Photography deals with two aspects of capturing light, art and craft. The artist prowls the world looking for material, sees an interesting subject in a special light, and imagines the best effect that subject has to offer. The craft practitioner gets that light onto the pixels of a sensor so that the imagined effect is realized in a permanent image. The following discusses photographic craft. Photographic art is both the more difficult and more subjective, requiring intuition, effort, a good eye, experience, and luck.
Current cameras require some exacting user setup to achieve a great image. But such images are still limited by the physics of both sensor and lens operation. A more capable camera that conveniently overcomes these limitations may be just around the corner, as explained at the end.
The camera/lens system controls the two most significant aspects of the imaging process: exposure and focus.
The digital camera manages three fundamental parameters related to light exposure: aperture, shutter speed, and light sensitivity. All involve light and time. Photography would be mechanically complex if there were not a simple relation between these three parameters.
Relations between physical properties are best illustrated by proper choice of measuring units. Hence, a special photographic unit called a ‘stop’ came to being (a term relating to the functioning of the earliest cameras). A stop is a relative metric of the intensity of light and the camera’s response to it. It defines a geometric scale, where adding a stop doubles the intensity of light.
Shutter speed has a range from seconds to thousandths of a second. Camera sensitivity is measured in ISO units, usually from 50 to the thousands. These parameters are assigned scales whose units correspond to stops; as expected, each scale value is a doubling or halving of the previous. Thus shutter speed (secs): 1, 1/2, 1/4, 1/8, 1/15, 1/30, 1/60, 1/125, 1/250, …; ISO sensitivity: 50, 100, 200, 400, 800, 1600, 3200, 6400, …
Aperture is the adjustable size of the ~circular opening through the lens that admits light. Aperture is measured in f/stops, the ratio of the aperture area divided by the lens focal length. It is a ratio because the actual amount of light per unit time striking the sensor is proportional to aperture and inversely proportional to lens focal length (lens focused at infinity). To avoid having a separate aperture scale for each lens focal length, a ratio is specified whose scale values represent doubling or halving of admitted light (independent of focal length): f/1, f/1.4, f/2, f/2.8, f/5.6, f/8, f/11, f/16, f/22, f/32. Moving to smaller apertures is called ‘stopping down’ the lens. Moving to larger apertures is called ‘opening up’.
Note that there is light lost to refraction and reflection within the many glass elements of sophisticated lenses, and that the light loss is greatest at very wide apertures. The motion picture industry observed this and replaced the f/stop system with stops corresponding to actual light transmission (T/stops). The digital camera manufacturers stick with f/stops, but internally adjust the sensor sensitivity (ISO) upward slightly at the widest apertures (f/2 and wider) to accommodate for inaccuracy of the f/stop scale at that end.
Every skilled photographer is well versed in these three scales, whose standard scale values correspond to relative stops. Beyond the scales themselves,understanding must extend to how these scales combine and relate to absolute exposure. The camera’s internal exposure meter, or a separate incident light meter, measures the light from/onto a subject and then advises about available settings for correct exposure. Many settings will achieve the same absolute exposure. The craft of photography involves selecting the right combination for the desired effect in the final captured image.
Heuristically, the sunny 16 rule anchors the scales to a standard daylight exposure without resorting to metering. On a sunny mid-day more than 2 hours after dawn or before dusk, set the aperture to f/16 and set the shutter to the closest speed corresponding to the inverse of the camera’s ISO scale setting. Then a ‘correct’ exposure will obtain. For example, if ISO 200 is the camera setting, then use a 1/250 shutter speed with f/16 aperture to take the shot. The aperture and shutter speed values are not discrete selections on most cameras, so that intermediate values can be selected.
The great convenience of stop increments on each scale is the ease with which changes to the three parameters can me made without affecting correct exposure. In the above example, say subject motion necessitates increasing the shutter speed one stop to 1/500s (allowing only half as much light from lens into camera). To compensate, one can either increase aperture one stop to f/11 (providing twice as much light through the lens) or increase sensor sensitivity by one stop to ISO 400 (making the camera respond similarly with only half the light). What could be easier?
Not all exposure settings are created equal. The smallest camera ISO settings provide best quality, because image noise increases with increased light sensitivity. The mid-ranges of the aperture scale are generally best. Wide-open lenses (lowest f/stop) will sometimes not provide a sharp image corner-to-corner. And more importantly, depth of field (see under ‘focus’ below) decreases as aperture increases (sometimes desirable). At the other scale end, small apertures (any f/stop over say f/11) may suffer from diffraction effects, lessening the acuity of the image (particularly true when close-focusing with extension). High shutter speeds freeze motion; slow shutter speeds blur motion. For example, a waterfall with a fast shutter will show individual drops of water; with a slow shutter, it is given an added dynamic, the appearance of flow.
One changes the camera’s ISO value least often; it is usually set at its lowest value, 100 or 200, for daylight shooting. If overcast, shooting in shadows, shooting fast action at high shutter speeds, or taking close-up shots where depth of field becomes critical, one may then advance to a larger ISO to enable more successful exposures with less light and/or a stopped-down lens. Once the ISO parameter is set, cameras can provide fully automated exposure, where the camera makes all the exposure parameter decisions. Nikon calls this Programmed Auto; point and shoot. This is convenient and usually entirely safe, but it is not the photographer who controls the image, it is the camera. As one grows into the art of photography, by imagining possibilities to exploit in a subject, it will be necessary to improve one’s craft by setting aside the automated safety and experimenting with the camera controls.
Sophisticated cameras allow user control of parameters, including shutter priority or aperture priority operating modes; the user selects the priority parameter and the camera adjusts the other accordingly. And then there is usually a full manual mode where the user selects all the parameters, based on special lighting effects one hopes to convey. Some digital cameras also are able to automatically vary ISO setting in select circumstances, which permits the camera to capture an image that otherwise would not be possible with the other two selected parameters.
One may rely on the camera exposure meter’s readings as base values for the exposure parameters and then modify them for the effect to be achieved. Subject tonality may often require exposure compensation. The meter is calibrated to measure the light reflected from a medium toned object (most objects have medium tonality). What if one’s subject is dark toned? Then using the camera’s readings, the subject will appear medium toned in the resulting image, resulting in general overexposure. To achieve a dark tone, one must cause the camera to underexpose from its meter readings. The opposite is true for a subject with light tone.
The amount of exposure change depends on the tonality. The camera’s total sensor bandwidth can be eight stops or more, depending on sensor quality. Thus there may be up to four stops from medium tonality to featureless white, and in the other direction, four stops from medium tone to featureless black. Where the subject’s tonality lies in this range will determine how much compensation is required, usually 1 stop or less in either direction. If one does not trust one’s judgement here, and there is a medium-toned object nearby in the same light, just meter that and apply those parameters to the subject image. Going even further, one may not necessarily want a literal reproduction of the subject’s tone. By modifying the camera’s exposure parameters, one can create any subject tonality.
Focus is achieved by movement of the camera’s lens elements toward or away from the camera focal plane. It is sometimes done by manually moving the focus ring on the lens (manual focus, MF). But more often today, the lens elements are moved automatically by a motor either in the camera or in the lens itself (auto focus, AF). Focus is the aspect of an image that one must perfect prior to shutter activation. While small exposure glitches can be easily compensated in post processing, a fuzzy image is always a fuzzy image, a throw-away.
Perfect focus occurs mathematically in a single plane perpendicular to the light path through the lens. But there is no perfect focus in reality, for a variety of reasons. Even a mathematically perfect lens will still map a point of light into a small disk, called the Airy disc, due to the diffraction of light through the lens aperture. Adding to diffraction effects, the optics (sometimes more than ten internal glass elements) of the best lenses will only be able to collimate a point source of light into an area of a small disk, sometimes called the circle of confusion (CoC). For a professional quality lens, this CoC is still too small for the human eye to resolve; it appears to us as a point. Which brings us to a final parameter of apparent focus quality, the acuity of the human eye. Any spot on an image smaller than ~0.2mm will appear to us as a point. This explains why most lenses will appear perfectly sharp to us; their CoC is smaller than our eye’s resolving capability.
Our eye’s imperfect resolving power permits a lens’s focusing mechanism a degree of latitude in producing images in apparent perfect focus. When a subject appears in focus, there will be some distance, before and behind the subject, that will also appear to the human eye to be in perfect focus. This ‘perfect focus’ band corresponds to the focus distance error within which the lens’s CoC is smaller than our eye’s resolving power. This distance is called the apparent depth of field (DoF). With enlargement of the image beyond the actual image size, the DoF may decrease because we can now resolve areas that were too small before (unless we also proportionately increase viewing distance from the enlargement).
When a lens is set at infinity focus, apparent DoF will extend from some near distance all the way to the farthest objects in the image. This distance beyond which all objects seem to be in focus is called the hyperfocal distance. It depends on focal length, aperture, and diameter of the CoC. At infinity focus, subjects closer than the hyperfocal distance will appear less well focused.
Achieving the desired DoF in an image requires balancing distance to subject, shutter speed, and aperture. DoF is inversely correlated with magnification; DoF is increased by smaller aperture, longer distance to subject, and smaller focal length (so long as you don’t move closer to maintain the same image size). By setting focus distance at the hyperfocal distance (HFD) from the camera’s focal plane (FP), maximum DoF (indicated by * below) is ensured from a near focus distance (NF=HFD/2) to infinity.
FP ————– NF *************** HFD ********************************** Infinity
Focusing at HFD is primarily useful in landscape photography. It ensures maximum DoF. Any other focus distance will throw away some useful DoF, either at the NF end or at infinity.
Build a Better Camera
In current cameras, best possible exposure and focus are limited by the user’s ability to control numerous factors related to the camera and its distance to the subject, by the camera sensor’s light bandwidth (range of light to dark shades that can be ‘seen’), and by the lens’s focusing limitations (limited DoF). To get better results and eliminate the ‘luck’ factor, one can manually take several exposures in sequence with varied focus and exposure settings (an image stack), and then either pick the best one, or use software in the post process to average the focus and/or exposure values across the stack, producing a composite ‘perfect image’. This takes considerable experience, skill, and investment to achieve consistent improvement, particularly with hand-held photos. And the process must be done multiple times, once for focus and once for exposure averaging (and perhaps once for sensor noise removal if that becomes possible). But what if the camera were smart enough to do this for us with a single shutter activation? Photographic Nirvana!
Cameras are already programmable devices with internal computers. Taking this further, Professor Mark Levoy of Stanford built a Frankencamera, a conglomeration of various camera and phone bits integrated with a computer on a chip that runs Linux. ‘Frank’ takes several exposures in a row, then averages focus and exposure to create the images we all want to achieve, allowing us to simultaneously exceed the sensor response bandwidth in exposure extremes, and to expand the DoF. It is also possible to re-focus images after the fact.
Frank began as a four pound monstrosity. Now, the programming has been squeezed into a Nokia N900 cameraphone that also runs Linux. Computational photographers with EE degrees can now build a similar camera themselves. The rest of us will have to wait until a major camera company grabs the bait.
Merge Still and Video Imagery in One Camera
Current digital still cameras have HD video capability as well. One does not have to look too far into the future to imagine a product that could extract still images from video that would rival image quality of those taken by the best still cameras today. As storage costs come down, shooting video with ability for extracting high quality still images will be affordable. And the ability to get just the right picture, rather than one a second too soon or too late, will make all the difference for most photographers. For after all basics are mastered, we still need to capture the exact moment of maximum expression from the subject in order to feel really good about the result. Such a camera will nearly eliminate luck as a factor in the image outcome.