When EV sounds like ET....
Written by Thursday, 26 March 2009 00:00
| Article Index |
|---|
| When EV sounds like ET.... |
| Page 2: Using Different EVs |
| Page 3: Scenes with fluctuating EV |
| Page 4: Tricks in the Field |
| All Pages |
Quite recently we had a long discussion with a fellow photographer on exposure and the use of EVs in photography – digital photography to be more precise. We were surprised to find out that although EV is used in our everyday life, very few photographers realize what EV stands for and what to do to get the most out of its - correct - use. These are the topics of discussion in the pages that follow. Our discussion touches on a number of other topics in order to clarify certain issues. These topics are by no means discussed in any detail; so further reading is strongly advised.
What does EV stand for? The million dollar question
Exposure value or EV actually refers to a specific amount of light. In our world, the world of photography, this usually means the amount of light emitted by any specific part of a scene, and has – of course – a unique value for this part of the scene. This amount of light is then measured by the sensor of our camera and is translated into an aperture / shutter speed combination. This combination should / will allow enough of this light to reach the sensor so that a picture correctly exposed for that particular part of the scene can be shot. The EV of the scene will not change no matter what settings we use in our camera. We may choose any aperture – shutter speed combination which results in the same amount of light reaching the sensor.
The term EV is usually used wrongly since what we really mean is the “average EV of the scene” (if using average metering) or “average EV of the part of the scene our meter is pointed at” (if using spot metering). Thus, we have two different interpretations of the term “EV” which further contributes to its poor understanding. Since this article is aimed for photographers, we will use EV to mean the “average EV of a scene”. We know that some may argue that this term is rather simplified (or even simplistic) but we will stick to it for the purposes of this article. For those wondering what a better definition of EV should be here is an attempt to give one: the EV is the average light emitted from a scene in the specific part of space our lens is collecting light from. In other words, it is only the reflected light that we are going to monitor that is of interest to us, while the scene obviously reflects light in all directions.
According to our simplified definition a specific EV corresponds to a specific light level reflected by a scene. A scene with a higher EV value reflects more light (or is being better lit). A difference of 1 EV between two scenes means that the scene with the higher EV value reflects or is being lit by double the quantity of light by comparison to the other scene. All camera meters are designed to capture the average light level, assuming that a scene has the same dark (below average) and bright (above average) points. As with statistics, the average may be the result of many combinations. Thus, we may have an extremely bright spot in our picture and some relatively darker areas here and there. In this case a lot depends on the relative size of the bright area compared to the dark one, its location in the scene, which part of the scene our camera meter is pointed at or what type of exposure mode we use (matrix, centre weighted, spot). If we have to take a picture of a wall painted half black and half white and equal parts of each colour are in the frame then going for the average means that the darker areas (black) will look a bit less dark (a dark shade of grey) while the bright area (white) will look less bright (light grey shade).
If we imagine a picture with a white little church lit by bright sunlight in the right side and somebody wearing black clothes sitting in the shadow under a large tree in the left side of the scene the average means that the church will look less white (very light grey perhaps) while the person in the shadows will be visible – to a degree. Let us assume that we, the photographer, want to focus on the person sitting in the dark rather than on the church. In this case it is obvious that we must somehow tell our camera that we want it to allow more light in the dark areas. This is done by overexposure, i.e. by changing the EV settings of the camera.
Thus, the term “overexpose by one EV” means to allow the sensor (or film) to collect “double the light” it normally would do from this scene. In the real world, this means that the church will be too bright, perhaps all the details in the white walls will be lost, the blue sky will be a pale tint of washed out blue but, indeed, the person under the tree will be much better recorded and that there will be detail on the trunk of the tree.
Similarly, “underexpose by one EV” means forcing the sensor (or film) to record the scene by collecting only half the light it normally would. In the imaginary photo we described earlier, the church would be more correctly lit, there would be a lot of detail shown on the walls although the church would look slightly darker than it did reality. The sky would now be a nice blue colour but (yes, there is always a “but”) the person sitting under the tree would not be visible. The whole area under the tree would be black without any detail.
This is, in short, what over- and underexposure by one or more EV will do to a picture.
Frank Panis, our team mate, sent us the following picture which was taken during sunrise. Frank was obviously after the interesting colours in the sky but also wanted some detail in the shadows. He took a reading from the middle of the scene to compensate for both ends, decided to use a slight overexposure (+0.3 EV) and came up with the following photo in which he managed to capture the nice colours of the sunrise along with the reflection of the trees and the sky in the water.
As you can see in the histogram Frank did not lose any detail in the shadows while the highlights were only partially blown. In this kind of situation it is almost impossible to get both ends right. Frank believes that he should probably bracket this exposure and then blend the two photos manually as opposed to using the high dynamic range feature of Photoshop CS4, which would show more detail in area which was in the shadows. The photo was taken with a Nikon D2x which is known to capture well detail in the shadows. Since Frank always shoots in NEF (which is the RAW equivalent for Nikon cameras) he made a different editing of the photo in Nikon Capture NX2, which is shown below. The tonal range has increased dramatically while the diffused light behind the tree line adds to the picture.
The histogram shows that a lot of information has moved from the extreme right (dark areas) towards the right side (bright areas). By comparing the two histograms we can see that the dark area has a much wider tonal range now, while the bright areas are not affected by this editing.
When discussing over- or under- exposure we sometimes use the expression “f/stop” instead of “EV” (e.g. “overexpose by one f/stop”). As we will see later on these two terms are in fact interchangeable since they lead to the same result. Indeed stopping down the lens by one f/stop (keeping all the other parameters the same) will result in half the light being recorded, therefore this outcome is equivalent to underexposing by one EV.
Modern cameras allow the photographer to handle EVs more accurately since they allow us to over- or underexpose in 1/3rd EV steps, a feature which offers a great amount of control on exposure. Before we take a better look at what EVs can do for - or to - our pictures, it might be useful to say a couple of things about the two settings that determine how much light will reach the sensor of our digital camera or the film emulsion for film cameras.
Aperture and shutter speeds
The amount of light that reaches the sensor through our lens is directly proportional to
- the surface of the opening of the lens which is formed by the aperture blades and
- the duration of the exposure
Clearly the larger the opening, or the longer it stays open (e.g. when we use lower shutter speeds) the more light will pass through to the sensor. It goes without saying that we can use one of many combinations that allow the same amount of light to get to the sensor. For example, a tiny opening will ask for a very long exposure time, while a larger opening will only need a proportionately smaller exposure time for the same amount of light to reach the sensor. Therefore, for the same amount of light i.e. the same exposure, we can increase one of the two parameters (e.g size of opening) and proportionally decrease the other (e.g. duration of exposure).
As far as shutter speed is concerned, things are very straightforward. What you see is what you get. For example, a shutter speed of 1/1000th of a second means that every part of the sensor (or film emulsion) will be exposed to the light for just 1/1000th of a second. There is a bit more to it but in general this says it all. Something you should know (we refer to it only because it affects flash photography) is that the sensor is not exposed to light on its whole if the shutter speed is higher than a pre-set limit, known as highest flash synchronization speed (usually 1/250, rarely 1/500). Up to this shutter speed the first curtain has exposed the whole sensor before the second one starts to cover it again. For any speed faster than that the shutter curtains will form a band which moves across the sensor, exposing each part of it for the pre-set amount of time (e.g. 1/1000th of a second). Though there are some variations to this rule the point here is that for most shots it is easy to determine which shutter speed (the one chosen by the camera or the highest flash synchronization speed is higher or faster; this is important when we chose our camera settings for a particular shot. (1)
Unfortunately, aperture is a bit more complicated than that. The aperture numbers we see on our lenses (usually in the form of these numbers : 1.4, 2, 2.8, 4, 5.6, 8, 11, 22) are not indicative of the surface but of the diameter of the opening of the lens, therefore to get the surface you have to multiply each aperture by itself. To make things even more complicated, the smaller the f/number the larger the opening. Thus, a small f/stop is numbered 22 (or 32) while a large one is 2.8 and a very large one is 1.2. This is because the f/number is actually the result of the division of the focal length of the lens (2) by the diameter of the opening which is formed by the aperture blades. Thus, when you have a 50 mm lens and an f/stop of 1 the opening of the aperture blades will be (50/1=) 50 mm (5 cm). When you have the same lens and an f/stop of 22 the opening of the aperture blades will be (50/22=) 2.27 mm – a far smaller opening.
We said earlier that when we stop the lens up or down by one f/stop we double the shutter speed or reduce it by half accordingly. The relation between aperture and shutter speed is based on simple geometry. If you watch those numbers closely you will find out that each f/stop number is the previous one multiplied by the square root of 2 (√2= 1.41424142 periodic). Thus f/2.8 is f2*√2. This is always true. In the same way f/8 = f5.6*√2, f/11=f/8*√2, f/16=f/11*√2 and so on. Therefore for f/ f5.6 on a 50 mm lens (= focal length / opening diameter) we have a diameter of 8.928 mm or a radius of 4.464mm. For f/8 we have a diameter of 6.25mm or a radius of 3.125mm. The equation for the surface of a circle is πR2 (pi*radius squared). Therefore, the surface in the case of f/5.6 is pi (π) *4.464*4.464 = 62.57 mm2. For f/8 we have π*3.125*3.125 = 30.66 mm2 – roughly half of that calculated for f/5.6. The reason it is not exactly double is that we have made some approximations when assigning the f/numbers and rounded some figures. The same is true if we assume the lens opening to be hexagonal or octagonal.
To conclude, the ratio of the surface of the blade openings of two consecutive aperture numbers will always be 2. Hence, stopping down (or opening) the lens by one f/stop means reducing (or increasing) the amount of light entering the lens by a factor of 2. Thus, for any EV level we have a good number of aperture – shutter speed combinations which allow the same amount of light to reach the sensor of the camera. We also have a good number of ISO settings (sensitivities) to choose from and each different sensitivity (ISO100, 200 etc.) will give us a further set of aperture / shutter speed combinations to use. All those combinations for a specific EV level are equivalent to each other and, once you know one combination, you can calculate all the others even without the camera. (3)
Photographers believe a picture to be worth a thousand words; it is time to show some photos by way of illustration and spare you the time to read the 20.000 words they stand for.
| < Prev | Next > |
|---|