality concern processing or exposure problems, copying, and the accumulation of physical damage, such as dust, scratches, and mold.
As with most moving picture media, film imparts the illusion of motion by displaying still images in rapid succession. In most cases, this rate is taken to be 24 frames per second (although some areas in Europe use a rate of 25 frames per second to maintain easy conversion between film and PAL video formats).5 For this reason, film footage is usually photographed at a rate of 24 frames per second.
In some instances, different frame rates are used. For example, time lapse and other fast-motion effects can be created by photographing a scene at a much slower frame rate. If, for example, a scene is photographed once per second, then the scene will appear 24 times faster when viewed. Alternatively, slow-motion effects can be produced by recording at faster frame rates, a technique that is used often for special effects or heightened drama. (There are other methods for changing the apparent speed of a shot, several of which are covered later in this book.) In terms of producing slow-motion footage at least, film remains the best option for capturing a scene.
3.5 Key Numbers
A typical feature film production will generate many miles of film footage that must be sorted through prior to editing. After editing, the final cut must be matched back to the original negative. This requires careful indexing and meticulous comparison of every frame of every strip of film. Fortunately, an efficient method of doing this has evolved—the use of “key numbers.”
Key numbers (or keycodes) are serial numbers printed along the edge of every strip of film. The idea is that every single frame of motion picture film ever produced has a unique number to identify it.
The numbers are necessarily long and can also be used to identify the manufacturer of the film material, as well as the stock type and batch number. Key numbers on most modern films are also machine readable, which is useful when film is telecined to video tape for editing purposes and when converting film footage to digital images in digital intermediate environments. (More information about the layout of key numbers can be found in the Appendix.)
3.6 Film Media Problems
Film media can produce images that are superior to many other imaging methods, and the infrastructure for film production is so highly controlled that it is easy to generate footage that can be viewed across the world on equipment that may be many years old. However, this comes at a cost. Film is notoriously expensive to work with, partly because of manufacturing costs and also because of all the support that is required to utilize it successfully.
Film is at constant risk of contamination by light. Unlike a video tape, which can be protected from accidental re-recording, film must be kept in absolute darkness from the moment it is manufactured to the point where the image is exposed. After the image is exposed, it must be protected from any additional light until it has been processed. Furthermore, film is physically fragile, susceptible to dust and scratches, and any number of factors that can damage it. All of this means it must be handled carefully, which of course, also costs money.
There is no degree of interactivity with film image formation, unlike with video formats, where the image can be viewed as it is recorded, allowing changes to be made, and the results seen instantly. With film, the actual results are not available until after a lengthy processing phase, which at best takes a number of hours to complete. The shooting process therefore requires a great deal of estimation on the part of the crew. It is possible to use a “video tap” to view a video representation of the image through the camera, but this method provides an incomplete picture, particularly because film’s generally predictable response to light breaks down at extreme levels of illumination (either very bright or very dark scenes).6
Finally, film quality decreases steadily with every copy made. There are many reasons for this degradation; it is a function of all analog media and is also a function of the accuracy of the chemical process. Film images cannot be adjusted as flexibly as video or digital images can. It is possible to change the color and shape of film images using a number of optical and chemical processes, but these processes are neither as versatile nor as accurate as other methods. In addition, they usually depend on the creation of one or more generations of the film.
3.7 Summary
Film produces higher-quality images than any other medium currently being used. The images are formed by groups of light-sensitive crystals, which determine the appearance of the final image, in terms of detail, color reproduction, and granularity. Film can record images across a great range of frame rates, which is useful for slow-motion and fast-motion effects, and generally produces better results than other formats.
Film must go through a lengthy development and processing phase before the recorded image is visible, and it must be protected from physical damage and accidental exposure to light. It may undergo additional color processing during development, but these methods tend to lack accuracy and may degrade the image quality.
Although different film stocks produce different results in the way a scene appears and in the length of time required to make an exposure, very strict definitions govern the placement and size of the image on the film material. Such strictures have made it possible for modern cinemas to present films made decades ago.
Film has the potential to capture and display moving pictures with unrivaled quality. It is limited in terms of editing and copying, which are processes that are easily accomplished on both video and digital images. (Digital images are especially easy to edit and copy.) For these reasons, film productions, with material shot on film and ultimately projected on film, benefit immensely from a digital intermediate process, which provides additional advantages to the film-production process. In the long term, the advantages of digitalimaging technology may overtake even those of film in terms of quality, speed, and expense.
1 The minimum density will never be completely zero because the base material inevitably provides some degree of opacity. In addition, there may be some slight light exposure to seemingly unexposed regions. For this reason, D-min is also referred to as “base + fog.”
2 Manufacturers may offer an “exposure index” rating in addition to (or instead of) an ISO rating. This rating is the effective speed rating under the most probable lighting and development conditions.
3 The exact dimensions are 24.89mm horizontally and 18.67mm vertically.
4 An interesting characteristic of shooting anamorphic images is that light reflections on the lens, such as flare, tend to be seen as large streaks of light when projected.
5 During projection in a cinema, each frame is actually presented twice in rapid succession, which reduces the visible flicker caused by the projector switching frames.
6 This condition is termed “reciprocity law failure.”