Limephoto.co.za - Tutorials and Writing

The Basics of Metering

or: The Problem of the Little Egret

What is metering?
Digital sensors, like film, require a certain amount of light to hit them before an image is recorded (or exposed). Too much light and the image is overexposed. This is sometimes referred to as ‘burnt out’ because the resultant image is too bright. Conversely, if too little light reaches the sensor or film the image is underexposed, with the result that it is too dark. As light levels are constantly changing the photographer has to use a combination of aperture, shutter speed and ISO to ensure that the correct amount of light exposes the image. The meter assists the photographer and the camera in determining the correct amount of light that should strike the sensor by calculating what the actual light level is and working out what adjustments (aperture etc.) are required to record the light without under or over exposing the image.

The complication with metering is that the meter, being part of a basic computer, cannot think. It only calculates. The photographer has to make the decision as to what to meter from and how to adjust the camera so as to correctly expose the image.

Film manufacturers (and later digital camera manufacturers) decided that on average light that exposes a piece of film should create an 18% tone of grey (In reality modern cameras actually meter for a 12-13% tone of grey – somewhat lighter than the traditional 18% grey). This is the same tone of grey that would be recorded on black and white film as green grass on a cloudless sunny day. It is also the same tone of grey that Ansel Adams chose as his zone 5. Some commentators have even argued that Adams persuaded Kodak to choose 18% grey on the basis that it was his zone 5 (I personally can’t verify this so put it down hearsay and trivia rather than fact). So the premise is basically that if the camera is adjusted to match the meter reading (i.e. aperture and shutter speed are set to what the meter regards as ‘correct’), the overall average tone of the scene being photographed should be 18% (in reality 12-13%) grey.

The problem lies in the fact that the natural world does not always present to us scenes of 18% grey. The camera, not responding to this real-world problem tries to make every scene that it photographs 18%grey in tonal value. Thus, if the scene is a brightly lit snow-scape, and therefore far lighter than 18% grey, the image becomes underexposed as the camera’s meter attempts to compensate for the brightness by making the image have a tonal value of 18% grey. Images with an overall light tone will therefore be underexposed by the cameras meter, while images with an overall dark tone will be overexposed by the cameras meter.

The complications with metering don’t only extend to situations where the overall tone is lighter or darker than 18% grey. There is also the issue of uneven tones in an image. For example, a portrait of someone with strong backlighting (the sun), or my favourite theoretical example: a white wading bird such as a Little Egret against a dark expanse of water and shadowy banks. Left to its own devices the cameras meter will meter for the dark background, attempt to reproduce it as 18% grey (lighter than it is in reality) and thereby overexpose the bird causing it to ‘burn out’. So how do we solve the problem?

Incident, Ambient and Through the Lens (TTL) Metering

First, we need to discuss the types of metering available to the modern day photographer in order to tackle the problem of the Little Egret. When we talk about metering a scene we are usually talking about calculating the reflectance value of the subject, i.e what is the intensity of light reflected off the subject towards the camera. This is known as incident metering. Prior to meters that were built into the camera, photographers used incident light meters (handheld lightmeter below right...this one known as the ‘Gossen Color Finder’) which were simple handheld devices with a selenium cell that measured light values. These light values were then applied to a simple calculator that would give the various aperture/shutter speed/ISO combinations. The problem was that the basic selenium cell was very sensitive to extremely bright situations and was also inadequate for situations where there were extreme shifts in tonal value.

Enter the ambient lightmeter (Below left, the popular Sekonic L-358). Unlike an incident light meter an ambient light meter measures the intensity of light falling on the subject. It’s use requires the photographer to move towards the subject and take a light reading of the light as it falls onto the subject. The meter looks very similar to the incident lightmeter but has a protuberance looking like half a ping-pong ball in which the lightmeter’s sensor is housed. The advantage of the ambient lightmeter is its accuracy in reading light values around the subject, even when there are large tonal differences.
Of course, to use a handheld lightmeter often requires some mental gymnastics in calculating exposure. The meter only gives a suggested aperture/Shutter speed combination for the camera and lens on its own. It does not take into consideration extension tubes, filters, focusing distance (light values actually diminish the closer we focus as it requires more space between the rear element of the lens and the focal plane itself), teleconverters and introduced artificial lighting.

When Pentax announced the Spotmatic camera back in the 1960’s it caused quite a stir due to its internal incident meter that measured light that came through the lens (TTL)*. This has formed the foundation of modern TTL metering systems ever since. Now, regardless of what accessories were placed on the camera lens, an accurate light value could be calculated (still with the basis that the average tonal value reproduced would be 18% grey).

* The Topcon RE Super actually preceded the Pentax Spotmatic to market by 2 years, but it is arguable that the more popular Pentax probably had a greater impact on the photographic community.

Metering Patterns
Cameras and their internal TTL meters have come a fair way since the Pentax Spotmatic. Not only are they far more accurate but there are numerous ways in which to overcome the problem of the Little Egret. Essentially, the camera measures the intensity of light by utilising a ‘pattern’ that measures light in a particular area of the scene. Some patterns are very simple (centre weighted) while others compromise of grids of sensors that even measure the colour and how it would affect the exposure (Nikon’s 1005 pixel 3D colour meter stands out her).

Centre Weighted Metering

Traditionally, centre weighted metering is the most commonly found metering pattern. In the 1970’s and early 80’s it was often the only metering pattern available on cameras featuring TTL metering. It simply comprises of a sensor (often a selenium cell) that measures the average light-fall in roughly 60% of the viewfinder concentrating on the centre of the viewfinder. It works best when there is the light falling on the subject is fairly uniform without huge differences between light and dark.

To the skilled photographer centre-weighted metering is useful when used in conjunction with the camera’s exposure lock. Simply choose the area in the scene that approximates the middle tone (remember that for digital cameras this is approximately 12% grey), meter, hold down the exposure lock, recompose and shoot

Camera Metering Patterns

One particular instance that I found centre-weighted metering useful was while photographing a group of students exercising in an open air-hallway for a bio-kinetics brochure. Bright light was spilling in from the edges of the scene and fooling the more complex matrix metering on my camera into underexposing. By switching to centre-weighted I was able to concentrate my metering on the broad central portion of the scene (where the composition was). The highlights were lost, but would have been lost regardless considering the vast difference between the highlights in the sky and shadowy hallway (see image).

The problem with centre-weighted metering is if there are extremely bright or dark patches within the area that is being metered from. When this happens the metering is biased towards the more dominant tone, even if this is not the subject to be exposed for. For instance: photographing a tree against a bright sky so that the tree is backlit. If the photographer wants to expose for the tree and points the camera at the scene, the centre-weighted meter will measure the light for the sky as its tone encompasses more of the scene than the tree’s does. The White Egret problem also rears its head. The lighter toned small egret only takes up a small portion of the image in relation to the darker water. The meter will bias towards the darker water, thereby overexposing the egret. Enter the Spot-meter.

Exercise image

Spot Metering

To overcome the problem of a too large area being metered from, camera manufactures concentrated the metering pattern to a small central circle equating to less than 5% (some are as small as 2%) of the image area. Early spot meter cameras had the spot locked to the centre of the image, although this is now able to be moved to any of the focus points on a modern digital SLR. This meter pattern overcomes the problem of metering off a small area, but does require a little more work in order to use it effectively.
Landscape photographers for instance will use the spot-meter to scan a scene, taking note of a range of measurements and then deciding on an exposure that balances the measurements. For example, if the photographer knows that there are about 7 stops of dynamic range that the camera’s sensor can record, he will take measurements from the shadow, middle-grey and highlight portions and choose an exposure that will encompass all of these (or alternatively decide on a using neutral density graduated filters or HDR imagery to the tonal ranges to a maximum of 7 stops). A lot of professionals actually prefer using a handheld spot meter as it means that they don’t have to point their camera in all directions to measure the light.
Sport photographers might use the spot meter to calculate the light from a middle grey tone on the sport field. Spot-metering off grass is one way of ensuring the correct exposure. Light to dark shirts that sportsmen and women wear can fool a meter, but by manually exposing after calculating the exposure through a spot-reading of the one constant – the grass – enables the photographer to shoot without worrying about the exposure.

Multi-Segment, Matrix and Evaluative Metering

What is obvious about the above metering patterns is that they require some work in order to get the best results. In the mid 1980’s camera manufacturers came out with the first multi segment metering patterns. These are variously known as Matrix (Nikon), Evaluative (Canon) and Multi-segment (Pentax, Sony). These are sophisticated electronic meters that split the image into segments where a measurement is taken from each segment and then tabulated against a database of images (some sources put it that as much as 100 000 images form this database) so that a similarly lit image will then guide the camera meter’s decision as to what the exposure should be.
Unlike centre-weighted and spot metering, the user theoretically doesn’t have to think about the exposure. Simply point and click. In reality multi-segment metering isn’t as fool proof as that. Once more the White Egret manages to fool the meter.

Partial Metering

Partial metering is unique to Canon cameras and involves the use of only the central segments of the Evaluative or multi-segment metering pattern. It is useful as a compromise between spot and centre weighted metering. I’ve noticed that it is quite popular among some landscape photographers as they are able to use it to measure larger tonal areas than a spot would measure, but still smaller and more accurate than centre-weighted metering would measure. Like spot and centre-weighted metering, it’s use requires some thought on the part of the photographer.

A Caveat and a Conclusion

With experience it is possible to overcome the problem of the White Egret regardless of which metering pattern in use. The experienced photographer will recognise that the water is considerably darker than the egret and compensate accordingly. She will already know that the camera will try to expose the darker water as roughly 12% grey. By simply dialling in a bit of underexposure she can overcome the problems that arise with both centre-weighted and multi-segment metering which would overexpose the water and blow out the egret. Even if she had opted for spot-metering off the egret itself, she would still have had to compensate for the fact that the egret is lighter than 12% grey and hence have to overexpose to obtain the correct exposure.

Getting the experience is the hard part, but it is made substantially easier with digital than it was with film. Knowing how to read a full channel histogram is the easiest way in which to obtain accurate exposure. Regardless of which metering pattern is in use, create a test exposure and look at the histogram – NOT THE IMAGE – to gauge whether the exposure is right for the image. Adjust any settings (compensation for instance) according to the histogram and shoot at these settings until the light or your or your subject’s position changes. Yes, knowing how to meter correctly can speed up the picture taking workflow, but by using the histogram you can learn how to anticipate what the meter will suggest and how to compensate accordingly.

Back to Tutorials