Photographers frequently refer to stops but what do they mean and how are the scales of aperture, shutter speed and ISO related to this concept?
A stop is simply short hand for changing the amount of light by a factor of 2, either doubling or halving the light. If I say something is underexposed by one stop it means that a proper exposure would have needed twice as much light hitting the sensor. Luckily we do not always have to manipulate the light in the scene to change the amount of light hitting the sensor, in the studio this might be appropriate, but for a landscape it would be much more difficult.
Cameras have three knobs that you can use to change the amount of light hitting the sensor, the aperture the shutter speed and the ISO (film speed, or sensor sensitivity to light).
Shutter speed
There are different types of shutter mechanisms, but their purpose is always the same, to expose the sensor for an exact length of time to light. Shutter speeds are typically stated in seconds for example 2, 1, 1/2, 1/4, 1/8, 1/15, 1/30, 1/60, 1/125, 1/250 seconds. Each shutter speed is a factor of 2 larger or smaller than its neighbor. Moving from left to right you must multiply by 1/2 to get the next value, you can see that we have 1/15 instead of 1/16 and 1/125 instead of 1/120, these are changed for human consumption, as are some of the 1/2 and 1/3 stop increments.
Shutter speed is probably the easiest to understand. If I let light hit my sensor for twice as long I get twice as much light, changing the exposure by one full stop. To change the exposure by 1/2 stop I need to let in 1.5 times more or less light, so the shutter speed must change by a factor of 1.5. If I want 1/2 stop more light than a 1 second exposure then the shutter speed needs to be 1.5 seconds.
ISO
ISO values are typically 50, 100, 200, 400, 800, etc… In the case of film they correspond to the physical size of photosensitive cells in the film, larger cells can collect more photons in a given length of time. Exposure of the cell is dependent on the number of photons collected so larger cells are more sensitive. They are also physically bigger so they reduce the fine detail in the photo adding a distinctive film grain aesthetic, which may or may not be desirable.
Digital cameras use silicon chips to record light, photons that hit the sensor create an electric charge on the pixel, more photons means more charge. By measuring the amount of charge you can determine how much light there fell on the pixel. These chips are quite fascinating and as an astronomer I know quite a bit about them, but it’s best to leave that for a possible future post.
The point is that their sensitivity can be changed… or rather you can read the amount of charge sooner and apply a larger gain to the signal. If you do this your measurement of the amount of light is less accurate since your statistical sample (the amount of charge) is much smaller. This creates noise, but not the kind of noise film has! This noise is known as Poisson noise, shot noise, or counting noise it is not considered aesthetically pleasing and should be minimized. Of course this is a purely cultural perception which you are free to challenge :)
To double the amount of light recorded in a given length of time at a fixed aperture you need to double the ISO. The numbers listed progress by factors of two 50, 100, 200, etc… So that an ISO of 1600 is one full stop more sensitive (or faster) than an ISO of 800. Many cameras offer 1/2 or 1/3 stop intervals which means you might see ISO values in between these numbers. The numerical values of these intermediate stops are perhaps a bit mysterious, the 1/3 stop values between 100 and 200 are 125 and 160. You might have guessed the numbers to be 133 and 166 but that’s not the case.
The reason is that the ISO numbers do not relate linearly to the amount of light sensitivity. To find the value of ISO for a 1/3 stop increase you must multiply the ISO value by 2 to the power of 1/3 or using the ‘^’ character to represent the exponent we have:
2^(1/3)=1.25
Likewise the 2/3 stop value is:
2^(2/3)=1.60
Now we see where those ISO values come from. ISO 100 times 1.25 is 125, which is 1/3 stop more sensitive than ISO 100. The 1/2 stop value between ISO 100 and 200 would be 2^(1/2)=140.
Aperture
Things get a little tricky here because people don’t usually understand what aperture is. The aperture (A) is defined as the ratio of the focal length (f) to the diameter (D) of the opening at the front of the lens. Or mathematically A=f/D. The values for this scale are a little more perverse (at first): 1,1.4, 2, 2.8, 4, 5.6, 8, 11, 16, etc…
The aperture changes the amount of light by changing the size of the opening at the front of the lens. With a constant focal length, if I want to change the aperture by a factor of 2 I need to change the diameter of the opening by a factor of 2. Smaller openings mean larger apertures (aperture is inversely proportional to the diameter) which means that the value f/16 allows less light to enter than f/2. That alone is useful to understand, but where do these bizarre numbers come from?
Changing the diameter by a factor of 2 does not change the light hitting the sensor by factor of two! The light hitting the sensor depends on the area of the opening, which is pi times the square of the diameter. That means if I want to double the area I need to change the diameter by a factor of square root 2, which I will write as sqrt(2). The square root of two is 1.4142… and now we start to see where the aperture scale comes from.
Let’s start with a lens which has a focal length equal to the diameter of the opening. In this case the aperture is 1, to half the amount of light we must divide the diameter by sqrt(2) which is the same as multiplying the aperture by sqrt(2), giving us f/1.4142 the next multiplication of sqrt(2) gives f/2.0, then f/2.828 etc… if we round off some of the numbers along the way to make the sequence more pleasing we find:
1, 1.4, 2, 2.8, 4, 5.6, 8, 11, 16, 22, 32, 45, 64
I find it useful to remember that this scale contains all the even multiples of 2, and between them is two stops worth of exposure, so that f/8 allows four times more light then f/16 and a quarter of the light at f/4. If I remember that f/4 and f/6 are two full stops apart then I can more easily remember that f/5.6 is between them, exactly one stop away from either. There are again 1/2 and 1/3 stop increments for aperture. In this case if I want to find f/8 + 1/3 stop I need to multiply 8 by sqrt(1.333), which gives f/9 (when rounded).
Usefulness
If it isn’t clear why you should try to understand these concepts then here’s an example of a typical problem. You are shooting at night, you expose the scene for 30 seconds at f/16 (for depth of field) and at ISO 400 (the largest ISO you’re comfortable using). You find that the picture is too dark…. now what!?! You have to switch to bulb mode and expose for even longer but how long? Here’s what you can do, set the ISO as high as it will go maybe 1600 and the aperture wide open, maybe f/2.8, here you find that a 2 second shutter is needed to properly expose the scene. This should be all you need to calculate the proper exposure time for the shot at f/16 and ISO 400.
ISO 1600 is 2 stops more sensitive than ISO 400, and f/2.8 allows 5 stops more light than f/16. By changing the settings to ISO 400 and f/16 we loose 7 stops of light, the only way to make this up is to increase the exposure by 7 stops. We multiply 2 seconds by 2 seven times, the result of each multiplication is: 4 sec, 8 sec, 16 sec, 30 sec, 1 min, 2 min, 4 min. So our shot should be 4 minutes at f/16 with ISO 400! This is very handy to know and probably faster and less frustrating than taking many incorrectly exposed shots trying to find the right shutter speed…. probably…