mcnicks
June 15th, 2008, 09:53 PM
Light meters are stupid and boring. They are stupid because even the most sophisticated matrix metering really has no idea whether that bright bit at the top of your scene is open sky or wallpaper. They are boring because they want to make everything grey. No matter how you use your light meter, you will always end up saying things like: "well the subject is dark so I'll want to overexpose a bit" or "there is some reflected back lighting so I'll underexpose a bit".
More often than not, unless you are photographing a really simple scene, you will have to bracket your exposure (that is, take some photographs underexposed, some on the button and some overexposed) in order to ensure that at least one photograph will be properly exposed after development.
That got me thinking about light meters. Why do we place an important consideration in composing a photograph in the hands of a dumb machine that we overrule more often than not, and that we do not trust?
Let's talk about light. In terms of light metering, there are two types of light: incident light and reflected light. Incident light is the light that falls on a particular scene. Reflected light is the light that bounces off a subject and into your camera's lens. Obviously, reflected light is the light that we are most concerned with in photography, but it is also the light that gives light meters a problem, because light meters know nothing about the subject. If the reflected light has bounced off a dark subject, the light meter will try to open up the exposure to make it grey. If the reflected light has bounced of a very light subject, the light meter will try to close down the exposure to make it grey.
But what about incident light? No doubt you will have seen professional photographers waving what looks like a tricorder with a half ping-pong ball in front of the bride's face at a wedding. These things are incident light meters: they tell the photographer how much light is reaching the subject. Incident light has an interesting property:
It stays the same for the whole scene.
Obviously there are exceptions if there is complicated spot lighting involved and such, but it remains a good rule of thumb. Regarding how much of that incident light will eventually reach our lens, we know that quite a lot of it will bounce of light coloured subjects, making them light, and that not much of it will bounce of dark coloured subjects, making them dark.
Wait a minute. Isn't that what we want?
The problem with incident light meters is that you have to get close to your subject to measure it. If your subject is covered in taffeta and standing ten metres away, that is not a problem, but if your subject is a large, mountainous object several kilometers away it becomes more difficult.
If only we could do without that incident light meter ...
Well, what happens if we do not have any light meter? Some of you will have heard of the "sunny 16" rule. The rule states that, on a bright sunny day, you can set your f-stop to f/16 and your shutter speed to one over your film speed to get the right exposure. So, if you were using ISO 100 film, setting 1/100s at f/16 would do.
The sunny 16 rule is an example of judging the incident light falling onto a scene without a light meter. If we could come up with a few more rules of thumb like the sunny 16 rule, then maybe we could throw away our light meters completely! Of course, we can make errors of judgement, so it would be wise to bracket our exposures but, as I pointed out at the start, we often do that anyway so we are not losing anything ... except the damned fool light meter!
Trying to remember f-stops and shutter speed settings is difficult. Luckily there is a simpler way to describe the sunny 16 rule using a number called Exposure Value (EV). The exposure value is a measure of the amount of incident light falling on a scene. Adding one to the EV doubles the light and subtracting one from the EV halfs the light. This is exactly the same thing that happens when we increment or decrement the f-stop, shutter speed or film speed, so EV really is a good match. We can rephrase the sunny 16 rule like so:
The light falling on a scene in bright sunshine will have an EV of 15.
This is much simpler than the 'one over film speed' stuff. We can also make up other, similar rules of thumb:
The light falling on a conventionally lit interior at night time will have an EV of 5.
And another:
The light falling on landscapes in twilight will have an EV of 10.
And so on. Once we know the EV of a scene, we can use it to calculate an f-stop and shutter speed for whatever film we are using. Of course, there will be a range of f-stops and shutter speeds that will work so we would need a large table to specify them all. That is, unless we can come up with a simple way of calculating it. It turns out that:
To properly expose ISO 100 film at 0 EV, you would expose for one second at f/1.
We know that an EV of 1 will have double the light, so we would have to reduce the f-stop or the shutter speed by one increment. In general, we can count off the EV and increment either our f-stop or shutter speed to arrive at the correct settings. For example, say our scene is lit at 15 EV:
f/1.4 would be 1 EV
f/2 would be 2 EV
f/2.8 would be 3 EV
f/4 would be 4 EV
f/5.6 would be 5 EV
f/8 would be 6 EV
f/11 would be 7 EV
1/2s would be 8 EV
1/4s would be 9 EV
1/8s would be 10 EV
1/15s would be 11 EV
1/30s would be 12 EV
1/60s would be 13 EV
1/125s would be 14 EV
1/250s would be 15 EV
So for ISO 100 film, we would expose for 1/250s at f/11. If you think about it, that is just the sunny 16 rule expressed with a larger aperture and smaller shutter speed. Try it yourself: count it off with your fingers. Its really easy once you memorise your f-stops, and it is even easier if you manually click your f-stop and shutter speed controls as you count off.
What about other film speeds? Well, we could come up with extra rules of thumb but that would complicate things. Instead, let us think about what a higher speed film actually does. As film speed goes up in increments, the sensitivity of the film doubles, which means that we can close the aperture or increase the shutter speed one stop and still get a proper exposure. So increasing the film speed is a bit like increasing the available light in a particular scene.
In fact, we can add the film speed to our estimated EV before we count off our f-stop and shutter speed. For example, for photographing a dinner scene (EV 5 according to the rule of thumb above) with ISO 800 film, we could count off:
ISO 200 is 6 EV
ISO 400 is 7 EV
ISO 800 is 8 EV
f/1.4 would be 1 EV
f/2 would be 2 EV
f/2.8 would be 3 EV
f/4 would be 4 EV
1/2s would be 5 EV
1/4s would be 6 EV
1/8s would be 7 EV
1/15s would be 8 EV
So, with ISO 800 film we could properly expose a dinner scene by setting f/4 for 1/15s.
Obviously, scenes are many and varied so there will be many exceptional circumstances. We can protect ourselves in two ways. First of all, we can come up with some additional rules of thumb to cover awkward lighting. Secondly, and most important of all, we can always bracket our exposures. Here are some additional rules of thumb:
Shooting in shadow? -2 EV
Shooting in heavy forest? -3 EV
Shooting into a light source (eg. lamp)? +2 EV
Shooting a silhouette (eg. sunset)? +3 EV
As long as your bracket your exposures and make rough adjustments for specific conditions, you should always end up with a reasonably well exposed shot.
Let's go back to why we are doing this. We know that meters are stupid and can never understand the true pattern of lighting and subject arrangement from scene to scene. But we are clever, so we can. We are not precise like a machine, so we have to give ourselves some latitude by bracketing but, in general, with experience, we have a much better chance of evaluating the required exposure for a particular scene. More importantly, we are much better placed to make an artistic judgement about a scene than any piece of technology.
So, my challenge to you, today, is to throw away your light meter.
If you have a digital camera, set it to manual for a while. Ignore what the little meter in the viewfinder is telling you and think about the scene. Count off the EV to come up with an f-stop and shutter speed and take a few photographs, then have a look at your results on your LCD. Are they awful? Did you count right? Were there areas of glaring overexposure? Go and experiment! Just remember to always bracket your exposures and you'll be fine.
If the above rules of thumb are no use to you, try making up some of your own. Aside from anything else, depending on the exact latitude you are at right now, the quality and strength of sunlight that reaches your scene will be different, and I am sure that the Scottish understanding of 'bright sunlight' is different from the Australian understanding of the same. Maybe your sunny 16 is actually 16 EV? Maybe mine is actually 14 EV? Over time, you will be able to work that out and, more importantly, you will take back control of your photography.
-- McNicks
Note: The inspiration for this post came from getting fed up with using a light meter with my fully manual Bronica. The exact calculations owe an great debt to Fred Parker's Ultimate Exposure Computer (http://www.fredparker.com/ultexp1.htm). I couldn't be bothered with tables so I came up with the 'counting off' idea, but I think its safe to say that most of this should be credited to him.
More often than not, unless you are photographing a really simple scene, you will have to bracket your exposure (that is, take some photographs underexposed, some on the button and some overexposed) in order to ensure that at least one photograph will be properly exposed after development.
That got me thinking about light meters. Why do we place an important consideration in composing a photograph in the hands of a dumb machine that we overrule more often than not, and that we do not trust?
Let's talk about light. In terms of light metering, there are two types of light: incident light and reflected light. Incident light is the light that falls on a particular scene. Reflected light is the light that bounces off a subject and into your camera's lens. Obviously, reflected light is the light that we are most concerned with in photography, but it is also the light that gives light meters a problem, because light meters know nothing about the subject. If the reflected light has bounced off a dark subject, the light meter will try to open up the exposure to make it grey. If the reflected light has bounced of a very light subject, the light meter will try to close down the exposure to make it grey.
But what about incident light? No doubt you will have seen professional photographers waving what looks like a tricorder with a half ping-pong ball in front of the bride's face at a wedding. These things are incident light meters: they tell the photographer how much light is reaching the subject. Incident light has an interesting property:
It stays the same for the whole scene.
Obviously there are exceptions if there is complicated spot lighting involved and such, but it remains a good rule of thumb. Regarding how much of that incident light will eventually reach our lens, we know that quite a lot of it will bounce of light coloured subjects, making them light, and that not much of it will bounce of dark coloured subjects, making them dark.
Wait a minute. Isn't that what we want?
The problem with incident light meters is that you have to get close to your subject to measure it. If your subject is covered in taffeta and standing ten metres away, that is not a problem, but if your subject is a large, mountainous object several kilometers away it becomes more difficult.
If only we could do without that incident light meter ...
Well, what happens if we do not have any light meter? Some of you will have heard of the "sunny 16" rule. The rule states that, on a bright sunny day, you can set your f-stop to f/16 and your shutter speed to one over your film speed to get the right exposure. So, if you were using ISO 100 film, setting 1/100s at f/16 would do.
The sunny 16 rule is an example of judging the incident light falling onto a scene without a light meter. If we could come up with a few more rules of thumb like the sunny 16 rule, then maybe we could throw away our light meters completely! Of course, we can make errors of judgement, so it would be wise to bracket our exposures but, as I pointed out at the start, we often do that anyway so we are not losing anything ... except the damned fool light meter!
Trying to remember f-stops and shutter speed settings is difficult. Luckily there is a simpler way to describe the sunny 16 rule using a number called Exposure Value (EV). The exposure value is a measure of the amount of incident light falling on a scene. Adding one to the EV doubles the light and subtracting one from the EV halfs the light. This is exactly the same thing that happens when we increment or decrement the f-stop, shutter speed or film speed, so EV really is a good match. We can rephrase the sunny 16 rule like so:
The light falling on a scene in bright sunshine will have an EV of 15.
This is much simpler than the 'one over film speed' stuff. We can also make up other, similar rules of thumb:
The light falling on a conventionally lit interior at night time will have an EV of 5.
And another:
The light falling on landscapes in twilight will have an EV of 10.
And so on. Once we know the EV of a scene, we can use it to calculate an f-stop and shutter speed for whatever film we are using. Of course, there will be a range of f-stops and shutter speeds that will work so we would need a large table to specify them all. That is, unless we can come up with a simple way of calculating it. It turns out that:
To properly expose ISO 100 film at 0 EV, you would expose for one second at f/1.
We know that an EV of 1 will have double the light, so we would have to reduce the f-stop or the shutter speed by one increment. In general, we can count off the EV and increment either our f-stop or shutter speed to arrive at the correct settings. For example, say our scene is lit at 15 EV:
f/1.4 would be 1 EV
f/2 would be 2 EV
f/2.8 would be 3 EV
f/4 would be 4 EV
f/5.6 would be 5 EV
f/8 would be 6 EV
f/11 would be 7 EV
1/2s would be 8 EV
1/4s would be 9 EV
1/8s would be 10 EV
1/15s would be 11 EV
1/30s would be 12 EV
1/60s would be 13 EV
1/125s would be 14 EV
1/250s would be 15 EV
So for ISO 100 film, we would expose for 1/250s at f/11. If you think about it, that is just the sunny 16 rule expressed with a larger aperture and smaller shutter speed. Try it yourself: count it off with your fingers. Its really easy once you memorise your f-stops, and it is even easier if you manually click your f-stop and shutter speed controls as you count off.
What about other film speeds? Well, we could come up with extra rules of thumb but that would complicate things. Instead, let us think about what a higher speed film actually does. As film speed goes up in increments, the sensitivity of the film doubles, which means that we can close the aperture or increase the shutter speed one stop and still get a proper exposure. So increasing the film speed is a bit like increasing the available light in a particular scene.
In fact, we can add the film speed to our estimated EV before we count off our f-stop and shutter speed. For example, for photographing a dinner scene (EV 5 according to the rule of thumb above) with ISO 800 film, we could count off:
ISO 200 is 6 EV
ISO 400 is 7 EV
ISO 800 is 8 EV
f/1.4 would be 1 EV
f/2 would be 2 EV
f/2.8 would be 3 EV
f/4 would be 4 EV
1/2s would be 5 EV
1/4s would be 6 EV
1/8s would be 7 EV
1/15s would be 8 EV
So, with ISO 800 film we could properly expose a dinner scene by setting f/4 for 1/15s.
Obviously, scenes are many and varied so there will be many exceptional circumstances. We can protect ourselves in two ways. First of all, we can come up with some additional rules of thumb to cover awkward lighting. Secondly, and most important of all, we can always bracket our exposures. Here are some additional rules of thumb:
Shooting in shadow? -2 EV
Shooting in heavy forest? -3 EV
Shooting into a light source (eg. lamp)? +2 EV
Shooting a silhouette (eg. sunset)? +3 EV
As long as your bracket your exposures and make rough adjustments for specific conditions, you should always end up with a reasonably well exposed shot.
Let's go back to why we are doing this. We know that meters are stupid and can never understand the true pattern of lighting and subject arrangement from scene to scene. But we are clever, so we can. We are not precise like a machine, so we have to give ourselves some latitude by bracketing but, in general, with experience, we have a much better chance of evaluating the required exposure for a particular scene. More importantly, we are much better placed to make an artistic judgement about a scene than any piece of technology.
So, my challenge to you, today, is to throw away your light meter.
If you have a digital camera, set it to manual for a while. Ignore what the little meter in the viewfinder is telling you and think about the scene. Count off the EV to come up with an f-stop and shutter speed and take a few photographs, then have a look at your results on your LCD. Are they awful? Did you count right? Were there areas of glaring overexposure? Go and experiment! Just remember to always bracket your exposures and you'll be fine.
If the above rules of thumb are no use to you, try making up some of your own. Aside from anything else, depending on the exact latitude you are at right now, the quality and strength of sunlight that reaches your scene will be different, and I am sure that the Scottish understanding of 'bright sunlight' is different from the Australian understanding of the same. Maybe your sunny 16 is actually 16 EV? Maybe mine is actually 14 EV? Over time, you will be able to work that out and, more importantly, you will take back control of your photography.
-- McNicks
Note: The inspiration for this post came from getting fed up with using a light meter with my fully manual Bronica. The exact calculations owe an great debt to Fred Parker's Ultimate Exposure Computer (http://www.fredparker.com/ultexp1.htm). I couldn't be bothered with tables so I came up with the 'counting off' idea, but I think its safe to say that most of this should be credited to him.