This is a long one. I have some stuff I wrote to use elsewhere, but never did, about how cameras work. Imagine an idiot’s guide written by an idiot. I’ll see if WordPress lets me file it as a resource or something, as I don’t want it to replace my eagerly-awaited weekly postings of nonsense and bad attitudes. Or for anyone to think this is a website about learning photography. Think of it as something I have to get off my chest and will feel a lot better for. I have a built-in need to explain: the only book I wrote is a manual on how to do a thing, rather than something I made up. Right, back to the plot.
My approach will start with the theory, as this explains why. That leads to what to do, which then leads to how to do it. When to use it is up to you: it’s a tool in your bag.
Photography records the way light reflects from a subject. A camera is a box that keeps out all of the light except that which reflects from the subject. The camera provides a method for focusing the light onto something that will record it, plus a method of capturing a specific chunk of light: a moment in time. This is where most people come in – they want to record what something looks like or to capture a moment in time. If you want to know how this all works so that you can change your results, read on.
Take a look at a scene or subject around you. Ignore the colours and look at how bright or dark things are. Your scene may include objects or shadows that are black right through to things that are white or shiny. The brightness of any part of the scene depends on how much light it reflects towards you. This can also depend on how much light falls on it: a white area of snow in sunlight can reflect a lot more light than a black dog in the shade.
Your camera contains a film or sensor that will capture and record the light that you let in. The first thing to know is that your film or sensor has limitations. The ideal that we strive for is to capture the full range of tones in the subject from shadow to highlight, while being able to see a little detail at both ends of the scale. We like to see just a hint of detail in the shadows or darkest parts of the scene so that they don’t appear as empty black spaces. We also like to see a bit of detail and texture in the highlights or brightest parts so they don’t appear to be glaring white. When you have learned to do this you can then creatively ‘ignore’ it, meaning that you can choose what you want the shadows or highlights to look like, how much detail you want to keep and how much contrast you want between light and dark. Look at something like Bailey’s portrait of Michael Caine. Not much detail in that suit, but there doesn’t need to be. Look at almost any Ansel Adams landscape and you will see every tone and shade the gods invented. So, what has this got to do with the limitations of the film or sensor?
Think of the light arriving at the camera as a stream of photons, like a stream of little bullets. The more photons that hit a single point on the film or sensor, the brighter that bit of the subject was. Obviously, the darkest thing that the sensor or film can record is a single photon. There is an upper limit too due to the physical nature and constraints of the sensor or film. Imagine the film or sensor is a row of buckets. Each bucket can hold up to 1000 photons. Any more than this and they overflow and run away. So the brightest thing you can capture needs to deliver 1000 photons per bucket and the darkest just 1.
This means that the slice of time you need to capture with the camera should deliver just the right amount of light to the film or sensor. Too much and all the buckets will fill-up so you will have a blank white picture. Too little and most of the buckets will be empty, so you get Michael Caine’s suit. At a basic level you have two controls over what the film or sensor records: how bright and how long.
Imagine taking a bar of chocolate out of the fridge and leaving it out in the sun. Leave it for a few seconds and you still have a tooth-breaker. Leave if for a few hours and you have a drink. The strength of the sunlight matters too. A minute in summer might be the same as an hour in the winter. So you adjust the two basic controls of the camera to allow just the right amount of light through.
So you want 1000 photons let through for the brightest bits. You can let them through in a brief wave if they stand side-by-side, or you can make them line-up in single file which takes longer. Imagine a bunch of people trying to get into a room. If the room is a hangar with a huge door you can open it wide and they can march in shoulder to shoulder. But you only need to open the door that wide for a brief moment. If instead they had to come in through a narrow gate they would have to file in, one behind the other, and the door would need to stay open for longer. So your basic controls of how long and how bright are linked together.
The shutter and aperture work together and can be changed in step, so that you can choose to have a wide group of photons walk quickly through the shutter or a narrow line of photons file through slowly. Holding one of them constant and varying the other controls the total number of photons that get through, making sure you have enough and avoiding over-filling any of the buckets. Setting the right combination to suit the subject and the sensor or film is called the exposure, because it is how much light you expose the film or sensor to.
So this explains the two main controls you have over a camera: shutter speed and aperture. The shutter controls how long the door stays open, the aperture controls how wide you open it. Why you might choose a narrow aperture over a wide one is something I will come to later.
How do you choose the right combination of shutter speed and aperture? Some cameras have an exposure meter built in that will tell you when the combination is right. Some cameras are automatic and will set the right combination themselves. But that is what they are doing: sorting out the size of aperture and the speed of shutter to give just the right amount of light.
Now, the thing that may seem illogical but is true, is that everything in photography works in powers of two. So the step between one shutter speed and the next is to double or halve the amount of light (depending on whether you change to a slower or faster speed). The aperture doubles or halves the area of the hole. This is how they match: doubling the size of the hole by opening the aperture can be balanced by halving the amount of light by changing to the next fastest shutter speed. Why doubling and halving? Just accept it for now – it’s to do with physics and the response of the film or sensor. What it means though is that both of the controls have the same effect: a single change in one of them can be exactly balanced by a single change in the other. You don’t need to do sums and try to work out what a 10% bigger aperture means: every major increment is twice or half, and the same applies to the shutter.
Let’s get back to the aperture and the lens. Imagine you are looking at a tall fence with a small knothole in the wood. If you put your eye right up to the hole you will be able to see almost everything on the other side of the fence. Stand further back and you will see a small patch of the scene beyond. Lenses and their apertures work like holes in a fence. Aperture is just another word for the size of the opening in the lens.
You will have seen that lenses have a size or distance printed somewhere on them as a certain number of millimetres (or inches if you have an old lens). This is called the focal length of the lens and is the equivalent of how far away the knothole is from the film or sensor. A small number means that it is close and you have a wide angle of view through it. A large number means far away, so a narrow angle of view. In photographer talk, for a wide angle view you want a lens with a short focal length. To isolate a small part of a scene or something at a distance, you want a long focal length. How long is long and how short is short? That depends on the size of your film or sensor. A great big wide sensor can capture right out to the edges of what comes through the lens. A smaller sensor or film can see only a central portion of what is available. So what might be an extreme wide angle lens for a big camera might give a much narrower view on a smaller one.
Don’t worry about it for now – the lenses or zooms that fit your camera will have been chosen to provide the range of wide to narrow angle of view. Where the theory comes back in though is that a hole close-by lets more light through than one further away. You can try this at home. Pull the curtains in a room during daylight but make a small hole at the join – a short length of toilet roll core perhaps. Look at the wall furthest from the curtains and you may see a dim upside-down image of the world outside. Hold up a piece of paper on that wall and then walk towards the curtains. The image or light on the paper will get brighter as the light that previously had to cover the entire wall is captured by a smaller piece of paper.
This could be a problem (but isn’t). If the amount of light reaching the film or sensor depends on the focal length of the lens, how do you calculate the exposure? You might get everything set nicely with a wide angle lens and then swap to a longer focal length (narrower angle of view, magnifies a small portion of the scene). The aperture is further away, so it will be darker. What do you do? Simples. A small hole close up is the same as a wider one further away. They are like the cows on Craggy Island. So instead of measuring the size of the whole, we express it as a ratio. The ratio is the focal length of the lens (how far away the knothole is) divided by the diameter of the hole. Very conveniently, this gives us a number that tells you how much light is getting through, whatever the focal length of the lens. So if the diameter of the aperture was 1/4 of the focal length of the lens, then every lens in the world that was set to the same ratio would let through the same amount of light. But because we are photographers and love jargon, we don’t call the ratio 1/4. Instead we call it f/4, usually shortened to f4.
So swapping my superwide 15mm lens for a superlong 400mm lens means that, providing they can be set to the same aperture number, I don’t need to change the shutter speed. And if the long lens is f8 while the wide lens was f5.6, I can compensate by slowing-down the shutter by one step. The biggest opening that a lens can make is also expressed as the same number and lets you compare two lenses. Think of a huge barrel of a lens with a large diameter bit of glass at the front. It gathers a lot of incoming light. Then think of a skinny wee lens with a small bit of glass at the front. It gathers and passes through less light. But we use the same number that we use for apertures to express how wide the lens is compared with its focal length. The lens will have printed on it the widest aperture it can provide. Using ratios to express the size of the aperture means that we don’t have to do sums.
So why the weird choice of numbers? For a start, think of the numbers as ‘one over’, so 8 really means 1/8 and 11 means 1/11. Nobody prints the 1/8 or 1/11 on the lens to save space. It’s part of the lore of photography and it makes us special. Shutter speeds are the same – the dial might say 2000 but it really means 1/2000. So the bigger the number, the smaller the hole or the shorter the time.
That still hasn’t explained the strange choice of numbers for apertures. What we are trying to do with apertures is make each setting double or half the area of the next one, because it’s the area of the hole that matters (it’s how wide we open the door). Think back to school – the area of a circle is pi times the radius squared. So to double the area you don’t multiply the radius of the hole by two, but by the square root of two (call it 1.4).
Imagine a lens that had the diameter of the front element equal to the focal length of the lens. We would call the maximum aperture of this lens f1. The next setting on the aperture that lets through half as much light would be (one over (f1 times 1.4)) or f1.4 (as it would be marked on the lens). The next smallest hole, letting through half as much light again, would be (one over (f1.4 times 1.4)) or f2.
Basically, the aperture numbers ascend in multiples of the square root of two. This means that each major setting is half or double the area of hole. There is no reason for choosing the particular ratios that give rise to the aperture values of 1.4, 2, 2.8, 4, 5.6, 8, 11 etc but the camera industry has mostly decided to settle on these. Older cameras may have different numbers, but the apertures will always be in multiples of 1.4 and the shutters speeds will always double or half. So in the picture of the old lens above, the biggest they could make it was f6.3 (so it has a diameter of 1/6.3 of its focal length of 105mm), but the rest of the aperture scale uses the normal range of numbers.
Basically, don’t fret about the weird aperture numbers. Each major setting is half or double the light. Every lens that is set to the same number lets through the same amount of light. See? Powers of two.
Back to exposure and the buckets. Why not blast the film or sensor with light to make sure that you capture enough light from the darkest areas? So the most each bucket can hold is 1000. As more and more of the buckets are filled, they can only show solid white. What you would (usually) prefer is that one or two buckets had 1000 photons, but there were some with 999, 998, 997 and so on. These will show in your picture as subtle tones and texture in your brightest areas. So you chose an exposure that only just fills the brightest bucket and lets the others give you tonality.
Same at the shadow end – you may only want a few buckets with a single photon in and loads of others with 2, 3, 4, etc. This gives you the detail and texture in the shadows.
Anyone with a digital camera can now feel smug. If your camera can display a histogram, what you are seeing is how many buckets contain various numbers of photons. By convention the fullest buckets are on the right of the graph and will correspond to the brightest areas in the subject. The height of the graph at each point is the number of buckets that contain that number of photons.
And this is how you decide on the right exposure with a digital camera. I could give this picture (above) a bit more exposure – a slower shutter speed or a wider aperture – because there is still some space at the right side of the graph. I don’t have a single bucket close to full. The nearer I can get to getting as many photons in the buckets as possible, the more detail and information my picture will contain. You can always choose to deliberately hide detail later (like Michael Caine’s suit), but if it’s not there to begin with you are stuck with only having that choice.
And if you shoot film? Experience and practice, but greater skill and rewards.
There is one remaining control, and that is the sensitivity of the film or sensor. Imagine if the buckets were wider – each bucket stands more chance of capturing photons and fewer of them would zip through the gaps between the buckets. So overall you would be able to let less light through but capture more of it. Film can be made with wider buckets so it is more sensitive to light. The down side of this is that the wider buckets are more visible in the final picture as grain.
Taken to the extreme, the grain gets in the way of fine detail. This is why you would choose a low-sensitivity film for pictures where you want fine details and smooth tones.
What happens with digital cameras is not so much the use of wider buckets, but using an amplifier to boost the signal. If the most that any of your buckets could capture was 500 photons, then using an amplifier to double the signal would get you back to a normal result. The down side of this is noise. If you ever had a home hi-fi with an amplifier, you may be aware of the hum you could hear if you turned the volume up when nothing was playing (OK – I only ever had cheap hi-fi kit). In a digital camera this is odd specks of colour in what should be a smooth tone.
Digital cameras are getting much better at this and now offer amazing sensitivity with very little noise.
The sensitivity of a film or sensor is (you guessed it) represented with a number. In this case it is the ISO number. The smaller the number, the less sensitive the film or sensor is to light. Less sensitive means it will capture smoother tones, with less grain or noise. More sensitive means you need less overall light to get a picture, at the expense of some grain or noise. 50 to 100 ISO is considered slow – meaning less sensitive. It needs to see the light for a longer time to be properly exposed, which is why we call it slow film. 400 ISO and above is considered fast – meaning more light-sensitive. You pick your film (or sensor setting) to match the conditions you are shooting under or the effect you want to get. And guess what? Film speeds come in powers of two. Swapping from a film or sensor at 100 ISO to one at 200 ISO means that you can close the aperture by one increment or speed-up the shutter by one.
So now you can play the combinations of sensitivity, shutter speed and aperture. An old-fashioned light meter might explain how the combinations work together.
The meter above shows the film speed set at 100 ASA. ASA is what we used to call ISO, so this is a slow 100 ISO film. The meter says there is a light level of just a tad over 80 – you can see it on the top of the dial. Setting the meter dial to match the reading of 80 gives a set of shutter speed and aperture combinations around the bottom of the dial. These are all equivalent, so I could choose 1/250 shutter and f5.6 aperture; 1/125 shutter and f8 aperture; and so on. If my camera could do it, I could use one of the extremes like 1/2000 shutter at f2 or half a second (1/2) at F64. And if the amount of light changed, or I swapped for a film or sensor with a different ISO, I would get a different set of shutter/aperture combinations to match.
Remember above when I said that I would cover the reasons why you would choose one shutter speed/ aperture combination over another? Right then…
Despite what the manufactures may say, no lens is perfect. They are designed to do the best they can to focus all of the light perfectly onto a flat film or sensor.
Look at that cone of light on the right of the lens, focused onto the film or sensor. The light comes down to a point and then widens out again. The perfect point of sharp focus is when the tip of the cone just touches the film or sensor. You focus a lens by moving it in and out, closer or further from the film or sensor. What you are actually doing is moving that cone of light so that the sharp point of it falls on the film or sensor.
Imagine that row of buckets again. The tip of that cone need not be perfectly sharp – it can be allowed to get as wide as the mouth of the bucket. Anything smaller than the bucket is still going to be recorded as one bucket’s worth. So there is actually a range of focus over which the picture will still look sharp. This depends on how big the buckets are, which is why low-sensitivity films with their small buckets can look sharper. It also depends on how large you make the final picture – your eye can only see down to a certain level of detail so anything sharper than that is lost. So this means that even if you focus the cone of light perfectly on the film or sensor, some of the parts of the scene that are not strictly focused still have light cones narrow enough to fit into a single bucket. This makes them look sharp too. What this means is that, depending on the film or sensor, the final enlargement and the quality of the lens, there will be a range of distances over which the subject looks sharp. This is called the depth of field.
The lens in the picture below is wide-open: it has a wide aperture. The cone of light focused onto the film or sensor comes in at a sharp angle and diverges again quickly. So the sharp point of the cone becomes a large disk as you move away from the point. The steep angle of the cone makes this happen rapidly as you move away from the sharp point.
Now imagine that you close-down the aperture of the lens.
The cone is now sharper, so it changes size more slowly as you move away from the point. This means that more of the subject appears to be in focus. This is why pinhole cameras, that have a tiny aperture, make it appear that everything is in focus (or as sharp as it can be for a pinhole camera). A wide-open aperture gives a shallow depth of field and can isolate your subject as sharp against a blurred background. A small aperture gives a large depth of field and can make everything in the picture look sharp.
So there you go: fast shutter speeds will minimise blur from your hands shaking or things moving. Open apertures give a shallow depth of field and closed apertures give a greater depth of field. Don’t over-fill your buckets but do get them as full as you can. And everything works in powers of two. You know I said a bucket could hold 1000 photons? It’s actually 1024 (two to the power of ten).