Chapter 1

The Physics of Light

Starting on the Right Wavelength

Light is a particular range of electromagnetic radiation that stimulates the optic receptors in the eye and makes it possible to determine the color and form of our surroundings.

Light has three properties that contribute to our perception of the things it illuminates:

Lighting directors must have a basic understanding of all three properties, from a scientific point of view, to make an artistic contribution to the productions they are lighting. In Chapter 1, we will deal with the aspects of color and quality; in Chapter 3, we will discuss the quality of light provided by various lamp and reflector types; and in Chapter 4, we will discuss aspects of intensity.

Color

We know that the light of the sun or of an electric lamp can be broken down into the colors of the rainbow. The common method for dispersing white light is by using a prism. As a young science student you may have learned the memory crutch Roy G. Biv to help you remember the colors and the order they fall in when white light passes through a prism and is projected on a white surface.

Primary Colors

Light has two components: luminance information and chrominance information. The luminance information deals with the amount of light intensity in lumens and is measured in foot-candles. Chrominance (color) information is subdivided into two factors: hue or tint, and saturation.

Hue defines color with respect to its placement within the spectral range, as shown in Figure 1.1. It is the basic color of the light. The term “tint” is often used interchangeably with hue in defining chrominance and in labeling the monitor control that adjusts that aspect of color.

Saturation is the property of light that determines the difference from white at a given hue. In other words, heavily saturated red might be described as fire-engine red. A poorly saturated red is closer to white in value and may be called pale pink. Unfortunately, most monitors just label the saturation control as color.

To understand saturation, think of color as a specific hue that gradually increases in intensity along a straight line from white at the left end to the pure color on the right. The pure color—for example, red—is said to be saturated, while its unsaturated hue is called pink. Adjusting the color control of a monitor or television affects the degree of color saturation in the scene.

Figure 1.1 shows that white light consists of at least seven distinct colors, known as the spectral colors. We know from observation that there are a great many other colors in the world around us. In the great scheme of things, these seven colors are of no particular importance except to illustrate the concept of refraction and that white light has distinct component parts—parts that can be measured. There are three primary colors, however, that are extremely important in understanding the physics of light and the transmission of color pictures by the television system. These colors, the capital letters of the name Roy G. Biv, are red, green, and blue. They are the primary colors of light. Various combinations of these three colors make it possible to reproduce all the other colors in the visible spectrum. Since this is true, we need only evaluate everything we see in terms of how much red, green, and blue light it reflects to reproduce its actual color. That is why the television camera has three pickup chips, each one reacting to the percentage of a particular primary color reflected by the subject. The display screen, a cathode ray tube (CRT), contains red, green, and blue phosphors that glow with an intensity relative to the signal generated by the corresponding pickup chips in the camera.

While the knowledge that white light could be broken down into seven colors in a particular order may have gotten us through our science test, we need to understand the color aspects of light in greater detail to understand how it affects the television camera (see Figure 1.2). Red has the longest wavelength of visible light. The farther you proceed toward the violet, or opposite, end of the spectrum, the shorter the wavelengths become. People commonly use the term “warm” to identify light in the red-orange portion of the spectrum and “cool” to describe light in the blue-violet end of it. These terms are subjective evaluations that relate to the perceived or psychological effect of these colors on the viewer. These terms should not be confused with the objective measurement of the actual spectral composition of a light source known as the “color temperature.

Color Temperature

The color temperature of a light source is determined by the wavelengths of light it emits. That is, how much red light, how much green light, how much blue light, etc. We know that as an object is heated it will first glow with a reddish color. If we continue to apply heat, it will give off a yellow light and change to blue and then violet as additional heat is applied. Because different substances emit different wavelengths of light when heated to identical temperatures due to their differing chemical compositions, a specific substance must be used to establish standards.

Such a standard is called a “blackbody.” This mythical body, or substance, is said to be composed of a material that neither emits nor reflects any light whatsoever. When it is heated to a specific temperature, it gives off a specific combination of wavelengths that are consistent and predictable. The temperature scale used is Kelvin (K), in which 273° Celsius is absolute zero. In theory, when we heat this body to a temperature of 3200°K, it will emit a certain combination of wavelengths through the yellow end of the spectrum. It is classified as “white light” because it contains sufficient wavelengths of all the colors of the visible spectrum which, when added together, form white. If we continue heating the body to a temperature of 5600°K, it will emit additional wavelengths nearer the violet end of the spectrum.

The important part of this definition is that the light emitted during the heating process progresses at a gradually increasing rate from the longer red wavelengths to the shorter wavelengths in the blue-violet end of the spectrum. This phenomenon occurs only when a tungsten filament is heated by passing a current through it. Incandescent lamps produce this gradual, predictable increase when current is applied to their filament.

For reasons both scientific and economic, tungsten is the metal of choice for manufacturers of lamp filaments. Since the melting point of tungsten is 3800°K, a working temperature of 3200°K has been chosen as the standard for tungsten lighting. At an operating temperature of 3200°K, a tungsten filament will have a relatively long life span and still produce a desirable spectrum. The tungsten lamps designed to burn at 3400°K for special photographic applications have greatly reduced life spans due to higher operating temperatures. The closer you operate a filament to its boiling point, the shorter its life.

Some tungsten-halogen lamps are rated at 5600°K, or “daylight.” Since the filament will vaporize at 3800°K, how can a color temperature of 5600°K be achieved? The answer is the application of dichroic filters. These are special optical coatings applied to the front of daylight lamps that reduce the colors complementary to blue and produce a pseudo-daylight spectrum. Most daylight sources, like the halogen metal iodine (HMI) lamps, are the result of specially designed discharge lamps that generate the higher color temperature without the need for special dichroic coatings that reduce output and lamp efficiency. No true blackbody source can produce the daylight spectrum, since no metal filament can be heated to 5600°K or higher, without melting.

Fluorescent lamps, on the other hand, do not produce light as the result of passing current through a filament. Instead, an arc of current passes through a combination of gases, excites them and causes a phosphor coating inside the lamp to glow. Such light sources do not produce a continuous spectrum and are very difficult to correct and balance with standard light sources. Sometimes you may see a correlated color temperature listed for some fluorescent lamps. Generally, such a listing will rate a cool white lamp at 4200°K and a warm white fluorescent lamp at 2900°K. Do not be misled by such ratings, however. They are not really scientific and do not provide satisfactory results when you add filters based on those temperatures to such a source. It merely means that a certain number of people have looked at this light source and agreed that it appears to the human eye to produce a blackbody color temperature that correlates with 2900°K or 4200°K. It is not actually 2900°K or 4200°K. No light sources, other than incandescent lamps, are true blackbody sources, measurable in degrees Kelvin.

Fluorescent lights, mercury vapor lamps, sodium vapor lamps, and various other multivapor discharge sources all produce a very erratic spectrum and cannot be rated in degrees Kelvin as blackbody sources can. The actual wavelengths produced depend on the composition of gases and the coatings on the interior of the lamps. The spectrum they produce does not provide a true white light containing known wavelengths from the red end of the spectrum to the violet end. Since they do not produce a true white light, they are very difficult to color-correct with filtration media (see the section later on in this chapter, “Working with Sources of Mixed Color Temperature”).

Color Rendering Index

A more scientific approach to the classification of the apparent color temperature of fluorescent and other discharge lamps is recommended by the International Commission of Illumination. That method is the color rendering index (CRI), in which eight standard pastel colors are viewed under the light source being rated and under a blackbody source of known color temperature. The color rendering index ranges from below zero to 100. A number on that scale is assigned to the rated source light based on how accurately it renders the pastel colors compared to the same swatches viewed under the blackbody source. The closer they come to matching the look of samples under the blackbody source, the higher the index number assigned to the source being tested. Cool white fluorescent lamps are given a CRI of 68. Warm white fluorescent lamps have a CRI of 56. Daylight (Daylite) fluorescent lamps have a CRI of 75. A special fluorescent lamp called the Vita-lite has a CRI of 91 and comes as close as possible to a natural or daylight source.

Light radiates from the source in waves. The length of these waves, when measured from peak to peak, varies with the color involved. As mentioned earlier, the longer wavelengths are near the red end of the spectrum. These are perceived as being warm in color. The shorter wavelengths, near the violet end of the spectrum, are perceived as being cool in color.

While the human eye is capable of adapting to a wide range of color temperatures and interpreting color correctly, the pickup chips of the television camera cannot. Television cameras are designed to produce accurate color when the scene is illuminated with light at 3200°K. Within a given range the camera circuits can compensate for slight deviation from the ideal 3200°K color temperature (see the next section, “Auto White Balance”). This color temperature is often referred to as “tungsten” light. The other general color temperature classification is “daylight.” It ranges anywhere from 5400°K to 6800°K. These color temperatures are usually found when shooting in sunlight or under specially balanced or color-corrected studio lights.

Camera Operation

Before plunging into a technical explanation of camera operation or, later on, proper setup techniques for a color monitor, let me say a word about why such topics are covered in a text about lighting.

To make valid judgments about your lighting efforts, you must be able to view the results through the system. A number of texts dealing with TV lighting state that your monitor should be properly adjusted before you can make a valid assessment of the scene. They do not, however, tell you how to adjust it properly. Understanding proper adjustment methods is important for both the independent video producer who must know some basic aspects of lighting and for lighting designers who work with monitors daily.

In Figure 1.3, we see that the light that passes through the lens is split up by the prism block by a charged coupled device (CCD) into three primary colors. Each CCD chip then produces a voltage signal that is relative to the amount of that particular color present in the image at any given location. For example, if we were shooting a primary red art card, the red chip would produce the entire signal, and the green and blue chips would produce no signal at all.

According to the National Television Systems Committee (NTSC) standards for American television, the camera should be set up to produce a 1-volt signal, from peak to peak, when it is properly adjusted. In the case of shooting the red art card mentioned earlier, that entire signal would be produced by the red chip. However, we rarely shoot a subject that contains a single primary color, so all pictures will be composed of varying voltages from each of the three CCD chips. Since white contains all the colors of the visible spectrum, we can reason that if we reproduce white accurately, we will automatically reproduce individual colors accurately. When white is reproduced on television there is a definite ratio among the three primary colors. In that ratio, red is 30% of the total signal, blue is 11%, and green is 59%.

Auto White Balance

To achieve the ratio for proper reproduction of white, and subsequently colors, the modern camera is designed with auto white balance circuits. These circuits are able to make adjustments in the output voltages of the three CCD chips so that their combined outputs form a 1-volt signal from peak to peak. The red signal will be 0.30 volts, the blue signal will be 0.11 volts, and the green signal will be 0.59 volts when the camera is aimed at a white card lit by 3200°K light. When this occurs on an indoor location and the operator pushes the auto white button, you can see the image on the monitor change from off-white to white as the three voltages are adjusted to the established ratio.

From aly technical standpoint, this voltage ratio does not directly determine the color information for the encoded signal. It represents the quantity of each of the three primary colors, but there is a far more complicated method of establishing the actual chrominance information of the NTSC signal. Since our purpose is not technical in nature, that information will not be covered here, but many excellent books on the subject are listed in the Appendix. The important concept for the lighting director to remember about color temperature is the ratio of the primary colors and how the camera is designed to establish them.

If white balance is achieved at the start of a shoot with proper voltage applied to the lamps and a voltage drop occurs later on, the camera will then produce pictures that are warmer in color. This is due to a decrease in the blue-violet spectrum that occurs, because the filaments are not as hot as they were during the white balance procedure. It is not due to an increase or abundance of red in the spectrum. The lamp filaments, operating at lower voltage, will burn cooler and produce less light in the higher end of the 3200°K spectrum, resulting in warmer colors. Note that the actual temperature at which the filament burns, based on the voltage applied, has an inverse effect on the color temperature. (For example, reduced voltage to a 3200°K lamp will cause the filament to run cooler. The cooler filament produces fewer wavelengths from the upper end of the spectrum, resulting in a warmer or more reddish tone to the scene being shot.) The actual amount of light in the red spectrum does not increase, but the reduction of wavelengths in the blue end of the spectrum causes an apparent increase in red. It is possible to calculate the color temperature by measuring the voltage at the lamps and using certain formulas.

From a practical standpoint, there is no time to do that, and there is no need to under studio lighting, since the camera is capable of auto white balance anyway. A color temperature meter may also be used to make quantitative measurements of the red, blue, and green light present from any source. In fact, use of such a meter is required when trying to determine the proper filter material to use to make color correction of discharge sources having unknown spectrums.

Black Balance

Before continuing our discussion about white balance, a brief explanation about the black balance function of the camera is suitable. The black balance is usually done before you white-balance, but some manufacturers recommend that you do it after you white-balance. Naturally, you should follow the recommended procedures for your particular camera. On professional cameras, the white balance setting and black balance setting are usually on the same toggle switch (at least on Sony models).

The reason for this seeming contradiction has to do with the design differences in the black balance circuits. More recent auto black circuits are able to modify the starting point of white balance voltages without changing their overall waveform, something earlier circuits could not do. It is generally best to cap the camera when making a black balance adjustment, although many cameras automatically do this for you. The diagrams in Figures 1.4A and 1.4B should help to illustrate the effect of and need for black balance.

In Figure 1.4A, the camera has been white balanced and the circuits are producing the proper voltage from each chip. The output of the red channel is 0.30 volts. The green channel output is 0.59 volts, and the blue channel is producing 0.11 volts. The problem is that each measurement does not start at the same baseline, resulting in “black” that is tinted.

In Figure 1.4B, the camera has been black balanced and the three voltage measurements start at the same baseline. In this case, the blacks will be a true TV black without any shift toward a color.

Selecting The Correct Camera Filter Wheel Position

As good as white balance circuits are, they have a limited correction range. To make it possible to achieve correction over a greater range of color temperature differences, cameras have filter wheels built into the optical path just behind the rear lens element (see Figure 1.3.). Most cameras have only four filter positions on a single wheel. There is usually a studio or tungsten filter, two daylight filters, and a fourth position that is not a filter but a cap to prevent any light from reaching the pickup chips. This position may be used for black balance procedure and should always be used when changing from one lens to another, removing a lens for shipment or when shipping the camera.

Since cameras are designed to operate under 3200°K lighting, the tungsten position of the filter wheel is really nothing more than a piece of optically clear glass. The two daylight positions filter out the blue-violet wavelengths that are in excess of those found in 3200°K light sources. Typically, they will be rated around 5000°K and 6500°K.

Some manufacturers assume that when the 6500°K filter is necessary, the scene will be so bright that you have to stop down to your smallest f-stop. For this reason, they combine the 6500°K setting with a 0.3 neutral density (ND) filter to reduce your depth of field somewhat. ND filters are designed to equally reduce the amount of light in the red, blue, and green spectrum. They do not affect the color temperature. A 0.3 ND filter reduces light by 1 stop. It is a nice touch, but it generally does not provide adequate control over your depth of field. For greater control, you should have a set of ND filters that can be placed in front of your lens to cut back the incoming light. The normal complement of such a filter pack is 0.3ND, 0.6ND, and 0.9ND, which reduces the light by 1, 2, and 3 stops, respectively. The filters may be used in any combination to give even greater latitude of control. They permit you to work with larger f-stops so not everything is in focus from 3 feet in front of the camera to the horizon. [Two sources for screw-on ND filters are Belden Communications, Inc. and Tiffen (see the Appendix). They also make a variety of other effects filters that will be discussed later.]

A few older studio cameras have two filter wheels, like the one illustrated in Figure 1.3. One wheel is for color correction filters only and one wheel for a complete set of ND filters and possibly some special effects filter, such as a star filter, diffusion, or low-contrast filter. Naturally, this makes it easier and faster to tailor the f-stop to your requirements. Not having filters built into your camera does not restrict you from using external filters to gain control of your f-stop setting. The end result will be no different whether the filters are internal or external. The external filter pack can be rented for a few dollars a day.

If the camera is white balanced in the studio under 3200°K lights with the tungsten (3200°K) filter in position, you will not be able to achieve white balance if you move outdoors and attempt white balance without changing the filter. The camera will not be able to bring the signals within the required ratios. That is because the color temperature of the light is probably above 5000°K and contains a great deal more blue light than the circuits are able to compensate for. It will be necessary to change the filter wheel to one of the daylight settings to remove the excess blue light before it reaches the prism block. If the color temperature is even higher, near 6500°K, as it would be under a blue sky on a sunny day, then it may be necessary to change to a second daylight filter position to remove the excess blue light present before the camera will white-balance properly.

Outdoor Color Balance

The sun’s color temperature changes throughout the day, so when shooting outdoors you should white-balance the camera often (as often as every 5 minutes when shooting early or late in the day). In that way, shot 39 made at 11:00 a.m. will match shot 40 made at 3:00 p.m. The change in color temperature is caused by the different wavelengths of each color and the filtration effect of dust particles in the atmosphere. The dust, moisture droplets, and pollution in the atmosphere act as a great filter and diffuser. These particles break up and absorb the shorter wavelengths near the blue end of the spectrum. The blue wavelengths that are diffused among these particles account for the blue color in the sky. The longer red wavelengths are affected less by the atmosphere and pass through it with greater freedom. It is this factor that accounts for the warm or yellow-orange look of the sun near sunrise and sunset. You may have noticed that sunsets always appear more vibrant (more red or orange) in color photographs than they do to the naked eye.

Our eyes are constantly “white-balancing” all we see; film and video cannot match that characteristic. At these times, when the sun is low in the sky, its rays strike the earth obliquely. At an oblique angle they pass through more of the atmosphere than they do at noon, when they are at right angles to the earth’s surface. The additional atmosphere filters out the shorter-wavelength, cooler colors, while permitting the warmer, longer wavelengths to pass through more readily. The greater the dust and moisture content of the atmosphere, the more spectacular the effect. Clear, cloudless days tend not to yield beautiful sunsets. Since the earth is constantly rotating, in effect changing the filter pack on the sun’s rays, the amount of blue light decreases and causes “warmer” pictures (see Figure 1.7).

After noon, density of the atmosphere increases with the passage of time due to the increasing distance between the sun and the subject. As the angle becomes more oblique, it travels through a greater portion of the earth’s atmosphere. To compensate for this ever-changing condition, repeated white-balancing is necessary if you wish to mask the time change.

Though great emphasis is placed on achieving perfect white balance so that whites are always white and the color in all the shots of a particular sequence will cut together during post-production, I must insert a caution about such a sterile approach. An old saying states that you have to know the rules before you are permitted to break them. It certainly applies here.

An example might involve taping an afternoon sporting event on a real-time basis using several cameras switched live. Watching the monitor, you would notice a change in the color balance of the shots as the afternoon progresses. You should be able to sense a warming in the color of the light and see shadows lengthening. That is normal and you expect these changes. It’s a natural phenomenon that shows the passage of time. The trick is to be able to recreate such an event out of sequence and maintain a convincing look indicating the passage of time once your shots are cut together.

That’s where the art of lighting comes into play. You need a feel for the scene and must decide early on whether you are going for complete realism or want an unnatural or artistic look. It would be very difficult to achieve and maintain a desired look if you did not understand how the video system reacts to light and how light interacts with scenic elements. Even interior shots may need to show time of day, and if they are of any length may require the use of some lighting techniques to show passage of time during the scene.

If you want a warmer look to all the exterior shots done throughout the day (as if you had been shooting around sundown), you can “paint” the picture with the camera controls as the day progresses to compensate for changing conditions and add the necessary red-orange to the picture. This used to require a very well-trained video engineer who had excellent color discrimination, but now almost any NLE system allows you to fine tune colors in editing. A much simpler and more accurate method is to fool the camera during setup. We said in the beginning that the assumption made by the auto white circuitry is that the subject is a white card lit with 3200°K light. When you are shooting outdoors, the daylight correction filter will compensate for the higher color temperature of the light and remove the excess blue before the electronic balancing process begins. The circuits assume that a white card is the subject and produce the required voltage for each of the primary colors.

Although vectorscopes are covered in greater detail in Chapter 2, let’s take a few minutes to look at the faceplate of a vectorscope now (see Figure 1.8). While the waveform monitor is designed to give you information about the luminance or brightness value of your scene, the vectorscope presents a graph of the color or chrominance content of the signal. You will notice that red is located at the 103° position on the display. The complement of red is cyan, and it is located 180° from red at 283°.

Remember, the camera always assumes you are shooting a white card when you white-balance. It will then produce the proper voltage ratios if the right color correction filter is in position. If you shoot a cyan-tinted card during the white balance procedure, the camera will compensate for the cyan by reducing the blue and green voltages (since cyan is composed of blue and green) to achieve the standard ratio that we have been talking about in order to produce a white field on the television screen. As a result of reducing the blue and green voltages, when the camera is focused on the scene, the picture will be warmer than usual. Since the circuits have been tricked into thinking there is more blue and green light in the spectrum than actually exists, the camera produces a red-orange field when it is aimed at a true white card after such a bogus white balance procedure. The circuits have not increased the red voltage output in this unorthodox white balance procedure, but the scene appears redder since they have reduced the complementary voltages.

Such a procedure will produce a consistent warm look for all of the shots taken throughout the day. Warning: Do not try this technique for the first time on an important shoot. Run some tests before you start your shooting schedule. Use various tinted chips to see what effect they produce. Keep accurate records of your tests and use the chip that your tests indicate will produce the desired result. This method will assure greater consistency throughout the shoot and will make painting much faster. Your test will show that the cyan chip you use for this white balance procedure should have very little saturation or the resulting flesh tones will be too red.

The same effect can be achieved under tungsten lighting conditions by white-balancing on a blue card. This fools the camera in the same way the cyan chip did in an outdoor setting.

This method will produce a technical solution to the problem of producing warmer-looking scenes, but do not forget the aesthetic considerations involving the shadows in your scenes. A colleague of mine once stated that he believes that the “blue card—balanced images” look “buttery” in that their warmed flesh tones have a more tanned, cocoa butter appearance.

Do not shoot wide shots at noon with the shadows falling directly beside their sources. The situation is bound to be noticed by your viewers. Close-ups and medium shots can be done throughout the day without drawing attention to shadows. Save your long shots for the beginning or end of the day when the shadows will fall naturally at oblique angles to the subjects, as they do around sunset.

Working with Sources of Mixed Color Temperature

Novices normally have great difficulty when they encounter mixed color temperatures on a set. These problems are at their worst when you are shooting indoors on location. A common situation is one in which you have to interview someone in an office lit by fluorescent lights. The subject is sitting in front of an undraped window, and you have a lighting kit with standard 3200°K lights. If you are lucky, you will come back to your edit room with footage of a person surrounded by a blue haze, with a strangely colored forehead, and an extremely hot background (see Figure 1.9). I say “lucky” because you may end up with footage of a talking silhouette. We will examine some solutions to this situation in a minute.

You are faced with two problems: (1) There are three different color temperatures of light involved; (2) there is excessive contrast between the subject and the background. It is very difficult to deal with if you try to light the subject well enough to overcome the intensity of the light coming through the window. To overcome these discrepancies, you must use special filter material called “color correction media” to color-balance the various light sources and ND sheets to control the excessive contrast. A variety of highly specialized materials is available to handle these types of situations (see Figure 1.10).

There are some problems to this method however. Suppose your window is several floors above the ground, how would you attach ND gel? A simple solution would be to gel the inside of the window. The gel will have to be cut much more accurately as not to be seen by the camera, but it may solve your problem (see Figure 1.11).

The other problem might be temperature, where it will be extremely difficult to attach the ND material to a building’s façade because the gaffer tape refuses to adhere itself to the structure. Other than warming the gaffer tape, use light stands and attach the tape to them instead of the buildings exterior. Use your ingenuity and you’ll come up with a solution (or avoid shooting and lighting in cold weather).

Failure to convert all of the light sources in any scene to a common temperature will cause expensive and sometimes impossible color-correction problems in the post-production phase. Never take the attitude, “I’ll fix it in post.” In many cases you cannot fix it, and when you can, it is usually very time consuming. You may end up either chewing up your time in the editing room or paying someone else to do it. Do it right the first time—when you are lighting. One hour of color correction on the NLE system will just about buy that gel material.

New Products to Solve Lighting Problems

One of the most encouraging developments in lighting during the recent years is the application of space-age technology to produce products that will solve these and other problems faced by the lighting designer. Experiments have been progressing with special coatings on the interiors of multidischarge lamps and fluorescent tubes to produce a spectrum that can be color corrected to standard photographic temperatures. Special reflectors have been developed that absorb ultraviolet and infrared radiation. Lamp and reflector designs are being perfected to fit in smaller, more efficient instruments, and lamp spectrums are becoming more compatible with CCD chips. There are now more durable and accurate filtration and reflector materials available as the result of new plastics and dyes.

Fluorescents presented a substantial problem. Though they were developed back in 1867 by Becquerel, they did not appear on the consumer scene until the 1940s. Even then they did not receive wide acceptance. One of the first wide uses was in supermarkets because of their need to light large areas evenly. Store owners complained to lamp manufacturers that fluorescents made the meat look blue, and that oranges and other produce looked bad. As a result, the tubes were redesigned to correct these problems. Had the first complaints come from a cinematographer rather than a butcher, we probably would not be cursed with the strange spectrums these lighting sticks produce today. These instruments are now commonplace on news sets and many other video shoots. We’ll go into much greater detail on fluorescents later.

Solving the Problems of Mixed Color Temperature

Let’s analyze that interview scene where the talent is seated in front of an undraped office window and the office is lit by fluorescents. We will achieve satisfactory results using some “inside” information about a variety of readily available accessories.

Since the primary rule of good lighting is that all the sources must be the same color temperature, you have to decide which temperature you want to make standard. Will you color-correct the daylight and fluorescents to match your 3200°K sources, or will you convert your studio lamps to daylight and leave the fluorescents alone, or correct them to daylight also?

You can do any of these things by using the correct color-correction media, and you must do something to bring all your light sources to the same color temperature.

If the window is behind the talent, and we will put it there for purposes of illustration, it may not be a wise decision to leave the window alone and place booster blue correction filters on your 3200°K lamps to convert their color temperature to daylight. While you can satisfy the criteria of matching the color temperature of sources this way, you will still be left with an impossible contrast problem. In all probability, if you expose for the window, the talent will end up in silhouette because your lighting kit will not be powerful enough to overcome the strong backlight. If you expose for the talent, you will “blow” the window. It will become a screaming white blob behind the talent. The excessive white level will not be handled by the white clipping circuits of your camera, and you will experience breakup on playback.

Your problem will be aggravated by the fact that you will also lose efficiency when you convert from tungsten to daylight by placing a booster blue filter in front of the source. If you require a full blue to convert 3200°K to a nominal 5500°K daylight, you will reduce your light output by 64%. The transmission of full daylight blue is 36%. If you can get by with half blue to convert 3200°K to 4100°K daylight, you will reduce your light output by only 48%. The transmission of half blue is 52%.

The easiest thing to do would be to close the drapes or pull the shade on the window, but you may want to see out the window to help establish location. You can accomplish this goal by placing a sheet of ND material on the inside of the window. This polyester-based material comes in rolls 100 feet long by 48 inches wide. It is optically clear, does not affect color temperature, and can be reused many times if properly cared for. Static cling will usually be strong enough to hold the sheet in place on a small window. Larger sheets can be taped in position.

Once you have reduced the incoming light sufficiently, you can set up your front lighting in the normal manner using booster blues in front of each studio lamp. As with the ND filters you screw on the front of your lens, ND sheets come in 0.3ND, 0.6ND, and 0.9ND to reduce the light source by 1, 2, or 3 stops.

If there are other windows in the room, off camera, their light can serve as general fill or can be bounced into dark areas of the scene with a reflector without any color correction. The natural light will cut down on the amount of color-corrected incandescent light you need to add to the scene.

Filter material commonly known as 85 converts daylight sources to 3200°K. Some 85 is combined with 0.3ND, 0.6ND, or 0.9ND filtration material to reduce daylight and change its color temperature in one operation. If you choose to correct daylight to tungsten so you can work directly with your 3200°K sources, it would be necessary to gel the off-camera windows or cover them so they would not add daylight to the scene. Obviously, it is better to be able to use the additional window light without having to color correct it.

The term “gel” is used here as a verb. Gel was the original material to which dyes were added for the purpose of coloring light sources in the theater. A solution of liquid gel, with dye added, was poured into large pans and permitted to solidify into thin sheets. It is still used as an economical alternative to modern plastic filtration material in low-wattage theatrical instruments, but it will not withstand the higher temperatures of quartz-halogen lamps. Gel is frequently used as a verb to describe the act of placing color-correction or ND material in the color frame of an instrument or on the surface of a window.

The fluorescent lights could be left alone if you use daylight as your standard. Their color temperature will blend in rather well with daylight. Or you can convert them to daylight by using a Minus-green filter. Fluorescents can also be converted to 3200°K by using a Fluorofilter filter in front of the tubes. You can place a color-correction sheet above the plastic lens or grid of the fluorescent fixture, or, if the tubes are exposed, you can purchase sleeves of filter material that slip right over the lamps. If you find yourself shooting repeatedly in the same office location, you may wish to leave your sleeves or sheets in the fixtures to save time on your next setup. A drawback to this approach is that the Fluorofilter reduces the light output by 64%. Its transmission is only 36%.

You can purchase fluorescent lamps rated at tungsten or daylight temperatures for installation in areas frequently lit by fluorescents. These lamps are more expensive than regular fluorescent tubes but they may be worth it if you shoot often in a given office area that is lit by fluorescents. If you are working on a tight budget, it may be possible to transfer the cost for these lamps to building maintenance. This would give you a threefold gain: there is no cost to your department to relamp the fixtures, you get a 3200°K source to work with, and you also save time by not having to gel the fixtures the next time you shoot in that area.

The Spectra 32 fluorescent tube from Luxor Lighting Corporation (see the Appendix) is rated at 3200°K and given an 82 on the CRI scale. The Luxor Vita-lite fluorescent is rated at 5600°K and given a 91 on the CRI scale. Such lamps are the best choice for lighting in a makeup room because they provide excellent color rendition of subtle makeup shading. Their color temperature will match the light sources on the set and they will not add heat to the makeup area. Do not use a string of household lamps around your makeup mirrors. Their color temperature makes critical evaluation impossible and their heat becomes uncomfortable as the actors work close to the mirror.

How Filters Work

The issue of using color-correction filters to correct for the strange spectrums of discharge lamps and fluorescents seems to be shrouded in an aura of mystery. While having lunch one day, two lighting designers were asked by an expert in the area of color correction, “What do you do to correct for fluorescents on location?” They looked at each other and one replied, “We start drinking early.”

Second case: The author of an excellent and very thorough text on filmmaking offers a scientific explanation of filtration formulas. In summing up things and trying to explain the reasons for selecting certain filters, he states, “Don’t try to figure it out; it’s like a game and those are the rules.”

The issue also seems to be made overly complicated. After numerous conversations with leading experts in the design and manufacture of filter materials, I find confusion of terms how to explain how filters work. It is not uncommon to talk of booster blues and imply, if not state outright, that they add blue light to the tungsten spectrum to balance it with daylight temperatures that contain more blue. That sounds fine. It even looks fine. If we view a light source that has been gelled with booster blue, there is a definitely bluer tint to the light that passes through it.

A pamphlet that is no longer in print, describing light control media, stated, “There are gels to add enough green to daylight sources to match fluorescent phosphors permitting an overall correction with a single lens filter.”

If the drinking water at your house has a funny taste, you buy a filter for your faucet and it makes the water taste better. Does it add good taste to the water that passes through it? No. It removes the chemicals that cause bad taste. That is all filters can do–remove things, whether they are objectionable chemicals in a water supply or objectionable frequencies of light. Why then do the experts talk of adding green or boosting blues? Actually, the people who make such statements are not the physicists and chemists who formulate these filters and who know such statements to be false. The problems are caused by copywriters and salespeople who try to simplify the technical aspects of their products. The result is greater confusion on the part of the users who are not better informed, but misinformed.

The Rosco Cinegel pamphlet is an excellent resource to have. It illustrates products currently available to control color temperature, and to reduce intensity and products that reflect light sources. The pamphlet is available free from Rosco (see the Appendix). It presents an overview of their products and gives practical application examples. You should also send for their swatch book that contains samples of booster blue, reflection media, light control Media, diffusion media, and daylight control media. Once you are familiar with these products, you will see that someone has already invented material to solve many of the problems you face daily in interior and exterior location lighting and in studio setups. Lack of knowledge about these and other helpful products available to the professional will cause you to come home with compromised video. There is no need to settle for unsatisfactory lighting, because the tools exist to correct the problems you face.

Think back to our efforts to fool the camera during white balance so that it would produce a warmer look. We did not add red to the picture. We balanced on a cyan chip. In effect, we were removing a color complementary to the red, namely cyan, to create the appearance of more red in the picture. The same is true of color-correction filters. They cannot add green as the text of the pamphlet stated, but they can remove magenta, which is the complement of green. That will produce an apparent increase of green in the light that passes through them.

Look again at the faceplate of the vectorscope in Figure 1.8. You will see that green is at 241° on the scale. Its complement, magenta, is 180° opposite that point at 61°. When we remove the magenta, the light appears greener because its complement is lacking. Light that passes through ND material is dimmer, not because the filter adds black but because it subtracts equal amounts of the red, blue, and green light. Filters can shift emphasis by removing complements of problem frequencies when they are present. They can also remove the offending frequency directly, if it is present in the spectrum at levels in excess of those needed to conform to the desired color temperature.

When a source, like the sodium vapor lamp, produces a wildly erratic output with many holes in the normally continuous spectrum of a blackbody source, filters are not able to compensate for the missing components. If you filter out the offending frequency, you create even larger holes in the spectrum and generate new problems for yourself. Since frequencies that complement the offending spikes do not exist, you cannot filter them out to shift the apparent output of the source. In short, completely effective color correction is a physical impossibility. Some color correction is possible, but do not consider yourself a failure when the colors are not perfect under such adverse lighting sources. Keep in mind that even the human eye does not perceive colors accurately under such lights and the viewer does not expect to see natural colors in such locations (see Table 1.1).

Quality of Light

It is possible to measure the color temperature and the intensity of light with the proper meters. I will discuss the aspect of intensity later, but for now I would like to deal with the third characteristic of light which plays a very important role in the aesthetic look or “feel” of the scene. That characteristic is the quality of the light. Quality can be judged by the density and sharpness of cast shadows. Granted it is a subjective evaluation, but on one end of our scale we have harsh specular sources that cast dense shadows with sharp, well-defined edges. The highlights are specular and the contrast range is high. On the other end of the scale we have soft diffuse sources that cast transparent shadows with poorly defined out-of-focus edges. The highlights are softened and the overall contrast range is lower.

The quality of the light accounts for the different look of two outdoor scenes. One take is shot on a heavily overcast day. The other is shot at the same location on a bright, sunny day. Despite proper exposure and color balance, the two scenes will look very different because of the quality of the light at the time of shooting.

As a lighting designer, you should not only be concerned with putting the right amount of light in the right places, but you should also evaluate and control the quality of the light you use. Fortunately, there is a battery of filtration products available to allow you to do just that. There are accessories to precisely control the quality of studio light and to change the basic quality of exterior light. These diffusion materials will be discussed later in the text.

The majority of lamps used today generate light when an electrical current is passed through a thin tungsten filament. The filament heats up as a result of its resistance to the flow of electrons. The heated filament emits photons of light that radiate in all directions from many points along its surface. One of light’s unique properties is that it does not require a medium in which to travel from place to place. Sound and heat cannot travel through a vacuum, but light can. Sound and heat are slowed down by travel through liquids, but light is not.

The harshest form of light is known as “point source illumination,” where all of the energy emanates from a single point and travels outward in a series of straight lines like spokes radiating from the hub of a wheel (see Figure 1.12). These straight rays cast crisp, well-defined shadows. There are no criss-crossed rays to soften the edge of cast shadows. This is the type of light present on a bright, cloudless day, like the one described in the previous example. The only artificial light sources that produce near-point-source illumination are the carbon arc and the HMI lamp.

Since the majority of lighting instruments use some configuration of a tungsten filament enclosed in a glass or quartz envelope, we will take a look at what happens to the light rays that are produced by such multipoint filaments. If we pick only one point along the filament to illustrate the properties of light, we can study the effects that lamp construction have on the visible spectrum (see Figure 1.13).

The ray that passes most directly from the filament to the atmosphere outside the envelope is an example of efficient transmission.

In the process of that transmission the ray is refracted, or bent, as it passes through the envelope to the outside air. Light will be bent whenever it travels from one medium to another of a different density. The same behavior can be observed in Figure 1.1, which shows light passing through a prism.

The next ray we observe does not pass directly through the envelope. It is reflected (3) from the interior surface one or more times before transmission occurs.

A fourth ray does not escape the envelope at all. It is an example of absorption (4). The absorbed ray is converted to heat energy by the envelope. The four characteristics of light shown in this figure are emanating from a single point along the filament. The fact that these behaviors are occurring at an infinite number of points along the filament in a random sequence produces a quality of light that is much less specular than that produced by single point sources. The random aspect of the transmitted rays can be altered and reshaped by the lamp housing, internal reflector and lens system of the instrument.

Lamp Output

While the electric light has been a great boon, a modern-day consumer advocate might argue that it is a device promoted and sold by utility companies to take unfair advantage of the public, to boost corporate profits and to bilk people out of their hard-earned money. How could such claims be backed up?

If you wade through a course in physics, you come up with a formula that equates a watt of electrical energy with 673 lumens of light. Therefore, it is reasonable to assume that a regular 100-watt light bulb would provide your living room with 67,300 lumens of light. What a deal that would be. The fact is, the box that my 100-watt lamps come in states that the initial light output of each lamp is 1535 lumens. That is the initial output, and it gets worse as time goes on. Not very efficient. This means that only about 2.25% of the electrical input is converted to visible light. The other 97.75% is converted to infrared radiation (heat) and invisible ultraviolet radiation. If you think that is bad, a lowly 6-watt bulb puts out only about 6.5 lumens per watt. The picture does get somewhat brighter as lamp wattages increase. A 1500-watt lamp produces about 25 lumens per watt. Still, it is nothing to write home about. As for studio lamps, they are also incandescent sources and bound by the same set of inefficient circumstances that plague the ordinary household lamp. They are no more efficient than household lamps. In Chapter 3, we will investigate some other, more efficient methods of generating light. It is generally accepted that tungsten sources produce 20% light and 80% heat.

Invisible Radiation

You have just read that some rays of light are converted into heat as a result of being absorbed by the envelope. If you look at Figure 1.2, you will notice that on either end of the visible spectrum there are forms of invisible radiation. At the top is infrared radiation and at the bottom is ultraviolet radiation.

Ultraviolet waves are extremely short and, fortunately, most of them are filtered out by the dust and moisture particles in our atmosphere. Like the gamma rays and X-rays that are shorter still, they are harmful to human tissue–they can cause skin cancer and other problems. They also have a negative effect on photographic film. Even though the human eye cannot detect ultraviolet waves, they have an effect on the eye and on the emulsion of photographic film. In the human eye, these waves can cause cataracts. In film and television cameras, they cause less accurate color rendition. Though the quality of visible light generated by HMI lamps is very pleasing on the set, you should know that it also produces a high volume of ultraviolet radiation that could be harmful after long hours of exposure, much like being out in the sun too long.

Infrared light is the longer of the two wavelengths, and far more of these rays penetrate the atmosphere because of their wavelength. They are not easily broken up by the particulate matter in the atmosphere. We cannot see them, but we feel them in the form of heat. In fact, as some of the incoming ultraviolet rays pass through the upper atmosphere, their short wavelengths bounce around the particles and are converted to heat energy or infrared radiation. Like ultraviolet radiation, infrared waves also affect photographic film and television cameras. In some cases, the effect is positive, such as infrared photography or night scopes, which allow us to see in the dark. Because of their long wavelengths, they can penetrate clouds and many other substances on earth, much as radio waves do. They are just slightly shorter than our lowest radio frequencies. While they do not cause harm to the human body, they do cause discomfort in the form of excessive heat.

HMI lamps produce very little infrared radiation, so the set stays cool.

Instruments like the Mini-cool, which is designed to use an MR-16 lamp, filter out the harmful ultraviolet radiation and render truer colors on film and tape, as well as reduce the potential illhealth effects of ultraviolet exposure. They do generate a lot of infrared radiation, but use a hightech reflector design that permits heat to pass through it and escape out the back of the instrument rather than be projected forward with the beam of light. The end result is a beam of light that is not hot. Though the heat is projected away from the actor, do not let the name “cool” light fool you.

The instruments themselves get very hot and should be handled with the same insulated gloves used to adjust or focus any other lighting instrument that has been operating for even a short period of time (see Chapter 11).

Falloff

Another property of light essential to understand is “falloff.” Most of us know the catch phrase, “light falls off at a rate equal to the square of the distance,” but we store it in that section of the brain reserved for useless information acquired in high school physics. It is not useless, however. It is extremely important to the proper placement of instruments and the selection of accessories. Perhaps an understanding can be gained most easily by looking at a couple of examples.

Example 1:You are shooting outdoors, using only natural light, with the sun located above and behind the camera. Your meter reading of the actor standing closest to the lens reads 5000 foot-candles. You move to a second actor standing 25 feet behind the first and take a second reading. It is 5000 foot-candles. You walk to a building 25 feet behind the second actor and take your third reading. Again, 5000 foot-candles (see Figure 1.14).

Example 2:You are shooting in a theater, using a 125-amp carbon arc spotlight to light the scene. It is in the projection booth above and behind the camera. You take a meter reading of the actor who is standing closest to the lens. It is 750 foot-candles. You move to a second actor standing 25 feet behind the first and take another reading. It is 490 foot-candles. You walk to the backdrop 25 feet behind the second actor and take your third reading. It is 330 foot-candles (see Figure 1.15).

In example 1 (Figure 1.14), there was no difference between the readings, though there was a 50-foot difference between the point of the first meter reading and that of the third. In example 2 (Figure 1.15), there was a 260 foot-candle difference between the first reading and the second, 25 feet away, and there was a 160 foot-candle difference between the second and third readings, again 25 feet apart. This was a total difference of 420 foot-candles over the 50-foot distance. Less than half as much light is reaching the backdrop as reaches the first actor. Why the difference? What principles are involved? And how should they guide us in placing instruments?

While math is not my strongest suit, it is necessary to look at a formula here for computing falloff: F = 1/d². It states that light falls off at a rate inversely proportional to the square of its distance.

Whenever I resort to anything as distasteful as mathematics, I do so to clarify a concept that must be fully understood to avoid lighting attempts that are doomed from the start. The test of the following information will not be whether or not you can compute falloff using the formula, but whether you position instruments in the most effective spots on troublesome locations.

For example, in Figure 1.16, if we place a light 4 feet from the head of a subject and square that distance, we will get a value of 16. This represents a given amount of light. The distance from that same instrument to the subject’s shoes is 8 feet. When we square that distance, the value is 64. Using the resulting fraction, 16/64, we see that only 25% of the light that reaches the subject’s head will reach the shoes. That is a 4 to 1 exposure ratio, hardly even lighting.

Using the same formula and placing the instrument 12 feet from the subject’s head, we find that 66% of the light that reaches the head will reach the shoes. That is a 41% improvement over the first case. Moving the instrument back 16 feet from the subject’s head means 75% of the light reaching the subject’s head would reach the shoes—an overall improvement of 50% in the falloff of light between the subject’s head and feet.

The example shown represents ideal mathematical structures. In actual practice, the results are not that precise, owing to design considerations of the light source and its lens systems. However, the lesson to be learned is that you should place instruments as far from subjects as possible to eliminate severe falloff problems.

You should understand that by moving the source farther from the subject you will be decreasing the overall light level by a factor that is inversely proportional to the square of the distance. However, that reduced level will be more evenly distributed over the subject. In Figure 1.13, there was no difference in the meter readings because the sun is so far away that falloff is not measurable over reasonable distance changes. If you were to take a meter reading on the top of a very tall building and at ground level, there would be no measurable difference between the two readings because of the extreme distance between the sun and the relatively close distances between the two readings on earth.

When working in homes and offices with low ceilings, there is a tendency to place the camera and lights close to the subjects to lessen the lighted area and reduce power requirements. This becomes problematical when there is any movement by the subjects. If they stand, their heads suddenly seem to fluoresce while their lower bodies are underexposed. Raising the instrument as high as possible and moving it back as far as possible will help even out the great differences in intensity between seated and standing subjects.

Once you have done what you can to let physics reduce the effects of falloff by moving the light as far from the subject as possible, you must turn to accessories to complete the elimination of the problem. There are three different things you can do to even out the light on a subject that moves closer to, or farther from, the light source during a scene.

The simplest solution, if your instrument is equipped with an accessory shoe, is to place a graduated scrim in it and turn the scrim so it reduces the light at the top of the beam. Graduated scrims are invaluable tools and come in many degrees and configurations. You can achieve the same effect using an ordinary window screen in a gel frame holder, and the shape of the reduced intensity area can be tailored to specific situations. Another quick fix is to place a section of ND filter in the top portion of a gel frame and insert it in the accessory shoe.

The third solution involves a little more hardware, but the philosophy is the same: Reduce the light output in the upper portion of the lighted area. If you are using an instrument without an accessory shoe, you can mount scrims, nets or ND material on a separate stand and place them in front of the source. This method actually offers the greatest control. The greater the distance between the source and the diffusion material, the more pronounced the transition will be. Examples of this method can be seen in Figure 1.17 and Figure 1.18.

A graduated scrim can be turned to reduce the lower half of the light beam from a key light to keep excess foot-candles off a white suit coat and still provide enough light on the face of a dark-skinned subject.

The rate of falloff differs from instrument to instrument depending on the make and type. Falloff from softlights and floods is greater than falloff from spots because of the diffuse nature of their rays, which dissipate more rapidly. Specular sources fall off less rapidly because their rays are parallel.

The hard part is over. If you have understood all of the information in this first chapter, the remainder of the text will be as easy as burning your fingers on a barn door.