If you use auto mode, aperture priority mode or shutter priority mode, you may have noticed something. Most of the time, your camera does a pretty good job figuring out how to expose a shot, but every now and then—maybe even more frequently than you’re comfortable with—you get a photo that’s really overexposed, or really underexposed. How can this happen in auto mode (or priority mode) and is there anything you can do about it? Read on to find the answer.
[ Top image keep calling, really : sacramento, california (2014) by Flickr user he who would be lost]
Don’t trust your screen … or your meter
But wait! These are the two most important tools you have for getting the exposure right, aren’t they? Well yes, and no. Your meter should always be thought of as a baseline, as a good guestimate for the exposure in any given lighting situation. But all it can ever really be is a guestimate. Let me explain why.
Your camera’s meter can’t evaluate a scene the way your brain can. Oh, it has its ways, but even very sophisticated meters don’t guess exactly right every time, because they are fundamentally limited. At the most basic level, your meter is designed to assume that everything within the frame averages out to the same middle gray tone. So in other words, it’s operating on the assumption that if you were to take the highlights and the shadows and all the tones in between and use that information to come up with an average, you’re probably going to be pretty close to that middle gray shade. And the truth is that that system is usually pretty accurate—most scenes do come pretty close to a middle gray average, and that’s why most of the photos you shoot with your matrix or evaluative metering system come out correctly.
But if you think about it, of course there are going to be scenarios where this isn’t going to be true. Let’s say your subject is a white cat in the snow. Or a black dog sitting in the mouth of a cave. The averages for these scenes are going to be nowhere near a middle gray—and even in less obvious examples, you may still be a stop or two off depending on how many blacks or whites or darker/lighter tones are in the scene. So yes, you can use your meter as a baseline, but you have to employ other tools if you want to make sure that your exposure is exactly where it should be.
Now, some meters are smarter than others. Many modern cameras (especially those in the higher price brackets) actually employ databases of exposure information for common scenes, which increases the odds of that correct guess, but there’s still no such thing as a perfect meter. And the accuracy of yours depends, of course, on which model you have and (unfortunately) how much you paid for it. So you need to also use some of your own knowledge to fine-tune your exposures.
That’s where your screen comes in—or does it? Logic should dictate that you take a test shot, check your screen and then use what you see there to determine whether or not you got your exposure right. And while that plan can also work as a baseline (you should be able to tell from your screen if you’re severely over- or underexposed), remember that your screen is tiny, which means that you won’t be able to see the details. How will you know for sure how much detail you picked up in the shadows? You won’t, and even zooming in isn’t always going to give you the right answers. Add to that the fact that it can be hard to see your screen at all in bright sunlight, and you really can’t depend on it as a way to confirm whether or not you got the exposure right. So what can you do?
Help! My histogram is eating me!
OK not really, histograms don’t eat people, but beginning photographers still find them pretty intimidating. In fact, I know intermediate photographers who find them intimidating, and even the occasional pro who just doesn’t want to go there at all. I mean, we didn’t need them back when we shot film, did we? Why do we need them now?
Just like the digital camera itself, the histogram is a tool. Yes it’s true that we didn’t have them back in the days of film, because in the days of film we did everything blind. You didn’t get to see your photo on a screen after you took it, so you could only use your camera’s metering system and your own experience with similar lighting situations to guess at the correct exposure for any given scene. So although it’s true that with enough experience anyone can learn to get a pretty good exposure without a histogram, in the meantime don’t ignore this simple little tool as a way to improve your photos while you’re on that learning curve.
Where to find your histogram
Cameras don’t all display the histogram in the same way. I have a Canon point-and-shoot, for example, that pops up the histogram before you actually take the photo, which I find infinitely useful. That allows me to make adjustments to my settings before I’ve even made a single exposure, so there’s no need for a test shot. My DSLR, on the other hand, shows me the histogram after I’ve taken the photo, so I can review it and then determine whether or not a second shot is required. Your camera may default to showing the histogram, or it may be something you have to switch on and off in your menu settings. I always hate saying that you should check your manual, because it seems like such a cop-out, but there are really too many variations between cameras for me to be able to make a generalization about how to find your histogram. But this is one thing that you really need to know, so it’s time to dust off your manual and look it up.
You can also find your histogram after the fact, in your post processing software. In Photoshop you’ll see it when you open up the levels tool. Now, this isn’t going to help you get the exposure right because by the time you’re in post-processing it’s going to be too late for do-overs, but you can learn a lot by viewing this larger version of the histogram and playing around with the three sliders under it (one is for shadows, one is for highlights and the other is for the midtones).
How to read a histogram
First, let’s talk a little bit about what a histogram actually is and how you can use it to improve your photos. In simple language, a histogram is a graphical depiction of how many shadows, highlights and midtones are in your photograph. Ideally, you want some of each—not too many shadows or highlights and a nice range of midtones. Let’s look at an example:
Butter Church (I) by Flickr user quas
In this photo, there are too many shadows and too many highlights. It’s actually pretty evident from the photo itself, which is the interior of an old stone building lit only by the light from the windws, but when you look at the histogram there are some obvious problems. You get a dip in the middle, with a lot of very black tones on the left and a few very bright tones on the right. Do you see how there isn’t a gradual tapering off of the tones on either side? That’s what we call “clipping,” and it means that there’s lost detail in both the shadows and the highlights, or detail that existed in real life but was not captured by the camera. The takeaway from a histogram like this one is that there was too much dynamic range in the scene, or a very broad range of tones that was more than your camera was capable of capturing. To fix a situation like this, you need to shoot in better light. Now that would be difficult to accomplish inside an old stone building, but generally speaking there may be times of the day (such as the early morning or late afternoon) when the light is streaming through the windows in such a way that it provides better overall illumination. And artificial light can help, too—a diffused off-camera flash might be enough to fill in those shadows and even out the scene overall.
You can get a similar problem just on the shadow side, or just on the highlight side. That’s easier to fix in-camera because it indicates underexposure rather than a dynamic range problem. If you do see clipping in the shadows, add some exposure compensation until it goes away. Likewise if you see clipping in the highlights, add some negative exposure compensation.
So what does a “good” histogram look like? If you have a very evenly lit scene with an equal number of light, dark and medium tones, it will look something like this:
Of course, that situation doesn’t happen very often, so don’t torture yourself trying to achieve that near-perfect bell curve. What’s important is that you have that gradual tapering off on either end of the histogram, which indicates that you don’t have any lost detail in the scene. And if the histogram is skewed to the left, that indicates underexposure, while a scene that’s skewed to the right indicates overexposure. The ideal histogram should have a nice balance of tones—it can have more shadows than highlights, and more highlights than shadows, as long as it doesn’t lean heavily towards one end or the other.
Now, that can be somewhat objective. If you have a very high-key subject such as a snowy landscape or a dozen white roses, you may have a histogram that is skewed to the right that is not, in fact, overexposed. So use your judgment. Remember that as long as you don’t have any clipping, you can make adjustments to your photo in post-processing in case you get home and decide that your image really is too bright after all.
After you start to get used to these ideas, it really will take no more than a glance at your histogram for you to know whether you’ve got the exposure right. Your histogram is a lot more accurate that just looking at the photo on the screen, because screens are really too small to give you a good idea of what the photo actually looks like (and what information they do give you can be impacted by how bright it is in the environment). So try switching it on, and take a few pictures without giving in to the temptation to view the images themselves on your screen. Instead, focus entirely on your histogram, and make any exposure changes based on that information. I think you’ll find that overall you’ll end up with a lot more perfect photos, and a lot less that are based on the guesswork of your very imperfect meter.
- Don’t trust your meter or your screen
- Meters work in averages
- Screens are too small to be accurate
- Your histogram
- Where to find it
- How to read it
- Use your own judgment
Most people think this post is Awesome. What do you think?