Lessons

Lesson Info

Histogram

Histogram is gonna tell us exactly , out in the field, whether we got the right exposure or the wrong one. You should be very familiar with how to turn the histogram on, and how to read a histogram in your camera. Because that is gonna provide you with absolute accurate information as to whether you got the right exposure. Generally speaking there is going to be either a display button or an info button on the back of your camera, and that'll usually cycle through different options on the viewing on the playing-back of your image. And if you cycle through there's usually two, three, four different options. One of the options is gonna have the histogram on it. A special little note to the makers of Nikon cameras: please stop turning this thing off and hiding it in the menu system. All your Nikon users have to go, in my class, I have to go over to their cameras and I have to go over to their custom functions and I have to turn on the histogram view so that people can see the histogram. S...

o Nikon, stop hiding it from people who buy your cameras. Little irritation on the Nikon camera because it's there, you gotta go turn it on and then you can see it. But it is there on all the cameras. So what is a histogram? It is a graph of the tonal distribution. Imagine a black pixel, a white pixel, and a whole bunch of gray pixels that are slightly different shades of gray. And we wanna know, how many are there of each of these different shades? Well what we do for a histogram is we put the black pixel over on the left-hand side, the left column. We put the white pixel over on the right-hand side, and all the different shades of gray from left to right, just like this. And so now, we can look at this simple graph, which is a shape. And you can look at a shape very quickly and determine what that shape looks like. And so that's what make these so easy to see out in the field. What we are looking at is about 256 levels of brightness. From pure black on the left, to pure white on the right. Now we're looking at the number of pixels on the vertical axis. And as we go along the bottom we have the darks over on the left, the shadow region, the mid-tones, and then the highlights over on the right-hand side. So we can see that this photo predominantly has a lot of mid-tones that are on the middle to slightly-darker side. Over on the very dark side we have very little and on the highlights, very little as well over there. Now, one of the things that you will hear is that there is no perfect histogram. And that's generally true. But I think this is about as nice a looking histogram as you're gonna find. And there's a couple things that I'm basing this judgment on. Number one: over here on the left-hand side that left column is empty. There is nothing in there. There are no pixels that are perfectly black, which means I've got a little bit of light from everything that I shot a photograph of in that particular picture. Over on the right-hand side, the pure white pixels are empty as well. There are no pixels that have been overexposed and have no color information at all. I have recorded the entire dynamic range of that particular scene. This is gonna be handy because in Photoshop or any other program afterwards, I can take these pixels, and I can make them a little brighter and I can make them a little darker. If you capture a pixel that is perfectly white, you can't really darken white. It doesn't have any color to go to. And if you have something that's perfectly black that has received zero light, you can't make black lighter. It just doesn't work once you are in post-production. And so I've captured the entire range of light. On my camera you might see this as a brightness histogram, which is just kinda a white hill. Or it might be a RGB histogram. In Adobe Lightroom it shows it to you in a slightly different manner. Different cameras, different programs will show this to you in a different way. But basically it's a curve, it's a graph, it's a shape that you're gonna be looking at. So let's look at what these exactly mean. This little perfect histogram comes from this tiger in India. And if you've ever ridden an elephant taking photos of tigers, it's kinda hard to see the back of your camera because you're moving around, and it's kinda bright out. If I had overexposed, I might not be able to look at the back of the camera and see the exact photo. But I can look the histogram and I can quickly tell that here's a photo that is very very bright in exposure. And I would be able judge and say "That looks wrong." "That looks off for the situation." And if I'd recorded the camera underexposed for some reason, I can look at the histogram and clearly see that something is way off because all of the pixels are shoved up against the wall. The main thing that you're concerned about are these graphs pushed up against the wall, alright? So we're generally looking at them more towards the middle. But every photograph has a different histogram. And it's a little like a fingerprint. Each one is unique and different. What you're generally looking for, very generally, is a mountain in the middle. If you have a mountain in the middle, you're doing pretty good. These are all decent histograms for completely different photographs. Now it's time to start the "quiz" section. This is the unofficial quiz. I want you to look at this photo and think about what that histogram might look like. Is it gonna be more to the left? Or is it gonna be more to the right? Now, does anyone here, would anyone like to elaborate on what they think the histogram will look like? Why don't you grab a microphone and explain what you think this histogram might look like. I think that the histogram would be more to the right. It that's the white portion. And you're basing that upon, what are you looking at here? The snow, the penguins, the background, there's barely any blacks except some of the penguins itself. So I would say it's more to the right. Right. So this is what we need to do, we need to be able to assess a scene and go "This is brighter than average, "and darker than average." And here's the actual histogram. Bonus question, see if you can get this right. Why do you think there is two different peaks here? What do these represent in the picture, why are they there? I would say the exposure itself. What sort of subjects, do you know? What's causing this brightness? What's causing this brightness? Would it be both penguins? No. The only other guess is the snow, I don't know. Well, notice this is water back here, I don't know if you can tell, but that's the water. And that water is a little darker than the snow. So the snow is probably causing this. And the water is probably causing this, because it's a little bit darker. Now the penguins are pretty similar in whiteness to the snow, so they might be kinda combined in here as well. Now the other little trick, and we'll talk a little bit about this, is you see how this blue channel got a little separated? When the blue is a little bit brighter you probably are gonna notice that color. You're gonna notice a little bit more blue here. And this very pure white where all of the colors are in the exact same spot. For the next one, I'm gonna show you the histogram. And come somebody else explain what this photograph is? Or what they think an attribute or what this photograph might have? What is this photograph gonna look like? It's gonna be very dark. How do you base that? Because everything's pushed to the left. Yeah, we have a huge spike. You see how this spike, it goes through the roof here? Don't worry about going through the roof. Going through the roof is totally fine. It's the walls you're concerned about. It just means there's a whole bunch that are nearly perfectly dark here. And there's a little bit up here, and a little spike there and there. So let's take a look at the photograph. So it's a silhouette. And this rock area is that dark area. Now normally, I would look at this histogram and say "I did something wrong, we gotta adjust the exposure." But for this type of photograph you go "Oh, wait a minute, now I understand and this makes sense." Because this is supposed to be really dark here. It would be very awkward to have this all bright and everything back here super super bright. That wouldn't look good. When I'm looking at histograms and I'm looking at photographs, it's like I'm a private detective. And I've got different clues I'm looking at. I'm looking at the photograph, I'm looking at the histogram, I'm looking at what I can see with my own eyes to see if this makes sense. So here's a photograph. And if I look at it right here, it kinda looks a little overexposed. You notice the sky in the background? It's really kind of blown out, there's no detail in that sky. And if you were there when I shot this you'd realize that this was happening right at sunrise and there's blue sky up there. Looking at the histogram I noticed that it was spiked over on the right-hand side. "Oh yeah, this thing is really bright." Let's try changing shutter speeds, apertures, or something to make it a little bit darker. And here's what I think that photograph should look like. Now this a little bit of a personal opinion of what I think a photograph should look like. But you'll notice I have darkened the photograph up, and now we can see some more color in the sky. Here's an example of a spike over on the right-hand side. Normally, that is a huge warning. That is bells "ding, ding, ding, ding, ding." Something might be up here. Doesn't mean you should change anything, but it's a potential that something's wrong. And in this case, overcast sky that has no detail in it is just going to be overexposed in this particular situation. And so it's perfectly fine. You can ignore the warning bells. Now some of your cameras will have a highlight alert, which will show you in blinkies and I'm gonna show it to you in red here. Shows you the areas of pixels that are completely blown out, which don't have any detail, don't have any color into them. And this would be a warning that you might be overexposing your photograph. If you are photographing a brides dress and the entire dress was blinking, you should probably adjust your exposure. The bride probably does not want a completely blown out, white picture of the dress. They probably want to see some detail in it. I gotta take that photo someday. I'm gonna take that photo and have it in next class. I gotta try that one. So this is showing you, this is something that you can turn on and off in your camera and normally it would be blinking at you. And it would show it to you either in the viewfinder for a mirrorless camera, which can be kinda handy. Or it'd show you in the review of an image afterwards. Actually, here's an example of what the highlight alert would look like in camera. So if we're overexposed, its gonna blink at you showing you the overexposed pixels. And as long as you understand that that might be a problem, it's not necessarily a problem, that it might be a problem. That's when you can look at the photo, you look at the light meter, you look at the histogram, to really try to get a better judgment as to whether that's truly a problem or not. Couple more photos. Alright, this is a tough photo. This is got very dark areas and very bright areas. We are pushing the latitude of what our camera and our sensor can handle. And we can see in the histogram that we have a fairly-big mountain over on the right and a bit of a little mountain over on the left-hand side. And this is just within the tolerances of this camera. Because it kinda just peters out right down there at perfect black and there's a few pixels in here that a perfectly white, but it's within the range. This is just right at the limits of what our camera can handle. There will be some situations which are beyond the dynamic range of our cameras. And in this case you can see that the left and the right-hand side have very sizeable areas that have gone pure white and pure black. And it's just because in this bright, sunny situation we've got some very bright light bouncing off of the water and the clouds, and there's some fairly dark areas in the shadows that we're losing information in here. By adjusting our shutter speeds and apertures, we can move histograms left and right. But what you can't do is take the ends and push them together. It's either left or right, you can't push, you can't stretch. That stuff is for post-production. That's Photoshop and Lightroom stuff where you can kinda play around in that regard. And so you are a little limited. So this is where you're right at the limits and how do you solve this problem? You can bring out gigantic flash units to fill in the shadows. Or you could wait for better lighting where it's not as intense during the day. Come back towards the evening, or come back the next morning. I enjoy graphics, I think this is one of my favorite all-time histograms. I think this is just a really colorful, fun histogram. And it's an equally fun, colorful photo. We have lots of colors in there, so we do get to see lots of colors in the histogram as well. So something unusual is going on in this photograph. Does anyone have an idea as to what is going on in this photograph? Remember when we talked about those blue? A little bit more off to the side. Well, red and green combine to form yellow. And so there's gonna be a lot of yellow in this photograph. And this might be the only photograph in the class that has Kenna as the subject. So there's lots of yellow which is causing that big old spike in the histogram over there. What do you think this histogram's gonna look like? There's gonna be a blue over to the right-hand side. Big old blue over to the right-hand side. A this shows up very prominently because that's a very bright color. I think this is kind of an interesting histogram. It's lots of little spikes going on. So what the photo is, is it's lots of different gradations of color. And each of these colors is kinda encapsulated in a small little area. Because if you were to just take this bottom little cut of the hillside here, everything's almost the same. And it's one particular shade of blue, and this is a different shade of blue, and on back. Each one of these is a different portion of that particular photograph. This is what I think is the worst-looking histogram that I've ever taken. It's two spikes, one on the left, one on the right. Now, do you delete photos simply because they have a bad looking histogram? No, you gotta look at the whole package. It's just part of the information. And the actual photograph now starts to make sense. The dark sky is the black spike. And that is a dark, dark sky. It's gonna be black. And fire is really really bright. And so those are gonna be blown-out, really hot pixels there. As the tree gets further and further away from the fire it gets darker and darker, which is this nice, even slope right along there. It's a great tool for judging if you went over or under exposure. But you gotta look at the photo, you gotta look at the scene with your own eyes and use your own judgment to figure out is it the right balance for what's going on.