Photographic light sources
INTRODUCTION
Photographs are taken by the agency of light travelling from the subject to the photoplane in the camera. This light usually originates at a source outside the picture area and is reflected by the subject. Light comes from both natural and artificial sources. Natural sources include the sun, clear sky and clouds. Artificial sources are classified by the method used to produce the light, including burning, heating, electric sparks, arcs or discharges and luminescence. Light sources differ in many ways, and the selection of a suitable source is based on a number of significant characteristics.
CHARACTERISTICS OF LIGHT SOURCES
Spectral quality
As discussed in Chapter 2, light is a specific region of the electromagnetic spectrum and is a form of radiant energy. The radiation from most sources is a mixture of light of various wavelengths. The hue (see Chapter 5 for CIE definition of hue) of the light from a source, or its spectral quality, may vary depending on the distribution of energy at each wavelength in its spectrum. Most of the sources used for photography emit what is usually termed white light. This is a vague term, describing light that is not visibly deficient in any particular band of wavelengths, but not implying any very definite colour quality. Most white-light sources vary considerably among themselves and from daylight. Because of the perceptual phenomenon of colour constancy (see Chapter 5), these differences matter little in everyday life, but they can be very important in photography, especially when using colour materials or where there is ‘mixed’ lighting. Light quality is described in precise terms. Colour quality may be defined in terms of the spectral power distribution (SPD), which is a plot of power versus wavelength throughout the spectrum. There are several ways this can be expressed, with varying degrees of precision. Each method has its own advantages, but not all methods are applicable to every light source.
Spectral power distribution curve
Using a spectroradiometer, the spectral power distribution of light energy can be measured and displayed as the SPD curve. Absolute SPD is given in absolute radiometric units (see Chapter 2), whereas relative SPD is the normalized power with respect to the power at λ = 560 nm. Curves of this type are shown in Figure 3.1 for various sources. Such data clearly show small differences between various forms of light. For example, the light sources in Figure 3.1 seen separately would, owing to colour constancy effects, probably be described as ‘white’, yet the curves are different. Light from a clear blue sky has a high blue content, while light from a tungsten lamp has a high red content. Although not obvious to the eye, such differences can be clearly shown by colour reversal film, which is balanced for a particular form of lighting.
Three general types of spectrum are emitted by light sources. Some have continuous spectra, with energy present at all wavelengths in the region measured. Many sources, including all incandescent-filament lamps, have this type of spectrum. Other sources emit energy in a few narrow spectral regions. At these wavelengths the energy is high, but elsewhere it is almost nil. This second type is called a discontinuous or line spectrum, given by low-pressure discharge lamps such as sodium- and mercury-vapour lamps.
Figure 3.1 Spectral power distribution curves of light sources. Typical SPD curve shapes of photographic light sources. P, relative power output. (a) Continuous sources. B, blue sky; D, mean noon daylight; T, incandescent tungsten filament lamp. Discharge sources. (b) Pulsed xenon lamp. (c) Xenon flash tube. (d) Fluorescent lamp.
A third type has broad bands of energy with a continuous background spectrum or continuum of varying magnitude, and is given by discharge sources by increasing the internal pressure of the discharge tube, e.g. a high-pressure mercury-vapour lamp. Alternatively, the inside of a discharge tube may be coated with phosphors which fluoresce, i.e. emit light at longer wavelengths than the spectral lines that stimulate them. Another method is to include gases in the tubes such as xenon or argon, and metal halide vapour.
Colour temperature
For photographic purposes a simple method of quantifying the light quality of an incandescent source is by means of its colour temperature. This is defined in terms of what is called a Planckian radiator, a full radiator or simply a black-body radiator. As described in Chapter 2, this is a source emitting radiation whose SPD depends only on its temperature and not on the material or nature of the source (Figure 3.2).
The colour temperature (CT) of a light source is the temperature of a full radiator that would emit radiation of substantially the same spectral distribution in the visible region as the radiation from the light source. Colour temperatures are measured on the thermodynamic or Kelvin scale, which has a unit of temperature interval identical to that of the Celsius scale, but with its zero at – 273.15°C (absolute zero).
Luminous sources of low colour temperature have an SPD relatively rich in red radiation. With progression up the colour scale the emission of energy is more balanced and the light becomes ‘whiter’. At high values the SPD is rich in blue radiation. It is unfortunate that reddish light has been traditionally known as ‘warm’ and bluish light as ‘cold’, as the actual temperatures associated with these colours are the other way round.
Figure 3.2 Spectral power distribution of full radiating sources (black-body radiation). P, locus of peak emission; V, visible spectrum; Wλ, spectral irradiance.
Colour temperature applies to sources with continuous spectra, but in practice is extended to others with an SPD approximating to that of a full radiator, termed a quasi-Planckian source, such as a tungsten filament lamp. The term is, however, often applied incorrectly to fluorescent lamps, whose spectra and photographic effects can be very different from those of full radiators. The preferred term describing such sources is correlated colour temperature, which indicates a visual similarity to a value on the colour-temperature scale but with an unpredictable photographic effect, particularly with colour reversal film.
In black-and-white photography, the colour quality of light is of limited practical importance. In colour photography, however, it is vitally important, because colour materials and focal plane arrays (FPAs) are balanced to give ‘correct’ rendering with an illuminant of a particular colour temperature. Consequently, the measurement and control of colour temperature must be considered for such work, or the response of the sensor adjusted, usually termed ‘white balance’. The latter may be done automatically in a digital camera or using a preset value chosen from a menu (see Chapter 14).
Colour rendering
With fluorescent lamps, which vary greatly in spectral power distribution, covering a wide range of correlated colour temperatures, the results given by two lamps of nominally the same properties may be quite different if used for visual colour matching or for colour photography. Various objective methods have been devised to give a numerical value to the colour rendering given by such sources as compared with a corresponding full radiator, or with visual perception. A colour rendering index (CRI) or value is defined based on the measurement of luminance in some six or eight spectral bands and compared with the total luminance, coupled with weighting factors, a value of 100 indicating ideal performance. Typical values vary from 50 for a ‘warm-white’ type to greater than 90 for a ‘colour-matching’ version.
Percentage content of primary hues
For many photographic purposes, the visible spectrum can be considered as consisting of three main bands: blue, green and red. The quality of light from a source with a continuous spectrum can be approximately expressed in terms of the percentages in which light of these three hues is present. The method is imprecise, but it is the basis of some colour temperature meters, where the ratios of blue–green and red–green content are compared to indicate an approximate CT. The same principles are used to specify the colour rendering given by a lens, as preferential absorption of light at the blue end of the spectrum is common in optical glass.
Measurement and control of colour temperature
In colour photography, the CT of the light emitted by all the separate sources illuminating a subject should agree in balance with that for which the process is being used. The tolerance permissible depends on the process and to some extent on the subject. A departure of 100 K from the specified value by all the sources (which may arise from a 10% variation in supply voltage) is probably the maximum tolerable for colour reversal material balanced for a colour temperature of around 3200 K. Colour negative material (depending on how it is CT balanced) may allow a greater departure than this, because a certain amount of colour correction is possible at the printing stage.
A particular problem is that of mixed lighting, where part of the subject may be unavoidably illuminated by a light source of markedly different colour quality from the others. A localized colour imbalance may then appear in the photograph. Another example is the use of tungsten lamps fitted with blue filters to match daylight for fill-in purposes, where some mismatch can occur. Electronic flash is suitable as a fill-in source when daylight is the main illuminant. A visual comparison of the colour quality of two light sources is possible by viewing the independently illuminated halves of a folded sheet of white paper with its apex pointing towards the observer. Any visually observable difference in colour would be recorded in a photograph, so must be corrected by filtration. A colour-temperature meter incorporates filtered photocells to sample spectral zones such as red, green and blue. A direct readout of CT is given, together with recommendations as to the type of light-balancing or colour-correction filters needed for a particular type of film.
A matrix array of several hundred CCD photocells filtered to blue, green and red light, together with scene classification data, can also be used in-camera to measure the colour temperature of a scene (see Chapter 14).
The CT balance of colour films to illuminants are specified by their manufacturers. The effective CT of a source may be affected by the reflector and optics used; it also changes with variations in the power supply and with the age of a lamp. To obtain light of the correct quality, lamps must be operated at the specified voltage, and any reflectors, diffusers and lenses must be as near to neutral in colour as possible. The life of filament lamps can be extended by switching on at reduced voltage and arranging the subject lighting, then using the correct full voltage only for the actual exposure. To raise or lower the CT by small amounts, light-balancing filters may be used over the lamps. Pale blue filters raise CT while pale yellow or amber ones lower it.
Because digital cameras have a fixed image sensor, it is necessary to adapt the response of the sensor to the colour temperature of the scene using white balance settings. This may be achieved in a number of ways, but results in the relative responses of the three channels (usually, but not always, from filtered pixels) being altered to match the illumination white point (refer to Chapter 14).
As conventional tungsten filament lamps age, the inner side of the envelope darkens from a deposit of tungsten evaporated from the filament, decreasing both light output and colour temperature. Tungsten–halogen lamps maintain a more constant output throughout an extended life, compared to ordinary filament lamps.
To compensate for the wide variations encountered in daylight conditions for colour photography, camera filtration is possible using light balancing filters of known mired shift value, as defined below. To use colour film in lighting conditions for which it is not balanced, colour conversion (CC) filters with large mired shift values are available.
The mired scale
The colour balance of an incandescent source is given by the mired scale, an acronym derived from micro-reciprocal degree. The relationship between mired value (MV) and colour temperature (T) in kelvins is:
Note that as colour temperature increases the mired value decreases and vice versa.
The main advantage of the mired scale, apart from the smaller numbers involved, is that equal variations correspond to approximately equal visual variations in colour. Consequently, a light-balancing filter can be given a mired shift value (MSV) that indicates the change in colour quality given, regardless of the source being used. Yellowish or amber filters, for raising the mired value of the light, i.e. lowering its colour temperature in kelvins, are given positive mired shift values; bluish filters, for lowering the mired value, i.e. raising the colour temperature in kelvins, are given negative values. Thus, a bluish light-balancing filter with a mired shift value of −18 is suitable for converting tungsten light at 3000 K (333 mireds) to approximately 3200 K (312 mireds). It is also suitable for converting daylight at 5000 K (200 mireds) to 5500 K (182 mireds). Most filters of this type are given values in decamireds, i.e. mired shift value divided by 10. Thus, a blue daylight-to-tungsten filter of value −120 mireds is designated B12. A suitable equation for calculating the necessary MSV to convert a CT of T1 to one of T2 is:
LIGHT OUTPUT
Units
A source can emit energy in a wide spectral band from the ultraviolet to the infrared regions; indeed, most of the output of incandescent sources is in the infrared. For most photography only visible light is of importance. Three related photometric units define light output: luminous intensity, luminance and luminous flux.
Luminous intensity is expressed numerically in terms of the fundamental SI unit, the candela (cd). One candela is the luminous intensity in the direction of the normal to the surface of a full radiator of surface area 1/600,000 of a square metre, at the temperature of solidification of platinum. Luminous intensity is not necessarily uniform in all directions, so a mean spherical value, i.e. the mean value of luminous intensity in all directions, is used. Luminous intensity was originally called ‘candle-power’.
Luminance is defined as luminous intensity per square metre. The unit of luminance is the candela per square metre (cd m−2). An obsolescent unit sometimes encountered is the apostilb (asb), which is one lumen per square metre (lm m−2) and refers specifically to light reflected from a surface rather than emitted by it. The luminance of a source, like its luminous intensity, is not necessarily the same in all directions. The term luminance is applicable equally to light sources and illuminated surfaces. In photography, subject luminances are recorded by a film as analogue optical densities of silver or coloured dyes.
Luminous flux is a measure of the amount of light emitted into space, defined in terms of unit solid angle or steradian, which is the angle subtended at the centre of a sphere of unit radius by a surface of unit area on the sphere. Thus, an area of 1 square metre on the surface of a sphere of 1 metre radius subtends at its centre a solid angle of 1 steradian. The luminous flux emitted into unit solid angle by a point source having a luminous intensity of 1 candela in all directions within the angle is 1 lumen (lm). Since a sphere subtends 4π steradians at its centre (the area of the surface of a sphere is 4πr2), a light source of luminous intensity of 1 candela radiating uniformly in all directions emits a total of 4π lumens, approximately 12.5 lm (this conversion is only approximately applicable to practical light sources, which do not radiate uniformly in all directions). The lumen provides a useful measure when considering the output of a source in a reflector or other luminaire or the amount of light passing through an optical system.
Illumination laws
The term illumination refers to incident light and depends on the luminous flux falling on a surface and its area. The quantitative term is illuminance or incident luminous flux per unit area of surface. The unit is the lux (lx): 1 lux is an illuminance of 1 lumen per square metre. The relationship between the various photometric units of luminous intensity, luminous flux and illumination is shown in Figure 3.3. The illumination E on a surface at a distance d from a point source of light depends on the output of the source, its distance and the inclination of the surface to the source. The relationship between illumination and distance from the source is known as the inverse square law of illumination. Light emitted into the cone to illuminate base area A at distance d1 with illumination E1 is dispersed over area B at distance d2 to give illumination E2. By geometry, if d2 is twice d1, then B is four times A, i.e. illumination is inversely proportional to the square of the distance d. Hence:
From the definition of the lumen, the illumination in lux from a source at distance d is given by dividing the luminous intensity by the square of the distance in metres. So the illumination on a surface 5 metres from a source of 100 candelas is 100/(5)2 = 4 lx. Recommended values of illumination for different areas range from 100 lx for a domestic lounge to at least 400 lx for a working office.
The reduced amount of illumination on a tilted surface is given by Lambert’s cosine law of illumination, which states that the illumination on an inclined surface is proportional to the cosine of the angle of incidence of the light rays falling on the surface. For a source of luminous intensity I at a distance d from a surface inclined at an angle θ, the illuminance E on the surface is given by:
The inverse square law strictly applies only to point sources. It is approximately true for any source that is small in proportion to its distance from the subject. The law is generally applicable to lamps used in shallow reflectors, but not deep reflectors. It is not applicable to the illumination provided by a spotlight due to the optical system used to direct the light beam.
Reflectors and luminaires
Most light sources are used with a reflector, which may be an integral part of the lamp or a separate item. A reflector affects the properties of the lighting unit as regards distribution and uniformity of illumination. The source, reflector and housing form a ‘luminaire’. Reflectors vary in size, shape and nature of surface, and can be flat or shallow, or deeply curved in spherical or paraboloidal form. The surface can be highly polished, smooth matt or even lenticular. Some intermediate arrangement is usually favoured to give a mixture of direct and diffuse illumination. The light distribution from a luminaire is given by graphing the luminous intensity in each direction in a given horizontal plane through the source as a curve in polar coordinates, termed a polar distribution curve (see Figure 3.4). The source is at the origin (0°) and the length of the radius from the centre to any point on the curve gives the luminous intensity in candelas in that particular direction.
A reflector has a reflector factor, which is the ratio of the illuminance on the subject by a light source in a reflector to that provided by the bare source. A flashgun reflector may have a reflector factor from 2 to 6 to make efficient use of the light output, which is directed into a shaped beam with very little illumination outside the primary area.
Figure 3.3 Photometric units and laws of illumination. (a) Photometric units derived from source S with intensity l of 1 candela (cd) at centre of sphere of radius 1 m. Flux emitted into 1 steradian (sr) is 1 lumen (lm). Area A is 1 m2 with illumination E of 1 lux (lx) and luminance L in cd m−2 with reflectance R. (b) Inverse square law of illumination. Point source L gives illumination E1 and E2 at distances d1 and d2, where E1/E2 = (d2)2/(d1)2. (c) Lambert’s cosine law of illumination. Illumination EX at point X from source S of intensity l is EX = l/(d1)2. At point Y, EY = l cos θ/(d2)2, where θ is obliquity.
Figure 3.4 Distribution of illumination. Illumination levels at 5 m distance from a light source. (A) A 2 kW spotlight in ‘spot’ mode. (B) In ‘flood’ mode. (C) A 2.5 kW ‘fill-in’ light.
Flashguns are often used with a large reflector of shallow box- or umbrella-like design and construction, called ‘softboxes’ and ‘brollies’. Various surface finishes and diameters are used to convert the flashgun from a small source giving hard shadows when used direct, to a large, diffuse source offering softer lighting when used as the sole illuminant, albeit with considerable loss of efficiency. The bare-bulb technique, where the flash source is used without reflector or diffuser, is sometimes used for its particular lighting quality.
A spotlight gives a high level of illumination over a relatively small area, and as a rule gives shadows with hard edges. The illumination at the edges of the illuminated area falls off quite steeply. The use of diffusers gives softer shadows, and ‘snoots’ give well-defined edges. A spotlight consists of a small incandescent or flash source with a filament or flashtube at the centre of curvature of a concave reflector, together with a condenser lens, usually of Fresnel lens construction to reduce weight (Figure 3.5). By varying the distance between source and condenser, the diameter of the beam of light may be varied. For a near-parallel beam, the source is positioned at the focus of the condenser lens. With compact light sources, such as tungsten–halogen lamps, where it is possible to reduce the physical size of spotlights and floodlights, the optical quality of reflectors and condensers needs to be high to ensure even illumination.
Small electronic flash units can vary the distance of the flashtube from the Fresnel lens to provide a narrower beam that will concentrate the light output into the field of view of lenses of differing focal lengths. This ‘zoom flash’ action can be motorized and controlled from the camera to change automatically to match the lens attached to the camera, or when a zoom lens is used.
Constancy of output
Constant light output and quality are necessary characteristics of photographic light sources, especially for colour work. Daylight, although an intense and cheap form of lighting, is by no means constant: its intensity and quality vary with the season, time of day and weather. Artificial light sources are more reliable, but much effort can go into arranging lighting set-ups to simulate the desirable directional qualities of sunlight and diffuse daylight.
Electric light sources need a reliable power supply in order to maintain a constant output. If the frequency and/or voltage of the mains supply fluctuate, variation in light intensity and quality may result. Incandescent lamps inevitably darken with age, lowering both output and colour temperature. Tungstenehalogen lamps largely circumvent these problems. Fluorescent lamps have a long life, but also gradually decrease in light output with age.
Electronic flash has reliable light output and quality. Acceptably constant output requires adequate recharging time between successive flashes. The ready-light indicators fitted to many units glow when about 80% of the charging voltage is reached, but the energy available for the flash is only some two-thirds of full charge. Further time must be allowed to elapse before discharge, to ensure full capacity is available.
Efficiency
The efficiency of a light source for photographic use is related to factors determining its usefulness and economy in particular circumstances. These include control circuitry utilizing design techniques to give a low power consumption and choice of reflector to concentrate light output in a particular direction. Electronic flash units are examples of efficient reflector design, with little of the luminous output wasted.
The photographic effectiveness of a light source relative to a reference source is called its actinity, and takes into account the SPD of the source and the spectral response of the sensitized material or photosensor array. The efficacy is the ratio of luminous flux emitted to the power consumed by the source, and is expressed in lumens per watt. A theoretically perfect light source emitting white light of daylight quality would have an efficacy of about 220 lumens per watt. Values obtained in practice are lower.
DAYLIGHT
Daylight typically includes direct light from the sun, plus scattered light from the sky and clouds. It has a continuous spectrum and colour temperature gives an approximation of its quality, which varies through the day. Colour temperature is low at dawn, in the region of 2000 K if the sun is unobscured. It then rises to a maximum, and remains fairly constant through the middle of the day, to tail off slowly through the afternoon and finally fall rapidly at sunset to a value again below that of a tungsten filament lamp. The quality of daylight also varies from place to place according to whether the sun is shining in a clear sky or is obscured by cloud. The reddening of daylight at sunrise and at sunset is due to the absorption and scattering of sunlight by the atmosphere. These are greatest when the sun is low, because the path of the light through the earth’s atmosphere is then longest. As the degree of scattering is more marked at short wavelengths, the unscattered transmitted light contains a preponderance of longer wavelengths and appears reddish, while the scattered light (skylight) becomes more blue towards sunrise and sunset.
Figure 3.5 Effects of moving the light source in a luminaire. (a) Principle of the spotlight. When light source L is in ‘spot’ position S at the focus of spherical mirror M and Fresnel lens FL, a narrow beam of some 40° results. At ‘flood’ position F, a broader beam of some 85° results. (b) A ‘zoom’ flash system. The flash tube and reflector assembly R move behind the diffuser D to increase light coverage from telephoto T to wide-angle Wsetting of a camera lens.
These fluctuations prohibit the use of natural daylight for sensitometric evaluation of photographic materials (see Chapter 8). Light sources of fixed colour quality are essential. For many photographic purposes, especially in sensitometry, the average quality of sunlight is used as a standard (skylight is excluded). This is referred to as mean noon sunlight, and approximates to light at a colour temperature of 5400 K. Light of similar quality, but with a CT of 5500 K, is termed ‘photographic daylight’. It is achieved in the laboratory by operating a tungsten lamp under controlled conditions to emit light of the required colour temperature, by modifying its output with a Davis–Gibson liquid filter.
Near noon, combined light from the sun, sky and clouds has a CT in the region of 6000–6500 K. An overcast (cloudy) sky has a higher value, and a blue sky may be as high as 12,000–18,000 K. The CT of the light from the sky and the clouds is of interest because it is this skylight alone that illuminates shadows, giving them a colour balance different to a sunlit area.
TUNGSTEN FILAMENT LAMPS
An incandescent photographic lamp produces light by the heating action of electric current through a filament of tungsten metal, with melting point 3650 K. The envelope is filled with a mixture of argon and nitrogen gas to allow operation at temperatures up to 3200 K. A further increase, to 3400 K, gives increased efficacy but a decrease in lamp life. A tungsten filament lamp is designed to operate at a specific voltage, and performance is affected by fluctuations in supply voltage. A 1% increase causes a 4% increase in luminous flux, a 2% increase in efficacy, a 12% decrease in life and a 10 K increase in colour temperature for a lamp operating in the range 3200–3400 K. Tungsten filament lamps include:
• General-service lamps used for domestic purposes in sizes from 15 to 200 W with clear, pearl or opal envelopes. The CT ranges from about 2760 to 2960 K, with a life of about 1000 hours.
• Photographic lamps are made for photographic use as reflector spotlights and floodlights. Their CT is nominally 3200 K, with a life of approximately 100 hours and power rating of 500 W. These have been superseded by more efficient, smaller tungstenehalogen lamps with greater output and longer life.
• Photoflood lamps produce higher luminous output and more actinic light by operating at 3400 K. A rating of 275 W gives a life of 2–3 h and 500 W a life of 6–20 h. Those with internal silvering in a shaped bulb do not need an external reflector.
TUNGSTEN–HALOGEN LAMPS
The tungsten–halogen lamp has a quantity of a halogen added to the filling gas. During operation a regenerative cycle is set up whereby evaporated tungsten combines with the halogen in the cooler region near the envelope wall, and when returned by convection currents to the much hotter filament region the compound decomposes, returning tungsten to the filament and freeing the halogen for further reaction. The effect is that evaporated tungsten does not deposit on the bulb wall and this prevents blackening with age. Filament life is also extended due to the returned tungsten, but eventually the filament breaks as the redeposition is not uniform. The complex tungstene halogen cycle functions only when the temperature of the envelope exceeds 250°C, achieved by using a small-diameter envelope of borosilicate glass or quartz (silica). The gas filling is used at several atmospheres pressure to inhibit the evaporation of tungsten and helps increase the life of such lamps as compared with that of conventional tungsten lamps of equivalent rating. The small size of tungstenehalogen lamps has resulted in lighter, more efficient luminaires.
Early designs used iodine as the halogen and the lamps were known as ‘quartzeiodine’ (QI) lamps. Other halogens and their derivatives are now used. Tungstenehalogen lamps are available as small bulbs and in tubular form, supplied in a range of sizes from 50 to 5000 W with CT from 2700 to 3400 K. A 200-hour life is usual with near-constant colour temperature. By operating at low voltages (12 or 24 V), a more compact filament can be used.
FLUORESCENT LAMPS
A fluorescent lamp is a low-pressure mercury-vapour discharge lamp with the envelope coated internally with a mixture of fluorescent materials or phosphors. These absorb and convert emitted short-wave UV radiation into visible light, the colour of which depends on the mixture of phosphors used, but can be made a close visual match to continuous-spectrum lighting.
Fluorescent lamps emit a line spectrum with a strong continuous background; their light quality can be expressed approximately as a correlated colour temperature. The colour rendering index (CRI) may also be quoted. There are many subjective descriptive names for fluorescent lamps such as ‘daylight’, ‘warm white’ and ‘natural’. Lamps are classified as ‘high efficiency’ and ‘de-luxe’. The former group have approximately twice the light output of the latter for a given wattage, but are deficient in red. They include ‘daylight’ lamps of approximately 4000 K and ‘warm white’ lamps of approximately 3000 K, a rough match to domestic tungsten lighting. The de-luxe group gives good colour rendering by virtue of the use of lanthanide (rare-earth element) phosphors, and includes colour matching types at equivalent colour temperatures of 3000, 4000, 5000 and 6000 K. The life of a fluorescent tube is usually of the order of 7000–8000 h, and the output is insensitive to small voltage fluctuations.
Colour images recorded using fluorescent lamps, even if only present as background lighting, may result in unpleasant green or blue colour bias, especially on colour-reversal film, needing corrective filtration by means of suitable colour-compensating filters over the camera lens.
METAL-HALIDE LAMPS
Originally the only metals used in discharge lamps were mercury and sodium, as the vapour pressures of other metals tend to be too low. However, the halides of most metals have higher vapour pressures than the metals themselves. In particular, the halides of lanthanide elements readily dissociate into metals and halides within the arc of a discharge tube. The ionized metal vapour emits light with a multi-line spectrum and a strong continuous background, giving virtually a continuous spectrum. The metals and halides recombine in cooler parts of the envelope. Compounds used include mixtures of the iodides of sodium, thallium and gallium, and halides of dysprosium, thulium and holmium in trace amounts. The discharge lamp is a very small ellipsoidal quartz envelope with tungsten electrodes and molybdenum seals. Oxidation of these seals limits lamp life to about 200 h, but by enclosing the tube in an outer casing and reflector with an inert gas filling, life can be increased to 1000 h.
The small size of this lamp has given rise to the term compact-source iodide (CSI) lamp. Light output is very high, with an efficacy of 85–100 lumens per watt. The hydragyrum metal iodide (HMI) lamp uses mercury and argon gases with iodides of dysprosium, thulium and holmium to give a daylight matching spectrum of precisely 5600 K and CRI of 90 with a high UV output also.
When operated on an AC supply, light output fluctuates at twice the supply frequency. The resulting variation in intensity is some 60–80%. This can cause problems when used for short exposure durations in photography, unless a three-phase supply or special ballast control gear is used. Ratings of up to 5 kW are available.
PULSED XENON LAMPS
Pulsed xenon lamps are a continuously operating form of electronic flash device. By suitable circuit design a quartz tube filled with xenon gas at low pressure discharges at twice the mains frequency, i.e. 100 Hz for a 50 Hz supply, so that although pulsed the light output appears continuous to the eye. The spectral emission is virtually continuous, with a colour temperature of approximately 5600 K (plus significant amounts of ultraviolet and infrared radiation). Lamp dimensions are small, and they have replaced traditional carbon-arc lamps. Power ratings up to 8 kW are available, and lamp life is 300–1000 h depending on type.
EXPENDABLE FLASHBULBS
The traditional flashbulb is now obsolescent and used only occasionally such as for lighting very large interiors or for high-speed recording where a series of bulbs fired in a ‘ripple’ give an intense compact source for a short time.
Flashbulbs contain fine metal ribbons of hafnium, zirconium or magnesium–aluminium alloy in an atmosphere of oxygen at low pressure, enclosed in a glass envelope with a lacquer coating to prevent shattering when fired. On electrical ignition, a bright flash of light is emitted as the metal burns, of about 0.01–0.02 s duration. The emission spectrum is continuous, with a CT of about 3800 K. A transparent blue lacquer coating converts this CT to 5500 K.
Trigger voltage from about 3 to 30 volts is from a battery-capacitor circuit. A battery charges a capacitor through a high resistance and the discharge of the capacitor then fires the bulb. Such a circuit may be used to fire several flashbulbs simultaneously in a multiple-flash set-up. Alternatively, slave units are connected to the individual extension flashguns so one flashbulb is triggered from the camera shutter contacts, and the light emitted by it operates photocell switches installed in the slave units to give near-simultaneous firing.
To avoid the necessity for batteries, alternative methods were used for firing arrays of bulbs in units intended primarily for simple cameras. Methods included a torsion spring striking a primer and a piezo-electric crystal and striker to give a firing pulse.
Flash performance is shown as a graph of emitted luminous flux emitted plotted against time, as in Figure 3.6a. Effective flash duration is a measure of the motion-stopping power and is the time interval between half-peak points. The camera shutter must be fully open at the point of peak output. The total light output (in lumen seconds) is given by the area under the curve.
A guide number or flash factor can be used to calculate camera exposure. A guide number (G) is the product of the f-number (N) of the camera lens and the subject distance (d) in metres for film of speed 100 ISO:
Guide numbers are subject to modification to suit the particular conditions of use, being influenced by the reflective properties of the surroundings of the subject.
Figure 3.6 Characteristics of flash sources. (a) Flash bulb. A, time to peak; B, useful output for 1/125 s shutter speed, ‘X’ synchronization; C, output for 1.125 s with ‘M’ synchronization and 15 ms delay – total output is area under curve; K, light output in megalumens (Mlm); t, time; W, effective flash duration between half-peaks. (b) Automatic electronic flash. E, effective flash duration measured between one-third peak power points D (≈ 1 ms); 1, full output; 2–7, quenched output to reduce output and duration. For 7, t ≈ 1/20,000 s. (c) Stroboscope. Flashing rate of 1 kHz. F, inter-flash interval of 1 ms; G, flash duration of 20 μs (0.05 ms).
Guide numbers are estimated from the following formula:
where L is the maximum light flux in lumens, t is the exposure duration in seconds, R is the reflector factor and S is the arithmetic film speed. A modification necessary for units with built-in reflectors such as Flashcubes is:
where E is the effective beam intensity of the bulb. This quantity is measured by an integrating light meter across the entire angle of coverage of the reflector and is not influenced by hot spots. It replaces the alternative beam candle power (BCP) measured at the centre of the beam only, but could be misleading.
Flash synchronization with the shutter is discussed in Chapter 11.
ELECTRONIC FLASH
Flash circuitry
In an electronic flashtube, an electrical discharge from a capacitor through a gas such as xenon, krypton or argon causes emission of a brief, intense flash of light. The gas is not consumed by this operation and the tube may be flashed repeatedly, with a life expectancy of many thousands of flashes.
The circuitry uses a capacitor charged from a DC supply of the order of 350–500 V, through a current-limiting resistor. This resistor allows a high-power output from a low-current-rated supply with a charging time of several seconds. The gas becomes deionized and the flash is extinguished after the discharge. The high internal resistance of the gas in the flashtube normally prevents any flow of current from the main capacitor, but by applying a triggering voltage (typically a short pulse of 5–15 kV) by means of an external electrode, the gas is ionized to become conducting, allowing the capacitor to discharge rapidly through it, causing a brief flash of light. This triggering voltage is linked to shutter contacts in the camera but a spark coil arrangement ensures only very low voltages pass though the contacts for safety. The trigger electrode may be in the form of fine wire around the flashtube or a transparent conductive coating.
The SPD of the flash discharge is basically the line spectrum characteristic of xenon, superimposed on a very strong continuous background. The light emitted has a correlated colour temperature of about 6000 K. The characteristic bluish cast given by electronic flash with some colour-reversal materials may be corrected either by a light-balancing filter with a low positive mired shift value, or by a pale yellow colour-compensating filter. Often the flashtube or reflector is tinted yellow or yellow–green as a means of compensation. The flashtube may be of borosilicate glass or quartz, the former giving a cut-off at 300 nm, the latter at 180 nm. Both transmit wavelengths up to 1500 nm in the infrared region. By suitable design of control circuitry, ultraviolet or infrared output can be enhanced, so an electronic flashgun can be a useful source of these radiations as well as visible light.
Flash output and power
The power output is altered by using additional capacitors, or the current may be divided between two tubes or separate heads or by thyristor control. The output can be ‘symmetrical’ or ‘asymmetrical’, e.g. 1000 Ws−1 divided either into 2 × 500 or 4 × 250, or into 500 plus 2 × 250 Ws−1 respectively. Alternative low-power settings typically offer alternative outputs from one-half to one-sixty-fourth of full power in halving steps. Extension heads may be used for multiple flash arrangements and each head may have its own capacitor. If the extension head shares the main capacitor output the connecting lead must be substantial, to reduce resistance losses. For greatest efficiency the connection between capacitor and flashtube must be as short and low resistance as possible. One-piece or ‘monobloc’ studio flash units are examples of this type. The use of slave units for synchronization facilitates multiple-head flash operation. Computerized control of outputs may be used.
The energy input Ei per flash in watt-seconds (W s−1), often still called ‘joules’ (J), not all of which is available as output, is given by:
where C is the capacitance in microfarads and V is the voltage in kilovolts. The energy output Eo is estimated by a formula for flash energy conversion such as:
where K1 is the conversion efficiency (some 50–60%), K2 is the percentage of spectral bandwidth used, K3 is the percentage of the emitted light directed at the subject by the optical system and K4 is an empirical factor.
Camera integral units are rated at 5–20 Ws−1 (joules), hand-held units for on-camera use are 20–200 Ws−1, while large studio flash units are available with ratings of 200–5000 Ws−1. The state of charge of the capacitor, i.e. readiness for discharge, is indicated by a neon light or beeping sound circuit, which is usually set to strike at about 80% of maximum voltage, i.e. at about two-thirds full charge. Several seconds must be allowed to elapse after this to ensure full charge. If the flash is not then triggered, various forms of monitoring circuitry may be used to switch the charging circuitry on and off to maintain full charge (‘top-up’) with minimum use of the power supply while conserving battery power. Some units may be equipped with monitoring circuitry that switches it off if not used within a set period. The stored energy is ‘dumped’ as a safety measure.
Changes to flash circuitry allow a flashtube to be operated in strobe mode, i.e. emit a series of short flashes as well as a longer single flash. This allows a series of ‘pre-flashes’ to be emitted for red-eye reduction or for scene analysis, or during shutter operation to give a multiple exposure strobe effect, or to allow high shutter speeds to be used.
Flash duration and exposure
The characteristics of an electronic flash discharge are shown in Figure 3.6b. The effective flash duration (t) is usually measured between one-third peak power points, and the area under the curve between these points represents approximately 90% of the light emitted, measured in lumen-seconds or effective beam-candela-seconds. The flash duration t, as defined above, in a circuit whose total tube and circuit resistance is R with a capacitor of capacitance C, is given approximately by:
So, for a 1 ms duration flash, the requirements are for high capacitance, low voltage and a tube of high internal resistance. The flash duration is usually in the range 0.02–2 ms. Units may have a variable output, controlled either by switching in or out of additional capacitors or by automatic ‘quenching’ controlled by monitoring of scene or image luminance (see later), giving a suitable variation in either intensity or flash duration. For example, a studio flash unit on full power may have a flash duration of 2 ms, changing to 1 ms and 0.5 ms when half- and quarter-power respectively are selected. Special strobe and microflash units may have outputs whose duration is of the order of 1 ms but with only about 1 Ws−1 of energy. The output efficacy (P) of a flashtube is defined as being:
where L is the peak output (in lumens). The guide number of a flash unit with reflector factor R used with a film of arithmetic speed rating S is given by:
The flash guide number for distance in metres for ISO 100 film is usually incorporated in the designation number of the unit, e.g. ‘45’ means G = 45.
The silvered reflectors used are highly efficient and direct the emitted light into a well-defined rectangular area with sharp cut-off, approximating to the coverage of a semi-wide-angle lens on a camera. Coverage may be altered to greater or lesser areas to correspond with the coverage of other lenses or a zoom lens by the addition of clip-on diffusers or by moving a Fresnel-type condenser lens in front of the flashtube to an alternative position. A moving tube ‘zoom’ feature may even be controlled automatically by the camera itself as a zoom lens is operated.
The use of diffusers and ‘bounce’ or indirect flash severely reduces the illumination on the subject and often needs corrective filtration with colour materials owing to the nature and colour of the reflective surface. This can be used to advantage, as when a golden-surfaced umbrella reflector is used to modify the slight blue cast given by some flash units, which may have colour temperatures of 6000 K or more. The main result of ‘bounce’ lighting is to produce softer, more even illumination. An umbrella-type reflector of roughly paraboloidal shape, made of translucent or opaque material, with white or metallic silver or gold finish, is a widely used accessory item, as are translucent umbrellas and ‘soft boxes’ to diffuse the light.
Portable units
A portable flash unit is attached to a camera via a hotshoe connection, or to a side bracket. The power supply is normally batteries, preferably alkaline-manganese. Alternatively, rechargeable types may be used such as nickel metal hydride and lithium-ion batteries. An LED display may show the state of charge or discharge of a set of batteries.
A range of accessories is available. The reflectors may be interchangeable from ‘bare-bulb’ to a deep paraboloid to give different beam shapes and lighting effects. Various modes of use include direct flash at full power, tilting reflectors for bounce flash, fractional output for close-up work or fill-in flash in sunlight, automatic exposure mode, programmed mode or stroboscopic mode. A data display panel gives a comprehensive readout of the operational mode in use.
Studio flash
A studio flash unit is usually used indoors but is transportable. Output is from 250 to 5000 Ws−1. The power supply is from the mains or a generator for location work, or even a large battery pack. The discharge tube may be in a linear or helical shape or possibly circular to give a ‘ring flash’. Studio units have the convenience of a ‘modelling lamp’ positioned near the discharge tube to give a preview of the proposed lighting effect. The modelling lamp may be a tungsten or tungstenehalogen lamp, often with fanassisted cooling, and its output may be variable, related to the flash output selected so as to facilitate visual judgement of lighting balance. The flash output may be selected manually by switches or by remote control, or by a cordless infrared programming unit used in the hand from the camera position. Output may be set within limits as precisely as one-tenth of a stop, a great convenience in balancing lighting rather than by shifting lights about, and allowing a smaller studio to be used.
The determination of the appropriate lens aperture setting is by the use of a hand-held integrating flash meter. Automatic in-camera methods use photocells to monitor the scene luminance or the image luminance on the film surface by reflected light to a photocell in the camera body during the flash discharge. This is termed off-the-film (OTF) flash metering. As a confirmation and to give confidence in the outcome, a sheet of self-developing film may be exposed to evaluate exposure and lighting, as well as to check that everything is operating satisfactorily. Digital cameras give a preview of the result from the photosensor array.
Automatic flash exposure
The automation of camera exposure determination using electronic flash utilizes rapid-acting devices called thyristor switches and has power-saving advantages.
A silicon photodiode monitors the subject luminance when being illuminated by the flash discharge, and the resultant output signal is integrated by circuitry until it reaches a preset level. The flash discharge can then be abruptly terminated by a thyristor switch to give the correct exposure, subject to some limitations. The switch-off level depends on film speed and lens aperture.
The discharging flash capacitor may have its residual energy diverted into an alternative ‘quench’ tube, which is a low-resistance discharge tube connected in parallel with the main tube, and activated only when the main tube has ignited so as to prevent effects from other flashguns in use in the vicinity. There is no light output from the quench tube and the dumped energy is lost. The effective flash duration, dependent on subject distance, reflectance and film speed, may be as short as 0.02 ms, with concomitant motion-stopping ability. A preferred arrangement saves the residual charge in the capacitor and thereby reduces recharging time per successive flash and also increases the total number of flashes available from a set of batteries. This energy-saving circuitry has the thyristor switch positioned between the main capacitor and the flashtube. The thyristor is closed when the flash is initiated, and then opened almost instantaneously on receiving a pulse from the light-monitoring circuit to terminate the flow of current sustaining the flash. Full flash discharge is possible in manual mode by taking the monitoring photocell out of circuit so that the ‘open’ pulse is never given. To allow for the use of different film speeds and lens apertures the photocell may be biased electrically or mechanically. Flash duration reduces as the flash is quenched more rapidly over a range of usually 30 or 50 to 1, i.e. from 1 ms at full power to 0.02 ms at minimum power. This is useful for capturing motion of a subject. Flash colour temperature may also change with output duration, tending to rise as duration reduces.
The photocell can be in a small hotshoe-mounted housing attached to the flash unit by a flying lead, which allows monitoring of scene luminance irrespective of the flash-head position, and allows the use of bounce flash. It is desirable with flashguns having an integral photocell that this is positioned to face the subject irrespective of the direction in which a rotatable or swivelling flashtube assembly may be pointed. Two or more automatic flashguns can be combined for multiple flash use with either connecting leads or ‘cordless’ infrared or radio triggering systems. The hotshoe on the camera has a single central ‘X’ flash synchronization connection and, depending on the camera system, other connections to interface the flashgun with the camera for features such as automatic selection of a shutter speed for correct synchronization when the flash is attached, a ‘thunderbolt’ readylight indication in the viewfinder, use of a short ‘pre-flash’ for autofocus use or exposure determination and control of a zoom flash feature. The flash synchronization may be selectable to be first blind or second blind type where it is triggered either by the first blind uncovering the film gate or by the second blind just before it starts to cover up the gate. This latter feature is useful for combined flash and short time exposures with moving subjects to give an enhanced impression of motion. The problems of flash synchronization to the camera shutter are dealt with in Chapter 11.
Control can be by a photocell inside the camera which monitors the luminance of the image of the scene actually on the photoplane while the camera shutter is fully open at the flash synchronization speed when the photoplane is fully uncovered. A segmented photocell provides a weighted analysis of the tonal range and distribution in the scene, termed ‘matrix flash’. For use in ambient light to provide fill-in by ‘synchro-sunlight’ techniques, this scene analysis can provide the right amount of supplementary flash to give a more pleasing image. An automatic flashgun may only operate fully with a particular camera, when it is called a ‘dedicated’ flashgun. The firmware of an automatic flash can be updated via downloads from a computer. A flow diagram for dedicated-flash operation is shown in Figure 3.7.
Integral flash units
Most cameras have an integral flash unit with a small but useful output. This can be just a fixed value as in single-use cameras or a system with a range of modes. In compact cameras the flash is automatically activated when the light level is too low for an ambient-light exposure. The flash output can be a single full discharge with the necessary lens aperture set to the required GN value by using subject distance information from the autofocus system used in the camera. This is a ‘flashmatic’ system. In other cameras of the SLR type, the flashgun is often fully automated with a photocell for through-the-lens and off-the-film operation. The output from a photosensor array may also be used. A menu of flash modes includes ‘automatic’, i.e. selected according to subject luminance, ‘off’ when no flash fires, ‘on’ when the flash fires irrespective of light conditions and always at full power, or ‘red-eye reduction’ to alleviate this disturbing effect.
Figure 3.7 Flow diagram of dedicated flashgun operation. The major stages of operational sequences and the features offered by such flashguns are shown.
Red-eye avoidance
An unfortunate optical consequence of an integral flash is that the flashtube is close to the lens so that the light is directed essentially along the optical axis. This direct axial lighting can be a problem with reflective subjects when a glare spot is given in the picture, but even more so when a subject is looking at the camera. Light enters the eye via the pupil, usually fully dilated in the dim light conditions requiring the use of flash, and illuminates the retina at the back of the eye, which is rich in red blood vessels. The result is a characteristic bright red pupil. This red-eye effect can be reduced in various ways. The ambient light can be increased to cause the pupil to contract, or a small projector in the camera or flashgun can direct a pencil beam of light at the face to the same effect or the flashgun can emit a number of short pre-flashes before the main exposure, when hopefully the pupil will have reacted. The subject can look away from the camera lens, but the best method is to move the flashtube away from the camera lens. This influences camera design and the integral flash unit may be on a short extension of the camera body for use. Portable flash units should be used if possible but on an extension cable. The choice and disposition of lighting to suit the ‘treatment’ of the subject is beyond the scope of this book.
The various forms of flash synchronization and output, whether full, quenched or pulsed, are shown and compared in Figure 3.8.
OTHER SOURCES
Other light and radiation sources find use in photography and digital imaging, for both image capture and illumination systems.
Light-emitting diodes
The light-emitting diode or LED is a very small solid-state device encapsulated in a housing with integral lens to direct and shape the emitted beam. Operation is by electroluminescence using forward biasing of a pen junction in materials such as gallium arsenide or gallium phosphide. Light output is monochromatic and, typically, red (649 nm), green (570 nm) or infrared (850 nm). The LED can be mains or battery powered and has very modest power consumption, but can still be a significant drain on the limited capacity of the battery in a camera. It has a principal use as an indicator light for equipment operational conditions or modes. The small source can be used alone or behind a patterned mask to give an alphanumerical or symbolic display, as in a viewfinder or flashgun data readout. The usual states are off, on, pulsing or dimmed (given by altering a flashing rate from a pulsed supply).
Figure 3.8 Flash modes and shutter synchronization. (a) First and second blind synchronization. 1, first (leading) blind; 2, second (trailing) blind; A, shutter triggered; B, reflex mirror time; C, blind travel time; D, flash triggered by first blind (X synchronization); E, flash triggered by second blind; O, film gate fully uncovered (flash synchronization speed); L, flash output; T, elapsed time; t, flash duration (nominal); K, main flash; Q, start of blind travel. (b) Quenched electronic flash. L2, reduced flash output; t2, reduced flash duration. (c) Red-eye reduction by bright pre-flashes. f1, f2, f3, pre-flashes. (d) Matrix metering of scene using faint pre-flashes. f1 to fn, weak pre-flashes; J, data processing. (e) Flash with high shutter speeds, e.g. to 1/12,000 second, using high-frequency strobe flashes. P, pulsed strobe flash. (f) Flash synchronization errors. M, incorrect use of M instead of X synchronization setting; W, shutter speed too short for synchronization; a travelling slit is formed.
To provide a more intense beam suitable for illumination purposes, LEDs are used in arrays of multiple sources, e.g. emitting in the near infrared as used for covert work to illuminate a scene for recording by a suitable infrared system. An illuminator operating from a 12 V supply can illuminate a scene up to 25 m distant. Illuminators can be rectangular, linear or annular in shape to suit the illumination task. A pulse rate up to 150 Hz is typical. The red monochromatic emission also finds use in the darkroom as a bright safelight for room illumination or as a portable unit for hand-held local illumination.
Diode lasers
The diode laser or ‘microlaser’ is a semiconductor crystal derived from gallium arsenide that emits intense coherent light at a few wavelengths by stimulated emission, as compared to the spontaneous emission of an LED that produces incoherent light with a wider spectral range. The astigmatic shape of the output beam requires optical correction to a circular form. The devices can be assembled in rectangular array, linear (bar) or single source form. Laser wavelengths of 750–780 nm are typically used in optical disc (compact disc, CD) reader systems with a single source. Wavelengths of 670 and 780 nm are suited to printing plate technology using digital data direct from computer image files.
These devices can have an intense monochromatic output suited to three-colour (red, green, blue) exposure systems in various forms of printer to provide a hard-copy output from digital files and may even use three different infrared wavelengths to expose the three-colour forming image layers. Display systems may use red (656 nm), green (532 nm) and blue (457 nm) modulated microlaser beams to form a direct write display.
BIBLIOGRAPHY
Carlson, V., Carlson, S., 1991. Professional Lighting Handbook, second ed. Focal Press, Boston, USA.
Cayless, M., Marsden, A. (Eds.), 1983. Lamps and Lighting, third ed. Edward Arnold, London, UK.
Edgerton, E., 1970. Electronic Flash, Strobe. McGraw-Hill, New York, USA.
Fitt, B., Thornley, J., 1997. Lighting Technology. Focal Press, Oxford, UK.
Jacobson, R.E.J., Ray, S.F.R., Attridge, G.G., Axford, N.R., 2000. The Manual of Photography, ninth ed. Focal Press, Oxford, UK.
Minnaert, M., 1993. Light and Colour in the Outdoors. Springer-Verlag, New York, USA.
Ray, S., 1999. Scientific Photography and Applied Imaging. Focal Press, Oxford, UK.
Ray, S., 2002. Applied Photographic Optics, third ed. Focal Press, Oxford, UK.