The Practicing Photographer

- This week on "The Practicing Photographer," we're going to get really nerdy. Now, I have to say that really nothing that I'm going to say here today really has any (laughs) practical impact on anything you're going to do as a photographer. That's not true. That's not true. If you do a lot of HDR, if you do a lot of astrophotography, or if you work with extremely high-end equipment, then what I'm going to tell you might have some practical import to you. But for the rest of us, it really doesn't matter, but I still think it's really interesting, and it will maybe teach you a little something about some color theory. So as you probably know, there are 8-bit images and there are 16-bit images. When we talk about 8-bit or 16-bit, we're simply talking about how big a number you can use to represent the color of a pixel. With 8-bits, you can count from 0 to 255. That means you can have 256 different shades of either red, green, or blue. So you've got 256 shades of red, 256 shades of green, 256 shades of blue. You mix all those together, and you can have beautiful full-color images. If you shoot in 16-bit images, you get thousands and thousands and thousands of colors. And you don't get a broader range of colors, but you get more granularity between colors. So what we say in general about the difference between shooting 8-bit and 16-bit is not that shooting 16-bit gets you better color or better images. It gets you more granularity, which means if you're going to do a lot of editing, your image will be able to stand up to that editing for more edits before you start seeing visible banding and things. Now, Photoshop allows you to work in 8 or 16-bit mode. And if you've ever really thought about the difference, then you may have encountered something a little bit weird, which we're going to look at right now. So I'm here in Photoshop, and I've got just this simple color image up here. And as I mouse around here with the eyedropper tool, if you look up here in the Info palette, right here where my mouse is, if you look up there, you see the different component values, the red, green, and blue values. And those are on a scale from 0 to 255. If I put it over something that's close to white, see, all three values are almost up to 255 each. If I put it on something black, they're almost all down to zero. So that makes sense. This is an 8-bit image. I can tell 'cause of that little 8 right there. So watch what happens if I convert this to 16 bits. Now colors are represented by a much, much, much broader range of numbers, but my Info palette is still showing 0 to 255. So what's happening internally? Well, what's happening internally is that Photoshop is actually dealing with much larger numbers, but for the sake of your sanity and ease of working with Photoshop, it's keeping the RGB values scaled to 0 to 255. So just because we're all... If you've done much editing, you're used to thinking in terms of 8-bit numbers. If it was actually to hand you 16-bit numbers, things would become much more cumbersome, and controls would become way, way too granular. Now watch this, if I pop this little menu open here, I can switch to 16-bit. Now it shows me the actual 16-bit values for each of these things. The other place this becomes relevant is in the Levels dialog. Here, I've got levels ranging from 0 to 255. And as I move these around, these change, again, on a 8-bit scale. I cannot change these to 16-bit numbers. And I think that's for the best. Again, if it was a 16-bit number, I would have to drag and drag and drag and drag to move this any appreciable distance. So it's important to know that this 0 to 255 scale that you see scattered around in Photoshop in the Levels dialog, in the Curves dialog, in the Info palette, internally, you're actually working with much bigger numbers. Now, here's the interesting thing about Photoshop's 16-bit color. It's not actually 16 bits. It's actually 15-bit color. This is where we're going to get really, really nerdy. 16-bit color was added to Photoshop in like version three, which was in the mid '90s. And at that time, computing power was not what it is today, and doing long division on a 16-bit number requires four clock cycles worth of instructions. Doing long division on a 15-bit number only requires one clock cycle's worth of instruction. So on a slower computer, that makes a big difference. At the time, Adobe decided, "You know what? 15 bits is big enough to hold any color data that we would ever want to hold, so that's fine. We'll just make 16-bit mode actually 15-bit mode, and we'll just call it 16-bit mode." The fact is, they were right. 15 bits worth of color is all you need for just about anything that you're going to capture. The typical digital camera today captures 12 to 14 bits of color per pixel, so a 15-bit container is still big enough to hold that. Where you get in trouble is if you were working with a device that actually captures 16 bits worth of data. Now, I can't tell you about a device that will do that. I don't know, but maybe there is high-end medical equipment or something like that that actually captures 16 bits. If you were working with that kind of device, you should never pass the data from it through Photoshop's 16-bit color mode because it'll actually toss out some of your 16-bit data. Where it does come up for those of us who shoot with more everyday technology is if you're doing HDR or any other form of post-processing that requires stacking. So maybe you're an astrophotographer, and you're stacking images to eliminate noise. At that point, that 15-bit container in Photoshop, that container that you think is 16 bits but isn't, is going to throw out some of the data that you produce through your stacking calculation. So if that's true, it's better to work in Photoshop's 32-bit container, which is how Photoshop's internal HDR processing works. I'm going to stay real general about that. All I mean is that if you do have Photoshop in your workflow and you do do some sort of stacking process, dig into the stacking software that you use. Find out if it's truly 16-bit. Find out of there's a problem with passing the results through Photoshop. So that's some nerdy 8-bit and 16-bit color detail. We now return you to your more practical programming.