Photography of shiny objects
Every modern camera is equipped with automatic exposure settings.
An exposure meter measures how much light is reflected from the scene and decides in some way what combination of lens aperture and shutter speed is appropriate, taking into account the sensitivity to light of the photographic material (film or electronic parts). The more sophisticated cameras not only offer autoexposure, but also incorporate facilities for multisegment metering, centre weighted measurement or spot metering, as well as the opportunity to set shutter or aperture priority.
The smaller the lens aperture (f-stop), the greater the depth of field, f. the area in front of and behind a focused subject in which the photographed image appears sharp. Changing the aperture by one full stop changes the amount of light transmitted by a factor 2. For example, changing from f 5,6 to f 8 (f stands for focus, the higher the value of f, the smaller the lens aperture) halves the amount of light transmitted per unit of time. For the same amount of light hitting the light sensitive material, the shutter speed should change reciprocally, f be doubled in this example. Shutter time and aperture should therefore be regarded as pairs. The longer the exposure time (and the longer the focal length of the lens, or the shorter the distance between object and camera when working with a macro lens), the greater the chance that your pictures will be unsharp due to camera shake. Apart from antishake in the camera body (Konica-Minolta) or image stabilisation in the lens, this calls for a sturdy tripod and shutter release via a cable. When taking photographs of shiny objects, the influence of camera shake is best diminished by lighting which is sufficiently bright to allow selecting an appropriate f-stop and sufficiently short shutter speed.
In various circumstances automatic exposure is less than ideal. This is particularly the case when the subject is very bright (high key) or predominantly dark (low key). A bright suibject is for example silver on a white cloth, a low key subject is for example an enamel piece with predominantly dark tones. In the first case automatic exposure renders the white to a medium grey, in the second case all colours will be systematically too bright. By understanding what an exposure meter was designed for, it is easy to use it appropriately in the above 'non-standard' conditions; this entails turning automatic exposure off, and/or manually off-setting the light sensitivity of the recording material (film or electronics).
In autoexposure mode the subject is automatically rendered as medium tone. This works well for the everyday snapshot of a landscape, family group and the like, where there is a spectrum of bright and dark tones, and assuming that the average is medium tone is quite appropriate. Most films and CCDs can reproduce a range of +2½ and -2½ f-stops from the medium tone. The range of light intensities in real life is often larger, and this will cause the darkest and/or the highest tones to fall outside the scale. However, except in unusual circumstances the range available to the photographer usually suffises to render the subject satisfactorily. Positioning the medium tones correctly is very important; if the medium tone is rendered too light, all tones will be too light. Conversely, rendering the medium tone too dark will make the whole picture look underexposed.
As long as we keep in mind that the exposure meter is designed to produce the correct exposure setting for medium tone, it is an easy matter to handle non-standard situations. To that end the photographer should familiarise himself with the zone system, developed to perfection by Ansel Adams for black-and-white photography. The bottom line of the system is to meter the brightest parts and the darkest part of a scene, which tells you whether the subject fits within the 5 f-stops alluded to above. In studio situations this tells us whether we should adjust the lighting ratio. Meter what you would like to render as medium tone, and all other tones will automatically fall in place. Be careful not to overblow the lightest parts, because subtle changes in those brightest will be lost: anything that falls outside the upper portion of the scale will be rendered white. You will enjoy the concise and excellent explanation of Ansel Adam's zone system on Norman Koren's website; he also explains how to apply it to colour photography.
Let us look at two examples. The first is a high key subject: a glass bottle filled with white powder, standing on white paper. The exposure meter will indicate an exposure setting which renders the subject to medium tone, f. the image will be underexposed. Obviously we will have to shift the tonal range upwards. From medium tone to the brightest white encompasses 2½ f-stops. Therefore if for the same shutter time we open up 2 f-stops there is still some leeway and the whites will not be overblown. If, however, we do that this means loss of depth of field. The alternative, then, is to increase the shutter time by a factor 4, 2 for each f-stop.
The second example is a low key subject. The dark subject will be rendered medium tone by the exposure meter. We should therefore shift the tonal range down. In the example illustrated below the subject contains pure black. If we wish to play safe shift the scene down 2 f-stops. Again, as we want to maintain the same depth of field, and hence do not wish to change the aperture, we must diminish shutter time by a factor 2 for each f-stop, and reduce it to ¼ of the time indicated by the light meter.
|The inserts illustrate the level with which darkest to the whitest tones in the image are represented. Left = black, right = pure white. The higher the peak, the greater the portion of the image represented by these tones. The graphs illustrate the results of tonal shifting.|
|Bottle with white enamel powder on white paper. Autoexposure, aperture priority resulted in 0.1 s shutter speed at f 10.||The shutter time was increased by an amount equivalent to opening the f-stop by a factor 1.3. This resulted in f 10 and 0.25 s exposure.||The shutter time was increased by an amount equivalent to opening the f-stop by a factor 2. This resulted in f 10 and 0.4 s exposure.|
|A dark brown statuette placed against a black background. Automatic exposure setting at f 10 resulted in 1.3 s shutter opening.
The image is severely overexposed.
|Same scene. Autoexposure setting overridden by manually settting a -1 stop adjustment (suggesting that the film sensitivity is twice the actual value, e.g. 200 ASA instead of 100 ASA).
Shutter time 0.5 s, f 10.
|Same scene. Autoexposure setting overridden by manually settting a -2 stop adjustment (suggesting that the film sensitivity is four times the actual value, e.g. 400 ASA instead of 100 ASA).
Shutter time 0.25 s, f 10.
It takes some practice to estimate how much exposure correction is required. If in doubt, use your camera's spotmeter and aim it at what you want rendered medium tone, and use the exposure setting your meter suggests. You can also take a number of pictures, varying the exposure setting. Many cameras provide facilities for bracketing. Another solution might be the use of a neutral test card with 18% reflectance, for sale in photo shops. The test card must be held straight up and down close to and in front of the subject, and facing halfway between camera and main light. A light meter reading should be made solely of the test card, and this will tell you the exposure setting for middle tone. It is generally recommended to use the card only for indoor scenes.
Generally 'expose to the right'. The reason for this is as follows. Let the span, expressed in f-stops, that a digital camera can record, be 5 f-stops, and let us assume that the output is an 8 bit image. The CCD is a photon counter, which implies that there is a linear relationship between its output and the number of photons ('the amount of light') hitting it. This means that all the information is contained in 28 different states or shades, i.e. 256 different states. From the widest lens opening to one stop down implies the amount of light hitting the CCD is halved. This one step change in lens opening takes one half of the available states, i.e. 128. The next change in f-stop takes half of what is left, i.e. 64 states. The third stop down takes 32, the fourth stop 16, and the last step is represented by only 8 states. Hence the darkest parts of the image are poorly represented, and by far the greatest part of the tonal information is contained in the brightest tones. Therefore, in order to preserve the most information in the darker parts, when not dealing with a 'low key' subject, don't waste the space available in the brightest tones, but beware of blowing out the highlights. Use the histogram that your camera shows when displaying the image, and adjust your exposure if needed. Magnify portions of the image that matter to make sure that highlights are not blown out. In many cameras areas of the image approaching or exceeding the shadow and highlight luminance are indicated by those areas flashing in the thumbnail on screen.
|Top: incident light measurement, the opalescent diffuser in front of the meter opening. Below: measurement of reflected light, the diffuser is pushed away from the meter opening.|
Reflected and incident light
Exposure meters in cameras measure the amount of reflected light. Dark objects absorb most of the light and reflect little, whereas bright objects reflect most of the incident light. The exposure meter aimed at the dark and the bright part of the scenere, respectively, will give very different readings, even though the amount of light emitted by the light source is the same. It might help to know how much light is emitted and not worry about how much is reflected. Incident light might in principle be measured by turning the camera 180°, facing away from the subject. This is not very practical, and we would still have to ascertain that the exposure meter measures the average lighting. In practice it is much better to use a separate exposure meter which is exposed to light via a diffuser; the latter is made of opalescent material that averages the incident light. Whilst this is a very good solution for our purposes where we create studio conditions, like the neutral test card this is not a panacea for field situations.
Flashlights are very useful replacements of tungsten lights; they emit very much white light. The considerations for lighting conditions are the same for any source of light. If you are lucky you can use studio flashlight. These are mounted on tripods and can be activated from the camera. To that end one can use a direct cable and/or a slave unit which detects a flash and without delay ignites another flash unit. Studio flashes have modeling light bulbs which allow you to preview the lighting. If you have araanged everything to your satisfaction one ignites the studio flashlight without activating the camera shutter and measures the incident illumination with a flashmeter aimed at the camera. Via manual control one then adjusts the f-stop of the camera, using the exposure setting recommended by the camera manufacturer for external flashlight. Alternatively one uses a set of electronic flashlights designed for on-camera use; see e.g. .