Lessons

Lesson Info

Understanding Bit Depth

In order for us to measure color, we need to contain color, using color models. Okay, so a color model really defines the way colors are described, mathematically. And there's a few main color models that we work with in photography. The main one, of course, is RGB. We've all heard of RGB, because that is, that is what we do, okay? Another one that you might have heard of, especially working through Photoshop, is L-A-B, not lab, but L-A-B. And then of course, CMYK is another one. And greyscale is another one, although greyscale, a lot of people think it's a great way to do a black and white, I'm gonna show you how that is not the case, a little bit later in this course, but RGB, LAB, and CYMK are the main ones, okay? So color models also may use 16-bit or 8-bit, we'll talk a little bit more about that later. But 16-bit color depth doubles the space needed to store values but gives you more color values in between colors. So from this color to that color, there's a lot more color in bet...

ween, with a 16-bit file, than there is with an 8-bit file, okay? So what's the difference, why is 16-bit better than 8-bit? Okay, let's have a look at a little bit of a description of what bit depth is. So bit depth basically quantifies how many unique colors are available in an image's color palette, in terms of numbers of ones and zeros. These are digital files, there can only be a one, there can only be a zero. These are called bits, okay, which are used to specify each color. Stay with me here. Images with higher bit depths can encode more shades of color, since there are more ones and zeros available, would you agree? There's more ones and zeros, equals more color, okay? Now every color pixel in a digital image is created through some combination, if you like, of the three primary colors, red, green, and blue. That's how the image is recorded on your sensor. Now each primary color is often referred to as a color channel. Later on in Photoshop, you'll see how we work with RGB channels and how we manipulate those channels to achieve that ultimate result. It can have any range of intensity values specified by the bit depth. So digital cameras that have 8 bits, if you are shooting in 8 bits and you're only shooting jpeg, you'll stop doing that by the end of this course, I can guarantee you. They can use a total of eight zeros and ones, per given channel, okay? So that is two to the power of eight, which is 256 combinations, translating into 256 different intensity values of each primary color. Not a lot of wiggle room, 256 from shades of a particular color or tone, from pure, pure zero to to a pure white, so pure black to white, 256 levels, okay? When we work in 16-bit, beautiful things start to happen. So 16, you think it's double, well it's a lot more. It's actually two to the power of 16, which gives us 65,536 levels of information per color channel. That's a lot of information, so we have a lot more color between color. You know, when we take these files into post-production, into Photoshop, and we start to play around with the naked image, what happens is, we start to get histograms that look like this. We've all seen them. But what does it mean when we start to see these gaps in our histogram? This is missing information. The minute you start to stretch that histogram, there's only 256 depths of information that I have. The minute I stretch it by pulling curves, levels, it doesn't take a lot of stretching to break that. These things here, in an image, especially when we print it, not so much, sometimes you see it on a monitor as well, banding, we've all seen that, and artifacting, and beautiful bands of magenta color sort of coming in, you go, wow, that's a special effect, because that's the only way you can sell it to your client. But it's not, it's actually the file saying, I don't have enough information to do what you want to do. 16-bit, I've got 65,536 levels that I can play with, between a pure black and a pure white, for each of the channels. There's a lot of wiggle room for me to pull that file any which way I go, to maintain that beautiful histogram intact. And of course, combining this and working with this in a non-destructive editing, that's it, you're starting to set yourself up for success. So the difference between 8 and 16-bit, which you'll hear a lot in this course. Let's get back to the color models, let's have a look at the RGB. I'm gonna jump in real quick. Before you move on, I feel like a lot of people just starting out doing this, or just starting out as a photographer, don't necessarily understand what that actually means. Like, as you're printing, as you're working with your images, what are the fundamental differences? And should people ever shoot in 8-bit or should they always shoot in 16-bit? That's a very good question, which we'll get to why we should never shoot in 8-bit, you should never shoot jpeg. The idea is to capture as much information as we possibly can and maintain the integrity of that information through to the edit and then through to the print. Now printing doesn't mean a lot of the times that we print in 16-bit. In fact, a lot of the times, we might even print in 8-bit. But the idea is to take that information, manipulate that information in the highest bit depth possible, okay, which is very, very important, and then, once we're done with that and no further editing can be done, we can go down the road of 8-bit, because it's basically taking a snapshot of that tonal range, and then we print that tonal range. But having said that, it depends on how big we're printing, as well, you know, printing in 8-bit for an album is fine, and we'll talk about that a little bit later as well. But if you start to do really, really huge prints, like 60 by 40 inch art prints, I wouldn't recommend printing in 8-bit, especially when you've got fine tone gradation of color and all that sort of stuff. You need to be able to preserve that. So that's in a nutshell, but we're gonna get to that a little bit later in this course. When we look at the RGB model as I said, when we mix the three, okay, we get white, okay? So if red's at 255, the green's at 255, the blue's at 255, it's white. If the red G, green and blue are at zero, then of course, we get pure black. That's how the RGB color model works. The LAB model, okay, works a little bit different. Where we have lightness up and down the central axis, so we're trying to map this color three-dimensionally, if you like, so we have lightness and we have, of course, saturation out on the very edge, okay, of a particular color, and we get two axes, an a-axis and a b-axis. The a-axis, your colors are running from red through to green, and in your b-axis, you've got your blues and your yellows. So blue, minus blue, and that's pretty much the way it's working. Okay so, imagine this in a three-dimensional format, we define color, you know, L being lightness or luminance of a particular image, okay, and then the chroma, otherwise referred to as the color, runs on two axes, okay. So we can place a dot here anywhere and it defines a particular color by numbers. If the dot's here, it takes on a particular shade of yellow, with a certain lightness and a certain saturation. Okay, if the dot's here, the same thing would happen for the red. If the dot is very close to the middle, up and down, it's a grayscale image, or still an RGB but it's monochromatic, because the saturation of the color is not much, in fact it's zero, and the color, as we move out of that central axis, we start to get color. Another model that we need to work with and understand, because this happens at the printing stage of course, is this thing of CMYK. Now CMYK, unless you're doing pre-press stuff, you're not working with it in Photoshop, per se. But CMYK is how your printer works, because it's using cyan, magenta, yellow, and black ink, cyan, magenta, yellow, and K for black. They could have used B, but then B would have been confusing because we already use B as a blue in RGB. So K, black, so CMYK. So CMYK is a subtractive color model, what does that mean? Each of these color rings is absorbed by a certain spectrum of light, okay, on a white piece of paper and it cuts that spectrum out. And if you combine all of those together and you start to add them on a bit of paper, you should get black, but it doesn't quite work like that, because the pigments and the dyes that we use in printing aren't 100% always pure. So you'd get a very dark, murky color. So we have to add a black ink if you like, or a black component, to have our nice, beautiful, clean blacks. So that's how a printer works. Now, in conventional printing that we do ourselves, through our Epson or our Canon printers at home, we're not CMYK printing. That's still classified as an RGB printer, because we're sending RGB information to the device itself. But this is how, you know, CMYK works. So we have color models of how we can define color, we also need to define then how these color spaces work, okay? And you've heard color spaces, you've heard sRGB, ProPhoto RGB, Adobe RGB, these are all color spaces that define a color in a three-dimensional way, in these color models, okay? So a mathematical model would simply describe the range of colors that you see or that are captured as numbers, okay? Color spaces are nothing more than containers of different sizes containing color, that's all they are. Yes, RGB is a container of color, or a box of crayons, if you like. Adobe RGB is a bigger box of crayons, okay, or a bigger container, okay? ProPhoto RGB is a huge box containing a lot of shades, with a lot of crayons and a lot of pencils. Okay, a lot of different shades, okay? So it's pretty much like your pots and pans, okay? You've got ProPhoto RGB on the left-hand side and you've got sRGB on the right-hand side. Then when we start to move around the world of, you know, color spaces, if I fill the bucket on the left here full of water and I tip it into the middle bucket, I'll have some water left over and we throw that water out. So we're converting from a big color space to a small color space, that's what happens. We throw that color out, or that color is converted to fit within the particular color space, to a certain degree. And then of course, if we go to RGB, we tip some more of that out, and we take, you know, some of it, and we throw it out. Now the property of the color doesn't change. The property of the liquid, as we go from one container to another, remains the same. Water is water, as we move from one container to the other. But the amount of water, or the amount of color, changes. Now one of the worst things you can do is, I've got a file in sRGB but the lab wants it, say, in Adobe RGB, oh I'll just convert it, change it. Well, you can change it but nothing's gonna happen, because you've lost color information. See, it's a one-way street. You can go this way but you can't go back, okay? And we'll talk a little bit later about, you know, where you should aim to be working in, okay? So when we look at, you know, the, the LAB color model again, as we have it here, okay, so there it is, and there's our axes and there's our lightness going up and down, okay, and we start with mapping out the sRGB color space, okay, and there it is. That's how the sRGB looks in, in the three-dimensional world, it's very pretty, very nice. And then, from that, we go into a bigger color space, which is the Adobe RGB color space, it's getting bigger. And then we go to something that's really, really huge, which is this one here, which is our friend, ProPhoto RGB. It sometimes can be a foe as well, we'll get to that a little bit later. So it's a very, very big color space, you know? So your camera, really, is capturing a lot of information. How much information? I'm gonna show you in a minute. So when we look at a two-dimensional representation of our color space world, okay? We have the spectrum of light, as we see it there. We have sRGB that sits inside of that. We have Adobe RGB, which is bigger again, okay, which is a big color space. Then you have this monster on the outside which contains colors that we can't even see. Okay, it's outside of the visible spectrum. It is a huge, huge color space. So we go from something that's quite big, in other words, we want to be able to capture as much color as we can, to work with, and then work our way from there, depending on the destination, it depends on where we're going, okay? Now, when we start to look at printer profiles, okay, so if this is ProPhoto RGB on the outside, on the inside here, I have a profile for one of my cans on platine papers. And it's a custom profile, and when you map that out, guess what, I'm actually printing colors that are outside this massive color space. So the paper in combination is able to give me a lot more information, which is a pretty cool thing. Okay, so it's being contained in there, but the profile is able to give me even more.