One important task of the camera is to determine the correct exposure time that needs to be applied when the photo is taken. This measurement of light intensity is called exposure metering. The camera then increases or decreases either the shutter speed, the aperture size or the ISO to ensure that the sensor receives the correct exposure to light.
Today’s light metering systems are the result of an evolutionary development of exposure calculation utilities. The history of light metering dates back almost 200 years. Knowing previous technologies may help to understand the current implementations and their characteristics. Here is a timeline of all major light metering systems applied in photography.
The very first support tools for photographers have been exposure tables that have evolved since 1844, almost 20 years after the invention of photography. The tables contained suggestions for exposure durations based on weather and light conditions, time of the year and time of the day, type of subject and film speed. As there were no powerful lenses used at that time, the exposures suggested by the tables typically ranged from several minutes in bright daylight to a good hour in cloudy conditions.
In 1888, the first exposure calculator – called Actinograph – was constructed by the scientists Hurter and Driffield based on their scientific investigations. The Actinograph and subsequent calculators were based on the exposure tables, but combined various slide rules and rotating scales. The calculators were the first approach to ease camera operation for photographers.
Actinometers were the first instruments not relying on pure experience but actually performing some active light metering. In an actinometer, a small test exposure was made on light sensitive paper. The usual procedure was for the paper to be exposed until its tint matches that of a comparison surface. The time to darken the test paper was then translated into the input values for the shutterspeed and aperture combination. Actinometers typically had the shape of a pocket watch and became popular during the 1890s and remained so through to the 1920s. Compared with the tables and calculators, the actinometer was independent from environmental characteristics and therefore profited from more versatile applications.
Another type of active metering tool was the extinction meter. This instrument did not perform a test exposure on photosensitive paper like the actinometer and presented some quicker readings to the operator. Through the viewfinder of the extinction meter, a photographer looked at a row of numbers, arranged behind a celluloid surface that is gradually increasing its opacity. Depending on the particular design, the highest or lowest number just barely visible was determining which light situation was given. This number could then be translated into the required camera exposure settings. Other versions of extinction meters had a pattern visible through the eyepiece, and a control dial varied the amount of light allowed into the device until the pattern could only just be seen. The position on the control then indicated the exposure value. These tools typically had a telescope shape and were rotable to adjust for pre-defined camera settings such as film sensitivity. Extinction meters have been very popular in the 1920s.
The first photoelectric exposure meters consisted of a single selenium photo cell. The sensor was introduced in the early 1930s. A selenium cell consisted of a steel support plate that had one side covered with selenium, and an ultra-thin layer of gold applied above the selenium layer. Light illuminating the top surface of the selenium cell generated a current to flow between the gold layer and the steel plate as soon as these were connected. The more light illuminated the photocell, the more current was generated. The current was tiny but it could be measured by an ampere meter. That meter could be calibrated and marked with scientific units like candles or lux which in turn could be translated to camera settings.
The light level is generally shown by a needle moving over a scale, a calculator is then used to combine the reading with the film speed to give the result of shutter speed and aperture. An implementation called the match-needle meter provided easier use to the photographer. A match-needle meter had a clockhand on the meter’s scale. The position of this clockhand could be adjusted by turning one slice of the analog calculator. When the clockhand matched the instrument’s needle, the exposure value was set right on the calculator.
The first selenium light meters were packed into separate handheld devices. However, those metering systems were soon to be on-camera systems. Therefore, new cameras around the 1960s had the match-needle instrument coupled directly to aperture and shutter speed dials on the camera instead to a separate analog calculator.
This was the most convenient way of match-needle instrument usage, especially when the meter’s scale is displayed in the viewfinder. This clip-on-camera implementation of the light meter was the beginning of exposure automation in photography. The original Canonflex camera (around 1960) is an example of such a camera system with a clip-on exposure meter that is directly coupled with the speed dials.
The Canonflex RM was released in 1963 which already had a selenium meter integrated into the camera case. A window at the top of the camera displayed the needle that pointed to the correct aperture to set depending on the shutter speed that was selected. In operation it was similar to the clip-on meters on the original Canonflex, but the reduced size was another major step in the evolution of exposure metering. With the rapid developments in exposure metering automation, most manufacturers stopped the production of handheld meters until 1970.
Selenium had several desirable properties as photocell material. As selenium cells induce their own current during exposure to light, no batteries had to be used. Furthermore, selenium cells were sensitive to wavelengths similarly to camera film, as opposed to the human eye. Therefore, these types of sensors reacted to light more like camera film and provided quite accurate results. On the other hand, selenium cells would eventually lose capacity over time. With the devices directly integrated in the camera bodies, they were much harder to replace. Around 1960, another material has already been found to compensate for these disadvantages – cells made of cadmium sulfide.
Unlike a selenium cell which created current during exposure to light, cadmium sulfide (CdS) cells were resistors that varied with the amount of light that fell on them. More precisely, the resistance dropped as the illumination increased. A battery was used to provide an electrical current, and a cadmium sulfide cell varied that current depending on the exposure. Again, an ampere meter was used to read the current. Compared with selenium cells, one advantage of cadmium sulfide cells was their high durability. The system relied on a current produced by a battery which could easily be replaced once depleted. Also, CdS cells were more sensitive and offered a more accurate reading in low light conditions. Their strongly reduced size made them even more suitable for an integration inside the camera body.
In particular, small CdS photo cells for the first time allowed for the implementation of a metering system that was able to perceive the exact image the main photographic lens was producing on the focusing screen. This technology was called through-the-lens (TTL) metering and brought exposure metering to a whole new level.
Being innovative and small, CdS cells also had some drawbacks. One minor drawback was that they reacted to light similarly to the human eye, which was not the same way most camera films reacted to light. Another major disadvantage was that CdS cells suffered from a temporary night-blindness – an effect where the cell was unable to quickly adapt from a bright scene to a dark one. However, in the early 1970s, camera manufacturers again used a new material for their metering sensors – silicon.
Silicon (Si) cells combined all the best properties of both selenium and cadmium sulfide cells: Just like cadmium sulfide cells they are small, they offer good performance in low light conditions, and they are driven by batteries. Like selenium they rapidly adapt to intensity changes and they react to a wider range of wavelengths. The Canon EF model, released 1973, used a silicon meter instead of a CdS meter. Since then, this new material has increasingly gained popularity and widely replaced its ancestors.
Up to the mid 1970s, electronic applications in single lens reflex cameras were still limited, for example, where the mechanical designs were largely responsible for much of the operations. But the Canon AE-1, introduced in April 1976, was the first camera in the world to incorporate a CPU (central processing unit) which enabled automatic exposure, memory transmission of signals, display, regulation of time and completion signal are all electronically controlled. This general principle of TTL exposure metering and signal processing by a CPU has not changed until today. All auto-exposure improvements achieved since the 1970 have related to the sensor cell layout.
Through-the-lens (TTL) metering is an approach for in-camera light metering that was introduced in the early 1960s, and has since been constantly improved and refined. TTL metering is still state of the art for all modern DSLR cameras. The concept is to have a metering sensor built somewhere into the camera body where it can use the light that has gone through the main photographic lens. Through-the-lens (TTL) metering is considered to be the most precise type of exposure metering as it takes into account individual lens transmission characteristics, stray light that is cought by a lens, and other factors.
The Canon F-1 SLR, introduced in 1970, has a CdS cell attached to the side of the focusing screen. A beam splitter built into the focusing screen directs a small fraction of light from the focusing screen towards the metering sensor. The beam splitter was designed so that it would only redirect a fraction of light from the central 11% of the focusing screen surface towards the metering sensor.
The Canon EOS series of cameras, introduced in 1987, have their metering sensor attached to one face of the pentaprism directly above the eyepiece lenses. In this position the sensor receives light that is emitted from the preview image. Depending on the camera model, the metering system can also have its position in the lower part of the camera body, yet the basic principle for TTL light metering is always the same.
The diagram shows the position of the metering sensor in a Canon EOS 1D Mark IV DSLR camera, and the rays of light reaching the metering sensor have been colored in cyan. Light metering is triggered when the shutter button is pressed halfway. At this stage, the reflex mirror is still in its lower position and reflects light from the photo-taking lens onto the focusing screen. From there, light is emitted towards the pentaprism where it not only reaches the eyepiece but also the metering sensor unit. It is important to understand that white and cyan colored bundles of light are not split by any type of prism or similar optical element, but they are light from the same image point on the focusing screen that fans out into various directions. Rays of light reaching neither the metering system nor the eyepiece are not shown. You can surely see why this is a very good position for the metering sensor to assess the exposure of the scene.
Light hitting the metering sensor is always of lower intensity than light hitting the image sensor of the camera. The primary reason herefore is the use of the semi-transparent reflex mirror that only reflects a certain percentage of the incident light towards the viewfinder area while the remaining percentage is intended to hit the autofocus sensor. Another reason for the decreased light intensity reaching the metering sensor is the focusing screen. As aforementioned, the focusing screen is required for the viewfinder optics to display a preview image to the photographer. However, the diffuse surface of the focusing screen spreads the concentrated light cone from the photo-taking lens into a wider cone, although the condenser lens can compensate this effect to a certain extent. With the light illuminating both the eyepiece and the metering sensor, intensity is further reduced. In order to let the camera find the right settings nonetheless, the sensitivity of the metering sensor is typically increased so it will react to light in the same way as the image sensor.
One substantial problem with the single cell layout was that it could only perceive an average light value of the entire scene that was projected onto the ground glass. Even with the most advanced materials used, a single cell could not tell whether the light intensity resulted from an illumination evenly distributed across the entire scene or a dark scene with a small and very bright subject. Although camera manufacturers have established an exposure meter calibration to give satisfactory exposures for typical outdoor scenes, the results have not been as desired if a scene included unusually large areas of highlight and shadow.
With the release of the Canon EOS 620 in 1987, Canon has introduced a new metering sensor layout. The single-cell metering system has been replaced by a new multi-zone sensor divided into six segments. The sensor is exposed to the light from the entire scene and sends this information to the camera processor. The processor then analyzes the different zones and compares their light intensities in order to identify unusual lighting conditions and to apply some sort of corrections automatically. The Canon EOS 1N, released in 1994, already featured a 16-zone metering system including a silicon photocell and integrated circuit. Furthermore, the Canon EOS 1N already used its integrated image processor to enhance the data retrieved from the metering sensor with additional information acquired by various other sensors such as the autofocus sensor. The illustration shows the sensor layout and the resulting metering zones across the scene.
Since then, metering systems have seen continuous developments on several factors. The most obvious development concerns the number of sensor elements. The metering sensor of the Canon EOS 7D, released in 2009, divides the image area into 63 individual zones. However, the true advantage of this metering system is that the sensor is designed to perceive color information to a certain extent. The spectral response of silicon light sensors is significantly higher to long wavelenghts than to short ones. In other words, metering systems with regular silicon receptors are more sensitive to red light than to blue light. This in turn has the effect that a camera will tend to underexpose red colored objects, and overexpose those dominated by blue colors. Certainly, this effect can be corrected by the application of a color filter system, but this can decrease metering sensitivity drastically. In order to reduce these types of miscalculations, the metering system of the 7D includes a dual-layer silicon sensor that can differentiate between long and short wavelenghts. This new type of receptor helps improve exposure accuracy in situations where the scene is dominated by colors at one end or the other of the color spectrum. The illustration shows both the sensor layout and the resulting metering zones of the Canon EOS 7D.
Again, this metering system combines light intensity readings with measurements of the autofocus system. During the exposure reading, the camera processor analyzes the data provided by the single AF points – regardless of the selected AF mode – in order to determine the areas that have achieved focus. Assuming that the focused AF points are likely the ones covering the subject, the metering system priorizes the metering zones around that area. This system is called iFCL and stands for ‘Focus, Colour and Luminance’.
The Canon EOS 7D Mark II, released in 2014, features the most sophisticated metering system at the time that this article was created. The 7D Mark II incorporates a metering sensor with a full color sensor consisting of approximately 150.000 RGB pixels. The processor divides the sensor array into 252 zones, each including around 600 pixels. Both the layout of the pixel array and the full color design are similiar to a main image sensor of a camera, only the number of pixels is less on the metering sensor. Still, the resolution on the metering sensor perceives enough detail of a subject to allow face detection. The diagram below shows the sensor layout and the resulting metering zones of the Canon EOS 7D Mark II.