Photography Concept 4

In this fourth post on basic photography concepts I'm going to explain metering. As always, I'm not going to attempt an exhaustive discussion on the topic, but rather I'll try to provide the basics in plain english, with hopefully just enough meat to build some understanding without reducing anybody to tears (of boredom or fear - take your pick).

If you've not been awaiting this post with baited breath (hah!) you can also check out the previous installments where we discussed the concepts of: exposure, focal length and lenses, and depth of field.

Back in that first post I described the concept of exposure. Exposure in photographic terms quite literally means how much light you let into the camera and for how long. I described that there are three ways to control the exposure: shutter speed, aperture and ISO speed. The first mechanism controls the duration of the exposure to light, the second controls the amount of light let into the camera and the third controls the light sensitivity of the camera sensor.

So why are we trying to control the exposure? In the most simplistic terms, we want the resulting photograph to accurately represent what we are seeing with our eyes (of course there are exceptions to this and quite wonderful ones at that!). If you underexpose, the photo will look dark. If you overexpose, the photo will be too bright and washed out.

So how do we know when we've got it right? The camera will tell us... hopefully. Modern cameras in fact make it quite difficult for us to get it completely wrong. But how does the camera know when it has it right (or at least mostly right)? Metering baby. Metering.

Measuring the amount of light in your scene (or more accurately the amount of light reflected off of it) is called metering. Now of course there are loads of different ways that metering can be accomplished. I'm not going to discuss all of them. Having you fall asleep at your keyboard is not my goal here. I'll briefly describe an older system of metering that I'm familiar with and we'll take a look at modern systems.

My first SLR camera was a 1977 Canon AT-1 given to me by my father about 5 years ago. That camera had TTL metering. The 'TTL' meaning "through the lens" metering. Older cameras had the photocell (the thing sensing the amount of light) on the camera body outside of the lens. On a TTL metering camera, the photocell was within the camera and measured the intensity of light coming through the lens - hence the name. This camera had what was called 'match-needle' metering. So the photocell would control a needle that you could see in the viewfinder. As the light intensity increased, the needle rose up and conversely, the needle lowered as the light levels dropped. There was a second needle which was controlled by the aperture and shutter speed adjustments on the camera. In order to get a decent exposure, you would make aperture and shutter speed adjustments that would bring the needle in line with the needle representing the light meter reading. You would 'match' the needles. This camera didn't have 'auto exposure'. The camera wouldn't make any effort to pick the right exposure settings to match the amount of light reflected off of the scene. This was left up to the photographer. In more modern systems with 'auto-exposure', the camera tries to make adjustments to the shutter speed or aperture settings (or some combination of both) based on the light coming in, in order to get a good exposure.

In a typical photographic scene, there are light areas and dark areas. So in order to judge what a proper exposure might be, the most basic system would come up with an average of all the brightness levels in the scene and then set the exposure settings (aperture and/or shutter speed) so that this average level was rendered an 'average' tone on the resulting photograph. An 'average tone' being a mid-grey (think of worn, sunlit asphalt). The key concept is that it would average all the brightness levels in the scene. So if there was one area that was very bright relative to the rest of the scene, it would throw the average off - the average tone would be brighter. So it would close down the aperture and/or increase the shutter speed to render this average tone a middle-grey. Imagine you were taking a picture of a person against a backdrop of snow. The white background would dominate the frame, providing a very high average brightness value. So the camera would try to render this high average value as a middle-grey, resulting in your white snowy background looking grey. Not a good solution.

So just averaging the brightness values of a scene is not a very good method. Unless your scene has a average tonal value somewhere close to middle grey. There had to be a better way. And there was.. actually there were several.

Most modern SLR (and DSLR) cameras have a variety of metering 'modes'. These modes take a much more complex look at the light coming into the camera from your scene. The most common metering modes are: centre-weighted, spot, and multi-segment.

It's important to remember that these modes were created to better analyze a scene so that the resulting exposure is more accurate. But don't forget that the resulting exposure still comes down to only 3 things: choosing a shutter speed, aperture and ISO speed. These fancy metering systems don't necessarily get you better pictures, they only make it easier (and more likely) for you to get correct exposure settings.

Centre-weighted metering is a system where the tonal values of a scene are averaged together, but the central area of the frame will get more weighting in that calculation. So if the your subject is in the centre of the frame, the brightness of that subject will have a more significant effect on the resulting exposure settings than if you just averaged the whole scene. This is still a viable method of metering. But it has been rendered somewhat redundant by multi-segment metering which we'll discuss in a moment.

Spot metering is a system by which the exposure is based on the tonal values read from a tight central area of the frame (say 5%) while any metering information from outside of that central area is ignored in the exposure calculation. This mode is good when you've got a dark subject against a bright background or a light subject against a dark background. Think of a bird against a bright sky, or an actor under a spotlight on a dark stage. This is a very useful mode. Not all cameras have a true spot-metering mode. My Canon Rebel XT for instance has what Canon calls a 'partial' metering mode. In this case, the camera takes about the central 9 or 10% of the frame and not the 4 or 5% present in true spot-metering systems.

Multi-segment metering goes by different names depending on what brand of camera you have. Canon calls it 'Evaluative metering', Nikon calls it 'Matrix metering', and Olympus has its 'Digital ESP'. Each manufacturer's technology differs of course, but the basis of each is the same. In a multi-segment system, the scene is divided up into many segments. Each segment is metered and the camera then tries to interpret this information into some logical exposure value. For instance, given a darker subject on a light background but where the subject is NOT centred within the frame, a multi-segment system could identify this and still expose based heavily on the darker subject, making the assumption that this is the primary subject of your photograph. (Don't forget that a typical spot metering system would not work here unless the subject was in that central 4 or 5% of the frame.). These systems can make getting a good exposure a piece of cake, however as with most technologies, they can be fooled.

With all this technical talk about whiz-bang metering systems, it's important to keep in mind that all you're trying to do is get the right shutter speed, aperture and ISO settings. A very capable photographer could get the same great exposures with only a match-needle metering system and a keen eye for tonal values. The high-tech systems just make it a little more foolproof for the rest of us.

A few quickie tips and comments:

- Try to look at your scene in terms of tonal values. That is, visualize your scene in greyscale. If you can do this, you can get a much better handle on when you're looking at a potentially tricky exposure.

- Experiment with the exposure compensation controls on your DSLR or digicam (not all have them). These let you override the in-camera metering. So if you think (or find) that the camera's high-tech system is not getting things exactly right, you can make quick adjustments to get the exposure brighter or darker. Don't be surprised. You're eyes and brain might be more than a match for the best metering system with a little practice.

- Make use of the histogram display on your digital camera. This can indicate quickly if you've got overexposed or underexposed portions in your photo. You won't always spot it on the lcd view of your photo, but a histogram will illustrate it much more clearly.

- Shooting in RAW format can give you significantly more leeway in terms of exposure. With my Rebel XT for instance I can adjust the overall exposure by +/- two stops. This is like having access to my exposure compensation controls after I've taken the photo. RAW format files allow you to make many significant adjustments to a photo that you would normally only be able to make on the camera. While you might not want to shoot RAW format for all your photos, you should experiment with it for shots that you feel might be tricky exposure-wise.

- I've only discussed TTL in-camera metering systems. There is a photography camp out there that would chastise me for not mentioning standalone light meters. So I'll do it here (I just have). These devices are a perfectly viable way of metering a scene - and in some ways are even better than relying on the in-camera systems. But this is a series of posts based on the layman, and I didn't feel it was appropriate to explore that whole avenue. For those interested, see here.