In 2011 I started corresponding about photography with a friend of mine, a fellow EE student when we were undergraduates, for whom my pictures had been an inspiration to pick up a camera for himself. This is a collated and edited version of some things I sent him concerning the basics of exposure. The occasional use of technical scientific terms betrays their origin as letters to an engineer, who could reasonably be expected to understand the jargon at first sight; I ask for the indulgence of non-technical readers in this regard.
Exposure is the total amount of light that reaches film or an electronic sensor when taking a single photograph. Two things determine that amount of light:
- Illuminance, which is the flux density (power per unit area) of light incident upon the film/sensor.
- The total time during which film or a sensor is exposed to light.
Mathematically, exposure is the product of these two values. One can decrease time but increase illuminance by a proportional amount, or vice-versa, without changing the total exposure.
Illuminance is controlled by the entrance pupil of the lens’ aperture diaphragm; a wider pupil increases illuminance, while a smaller pupil decreases it. In photographic jargon, this control is simply called “aperture.” Time is controlled by a shutter, located either within the camera body (SLRs and most interchangeable-lens mirrorless cameras) or within the lens assembly (some compact cameras as well as many medium- and large-format cameras). In photographic jargon, this control is called “shutter speed.”
Shutter speed is calibrated in seconds or fractions of a second. Aperture is calibrated in units called “f-stops”, so called because they are computed as the ratio of the lens’ focal length (f) to the diameter of the diaphragm opening. F-stops are dimensionless; the same f-stop on any lens indicates the same illuminance, irrespective of focal length.
The progression of settings on the controls for exposure is arranged such that moving from one value to the next changes exposure by a factor of 2. In photographic jargon, a factor of 2 is called a “stop”; “opening up a stop” means doubling exposure, while “closing down a stop” means halving exposure.
Moving from one shutter speed to the next changes the exposure time by a (sometimes approximate) factor of 2. Because illuminance is proportional to the area of the entrance pupil, f-stops are arranged in a progression by the powers of the square root of 2, like so:
f/1, f/1.4, f/2, f/2.8, f/4, f/5.6, f/8, f/11, f/16, f/22, f/32, f/45, f/64, f/90, f/128….
In most cases, these numbers are rounded from the exact values to make them easier to remember. A good-quality camera should allow the photographer to change aperture and shutter speed in fractions of a stop; 1/3 stop is the most common value for professional equipment.
Film Speed/ISO Rating
As with “stop,” the concept of “speed” has multiple meanings in photography: it can refer to the maximum aperture available on a particular lens (a “fast” lens has a large maximum aperture (small f-stop) while a “slow” lens has a smaller maximum aperture), or the duration of time for which the shutter is open, but it can also indicate the photographic medium’s sensitivity to light.
The original systems for representing film speed are an arithmetic scale promulgated by the American Standards Association (ASA), and a logarithmic scale promulgated by the Deutsches Institut für Normung (DIN). In the ASA system, doubling the rating gives a doubling of sensitivity; in the DIN system, adding 3 to the rating gives a doubling of sensitivity. The ISO standards for film speed combine both scales into a single rating, although in common parlance only the arithmetic (ASA) scale is used.
Film speed is determined by measuring the exposure necessary to yield a specific image density (opacity, for negative film). A faster film will require less exposure than a slower film, and will thus be better suited for fast-action or low-light photography. Speed comes at a price, though: the fast film requires larger crystals of silver salts than slower films do, and the resulting images display coarse grain. For applications where image quality is paramount, slow films with finer grain are required.
A digital sensor’s responsiveness to light is governed by the size of the photosites and the depth of the quantum well; an arbitrary sensitivity may be established by adjusting gain. ISO defines a standard for measuring the parameters of a digital sensor and converting it to an equivalent film speed rating. Image sensors have achieved sensitivities far in excess of anything possible with film. Without special processing, film rarely achieves speeds above ISO 800, but, to take one example, the Nikon D3S may be set to an equivalent sensitivity of ISO 102,400. Greater sensitivity also comes at a price: just as fast film has coarse grain, turning up the gain increases image noise. It also reduces the dynamic range that may be captured in a single exposure; for instance, a digital sensor may have 9 stops of usable dynamic range at base ISO, but at heightened sensitivity it may only have 5 or 6.
The Digital Exposure Triangle
When photographing with film, sensitivity is fixed for a given film stock. The photographer chooses his film for his application, and then is left with only shutter speed and aperture to control the image as it is made. Digital changes that: the sensitivity of the sensor may be varied with each image exposed. We thus have the “exposure triangle” for digital imaging, balancing shutter speed, aperture, and sensitivity. From our previous discussion we can see that there are tradeoffs among these three:
- Faster shutter speeds freeze motion but require wider aperture/higher ISO/more light. Slower shutter speeds allow lower ISO/narrower aperture but require that the subject be still and the camera be well-supported.
- Narrower apertures increase depth-of-field but require higher ISO/slower shutter speeds/more light. Wider apertures allow lower ISO/faster shutter but less of the subject is sharp.
- Lower ISO decreases noise and increases dynamic range, but requires wider apertures/slower shutter speeds/more light. Higher ISO allows narrower apertures/faster shutter speeds, but increases noise and decreases dynamic range.
Note that “more light,” i.e., supplementary light provided by the photographer, mitigates the tradeoffs somewhat: a powerful flash unit, for instance, allows the use of fast shutter, small aperture, and low ISO all at once. The use of supplementary light is a subject all its own, for a separate essay.
Light Meters and 18% Gray
As photography refined its craft with the help of physical sensitometry, it became more and more common to establish exposure through measurement of the scene to be photographed rather than through experience or by trial-and-error. This led us to a standard and a device for measuring the standard: 18% gray and the electronic light meter.
On a geometric progression from black to white, the halfway point may be found at 18%. Ansel Adams’ use of 18% as the central zone (Zone V) in his Zone System encouraged its adoption as the fundamental unit of measurement for exposure, and today cardboard or plastic cards manufactured with a surface of 18% reflectivity are standard equipment for most photographers.
Light meters come in two flavors: incident and reflective. Incident meters are used more often in studio work where the photographer controls all the light sources; they are designed to be held next to the subject to measure the collective subject illumination. A reflective meter is designed to measure from the camera location, reading the reflective luminance of the object at which it is pointed. The most effective variety of reflective meter is the spot meter, which measures a cone of 1 degree of arc; it can isolate objects or regions of objects for comparative measurement.
When using a spot meter to measure the reflective luminance of an object, the meter will show an exposure for rendering the object as 18% gray.* Not all objects are of 18% reflectance, however; if I use my spot meter to measure snow in direct sunlight and then expose as given, I will underexpose my image. A correct exposure for such a scene would be 1.5-2 stops more than the 18% reading off the snow. Similarly, if I measure a black object and use the meter’s exposure without change, I will overexpose. In challenging situations, measuring a known 18% gray surface placed in the scene to be photographed can help establish a correct expsoure quickly. Using an incident light meter is similar to using an 18% gray card; the light sensor is surmounted by a hemispherical diffuser of known transmittance, which achieves a similar result to measuring a gray card.
*Ideally, this is the case; however, the ISO standards for calibrating meters are defined in terms of luminance, not reflectance, and the recommended luminance converts into 12% reflectance registering as medium gray. The standards do allow for some latitude in calibration, though, and if the manufacturer chooses the higher end of the allowed range they will get 17.6% gray instead. Nikon and Canon internal meters are calibrated to 12%, and handheld meters from Sekonic (which this writer thinks are the best currently available) are also calibrated to 12%. The popularity of the 12% calibration has induced Kodak to print a cryptic note on their 18% gray cards that the photographer should meter off the card, and then open 1/2 stop (the difference between 12% and 18%) to avoid underexposure.