The Big Hullabaloo about Exposure Compensation in DSLRs – My Point of View
As someone who is passionate about photography, I tend to read a bit and also put some of those ideas into practice in the field, to experiment with settings and ideas. We do tend to get into debates as well sometimes on shooting techniques, camera settings (other than the eternal favourite – Gear) etc. Things like using High ISO, shooting in Aperture priority vs Shutter priority vs Manual, evaluative metering vs spot metering etc. Every photographer has their comfort zone I believe, with respect to settings… which is why the variety of opinion.
As photographers using DSLRs to take images, apart from the critical aspect of composition, we strive to work on the exposure triangle, i.e. Aperture, shutter speed and ISO values to get the perfect exposure for our images.
That brings us to an interesting point. Why exposure compensation? When we did all that we did to get the camera to get us the best possible exposure values for the scene in the first place, why would we want to go and change those values to get some other exposure value? That defies logic at first thought.
Essentially, depending on one or more parameters that we set (could be Aperture or shutter speed, and ISO, generally a combination) the camera’s light meter measures the light and the tonality available in the scene and based on the same, changes one or more required parameters to give us the best exposure value for the scene as it found out. DSLRs are pretty intelligent and accurate today to do a good job on this, in a variety of situations. Particularly when there is more or less uniform lighting /tonality across the scene (it could be bright light or low light but mostly across the scene), most of the times the camera’s metering can give us the ideal exposure. Hence, again the question – why is there a need for exposure compensation / bias? And what is exposure compensation by the way?
Exposure compensation is a function in your camera that tells the camera to override the exposure values that the camera has metered and go by the values that the user is instructing it to go with. How does it work? Why would you as the user do that and when would you do that?
Let’s understand how the camera’s metering works. The cameras are manufactured with 18% grey color tone as their reference point (universally chosen based on ICC color profiles…that’s a separate topic of discussion) and when the camera meters a scene, it sees all the tones in the scene in reference to that 18% grey tone. Based on how far off or how close to that value the tonalities in the scene are, it calculates (again there are various modes of metering that will have a difference in how the camera calculates the area of coverage of the tonalities.. that’s a topic for another day) and arrives at the exposure values, hence it adjusts one or more parameters of the camera’s exposure triangle settings (either shutter speed or Aperture or both, for a given ISO value) to provide the required exposure, which again I say, it does pretty well in most of the uniform lighting conditions.
However there will be many times when you will not be faced with the ideal situation and in challenging conditions, the camera’s metering can be fooled. For example if you have a subject (say a person) to shoot on a very sunny day, say on a field of white snow, or backlit by the harsh sunlight, if you go by the camera’s metering, your subject will be severely underexposed. This is because the camera sees a majority portion in the scene as very bright tonality (white snow or the harsh backlight) and calculates the value based on that, and in order to avoid clipping the highlights, underexposes the image to the calculated extent. As a result the subject (the person) is severely underexposed and looks dark in the image. Conversely, in the case of the subject being placed against a large dark background (say forest in the background), the camera sees a majorly dark tone and to avoid clipping the shadows, overexposes to the calculated extent to get the correct exposure as per it. As a result the subject gets overexposed.
Here is an illustration of how the subject was underexposed in a sunny snowy background, the photographer here going by the camera’s metering:
Image shot based on camera’s metering – subject comes out underexposed
These are challenging situations for the camera and hence the photographer in you has to step in and override the camera’s metering values and instruct the camera to go by the values you recommend, using the camera’s exposure compensation function. This is essentially the crux of the story.
This below illustration shows how the photographer should have overexposed to get the right exposure for the subject (background would come out overexposed, but based on requirement could be corrected in post processing):
The subject being visible clearly in the above image, now the background exposure can be corrected in post processing, as illustrated in the image below:
Background overexposure corrected in post processing – now you have an image with a correctly exposed subject and a good background.
Exposure compensation (EC) works to alter different functions based on the mode that you are shooting in. EC works in the Av, Tv and P modes of the camera (I will focus of Av and Tv mainly as the maximum used modes by photographers) and does not actually work in the Manual mode (the metering only gives the value, you will have to change the parameters yourself in manual mode). One point to note – I am a Canon shooter so my jargons maybe aligned to canon vocabulary, Nikon users please use the corresponding Nikon jargons.
Let me first get over with the manual mode. Many people will probably dispute the fact that EC does not work with manual mode. The fact is that in manual mode you set all the 3 parameters (Aperture, shutter speed and ISO) yourself, you control all the aspects to get the desired exposure. Hence there is no point for EC here, you will yourself always change the values of either shutter speed or aperture or ISO to determine the best exposure for you.
Having said that, more recent cameras from the Canon lineup (1DX2, 7D2, etc) or many Nikon cameras do have the exposure compensation functionality in manual mode when you use Auto ISO. (Point to note, Canon 5D Mark3, 6D etc. do not have option for exposure compensation in manual mode at all) A case where this is useful – for example when you are shooting a bird in flight and you want a particular shutter speed and a particular Aperture to remain constant (that’s why you chose manual) but there is lot of harsh light back-lighting the subject, you have the option of changing the ISO value based on change in exposure compensation. In essence it is then not strictly manual, but helps you achieve your objective.
Exposure compensation is best used in the Av or Tv mode. For example, If you are in the Av mode and decided on an Aperture value of 6.3, and you have set an ISO of 800, the camera is giving you a shutter speed based on its metering, of say 1/200. If you feel that subject in a darker tone compared to the background, and you dial in a EC of +1 stop (EC values can be modified by 1/3 or ½ stops by most cameras), you are essentially telling the camera – I understand how you have metered but I want you to overexpose by +1 stop, since I feel that will expose my subject correctly. The camera then being on Av mode, will keep the f 6.3 constant, but reduce the shutter speed to 1/100 (effect of 1 stop halves or doubles the values accordingly), allowing a full additional stop of light to overexpose the image. You get the subject rightly exposed (you can correct the overexposed background in post processing if that is important). The reverse will happen if you dial in the EV button for exposure compensation, underexposing it by -1 stop (the f6.3 will remain constant but the shutter speed will double to 1/400 to reduce one full stop of light), this will allow your subject in a dark background to be exposed correctly (and the background can be corrected in post processing if required).
If you are in the Shutter priority mode (Tv) and have set the shutter speed to say 1/200, for the same example as above, for +1 stop exposure compensation, the camera will over expose by 1 stop, here the shutter speed will remain constant at 1/200, but the Aperture will increase to f5.6 to allow one full stop of light into the camera, thus exposing the subject correctly (bg may need to be corrected in post processing if that is required). Similarly for -1 stop underexposure, the shutter speed will again remain constant at 1/200 but this time the aperture will reduce to f 8 to reduce one full stop of light, thereby exposing the subject again properly.
I will provide another set of illustrations, this time the subject in a backlit background, which should provide you clarity on this issue.
In the below image, the photographer went by the camera’s metering (which sees a lot of light in the background and meters the scene accordingly);
Image shot as metered by the camera
This next image is where the photographer overexposed by +1.3 stops to get the correct exposure for the subject.
Image shot with +1.3 stops overexposure, subject comes out correctly exposed
This next image is the final output by correcting the background overexposure, where you get to see the background as well as subject as they should have been seen in that scene, correctly exposed.
Post processed image with the right exposure for subject and background…
depicts the scene as the photographer actually saw it while shooting the image.
A similar illustration, now for a bright subject against a dark background.
The image below is shot at how the camera metered the scene (in a home studio, everything other than the subject was dark and subject was well lit). You can see that a part of the hat is completely blown out due to over overexposure, and even the subject on the whole is quite bright.
Image shot based on how the camera metered the scene
This next image below illustrates the same image shot -1 stop underexposed to compensate for the bright subject. Now when you see that Hat, you will see it is perfectly exposed, but you will also see the blacks go even more dark due to underexposure (if you see the model’s hair on the left part you will notice that its almost merged with the black background). The shadows on the model and the overall background can be restored to the correct exposures in post processing, but in this image you do not have any blown out highlights.
Image shot at -1 stop underexposure
This next image depicts how the scene actually was. It reflects the correct tonality of the model and her hair, the right background as intended, and no blown out highlights in the image.
Post processed image with the correct exposure for the subject as well as background
This is in nutshell, how exposure compensation will help you achieve the desired exposure in challenging light conditions.
This brings me to think about another pertinent question… Exposure compensation vs ISO, often a topic where one is lost. In low lighting conditions do I increase the ISO or dial in my exposure compensation to get the best exposure? Same applies to brightly lit conditions. Changing either of these could alter my other parameters, namely Aperture or shutter speed or both.
Exposure compensation on the other hand, is not the 4th physical attribute, it is more of an instruction to the camera to override the camera’s light metering and use the given values. It affects one of three physical attributes mentioned above to achieve the desired result, changes the shutter speed in Av mode, changes the aperture value in Tv mode and can change either Aperture or shutter speed or even ISO in the P mode.
ISO (increase or decrease) does not impact camera’s metering, hence it does not impact over or under exposure. It only helps you alter values of Aperture and/or shutter speed that you might want for any particular scene or action.
So essentially in a scene, you would use ISO only to compliment the achievement of a desired Aperture or shutter speed or both, and not based on whether the scene / subject is underexposed or overexposed. You would use exposure compensation to do that.
Yes, it is important to understand the fundamental reason for you to make a change in settings in your camera and what it actually impacts. This has to do with learning the right shooting technique… and not be always dependant on the gear to carry you through. I say this because the high-end DSLRs do come with lot of bells and whistles…..and the tendency of the photographer is … because I have this feature let me use it … not judging whether it is appropriate or whether the scene needs it.
A classic example of that is the tendency to use very high ISO with high-end cameras. Since these cameras handle digital noise very well even at very high ISO levels, the tendency is to bump up the ISO very high to achieve a high shutter speed. Reason – I am preventing motion blur or any blur due to camera shake. Example, you are shooting in Av at f 6.3, you bump up the ISO to 3200 to get a shutter speed on 1/2000. What are you shooting? A tiger sitting idle under a tree (maximum it does in a while is a head turn or a yawn, which again you are shooting at 14 frames per second). I would say your fundamentals are not strong, you are dependent on your gear to a large extent to get the images that you are getting. What we need to understand is that not all photographers can go and afford a full-frame camera, let alone a 1DX2 or equivalent in any brand. Does it mean he/she will never get the perfect and sharpest shot of the tiger? The high-end camera guy will probably not have the answer, because he has not done it (this is the general scenario mostly in amateurs, of course there are the professional photographers who use the highest end equipment but their basics are very solid).
What will happen if I am forced to use a gear which can tolerate ISO at 1600 at best, better still if I can use ISO 800 (without getting any digital noise)? Will I be able to get away with a shutter speed of 1/500? In the example given above, most likely yes. Hence the need to know these fundamentals well, and then if you graduate to a high-end equipment it is your discretion on the best possible way to use the parameters.
While the above paragraph uses ISO as an example, so is the case with exposure compensation. We tend to underexpose the moment we see a little bit of shadows, irrespective of the tonality of the subject in reference to the background. If the subject is brighter than the background I can understand underexposure maybe required, but mostly I see people underexposing even when there is a uniform lighting situation in the scene….in a uniform lighting scenario, your camera is anyways metering correctly, why do you need to override it? If your fundamentals are not correct on lighting and tonality and you cannot determine approximately how much difference is there in the different tonalities in the scene, how do u expect to set the correct underexposure (-1/3 or -1/2 or -1 or more)? The same holds good for overexposed scenes as well. When you know these fundamentals by going out in the field and experimenting (that’s the only way you get first-hand experience and expertise), you can even mix and match ISO and exposure compensation to get a particular shutter speed or Aperture as the case maybe, without having to go to the extremes / high values of the same.
I hope I have been able to make sense on the subject, and though slightly technical, is easy enough for you to grasp.
© Atanu Chakraborty
Atanu is an avid traveller and a passionate photographer. He is a wildlife photographer and also loves to shoot landscape. He is based in Bangalore and is always happy to share his knowledge on photography and more. You can reach him at firstname.lastname@example.org.