In this article I’d like to show you how switching to the manual (M) mode in your camera is the single most important factor in getting consistently well-exposed photos, especially if shooting outdoors. Please note that if you're not familiar with the different modes (M / A / S / P) or how a combination of ISO / aperture / exposure time determine the final exposure, read this article first!.
Before I started using this technique my process was always as follows: 1. take a photo 2. review the histogram (or check the under- and over-exposed areas). I could never be sure to have taken a good photo without doing the histogram review. This was incredibly irritating because of the amount of time needed to take a useful photo (more for the review than the shooting) and also because this meant that many “decisive moments” were lost.
So, what’s the problem with the other modes (aperture-priority, speed-priority or program)? As always the problem is that we tend to be lazy and we think or hope that the camera will do the job for us. Well, no it won’t. The manufacturers have spent lots of time and effort developing ever smarter algorithms to determine the correct exposure but they will never substitute the photographer (this is why photography is considered art). It’s true that in many cases they’ll do a relatively decent job, for example on a cloudy day, or on a noon of a sunny day where you have few shaded areas. Fantastic, but first of all such conditions don’t always happen (and for landscape photography, for example, are quite boring), and second of all, even with a uniformly lit world around us you may still want to change the exposure for a special effect.
Autumn photos of the Verzasca valley, © Rafael Zwiegincew. In my opinion the first shot is more interesting because of the shaded areas, the second one still nicely shows a beautiful autumn but is not as “deep”.
More importantly, though, let’s take a look at the best conditions for outdoor/landscape photography: some sun and lots of shaded areas. These are the conditions that often make a difference between a boring landscape photo and an interesting one.
Here are some shots which in my opinion work only because the scene is partially sunlit and partially shaded.
Photos of the Verzasca valley with lots of shade, © Rafael Zwiegincew
The same is true for different types of photography, e.g. in the city. The following shot would be badly overexposed with the light meter determining the exposure. It would also be boring if not for the shade.
Here are some other typical conditions where automatic exposure would likely fail, in night photography. With so little available light you want the shots to be generally underexposed, otherwise you’ll get camera shake or will need to boost the ISO too much, ruining the picture. Underexposing night photos like these is ok and it gives the photos a nice atmosphere:
Night photos of Siena, Italy, © Rafael Zwiegincew
Winter with lots of snow is another example where the exposure could be badly guessed by the camera (this time, tending to underexpose). These are very delicate conditions, you want the snow to be white and very bright, but not overexposed.
Winter photo, Blenio Valley, © Rafael Zwiegincew
Especially with such mixed lighting conditions the cameras’ light meter will be very inconsistent. Depending on the scene and the position of the sun, the exposure suggested by the camera will be random across a series of shots. For example, in areas with lots of shade, you’ll get ugly, overexposed sunny areas. Other shots will be exposed perfectly. This is why you’ll find yourself checking the exposure after each shot.
The best solution, in whichever lighting conditions, is to use the manual mode. Here you simply need to set the aperture/exposure time once for given conditions. It’s best to do your test shot on a very bright area, e.g. a building lit by the sun, making sure you don’t overexpose any large area. Don’t do your test shot pointing at the sun because this will make all your photos underexposed. Now that you’ve set the aperture/exposure time so that very bright areas will be also very bright in the photos but not overexposed (barely), you can walk around and take shots of both bright, dark, and mixed areas. They will all turn out very interesting and correctly exposed! Of course if you want to take a shot e.g. in a narrow street with not even a small sunlit area, the shot will be all underexposed and ugly, but either way in such conditions it’s difficult to come up with an interesting photo.. I suggest you then temporarily switch to the A / S mode. Note that you will need to check and change your settings (aperture/exposure time) only after the lighting changes, perhaps after half an hour or an hour, depending on the time of the day. Until then, you won’t even need to verify your photos..
One potential problem that will come up: what if you want to change the aperture or exposure time to accomplish a specific effect, e.g. blur the background? Well, it’s true that in such case a dedicated mode (A / S) would make sense. Still, considering the exposure problems mentioned earlier, you may find that sticking with the M mode will continue to be a good idea. You can do one of the two things:
1. Count the steps of the aperture you’re changing and then change the exposure time by the same number of stops, e.g you go from f/8 to f/2.8 which means 3 stops. To compensate, you’ll need to shorten the exposure time also by 3 stops (e.g. from 1/60 sec to 1/500 sec). The better the camera, the easier this technique will be. The best are professional cameras with 2 dials. You can then turn the dials at the same time, feeling the number of clicks they make. Even if each such click is set to half a stop or 1/3 stop, this will work perfectly and let you change exposure time and aperture freely without changing the resulting exposure. I have to admit that with entry-level cameras with a single dial, this technique will not be convenient.
2. You can point the camera to a specific scene and keep it fixed there. Now, check the light meter in the camera, and change your aperture to the desired value. Now, start changing the exposure time and keep going until the light meter shows you the same exposure. E.g. if you started with the light meter telling you you’re underexposing by 1 stop, then after your change you’ll also need to end up with 1 stop of underexposure.
I tend to use the first technique because the cameras I use always have 2 dials (this is for me one of the most important criteria in a camera).
Now, to demonstrate the point, I took shots from the same place, looking in two opposite directions. The first set was taken in the manual M mode, aperture was set to f/8 and exposure time to the value where the sunlit areas are very bright but not completely overexposed.
The second set was taken using the default settings (Nikon D700) and the aperture-priority A mode.
As you can see the first shot was badly overexposed, the camera was trying to record details in all the shaded areas and the result is really bad. The second shot is also overexposed although not as badly as the first one, even though the shaded areas are even greater. It’s difficult to find out when the algorithm decides to do what, which is the whole point of this article. As you see, setting the exposure compensation to e.g. -1 or -2 won’t help because the results are unpredictable. Still, if you look at the bright house in the second photo, you’ll see overexposed areas so this is not a well-done photo.
To prove the point further, I then took another set of shots some 100m further down the road (as before, single position, just looking at different directions). The first (manual) set of course won’t show any surprises, the exposure is always as it should be, no need to change any settings.
Now take a look at the same shots using the A mode.
The first two photos managed to get a good exposure (the second shot is only slightly overexposed) but the third shot is really badly overexposed. You may not like Nikons but any other brand (or later models) will have the exact same problem: unpredictability. You’ll always end up with some good shots and some bad ones. Usually the problem will be overexposure although traditionally in the winter snowy scenes cameras tended to underexpose because of all the snow tricking the camera. I imagine some of the latest cameras will handle snow ok in the A/S modes but still, why go for this randomness? The more you control the better the result will be. The only exception is when you can manipulate the photo in post-processing. For example, the white balance chosen by the camera is also often off but it doesn’t matter at all as long as you’re shooting raw and not jpeg — in post-processing you can always adjust the white balance. One can’t say the same about the exposure: if you overexpose, the details in these areas are lost forever. If you underexpose, you’re ok as long as you underexpose by a bit (say, no more than 1-2 stops, and even less with high ISO).