Art 100

Introduction to Digital Photography

The Digital Camera I: Exposure

Basic Control of the Camera

The degree to which you are able to control your camera will depend, of course, on what kind of camera you have. In these notes, I will offer a basic description of an SLR digital camera (closest to the operation of a 35 mm SLR camera that would normally be required equipment in a photography class) but I will add comments when the treatment of a digital point-and-shoot will be different. Some point-and-shoot cameras, of course, are virtually automatic and offer you no control.

Basic Adjustments. All films have different sensitivities to light and color films have different sensitivities to the chromatic characteristics of light (so-called “color temperature”). The sensitivity to light intensity itself is measured in a standardized system (ISO); film ISOs usually run from 50 (rather insensitive) to 400 and above (very sensitive). Digital cameras allow you to select an ISO value; what you are actually doing is selecting the electronic sensitivity of the CCD sensors --- rather like the “gain” on an amplifier. High ISO films are “grainy” and highly amplified electronic signals are “noisy”. Thus, noise is the digital equivalent of grain, though it is not quite that simple. To see noise in a digital image, closely examine larger areas which should have little color variation; digital noise will appear as variable color blotches.

Since you usually want to create a nice sharp image, you would usually select the smallest ISO setting possible. Unfortunately, when you try to take pictures in darker areas, you quickly find yourself caught in between a number of options all of which have their down sides. Opening the lens aperture reduces the depth of field, and slowing down the shutter will blur moving objects. How you choose between these options will depend on how all of these factors weigh into the artistic idea that you are pursuing. [Generally speaking, you are well to leave the ISO setting as low as possible.]

The color temperature is a measure of how light intensity is represented across the color spectrum. “Cold light” will have a lot of blue but much less red. As the color temperature increases, the middle colors of the spectrum increase in intensity. Really “hot light” is very red, almost pink. Different lighting situations, both natural and human-made, represent different color temperatures. So color films were developed for daylight, incandescent light, and fluorescent light. These compensate for the different spectral characteristics. Similarly, digital cameras allow you to change the “white balance” to compensate. Usually, you can either allow the camera itself to make this judgment or you can set the white balance to the conditions that you think apply. Digital cameras have two advantages over film cameras in this respect --- first, they offer a larger number of color temperature options and, second, you can re-set the white balance to appropriate values for individual images on a given memory card rather than selecting a whole roll of film with one color temperature sensitivity. [The latter advantage applies to the ISO setting as well.] One further advantage with raw format images is that you can change the color temperature sensitivity for a single image in your capture software after the picture has been taken.

Other basic settings on the digital camera deal with focusing, framing, and light metering. Most cameras offer the option of manual focus even though virtually all contemporary cameras offer automatic focus as the default focusing method. Manual focusing may be essential in situations where the image is too complex for the auto-focus technology to choose correctly. (An example is where you want to take a picture of the rocks in the bottom of a stream though the auto focus will try to focus on the water surface.) The default setting for framing is single framing --- one frame for each depression of the shutter release. However, there are situations (like sporting events) where multiple frames should be taken so long as the shutter release remains depressed. Another option on some cameras is “bracketing.” This automates a classic technique used by film photographers for years --- taking three frames for each image, one at the exposure selected, one slightly underexposed, and one slightly overexposed (see the next section). Finally, all modern cameras have internal light metering devices. The best cameras even offer you a selection of which metering pattern can be used. Most high-end cameras now offer a matrix metering pattern that divides the image into a number of sections and weights the light measured in different sections differently in order to allow “the best exposure” possible. [This means that a bright reflection from a mirror in a corner of the frame will probably not have much impact on the exposure selected.] A simpler pattern is “center weighting” which simply gives more importance to light coming from the central area of the image frame. Finally, a “spot meter” measures only the light in a small central area of the image frame. This is crucial if you are trying to apply the Zone System to your work or in situations where you want to exactly expose some small part of your picture (like a face) and are willing to let other parts of the image fall where they may. (see the next section.)

Finally, before you start shooting pictures, you need to select the size and quality of the images that you want your camera to save. Most digital cameras will save image sizes small, medium, and large; this corresponds to how many pixels will be saved along each axis of the CCD rectangle. The quality adjustment corresponds to the file format in which the pixel information will be saved. A tiff file retains all of the pixel data collected. A jpeg file compresses the pixel information to different degrees, usually called basic, normal, and fine, basic being the most compressed and fine the least compressed. When compression is executed, the software compares neighboring pixels and decides when they are “essentially the same,” storing only one of them when it determines that the rest are the same. The stronger the compression, the less critical the software is about sameness! In other words, when a file is compressed, data is thrown away under the assumption that it is redundant.

You must select both a size and a quality for your pictures, though you can change this selection within a given memory chip if you see a picture that you want to record differently (another advantage of digital cameras). But what does all of this matter? With a Nikon D100, choosing a large tiff file means recording each image as a 17.5 mgb file. If the memory medium (Compact Flash, in this case) is 128 mgb, you can only take 7 pictures! On the other hand, if you choose a large fine jpeg file, the image will be 3.5 mgb and you can fit 38 pictures on your memory chip. Obviously, you can fit even more pictures on your memory chip if you choose a small basic jpeg file. Why not just go for it, then? Well, it depends on what quality prints you want. Large high-quality prints require lots of pixels and come from large image files. [In this discussion, raw files are pretty much the same as tiff files, in terms of quality, but actually take up less room --- about half the space of a tiff file.]

Exposure. The basic concept of “exposure” is to admit enough light to the camera but not too much. Without enough, the image is underexposed; with too much, it is overexposed. But what does that mean and does it mean the same thing for digital as for film cameras.

When black&white film is underexposed, not very much of the silver halide is activated and, in normal development time, very little metallic silver is deposited in the negative emulsion. As a consequence, when you attempt to print the picture, the negative allows a great deal of light through it and the resulting print is very dark. Try as you might to compensate for this in the darkroom, the negative simply does not have enough “data” (microcrystalline silver deposits) on it to provide for a picture of normal contrasts between blacks and whites, with intermediate grays. The situation in overexposure is similar except, in that case, too much silver overall is deposited in the negative and the print is very white.

The basic rule of exposure in black&white photography is to set your exposure by the darker shadows in the image. Black&white film still offers some texture detail even when it is under-exposed by two "stops" (see aperture settings below). Under-exposure by more than two stops will produce solid black shadows with no detail. A well exposed black&white photo will have elements that are overexposed by up two stops as well as others that are underexposed by up to two stops, as well as everything in between. (see discussion of the Zone System below.)

In digital photography, the dangers are reversed. Since the photo-sensors are like little buckets that you pour photons into, the danger is that they will all get filled up and look identical to each other. The danger, in other words, is overexposure. Rather than looking critically at the shadows, you need to look at the highlights in an image and make sure that the highlights are not overexposed. Anything that is overexposed will appear uninterestingly white (burned out) in the final print.

When the camera is turned on and you have framed a subject in the viewfinder, you can measure the exposure by depressing the shutter release half way if you have set the camera in "manual" shooting mode. In other shooting modes, the camera will give you very little exposure information. With manual shooting mode selected, you are in complete control and the camera gives you an exposure chart in the viewfinder. The chart usually ranges from about two stops underexposed to two stops overexposed, with a central line marking the optimum exposure. The exposure being read is indicated by vertical bars to the left or right of the central line. If there are no bars, the exposure is "optimum" in terms of the camera's shooting system. You can watched the measured exposure change as you use the camera's adjustment system to change the aperture setting or the shutter speed or the ISO. If you use the camera’s spot metering system, you can even measure the exposure at small parts of the picture. In particular, aiming the spot meter at highlight areas, you can make sure that you do not overexpose the highlights. When you have made these adjustments and reached a satisfactory exposure reading, you can depress the shutter release the remainder of the way. [On most cameras you can simply turn on the metering system by depressing the shutter release momentarily. When the camera is fully adjusted, you can then press the shutter release again (fully).]

If you have chosen “automatic” or “programmed” as your shooting mode, the camera will not give you any exposure data unless it is simply impossible to take the picture under these circumstances, in which case it simply refuses to operate. [The new Nikon D70 will even pop up the built-in flash and use it if it has to.] Under automatic shooting, the camera is measuring the exposure and resetting the aperture and shutter speed to achieve what it is programmed to judge as an "optimum" picture. Equally well, if you have chosen “aperture select” or “shutter select” modes of shooting, the camera will not give you exposure data but will select whatever other settings it needs in order to complete the picture. In "aperture select" mode, for instance, it leaves the aperture as you set it but changes the shutter speed until it judges the exposure as optimum. These can all be useful settings, but it is very important to understand what the camera is doing. [If you set too small an aperture for the lighting situation, for instance, you will force the camera to choose an extremely slow shutter speed that will blur the resulting picture.]

Understanding Exposure Issues. The exposure problem is just the beginning of course. The next question is what you are doing to your picture when you open the aperture or slow down the shutter speed or set a higher ISO. Once you understand these variables, you will see that letting your camera do all of the work in automatic mode is making you miss most of the creative opportunities in photography. Basically, this is because the programmed automatic mode is a very conservative exposure combination that sets a fast shutter speed to avoid blurring due to hand motion and a wide open aperture to allow enough light into the camera.

What is the aperture and how does its adjustment affect the image? Every simple lens has a fixed effective diameter through which light is admitted to the camera interior. The simple lens also has a fixed focal length --- the distance from the lens plane to the plane where infinitely distant objects focus. The aperture of a simple lens is the diameter of the lens divided by the focal length. This is noted as a fraction "f/A" where "A" is the aperture. Thus, an 80 mm lens has an aperture of f/4 if the diameter of the lens opening is one fourth of the focal length or 20 mm. [20mm/80mm = 1/4] Notice that the descriptive number for aperture is in the denominator of a fraction so that a larger number means a smaller diameter of lens opening and less transmitted light. Notice also that in order to build an 80 mm lens that allows more light into the camera, say an f/2 lens, you would have to increase the diameter to 40 mm. This is called a "faster" lens because it admits more light, but it requires significant increases in size, weight, and cost.

A modern compound lens has a diaphram mounted inside. The diaphram is constructed with interlacing leaves that can close to a very small diameter; thus, the effective lens diameter can be changed. The classic 50 mm lens mounted on 35mm cameras can usually close the diaphram down to an effective diameter of about 2.3 mm for an aperture of f/22. The diaphram can be opened to the full diameter of the lens so the wide-open aperture depends on the size of the lens --- f/2 requiring a 25 mm diameter. [Next time you watch a baseball game on TV, watch the huge lenses that sports photographers are using from the sidelines. Since they have to use long-focal-length lenses in order to "get close" to their subjects, they have to have very wide lenses in order to keep the aperture reasonable and admit enough light, especially since their subjects are moving and the shutter speed must be fast. A 600 mm focal-length telephoto lens could be opened up to an aperture of f/6 if it was 100 mm in diameter. Such lenses cost thousands of dollars.]

If you are up for a little paper&pencil experiment, draw the following diagram. Imagine that you have two lenses, one with a focal length of 40 mm and one with a focal length of 80 mm. Further, imagine that both lenses have the same aperture, say, f/2. This means the diameter of the former lens is 20 mm and of the latter is 40 mm. Draw the two lenses on top of each other on the left side of the page with the smaller lens in the middle of the larger lens. Draw rays of light from top and bottom of the smaller lens to its focal plane 40 mm to the right. Then draw rays of light from top and bottom of the larger lens to its focal length 80 mm to the right. Note that the rays are parallel. This is an attribute of aperture and it means that neither lens, so long as they have the same aperture, intensifies the light image more than the other. No matter what the focal length of a lens, it will admit the same light intensity as any other lens when set at the same aperture. This is a very important property when it comes to measuring exposures.

On standard compound lenses with variable apertures, the diaphram is set up to move open or closed in increments rather than moving continuously. These are called "stops." Each stop closing down cuts the amount of transmitted light in half. The stops are designated by the calculated aperture for the effective diameter. Thus, for an f/4 lens the first stop is at f/5.6 which means that the effective diameter has been reduced enough to cut the transmitted light in half. [As above, if the lens has a focal length of 80 mm, its wide open diameter is 20 mm. At f/5.6 its effective diameter is 14.3 mm. Since the area of a circle is proportional to the square of the diameter, we can note that the square of 20 is 400 and that the square of 14.3 is approximately 200, being half as great.] You must keep in mind that aperture is a fraction so that larger numbers (denominators) mean smaller actual size. When we talk about closing a lens, we are making the effective lens diameter smaller, allowing less light through, and this means moving to nominally larger lens stops.

But what does it matter? Since a lens is a big chunk of glass that is re-directing the rays of light, the larger the amount of glass through which the light passes, the larger are the strange aberrations that will affect the image. Some of these are chromatic and others are focusing issues. While large wide-open lenses are fun to take pictures with, they usually produce images of reduced quality. You can limit aberrations by using less of the lens or a higher aperture value, say, f/22. [Weston, Adams, and other West-Coast photographers of the '30s formed a club called "Group f/64" signifying their interest in extremely high grade images.]

For our purposes, the chief effect of varying aperture is called "depth of field." The camera’s lens focuses an image on the film plain or the CCD surface at a constant distance from the back of the lens. In order to do this, the focus ring on the compound lens moves portions of the lens relative to each other. Moving the lens’s focus ring manually, you will see that you can make objects close by focus and then move the plain of focus toward infinity. In other words, if you imagine the three-dimensional space in front of the camera lens as a cone that reaches toward infinity, the focused image is a circular cross section of this cone at any arbitrary distance along the axis of this cone, from very close to infinity. Now, for the final wrinkle, the eye always perceives more than just this cross sectional plain as being in focus; in fact, the lens apparently focuses all objects within a volume of space equal to this section by some width along the axis. This “width” of apparent focus is called “depth of field.”

The depth of field changes with the aperture setting of the lens because it is a function of how much glass the light passes through. In particular, at a large aperture value (very small lens opening), say f/22, you have a large depth of field and it may be that everything from, say, 10 feet off to infinity appears to be in sharp focus. With a wide open lens, say f/2.5, the depth of field may be so narrow that focusing on a person’s nose will leave the ears somewhat out of focus. You can begin to see why it is not always in your best interest to let the camera automatically set the aperture.

What is the shutter speed and how does its adjustment affect the image? In an SLR camera, the light path to the film plain or the CCD is closed by a shutter most of the time. The "shutter release" is just that, a mechanism that momentarily opens the shutter so that the focused image falls on the CCD. Since the length of time the shutter opens is usually very small (fractions of a second) shutter designs are very technical, allowing equal exposure of the image across the CCD. One of the biggest differences between SLR digital cameras and point-and-shoots lies here. The point-and-shoots usually do not have a mechanical shutter and, in fact, the focused image falls on the CCD so long as the camera is turned on. The "shutter release" on a point-and-shoot is simply a switch that signals the camera's computer to save a copy of the current contents of the CCD. This creates a lag time between the photographer's desire to "snap the picture" and the computer's response. So, occasionally you will find that you failed to capture what you had hoped for. [You can see why some point-and-shoots merge almostly seamlessly with video cameras.]

At shutter speed 1/60 of a second, most people can hold a camera sufficiently still so that hand movement does not produce noticeable blurring of the image. From this central point, shutter speeds are graduated into "stops" just like apertures and each stop either doubles the amount of light or divides it in half. A shutter speed of 1/30 admits twice as much light while 1/125 admits half. This calibration is very convenient since it means that aperture and shutter speed settings work with each other. If an image is optimally exposed at 1/60 and f/5.6 but you are worried about blurring, you can speed up the shutter by one stop --- to 1/125 --- and still maintain the exposure by opening the lens one stop --- f/4.

Blurring is the chief issue with shutter speed settings. At 1/60 and higher speeds there is no problem with blurring images unless objects in the image volume are moving. At shutter speeds slower than 1/60 the camera should be placed on a tripod and the shutter should be released by cable to minimize the danger of hand movement. If the image does include moving objects, you need to decide whether to "express" their motion by allowing blurring or whether to capture the moving object "frozen" in transit. Speeding up the shutter to, say, 1/125 will often avoid blurring in normal situations but will probably not freeze a track runner or a racing car. Similarly, at 1/250 the camera will probably freeze the droplets of water in a water fall or fountain, but you may want to slow the shutter down a stop and express the water's drop by blurring it a little bit. Obviously, it requires some experimentation to determine the right settings for the effect you want.

Finally, lets consider the ISO and noise. ISO ratings are incremental in light sensitivity by a factor of 1/3 of an aperture stop. Thus, the standard ISO scale is 64, 80, 100, 125, 160, 200, 250, 320, 400, . . . If you replaced ISO 64 with ISO 125, which is three steps more sensitive, you would be able to close the lens aperture by one stop, reducing the incident light. If you are taking pictures in a dark place, you are forced to increase the ISO sensitivity so that you can maintain a more-or-less normal combination of aperture and shutter speed and still achieve a reasonable exposure. With films this produces grainy pictures simply because more light-sensitive emulsions do produce larger grains of precipitated silver. The apparent affect is a softer focus of objects and it is not always contrary to one's artistic wishes. With digital cameras the higher ISO sensitivity produces more electronic noise or scatter of color/intensity values from pixel to pixel. This effect is almost always contrary to artistic interests and may appear downright unpleasant. As a consequence, there are growing numbers of computer programs that will help you try to clean noise out of digital images. Indeed, most digital cameras have noise filtering already built in.

The Trade-Offs

Let's assume that you want to take a picture in an outside setting with fairly strong sunlight. You should be able to set the ISO on the lowest value provided and the white balance on either auto or sunlight. You should set the camera on manual operation if that is possible --- even manual focus if that is possible. Select a subject that has both color interest and contrasting brightness. Before pressing the shutter release you need to adjust aperture and shutter speed (and possibly focus) in order to create an image with acceptable exposure.

From this basic starting point, you need to consider what qualities of the image are desirable. If you want a large part of the image to be focused no matter at what distance from your camera, then you will have to set the aperture at a large f-stop (small lens opening). Since you are reducing the admitted light by one half for every stop, you will have to admit more light by lengthing the time the shutter remains open. When you pass 1/60, however, you need to put the camera on a tripod. Spontaneity is out the window! And the scene better be still. If you didn't bring your tripod [You do have a tripod don't you.] you will have to stop at 1/60 and make up your exposure by increasing the ISO. But now you have more noise. All fine art photography is a matter of struggling with trade-offs like these.

Neither films nor CCDs have the amazing tolerance for contrasts possessed by the human eye. A scene may look perfectly normal to you but the print or image produced may be quite unpleasant. Skylight is one of the worst offenders here. The normal exposure of skylight is dictated by the rule of "Sunny 16". The proper exposure is f/16 and 1/(ISO) shutter speed. In other words, if you have set your camera with ISO 125 and you have a bright sky in the picture, you can set your aperture at f/16 and your shutter speed at 1/125 of a second and expect to expose the sky appropriately. The problem is that most scenes include principal subjects that are much darker than the skylight and your eye does not detect this. If you use the camera's spot meter and check exposures at different parts of the scene, you will begin to see the problem. Just exposing a lawn or tree may require a couple additional stops of light and dealing with shade will push this to an extreme. In fact, setting the exposure to correctly capture people and other objects will probably leave the skylight entirely burned out (overexposed) --- something you might not want to do if there are interesting clouds in the sky. In many situations, however, the best solution is to avoid strong contrasts altogether, perhaps cropping the sky out of your image completely.

Even computers need time. You can set a film camera on rapid multiple framing and hold the shutter release down, listening to the grand sound of your SLR clicking away at rapid fire. The result is limited only by the speed with which your camera can crank the film and the length of your roll of film. Not so with a digital camera. When a digital camera takes a picture, the data is immediately moved to a buffer where the computer begins to work on converting it to a saveable file. Once the data is in the buffer (somewhat comparable to the cranking speed of film), the camera can take another picture. But this has its limits --- in particular, the size of the buffer. When the buffer is full of data being processed, the camera will shut down. You can get more frames per second only by buying a better camera with a larger buffer and faster processing speed or by setting your file saving to smaller files.

[One final note. Precisely because of the buffer processing, noted above, you must not turn your camera off right after taking a picture --- something that film photographers get used to doing in order to save battery life. You will lose all data in the buffer which remains unsaved.]