In the previous chapters of Photography Basics we saw the three pillars that affect the exposition: Aperture (I and II), sensor sensitivity and shutter speed. We also saw the implications that they have on the picture style when the values of those parameters are changed. But if you recall the examples that were given for all those cases, all of them started assuming that some correct exposition values had already been chosen and we just modified them to achieve the desired effect. Never was explained how to achieve those initial values. That’s what this post is about.
For this post I’m assuming that all work is done on M mode (full manual), which means that the photographer is in charge of everything. For all other modes, semiautomatic and automatic, part of this work is done by the camera with all the limitations that implies. Only in M mode the photographer is completely responsible of acquiring the correct exposition for the picture he wants to take.
Every modern camera (and by modern I include even film cameras back in the 80s or earlier) has a device called a photometer. As the name implies the function of this tool is to measure the amount of light that enter the camera through the lens. In modern cameras an array of photodiodes produce a tiny electric current that the camera can measure. When those diodes are illuminated by light they change the current in a predictable way, so the camera can calculate the amount of ambient light and show us on the dial. In analogical cameras, the light changed mechanical resistance in a spring, which makes a needle move over a graduated scale.
The dial the camera shows is similar to the one depicted on the right. It has a scale that goes from -2 or -3 to +2 or +3 (depending on the dynamic range of the camera, i.e. the maximum number of illumination tones that can distinguish between pure black and pure white). Under it, a dynamic mark moves around depending on the illumination.
Cameras, being as dumb as any other automatic device, need a standard to measure against. Without it the camera wouldn’t know if a specific amount of light is good enough for a situation. The pattern used by the camera is called grey-18, a grey color that reflect the 18% of the light that incises over it, which is very similar to what we would call a “50% brightness grey” in any software (not to be confused with 50 Shades of Grey, cameras and handcuffs have little in common). So what the camera does is to evaluate the amount of light that is entering through the lens and calculate, using the current parameters that are set in the camera, how the exposition of the picture would result compared with a grey-18. If the value is 0, it means that the amount of light that will be recorded will lead to a correct exposition for a subject that reflects 18% of the light. A value of +1 means that the camera will record the double of light that is needed, and -2 means that the camera will record only one fourth (2 ^ -2 = ¼) of the necessary amount of light.
If the picture is taken with the mark in the “+” side, that will lead to an overexposed picture, bright and with washed colors as we are acquiring more light than necessary. If the picture is taken with the mark on the “-“ side, the picture will be underexposed, resulting in a dark and noisy image because we are recording less light than necessary. An example of those situations can be analyzed in the picture below. So if we are photographing a grey-18 likish subject, and we are in the +1 point (double of the light needed) we will have to reduce 1 stop ISO, speed or aperture in order to achieve a correct exposure (mark on 0). If we have the mark in the -2 value, we will have to increase 2 steps speed, ISO or aperture, in any combination of them, to reach the 0 mark. So, as a summary, correctly exposing for a grey-18 is as simple as leading the mark to the zero point.
But… hey! My picture is not grey…
And now the complications begin…
I tried to make it simple and easy, but you just couldn’t be satisfied with photographing grey walls, and grey skies, and even grey cats… You want also to capture the bright colors of the rainbow, and your family in that trip to the snow, and even that black cat crossing the black asphalt in a dark night of that mysterious black town where your car engine decided to go on strike without previous warning (apart of those little squeaky sounds that happened for the last 300 km).
It happens that the world is not grey, but the camera has no way to know that. The photometer can only compare to that illumination value, so is our responsibility to correct it. For example, we are traveling with our family to that fancy snowed mountain where people ski happily while on holidays. We take our camera from the bag, move the dials until we get our mark to the zero mark, we ask everybody to smile, we ask again hoping that this time they will listen to us… and finally push the shutter button. But the picture that we obtain is not as clear as the snow really is, and instead we get a dim grey snow and an overall dark image. That’s because snow reflects more than the 18% of the light it receives (usually is near 30%, but depends on the kind of snow and how the light reflects on it). The camera assumed that the snow reflects only the 18% and optimized the parameters for that amount. As a result it let to pass less light than necessary (you get grey snow as it had reflected 18% of the light) and you obtain less light than needed. The way to compensate this is to expose not for the zero mark, but for the +1, which means in “camera language”: Capture the double of light than what you would consider the normal. As 30% is approximately the double of light than 18%, this adjustment will work in a much better way. If still the picture is not bright enough, you can expose higher until you are satisfied with the result, always taking care that the snow doesn’t reach a pure white color, where it loses all the details.
The opposite is also true. We are walking in the park and suddenly a black cat appears. Fascinated with it (I don’t know why, but our example photographer gets fascinated by such cat) he takes the camera, exposes to zero and pushes the shutter button. The picture we get is overexposed and brighter than necessary, and our cat appears washed and greyish instead of the real black color. What happens is that the cat reflects less than 18% of the incident light (around 8-10%), but the camera allowed light to reach the sensor until achieved what it considered a 18% grey would be, which is too much light in this situation. In this case, as 10% is approximately half of the light than 18%, we should adjust our exposition values to the -1 mark to get the appropriate exposure.
In order to learn the correct values for many situations the best method is to practice and take loads of pictures, as experience will lead you to the optimal values for your camera. Anyway, approximate values can be deduced by observation. A grey-18 color is easy to memorize. Anything that is darker than it will get a lower exposition value. Black animals usually get -1, dark skinned people usually need a -2/3 exposition, or even -1. We will need to go down to -1&1/3 if they have really dark skin. Asphalt ranges from -1 to -2 and some black fabric can get up to -2 easily. White skin usually exposes correctly from +1/3 to +2/3, snow requires +1 to +2 and a white building wall or the clouds can definitely get up to +2. Memorizing all those values will eventually happen, but meanwhile they can be estimated just by looking at the scene and comparing it to a mental grey-18 representation of the color.
That’s too complicated… I’m sticking to semiautomatic.
Even in the semiautomatic modes this problem arises. In any semiautomatic mode is the camera who decides at least one of the parameters for the exposition. And it will adjust that parameter assuming that what you are taking the picture to is grey-18. So a snow picture with a semiautomatic mode will still yield a grey and underexposed image (except on special “snow” modes, that are programmed to compensate for this particular situation, but not others).
The way to solve this problem is called exposure compensation. When you activate it, what you do is to place a mark on any point of the exposure scale, and the camera will assume that is to that point to which it has to adjust the parameters instead of the zero mark. So in the case you want to take a picture in the snow you will have to adjust the exposure compensation to the +1 mark, and the camera will automatically adjust the parameters to it instead of the zero mark.
If you changed to semiautomatic for not having to learn the values of the different situations, you should return to M mode now, as you’ll have to learn them anyways to get nice expositions. This doesn’t that mean semiautomatic modes aren’t appropriated for other situation, but you will still need to know how to expose correctly.
After all this explanation you’re ready to go outside (or stay inside if the weather is bad, you have my permission) and start practicing how to expose correctly. In the next post some more advanced techniques will be explained that imply to control how the camera measures the light. But in order to understand those concepts, understanding how to correctly expose is necessary, and the funniest way to do it is taking pictures.