When we use our DSLR to view a scene, it does something called metering. Put simply, this is a process where our camera measures the amount of light being reflected off a subject (or several subjects, or your entire scene), and feeds back an exposure value it feels will adequately expose the image.
To really make sense of what metering is, you need to understand firstly what the meter is trying to do with the readings it gets and then the different metering modes available in your camera (the different ways it can get those readings in the first place).
I’m trying to keep the definitions of how each mode works and what they do as broad as possible, as every manufacturer does things just slightly differently in order to achieve the desired results. Even different models of camera from the same manufacturer can work slightly differently – see your manual for details.
Ok, so you say my camera looks at the light reflected off a subject?
Yes, the meter inside your camera is what’s known as a reflective meter. This is pretty much as it sounds, it takes a reading based off the amount of light reflected off either part or all of your scene. Different objects reflect back different amounts of light. For example, a black T-Shirt is going to reflect less light than a white one.
What it then tries to do is figure out an exposure value that would allow the camera to represent all this light at a brightness level of 18% grey (which is the mid point between pure black and pure white).
It works this way in order to try to have some consistency in the brightness levels from shot to shot, under potentially very different lighting conditions, and to try to retain as much detail as possible in both the shadow and highlight areas of the scene. But, it can be confused depending on what you’re shooting.
I’m with you so far. So what are metering modes?
Metering modes are the different ways you can tell your camera to “see” the light and determine its readings.
There are three main metering modes typically available on DSLRs; Spot metering, Centre weighted metering, and Matrix metering (or evaluative metering, depending on your particular brand of camera – but for all intents and purposes they’re the same thing).
Matrix metering, I’ve heard of that one!
Let’s have a look at matrix or evaluative metering first then. This is the one that most people will probably use for the majority of the images they shoot, and it’s typically the one your camera is set to by default when you purchase it. In this mode, your camera basically looks at the entire scene, it then tries to figure out an average value for the brightness levels represented in your scene, and determines an exposure value to make the scene’s overall brightness average out to 18% grey. This is a simplified explanation, but it demonstrates the point. There’s a little more information on this down at the bottom.
As an example, let’s say you’re spending the day at the park with your family. Your significant other is wearing a rather fetching bright yellow number, your toddler is with you wearing his spiderman suit, it’s a glorious day with a few white clouds darted about the sky, there’s grass and trees in the background, and they’re on a near black footpath. Then, you bring up your camera to take a shot.
The camera will look at all the brightness values of each of these areas around the frame, average them all out, and then determine an exposure value that will make this average brightness level become 18% grey. This way the camera can try to retain as much detail as possible in both the bright highlights (the bright yellow shirt/dress/whatever, the fluffy white clouds, etc) as well as in the darker areas (the dark footpath and the deep reds and blues in the spider suit) when you actually press the shutter.
If you’re shooting in one of the fully or semi automatic modes (P, A & S for Nikon or P, Av and Tv for Canon) it will try to set what it thinks is the right combination of aperture & shutter speed, based off your ISO, in order to achieve that exposure value. In manual mode, the meter just tells you if you’re over or underexposed, by how many stops or fractions of a stop, and lets you determine whether you want to adjust your ISO, aperture or shutter speed in order to “zero it out” and get the exposure value the camera tells you is the “correct” one.
The majority of the time, matrix metering (or evaluative metering) is the method that will get you the best overall result, but this is also the metering mode which can get confused easily if you’re shooting very contrasty scenes or shooting a scene that contains an unusually high amount of particular bright or dark areas. This is why you may find that the nice portrait on the beach turns into a beautifully blue sky & ocean with your subject a pure black silhouette, as in the example below of the lovely Collette Von Tora (not that you can tell in this image).
Alternatively, if you’re shooting against a dark background that is in shade and your subject is in bright sunlight, it may give you a beautifully exposed background while turning your subject pure white.
Ahhh, ok, so what can we do to fix this?
In the sample image above, the only way to really get a good exposure on Collette while maintaining the exposure in the sky is to pop her with flash (which is what we did on that shoot).
But let’s say we’re not that bothered about whether the entire scene is well exposed, and we’re just concerned with our primary subject and we don’t have a flash with us, or it’s impractical. Well, that’s where centre weighted metering and spot metering come in.
In spot metering mode the camera looks at a very tiny section of the image, and reads only the light reflected by whatever that very small section is covering. Spot metering is generally associated more with the Zone System, developed by Ansel Adams based off the work of Fred Archer. While we’re not going to get deeply into the zone system here, there are other times when spot metering can be very handy. The above possibly being one of them if we weren’t bothered about having detail in the sky.
Traditionally, in spot metering mode, the camera just looks at the brightness levels in a small circle covering about 3mm diameter of the image in the dead centre of the frame, and then formulates an exposure value based entirely on the reading of that small area. On more modern recent cameras, this 3mm circle will follow your autofocus point. So, if you’re using a single AF point other than the centre one, in order to lock focus onto your subject, the meter will be looking at the same part of the frame to determine what brightness levels it needs to pay attention to. When you change your AF point from the centre, that 3mm circular area moves along with it.
This is a handy advancement in the way things work, as you typically want to use the AF point nearest to the part of the frame you want to focus on, and as it’s the point of focus, that’s what you’ll want to be metering too.
In this example, of a pleasant old chap I met at Blackpool Zoo, I encountered a similar problem to that which I faced with Collette. Bright cloudy sky (although not quite as bright as the clear cloudless sky in image above), and a dark subject that appeared near black if I’d allowed the camera to expose the scene all by itself using matrix metering mode, and as he was around 35-40ft away from me (and about 10-15ft above me), there wasn’t much chance of throwing out some flash.
Our primary focus here was the orangutan, not the environment around or behind him. As such, it didn’t really matter if our sky was blown out slightly (although, as I was shooting RAW, I was able to pull back a tiny bit of detail so it wasn’t completely white) and I wasn’t too bothered if the wooden frame of the climbing apparatus was adequately exposed either (as it was also slightly brighter than the orangutan, I didn’t want it to contribute to the exposure either).
Spot metering was the answer here. In spot metering mode, I placed my autofocus point directly over one of the areas of the orangutan that was lit by sunlight. I figured this was the same approximate brightness as 18% grey. With my ISO at 200, and my aperture set to give me the depth of field I required, this allowed me to get a shutter speed that would give me adequate exposure on the subject, regardless of what else may be in the scene.
With my ISO, aperture and shutter speed set, I was now free to compose the image how I wanted, and be able to make the photo knowing that my exposure for the subject would be good.
That makes sense, so what’s centre weighted metering?
Centre weighted metering is a bit of a hybrid between matrix/evaluative metering and spot metering. Centre weighted metering does look at the entire scene, however it primarily focuses on an area in the centre of the frame (or, again, following your AF point), and gives that area more “weight” than the outside of the scene when determining how to expose.
The camera I primarily use is the Nikon D300s. On the D300s, the camera has a default value of 8mm set for the diameter of this circle (which also moves along with the AF point), although this setting can be changed via the camera’s menu system to be 6mm, 8mm, 10mm, 13mm or even average the entire frame. The default setting, and whether you can even change it or not will vary depending upon the make and model of your camera – again, see your manual.
Anyway, let’s go back to our day out at the park. There you are shooting your son, but instead of a spiderman outfit, he’s wearing standard blue jeans and a grey t-shirt. We have a similar arrangement with bright blue sky and dark ground, but matrix metering isn’t quite giving us what we want. Perhaps the sky is too bright, or the ground too dark, and it’s throwing things off a little bit.
Here spot metering may be impractical as we’re not entirely sure how closely the different areas of our subject or his clothes represent an 18% grey level of brightness, and we really don’t need to be that precise anyway. We just want to have a nicely exposed image of the boy, and it doesn’t matter if our sun is slightly overexposed, or our ground underexposed.
If we set our camera to centre weighted metering mode, and set our AF point to somewhere around our subject’s face, when we now bring up the camera and frame it on our subject, our camera is still metering the entire scene, however the areas contained within that circular area around our AF point are receiving about 75% of the weight (influence) of the whole shot. The other 25% is the rest of the scene outside that circular area.
In theory, this gives us a nicely averaged exposure on our subject, without giving the outside areas so much importance. This was generally the most common method of metering 35mm cameras for general use until matrix metering came along.
You said that centre weighted metering can be configured to look at the whole scene, so how is that different from matrix metering?
Well, matrix metering has come a long way over the years. Initially it pretty much was just looking at the entire scene and averaging things out, as I would get on my D300s if I configured CW metering mode to cover the entire viewfinder. These days matrix metering takes a lot more of the scene’s information into account in order to determine an accurate exposure.
Nikon’s current matrix metering system, “3D Color Matrix Metering II”, takes many other factors into account in order to determine what it is you’re actually trying to get the exposure of, and uses what it perceives to be in focus to help determine a valid exposure. Things like the distance at which your lens is focused (with compatible lenses – which is most modern Nikon glass), the actual colour of items within the scene (one of the drawbacks of spot & CW metering is that they only “see” in simple grey brightness levels), exploring individual RGB values of everything in the shot. The meter then accesses an internal database of over 30,000 images in order to compare and try to give you the most accurate exposure possible (and it does this almost instantaneously).
Evaluative metering, as it is called on some other camera systems, has similar methods that it utilises in order to try to determine the correct exposure, but, as mentioned above, it can still get confused in scenes that contain a very high dynamic range, a lot of contrast or when a major part of the image is an extremely bright or dark area.
So my camera’s pretty smart, huh?
Generally speaking, when it comes to metering, yes, it can be.
As I stated at the top of this article, matrix or evaluative metering is the mode you’ll want to be in for the vast majority of your shooting. On the current generation of bodies (and those newly announced ones waiting to hit the streets) the metering systems are very advanced, and will often just know what you’re trying to do, even if you don’t.
Unless you know you have a specific need to jump into spot or centre weighted metering modes (and only experience can really tell you when this is), you can pretty much stay in matrix metering mode the whole time and not have to worry about it.