Photography 101: Understanding the histogram, Levels and Curves (part 1)

This two-parter is an accompaniment to the Photography 101 posts on the basic concepts of photography. It’s not a post that will help you take better pictures; rather, it will help you understand the pictures that you have taken, and make more meaningful edits.

This two-parter is an accompaniment to the Photography 101 posts on the basic concepts of photography. It’s not a post that will help you take better pictures; rather, it will help you understand the pictures that you have taken, and make more meaningful edits. It’s fun and easy to mess with different filters in apps, but if you ever wanted to understand the mechanics of a photo and cut down on the ‘I wonder what will happen if I use this filter’ time you spend editing photos, this is a good place to start. This is the theory; Part 2 is the practical.

You know when you post that artistically-posed lingerie thirst trap on Instagram? What the photography neckbeards are thinking when they’re looking at your photo is, “what a terrific bar chart!” You know why? Because a bar chart is basically what they’re looking at.

Each pixel in a photo has a specific tone – that is, a certain amount of brightness or darkness. Each photo will have some pixels that have 100% brightness (which look like pure white) and others that have 0% brightness and look like pure black. In between, you have some pixels that are all the other values. Once you have a collection of numbers, you can express them as a bar chart. Et voila:

This bar chart represents the tonal values in a fictional photo. On the far left and far right are the two extremes – brightness and darkness. Or, as normal people might say, shadows and highlights. The height of the bar represents the number of pixels that have those values – nothing to do with the relative brightness and darkness, in case you are wondering. In our fictional photo, you see that the highest bar is on the right, the 100% brightness. That means, a lot of pixels in our photo will be pure white.

The second highest bar is the 0% brightness, which means that quite a few pixels will be pure black, but not as many as will be pure white. The bar in between the midpoint and 100% brightness is also quite tall, so what do you think that means? Exactly – overall, this fictional photo is likely to be quite bright because a lot of the individual pixels are more than average brightness.

The name of this bar chart is the histogram. I appreciate this is a bit abstract so let’s look at some example photos with our simplified histogram. I’m going to use black and white photos, but the same thing is true of colour photos. Remember, we’re talking about the brightness of each pixel, not the colour.

The histogram for a low key photograph

There are a lot of shadows in this photo. Almost everything in the bottom half of the photo is pure black. There are a few highlights on the left, and in the sky. There aren’t many midtones either. So in the histogram we’d expect it to be heavily biased towards the left side, the dark side:

The histogram for a high key photograph

In this one, there are some pure black areas, but not many. There are some midtones, but the vast majority of pixels are basically white. So, we’d expected a histogram biased towards the right side:

A extreme example histogram

This one has lots of pixels that are pure black and lots that are pure white, but almost none in the midtone range. This was a conscious creative decision to produce this extreme, abstract, graphic image (which is of the side of Corporation nightclub in Sheffield). The histogram it produces is similarly extreme:

One final point. Obviously the diagrams above are simplified (IKR, hard to believe that they weren’t produced by a sophisticated computer the size of Birmingham Bullring) to make them easier to understand. There aren’t just 5 bars in a real histogram. There are actually 256 bars, numbered 0 (black) to 255 (white).

This is all very well but what actual use is it to me? and similar questions

The thing about pure black and pure white is that there’s no detail in them. Think about it – if you draw with a black pen on black paper, what can you actually see? So, the more pixels you have at the very extreme ends of your histogram, the more detail you have lost from your photo. And in some cases (like the last pic above) that’s a justifiable creative choice. But if that wasn’t your intention, then you’ve just lost detail.

If you turn your camera histogram on – sometimes you can have it in the live preview before you take the shot, sometimes you can only see if after you have taken a photo – you’ll be able to use that to judge your exposure, and adjust it if not.

But we can use the same data to help correct and edit our images afterwards using Curves and Levels – in Part 2, coming up soon!

Homework assignment

  • An easy one today – just work out whether your camera can display the histogram, and if so turn it on. If you can see the histogram for pictures you’ve already taken, take a look at some.

Quick snaps

  • Each pixel in a photo has a brightness value, called the tonal value.
  • The tonal value for a pixel goes from 0 (pure black, 0% brightness) to 255 (pure white, 100% brightness).
  • The camera creates a bar chart of how many pixels in a photo have each of the 256 tonal values, called the histogram.
  • The histogram can be used to help you judge the right exposure for your photo, amongst other things.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s