The Histogram is Your Best Friend

One of the biggest things you can do to improve your exposures in a hurry is by getting to know your camera’s histogram, and what exactly all those peaks and valleys mean to the outcome of your picture.

Most all cameras now have an option to show you the histogram when review an image on the back of the camera. It will look something like this….

 Histogram on the back of the camera LCD.

Histogram on the back of the camera LCD.

If you don’t see it when reviewing an image on your camera, you probably need to push the Info button a time or two until it displays on the screen. Lets break down what it is that you’re looking at with this cluster of peaks and valleys.

Here is a demonstration of how the lines of a histogram work (please, not everyone all at once ask me for graphic design work).

 Horizontal axis represents brightness from solid black on the left, to solid right on the right. Vertical axis represents the density of pixels at any point along the brightness line. 

Horizontal axis represents brightness from solid black on the left, to solid right on the right. Vertical axis represents the density of pixels at any point along the brightness line. 

 

As you look at the histogram, the horizontal line on the bottom would be the overall brightness. The far left corner would be pitch black, and the far right corner would be pure white. Both ends of the histogram represent where there is no detail left at all. Just pure black or white. This means that if you want to recover any of the exposure in post processing, even after bumping the shadows up, or bringing highlights down those would still be solid black or white. The goal is to have the mountain not touch the edges on either end, know as clipping, as this beautiful example above represents. That means that none of the image is completely lost. If any part of the image gets clipped, there is no bringing it back.

The vertical lines represent the amount of overall pixels that are of the corresponding brightness level. So pick a point along the bottom line and go up straight up, and that will show you how much of that brightness is contained in the image. Look at the “darks” label on the below picture. The histogram is a good ways up at that point (represented by the long red line) meaning that a fair amount of the image has a dark tone to it. That’s not a bad thing, as long as the histogram doesn’t go all the way over to the left side (pure black). Now look at the “pure white” label. Technically pure white is the furthest point to the right, but in this example the histogram doesn’t quite stretch to the edge, again a good thing. But, look straight up from the word “white”. The histogram is still coming down and there are some pixels that are very bright (represented by the short red line), but not many. And none that are pure white. Yay!

 This photo has a good blend of brightness throughout the scale. Neither side of the "mountain" touches either edge of the scale, which means everything in the scene will be usable and not blown out or underexposed too the point of no recovery.

This photo has a good blend of brightness throughout the scale. Neither side of the "mountain" touches either edge of the scale, which means everything in the scene will be usable and not blown out or underexposed too the point of no recovery.

 

So the goal of the histogram as a whole is to have almost a bell curve in your image. That is absolutely NOT science, because every scene will vary, but a good rule of thumb anyway. You generally want the most amount of pixels to be focused on the areas between the “darks” and “lights” area kind if like this….

 A proper, well exposed bell curve on the histogram.

A proper, well exposed bell curve on the histogram.

 

An image with a proper bell curve image may look something like this…

 Nothing here is too bright, or too dark. Everything is properly exposed!

Nothing here is too bright, or too dark. Everything is properly exposed!

 

See how the bulk of the highest spots on the histogram lie in the middle three sections (darks/midtones/lights) and tail off on each end all the way down the the bottom of the graph? That would be a great exposure. Neither side of the histogram reach the edges so no details are lost to clipping. Great job! In the example image, there is a great blend of darks and lights with none of them being completely black or white.

So, what does a bad image look like on the histogram? If you underexpose an image it may look a lot like this…

histounder.jpg

 

And an example of what an image with that histogram may look like.

 This photo is clearly underexposed, but sometimes it is okay if that is what the photographer is intending. If you wanted a nice bright image of this staircase, it obviously would be much brighter than this.

This photo is clearly underexposed, but sometimes it is okay if that is what the photographer is intending. If you wanted a nice bright image of this staircase, it obviously would be much brighter than this.

Notice how the bulk of the histogram is concentrated on the left side of the graph and falls completely off before getting even close to the pure white side of the graph. This means that there are very few pixels that are in the sweet spot between darks and lights, and most of them are from pure black to dark. Also note that the highest point on the graph is against the edge on pure black. That means that there are a ton of pixels that are completely lost and will just show up on the image as solid black. Now, the good news is that in post processing (assuming you shot in RAW) you can recover this image still, although you’re going to lose the totally clipped parts. They will remain black. Grab that exposure slider in lightroom and drag that sucker to the right and watch the histogram balance out! You’d be amazed at how much you can actually recover from the unclipped parts of the image!

Now on the opposite side of things, an overexposed image may look something like this on the histogram….

 Notice the bulk of the histogram pushed against the right side! 

Notice the bulk of the histogram pushed against the right side! 

 

And an image that may have a similar histogram…

 This is overexposed to the point of being washed out. As with the underexposed staircase, this can be a creative decision by the photographer that can work when done right.

This is overexposed to the point of being washed out. As with the underexposed staircase, this can be a creative decision by the photographer that can work when done right.

 

Just as the opposite in the previous example, this time we are up against the edge of the graph again, just on the right side this time. There is not a whole lot of dark in the image at all. Even the darkest parts are still relatively bright. Now, in the example with the palm trees, this is actually a pretty nice exposure, and a good example of what I mentioned earlier about how every scene differs. This is one time where a standard bell curve histogram wouldn’t get you the same effect that this photo is going for. Blowing out the background on that image works in this composition.

A couple of examples of when you will intentionally break the rules are when you’re shooting a silhouette photo, which will be mostly underexposed and may look like our example above, and when you are shooting that dreamy, sun-flare filled portrait in a field with a kid on a couch with a straw hat on. Okay, maybe I went a little too specific on that last one, but those kinds of portraits are all the rage right now, and rightfully so. Everyone loves a nice blown out background full of rays of light and a beautiful shallow depth of field. I’m guilty of taking many of these type of shots myself. And almost always the histogram looks a lot like our overexposed example above. So there are definitely times where breaking the bell curve rule is acceptable. But for your typical exposure where you aren’t going for a specific creative look, try to stick as close to the curve as you can.

So, with all of that said, learn to love your histogram. The problem with the screens on the cameras we use is that they can be very deceptive when looking at them in different lighting conditions. If you have the brightness bumped up on your display you may inadvertently start underexposing your images because the screen is making it look brighter than it actually is. If you have the histogram enabled that won’t matter anymore. You don’t use the actual image review to check your exposure anymore, just that friendly black and white graph called the histogram! Obviously, still check your image for focus issues and the like, but for full on exposure correctness always use the histogram. Always.

Most of the modern mirrorless cameras with Electronic Viewfinders (EVF) actually allow you to display a live histogram onto your viewfinder display in real time so you can see the exposure BEFORE you take the shot, rather than on review after pressing the shutter. The DSLRs with optical viewfinders unfortunately do not have that option. You’ll have to shoot, then check the histogram when reviewing the image afterward. This is one of the many reasons I finally moved from my tried and true Canon over to the Sony and Olympus systems. I’ve grown very fond of the EVFs and all of the options that they allow for. But that’s for another post another day.

The bottom line is, don’t shoot another picture without enabling the histogram on your camera. It will become your BFF and you’ll never want to be without it ever again once you see just how much of a difference it can make on your photos. Don’t trust that sneaky LCD screen to determine if your image is exposed properly. It’s full of lies! But the trusty histogram always tells the truth.