Lessons

Lesson Info

Compositing

compositing to me means a couple of different things. I think many people as soon as you hear compositing, you think, take a photograph of a person and put them on a different background. And that's a very popular method of doing that. However, it's not the only thing. So to me, compositing is anything where you're ending up with an image you like by combining images. So as an example, skill to my compositing folder. So I was very happy to visit Italy a couple years ago. The problem was, the time we were there was in November and every single day the sky was great. It wasn't a hint of color, There wasn't anything. It was just so we're in this beautiful place, and every photograph I took had a sky that looked like that was like awesome. So I got some photos that I was otherwise very happy about. Excuse me, but I wanted to print somehow framed them. I thought so. Some people, when they take a photograph, they say, I want to capture the moment, period, and I'm like, but the moment wasn't ...

very interesting looking. I want tohave a nicer sky in there, so I have no problems whatsoever with putting a new sky in so that it looks nicer to me because my wife and I were the only two people love in the gondola guy at the back who was steering this thing that knew That's what this guy look like. So as long as it looks realistic to me, ends up looking right, then I have no problems whatsoever in enhancing my photograph because I'm not. I don't work for Time magazine, where I have to be a photojournalist and say this is exactly how it looked. I'm trying to create an image that I like, and that's I think an important concept to embrace is to say, if you're trying to create something that either you like or that your client will like better than what's problem with enhancing it by doing something that was never there. So you may remember possible that once or twice I talked about this thing called a smart object you mentioned a couple of times. This is one of the reasons why another reason why use smart objects. Ultimately, I want to open this photograph and put a new sky in that existing one now right away. There's some pretty good contrast here so that I can see the edges of where the sky and the buildings are. But since I haven't set up is a smart object. If you didn't see that part, this is this option that's called opening photo shop. A smart objects. Then what I might decide to do is I want to make the job of the quick selection tool, which I think of us the edge detection selection tool. So because of the edit herbal nature of this, I might decide to change the look of the photograph quite dramatically. And all I'm really doing is looking up here at the edges to make sure that I'm accentuating the edges. So now I hit open object. This would also assume that I've chosen the correct size that I want and all that good stuff from then I have another photo of a sky. I'm lucky, because in Florida this is like every day you see clouds like that. So I have this ridiculously enormous folder of cloudy skies that I've photographed thinking I might need this at some point. So I'm just gonna make a rough adjustment. I think I want just a little more vibrant, perhaps And again I hit open object. Now, ultimately, I'm gonna lose the, um, smartness of this at first, at least because of the simplest way to combine these two photographs. You can't do it as a smart object. Aiken end up that way. But I can't do it initially, which will make better sense when I show you. So I switched to my quick selection tool, which has thesis impolite to remember shortcut W for quick selection tool because it shares the same spot as the magic wand. So it's W for one w for quick selection tool. The beauty of this tool is you can see where my mouse is. It's right in the middle of the sky. I just drag like a hair and it selected the whole sky because what it's doing is looking for edges. Personally, I prefer this much more than the magic one because the magic wanted. Some of you may have encountered this. My biggest beef with the magic wand is you'll click in the sky and your goal. Go look, it selected the whole thing. Then you do the next step. You see little tiny patches where didn't quite get it because it's based on color shades. So if there's even a slightly different shade, it might not get it. The quick selection tool basically ignores that and looks for edges in a case like this is, I don't care what? How many colors? Aaron. There are different slightly shades of this, this white. I'm gonna push right to the whatever edges I confined in this case, I had barely move my pen or mouse to say, Get the whole sky. What I want to do is, I want to take this sky and put it over there. Now the long way would be to drag this over and position it and add a layer mask, which to preserve the smartness. That would be one way. But I'll show you the the other option, which you can. If you don't want to alter this guy very much, you can do this and I'll show you kind of a secondary plant. So I'm gonna copy this using the keyboard. So I've copied it over here. I still have my selection now. If I just did a regular paste, it would ignore my selection and put the sky on a new layer. What I want is to have Photoshopped do all the work for me and say Make a new layer. But make a mascot the same time based on the selected sky, and the way you do that is with a command under pay special. It's called Paste into and has the very easy to remember entire left side of the keyboard. So and what it's done. I see there's the layer and it has a mask already. And there's one other thing that did that was very important, so I don't have to do it. The layer and the mask are not linked together. Normally they are. But if they were linked together and I tried to move it, it wouldn't make any sense because the mask is moving with the sky. So when you do paste into one of its advantages is it's automatically unlinked. So now I can decide where I want to put the sky, So even though it looks like I'm moving the sky behind the buildings, it's really in front in the layers panel. But that mask that staying put is making it look like I'm moving around behind. Now the only down side to this is I can't edit the settings of this sky using camera raw, because when you copy and paste, it's not smart. So what I would probably do just after the fact of I decided Maybe I do want to be able to edit that Skye is now. Take this whole thing, drag and drop it and then just move, mask and delete this one. So now, effectively, I've got to the same place. But now the added bonuses, I can double click on this one and saying, Maybe I don't want it quite so bright because that day was for it to match in a little better might look better like this. Okay, so one of my favorite expressions, as you've heard me say multiple times, is always thinking end up with whatever it is in this case. I want to end up with the sky that I can edit inside the existing sky. Now the problem is, the overall photograph is too dark. Remember, I did that temporarily to make it easier to make a selection. So now I would double click on this camera smart object and say, Let's put it back to the defaults. Maybe just changes a little bit. And now when I click OK, it'll be Maurin context. So now I'm looking at it in context. OK, I think that's guys a little much. Sometimes like in this case, there's little details in here like there was. If I hide this for a second, you see there were like wires or antennas or something. Well, sometimes you can make the sky a little bit more intense than it might should be. Maybe should be and then lower the opacity, even just a hair. And I said, Now you can see the little antennas starting to show again. So I don't have to worry about being an incredibly accurate selection where I'm actually painting around every little antenna. It's just saying, Put this guy in there and lower it down a bit. So lower the opacity. Once I've lowered it, I might make one more trip back and say, Well, now that I've lowered it, maybe I do want toe make it just a bit darker. And this is the beauty of the whole camera. Raw interaction is you can do that. You don't have a Well, I've already pasted my sky in there and now I'm out of luck. So this way you're like, I like that now. One of things that that we have to always realize about this kind of thing is I firmly believe and I know I was the case with me, and I've talked to other people that say the same thing. We are our own worst critics. So because we know what the original photograph looks like, we're looking going. I can see this. You have toe Stop that. I mean, you need to be conscious about to make sure you didn't miss a part, But let's put it this way. I have I print much like this that I'm hanging on my wall and no one's gonna coming up in going that chimney over there because no one does that. I mean, well, some people might, but the average person doesn't because they just go in thinking, Oh, there's a nice photo. They don't look at the layers panel and go, Oh, I see you added another layer. So if it's believable to the point where I mean, I wouldn't put a you know, a nighttime sky when the photo looks very tight daytime, so you have to think about that kind of thing. But in terms some people knit, pick and zoom into a 1,000,000% and go. I need to get this little part here. While no one's gonna look up that close with a magnifying glass, one is hanging on your wall again. Less. You have really weird friends that do that kind of thing. But if you do, then you know to prepare for that. But the average one doesn't. So the worst thing we can do is be constantly going before after before after and making a comparison where at a certain point, you just have to look at the after and go. Does that look believable? So the two main things about this example and you'll see this as a recurring theme is this is the other reason why use camera smart objects is because I can deliberately change the look of my photo toe, make it easier to make a selection, then change the settings back so they blend in better and everything looks more realistic. Okay, so that's kind of our first point, and the other related point is, any time you see a cool sky Take a photograph of it, stick it in a folder called My skies Because you just never know where you're gonna like Be somewhere really cool otherwise, but the problem is this guy just doesn't look the way you want it to. Okay, so here's now that the main thing we're gonna talk about today is compositing in the sense that more and more people are thinking of which is take a person on a plain background or some background. You have and put, um, on something more interesting. And what I used to do was open the photograph of the person and then try and make a selection and make a mask. And this case, I'm starting off with a, frankly, an easy example, because this fellow doesn't have a lot of hair going on either. But the concept will be the same. But so here's my little story about how one simple idea has saved me. I can't even tell you how much time I wish there was a way to calculate. And it was a case I was working on where I had a photograph where there was someone whose hair was blowing in the wind like wind machine in the studio and particularly on one side of her. So I opened the photograph, and I pretty much obsessed over getting every single hair, took me a long time. Then I dragged her on a new background and realized when I positioned her you couldn't even see half of her hair because it went out of the frame and I realized it spent all that time working on that hair that was completely not necessary. So I suddenly thought, Why don't I drag the person new background first, Make that step one, because that way every other decision I make is in the context of the new background. And that was one of those lightbulb moments that for me has changed the game completely because I realized I've cut down dramatically on how much time I take try to extract someone because I'm extracting them in the context of the new background. And there been times I'm like, Wow that his hair is kind of dark and so is the background. So it looks fine. Why would I take another 20 minutes to try and get every little hair when I don't need to. So I would say, in the world of compositing, that's without question. Rule number one put the person on the new background first. Now some people say, What if I don't have a new background? I don't know what background to use, I would say all kidding aside, then wait, Why would you extract someone and have them nicely extracted without really knowing yet so Or find a photo that's a similar as you can to the background. But the bottom line is, I want to not just work on the person I want to say, and I'm gonna put him on this background. So the first thing I do is say, Let's just get him over here And depending on sizing, you may have to go. Okay, is little large. And that's the other reason is because sometimes when you look in the new the context of the new photograph, you're like, why am I spending all this time when they're going to be this big? So the reason that the main reason we do this is I heavily used this wonderful command called Refine Edge and one of the options for refine edges, how it displays so For example, I take my quick selection tool and just kind of drag around the edges of him. Looking seems like it's pretty good so far, but want my without even thinking about it now when I think I'm getting close to a good selection, I just press the letter. Q. That goes into this mode. Call quick mask, and it shows my selection with an oak color overlay. And very often you go hope. Didn't notice. That helps you spot things to go. I missed a bit there. I missed a bit there, and they'll be times where you're trying to make a selection, the quick selection tool, and we'll see an example later on. Where there's a fellow, I have a photograph, and he has like writing on his shirt. And the first time I did, I completely missed the fact that it missed one of the letters because it's on edge. But as soon as I hit, quick mask was like, Well, my eyes was went Z and I saw exactly where the problem was. Now you could edit in quick mask using layer mask type functions like black to hide white to show that's fine, too. But for now, I'm just going to use it as a method to kind of highlight the areas that I need to add. So let's get a smaller brush and just add these couple little bits that it was missing looks better. So here's the thing with refine edge. The name is very descriptive, refined edge. It's not called fix edge. So if you go into our financial, you realize I miss part of his elbow hit, cancel and make a better selection, so refine and the whole theory of it is I've got very close to the edge I want. I want to refine it and make it better. So if you realize you missed a bit or you got too much, don't go to refine edge yet, or if you do go to refine edge, step back out of it, and I need to make this better before I go to refine it. So Refine Edge is much closer to the last step that it is an early step, so trying to make a good selection to begin with and we'll talk about a couple of tips of making sure you're getting a good selection to begin with then I hit refined edge. Now, in a financial, I have all these options for how it's viewed, each of which have their own benefit. But here's a good example. When I switch it to on black right away, my first reactions. Oh, look at that friend is horrible. And I'd spent all this time fixing it. But he's not gonna be on a black background. He's gonna be standing in some photograph I've chosen. So this is a choice, which I would say, probably isn't the best one cause gonna distract you as opposed to help you. So the one that I use honestly is this one on layers that's showing the background layer. So now all that little white French thing that kind of distracting and ago I hardly see it at all. And if I move him to a different part of the photograph, it might be even better now if you do see areas that yeah, you need to fix, you need to get the hair. But so the two views that I use the most in this refine edge dialogue box are on layers, which means show me the actual background and then because ultimately My goal is to turn this into a layer mask than the other view I use is black and white. Because that's kind of a preview what your master gonna look like and as well see, there will be times where you think I'm not sure if I've got those little hairs, but in this black and white view, you'll see just enough to realize, OK, they actually looks better than I thought when I'm on layers. And if you're a keyboard short kind of kind of person l for layers K for black and white. So as you're working, you could just press those shortcuts and then be working and kind of switch back and forth. So while this dialogue boxes open, everything is live, so you can try different things in this particular case. Overall, it doesn't look too bad. There is a slider here, smart radius and radius. Now, some people's first reactions to turn that on and move this radius. And the the theory of that is I will be smart. I being refined edge will be smart and go well. They're soft edges here, and there's hard edges here, and some of the time that works. My concern is if you push that slide a little too far, you'll suddenly go. Oh, his hair looks great. And then you realize that now here, the edge of his arm looks a little soft, so there's a very fine line. I've yet to find the formula that says These are the exact settings because it's different all the time. My personal feeling is I want to use these as little as possible. If I need to use them, I will. But as an example, let me to show you what I mean by that. Looks to me like it. It found a little bit more. But when I go to that black and white view, I can see some areas like this wasn't there before and there's some problems here, so it's actually introducing more of a problem than helping me, so it's very image dependent. But I would say I would not use that very much. The tool I do use a lot is this one, which is automatically selected. You don't have to click on it. Refine edge tool allows you to go around areas like hair. At first. When you paint like this, you'll see the original color of whatever the person is on. When you let go, it will attempt to find Mawr little hairs. Now, in some case like it is, it doesn't look like it's made that big a difference. But if I go into that black and white view, I can see it's found a little fume or hairs that I can accentuate. Now again, I'm deliberately starting with one where there's not a lot of hair just to kind of show you the thought process I would go through. I would not feather here. Some people have this philosophy that they always feather the edge because that makes it blend in better. That's fine. Don't do it here because here, once he had feather, there's no DeFede Arise command. It's gonna be soft edge. And if you decide later on, I wish it wasn't. That's one of the hardest things to fix in photo shop. If you feathered and made a layer mask and you have a slightly soft edge in the mask, really tough to easily change it back again, so I would instead use a different method where, if I do decide to feather, I'll do it later to the mask, not to the selection. So personally, I would say feathers out of the picture until later. Not to say you don't use it at all. I don't that often, but when I do, I don't use it here. So in the early versions of refined edge, these were the main controls we use now. I don't use them very much at all. With the possible exception of going shift adds just a little tiny bit, because if there is a very slight fringe, just kind of compacting it or contracting it, just a hair sometimes can help. There is an option called decontaminate colors. Now this one is can be useful, but you have pay a price. So say you had a photograph of someone on a white background or a colored background. No matter what you did, you were seeing a little fringe of that original background. When you hit decontaminate colors, it attempts to remove that fringe. But the problem is, and it may not be a problem, but I want to make you aware of it. It actually alters the original pixels around the outside of the person, so if you ever decided to go backwards a little bit, You're not gonna have a solid white background. You're gonna have this weird color thing that that software does to try toe decontaminate. So if you know well, there's no reason I would ever go back to the him standing on a white background than decontaminate colors works fine down here at the bottom. Says, How do you want to help with this? With very, very few exceptions, I output to a layer mask, and one of the reasons you'll see in a moment is it gives me a second shot at this. So if I make a selection, the only reason I don't usually make a selection is to turn into a mask anyway, so it just is easy to choose. Layer mask. Some people choose new layer with a layer mask, and that's just to preserve the original. But to be when you make a layer mask, you are by nature preserving the original cause. You can always throw away the layer mask so and this is kind of interesting. Remember settings you could check that it might remember, but more importantly, the chances are you're not going to use the same settings all the time anyway, So I would suggest that this perhaps wouldn't be the best choice because every photograph you're going to alter the settings slightly anyway, so click OK, and it has the layer mask. Now, I can still see some areas that I want to tweak. We talked briefly the other day about the whole don't get caught up in zooming into 1000% and making decisions there because no one else is gonna look at it that way. So remember that when you're evaluating, how does that is? Starting to look realistic is looking ATM or the way people would be looking at it not, you know, in this context, zoomed in. Also, you may decide later on. Actually, this might look better if he was over here. Now that changes things cause I might have worried about. There's a little white fringe right there. But now doesn't matter because there's a white thing behind it. So I'm all about cost saving in terms of time. So if sometimes you go, if I move them over there a bit, then I don't have to do that side anymore. So nothing wrong with that. Okay, Now, What you may find is that there still areas where, like I am still not convinced about this. I want to just this a little bit. Whatever it might be, one of the advantages of making a mask is refined. Edge is now out of the picture. I can't use refine Edge anymore because refine edges really refined the edge of my selection. I don't have a selection anymore. I have a mask. Luckily, the last couple of urging the photo shop in photo shop CS five was called the Masks Panel. Now it's in properties. When you have a layer mask and you click on it in the Properties panel, you see masked functions including this. If I did decide to feather, which I may or may not, this is where I do it because it's non destructive feathering. If I put a feather of two and then later on, look at it go. That was too much. I could go and change it back, whereas in refined edge, if you feather, there's no going back. So this is a better place to feather. But the other thing that's in here, which is incredibly valuable, is this button that says Mask, edge, this is like gives me another opportunity to make the edge look better because it's exactly the same as refined edge but for the mask. So now it's giving me a second chance of going well. Overall, it's not bad, but let me do a little bit of further tweaking. So this is a place where if you have done major selection, you've done refine edge overall, looks really good, but you're concerned about a couple of things. This is like a second chance. I don't use this all the time. Sometimes it works fairly well, but it's nice to know you have this backup plan or an additional plan to say Yeah, still not sure about this. Let me go in and try and refine this edge a little bit and see if I can make it look a little better. So but what is doing as it's doing it to the mask? So it's adjusting the mask. So the combination of those three things dragged the personal new background first, Making accurate selection as you can and then refine edge turn into a layer mask is actually four things and then, if necessary, refine mask That combination has shaved off a huge amount of time for me in doing compositing and kind of foreshadowed a little earlier towards the very end of this class. I'll show you another trick, so you don't have to obsess over every single little hair. There's what I think is in many ways, a better way. It's a little trick, but it actually works like a charm a lot of the time. So that's kind of the basic concept of the steps were going to do it again. I deliberately started with one that was relatively simple just to give you the idea. The reality, of course, is we don't often have ones that are quite that simple. We want to do ones that are a little more difficult. So let's do this as an example. So here's a studio shot, and I want to put her on to that other background. It's a camera raw file, so I want to preserve the camera raw, smart object ability and the simplest way to do that when you're starting in bridge as I am or many bridges to choose place in photo shop. And that does two important things. First of all it opens in camera because it is a camera off while and at this point, we still have to use kind of the thing. Same thought process is I'm not at this stage tryingto make her look perfectly exposing everything. I'm trying to look at edges, so I would look at this strictly from a how can I make the job of the quick selection tool easier? And it may be that you adjust in some way where you're doing bad things to the interior, but that's OK, because I'm looking at edges, so I click. OK, the other advantage of doing this place is that automatically scales the camera file to the dimensions of your photo. So if you had doing it normally, what would happen is if you drag and drop, it's gonna drag at the original size, which may be huge, and then you have to scale it down this way, it automatically scales that handles down to the size you want, and then, if you want to, you can of course, scale it down even further, and we always have the ability to lower the opacity to kind of scale in context. So we're going what would make sense from a sizing standpoint, as toe have her maybe around that size, perhaps hit, enter and put the opacity back up again. Now here's a little more challenges. She's got a bit of hair, and there's a few little hairs here and there. So here's the other thing that has the approach that I've done that saved me a ton of time, and it's one of my favorite expressions. I want to end up with her looking extracted properly. But what it occurred to me. The first time I tried this, I made a selection of all of her body arms, hair and tried to get refine edge to do a perfect job on her hair and a perfect job on her arms and body. And the problem was, as good as the technology is, it's still not perfect where it'll miss part of her hair or will start to make the edge of her body a little too soft. So at one point I thought won't Why don't I just do it into pieces? Then take all the parts that are hard edged like body and arms and do that in one shot and then take her more difficult areas like hair, and do that in a second shot. And as soon as I started doing that, I was like, Works like a charm People look at and go, but okay, I mean, at first you're like, but I it's still gonna look like she's extracted completely. You just have to be prepared that at first it's gonna look really weird. So I'm going to duplicate the smart object, so I have two copies of it because I duplicated it. That means later on, when I goto adjust the camera's settings, it will affect both layers. The same. So that's one of the reasons that I'm able to do this is because I'm making two copies so I can extract her body on one and instructor hair on the other, but ultimately look like she's still there. So let's do the easier one first. So again quick selection tool. At this point, I'm just worried about the things like her arms and body to get rid of a selection. I'm holding down Option or Ault and but here now, a couple of tips about making selections. One of the benefits of the quick selection tool is it finds edges. The other benefit is as I start working on it. I don't have to hold down any keys to add to the selection, because the quick selected was one of the few that does that automatically. The only time I do have to hold down the keys of UN removed from the selection. I hold down option or old, So it here. I know if you can see it, but there's like a little minus sign in the middle of the quick selection tool, and then I can start making the edge a little better. And this is certainly where I would zoom in closer, because when you're zoomed out, it's fine to sort of do a global shot. I don't really know how accurate is until you get closer and because of the way. And I do this like here. When I did that normally think, Oh no, the quick selection tool missed all her hair. That's okay. Hair is part two, so part one is making the rest of her look good. If you decide that you're tryingto ADM or to the selection and you can't get it toe work, it is always possible to switch to any other selection tool. For example, I'd go the regular lasso tool I need to add apart here. As long as I hold down the shift key, it's gonna build on whatever I have. So there's no rule that says you must stick with the same selection tool. The whole time, in fact, is exactly the opposite. Use whatever tool you need to to get the result you want all the time. As I'm doing this, I'm pressing queue for quick masters to do a quick look at it in a different way to see how does it look? And I think for our purposes right now that lets looks pretty good to me. So now I would go to refine edge here. I would definitely not move this slider at all because I don't want to do anything toe potentially soften the edges of her body and arm. I just wanted to look like this. If anything, I usually just put this minus shift as just a hair just to make sure that there is any kind of little fringe thing. I'm not going to see it that looks pretty good other than the fact she has no head. But that's OK because we're not finished it. If we stop now, it would be very strange, but we're not gonna stop yet, so I click. OK, and now I have her body extracted looking pretty good. So now I would hide that and go to this other one. Now I take the exact opposite approach, which is now let's focus on the more challenging part, which is her hair. So I switched to the back too quick selection tool and the quick selection tool is not gonna find every single hair here. But that's okay. I don't really expected or wanted to. So I'm trying to get close to where I know there's hair. And then there's some parts where I'll probably just have to get it myself. So I'm not gonna, like zoom in and try to select around everywhere. I mean, you could, but I would first try the way that the refined edge will attempt to find it for you. Before I would put a huge amount of effort. I used to be that guy that would zoom into 1600% and gold. You know, like I get every little hair that's kind of the hard way, and you may still do that occasionally. But towards the end of my show, you my little hair trick you'll find made change your opinion about Do I really want to zoom in and try and get every single little hair? I'm not sure yet how it's gonna look. I go to refine edge, just you have to get past the weirdness of floating head in the middle of a photo. We just have to get past that part. Just look at it for what it's really worth. But here's the difference. Before I said, I'm reluctant to move this radius slider up too high to make her hair look good because it affects her body. Except in this case, there ain't nobody, so it doesn't matter. I can push this slider up really high because I'm not worried about that. See what just happened? It just found hair. Now, if I had done that to the whole photo, the problem would have been The hair would look better, but her arm might look a little soft, but because I'm Onley focusing on her hair. Aiken, try pushing this really far and see. Look, it look at this. Finding all these hairs now. I would very quickly switch Kafer this mode. You can see it's still got a few problems, but I'd much rather have that and have to pull back a bit than the opposite. So I probably went say hair too far. Sorry. A little bit too far. That was completely unintentional. But I'm gonna use that again. It was definitely use that another time. I want a hair too far. So as you'll see in a moment, I'd rather have a little bit too much of this. Great, because that I can fix. But it's done a wonderful job down here and over here, so I'm quite happy with that. So this point, I would once again output to a layer mask. So now I have hair layer body layer. Now it's already to me looking pretty good. I'm not finished yet, but I mean, when someone considering how long it would take most people to do that, I think that's relatively fast and it looks pretty good considering the amount of time and to say the same thing again. It's also because I'm narrating, so if you were just doing it wouldn't take that long toe after you wouldn't be explaining it to yourself. You just be, you know, moving sliders. So it's pretty good, but it still needs a little bit of work, especially in some of the hair. Is. But caution as bad, isn't it? I mean, look at that. That's it's actually pretty good. Look at those hairs right there. I mean, I would have sat there for way longer than that, trying to get those hairs and it pretty much found them. And here's the other thing. If you spend too much time looking at the mass like, Oh, I don't really want all that gray there. Keep your focus right here. So this is what it looks like on the mask. This is what it looks like in the photo. I don't really see it. So if you spend too much time looking the masculine Ooh, I can see stuff there. Can you really look it in the context of the new file? And maybe you won't, because where that problem is is right where her armies. So it's selected anyway. So the problem I found over the years is people say, OK, I've got to really make my mass look better, and they spend their entire time looking at this and trying to go wanting to fix this. No, you don't actually, because it's already on that other layer. So while it's good to look at the layer mask and we'll talk a little bit about some techniques you can use, don't be swayed by this too much. That's why going back to square one step number one was dragged her on the new background. If I hadn't done that, that I might have been looking this going do looks really batting to fix that, but it doesn't. I don't worry about it right now, because when I look at this in the context of new background, I mean there's still a couple of areas. But what I'm saying is that what looked like a problem looking at the mask isn't really is a big a problem. As I thought. Question is that gray showing where the two masks are overlapping know that with the way that what we're seeing here is when we look at the mask, white is what's being revealed. Black is watch hidden and gray is sort of kind of showing kind of hidden so normally this would be a problem. But because what the reason is seeing Gray is it sort of started selecting her arm. But because I have this other mass where the arm is selected, it's gonna basically override that where you won't see it anyway. Okay, now, there may be still some areas where I do need to worry about Gray like that part there. That's part of our arm twos. That's OK. But up here at the top, that little part there. And I don't really want to see gray in the middle of her hair because that should be white, the very edges like here. It's okay to have little tiny bits of gray because that's a little fly away hairs. But I don't really want to see that. I want that to be white. So here's the best tip ever invented. I wish I could take credit for it, but I didn't. And I wish I knew who I want to say. It was Catrine Iseman. That was the first place I ever saw this. But I thought this was like the most brilliant thing ever. So I want to make this part of her hair white So hard to do that. One option would be to take my paintbrush, zoom in to 4000%. That would be the long way. So here's the problem. I'm gonna try and get a small brush. I'm gonna start painting and I'm doing a pretty good job. But all of a sudden I sneeze and I'm like, Dang it. So I don't want to even enter that into the equation. So there's a very simple solution. And as the best thing ever invented, you change the blend mode of the brush toe overlay overlays like this magical mode. I have white as my paintbrush right now, and I know you can't tell this, but believe me when I'm say I'm actually changing it. Overlay, Don't just say overlay. Okay, so I'm attempting to paint here in overlay mode. White cannot touch Black just doesn't work. It said something about comparing the colors that says white cannot paint on top of black when you're in overlay mode, what it can do is take very light gray and make it white, and I don't have to worry about staying inside the edges because it does that for me. It's just like the most magical thing. I remember the first time I did that. I actually had a bruise on my chest from the impact of my jaws for like, two weeks. I was just like, Seriously, that's just saved me so much time and it works the other way around. If let's pretend well, there actually is a bit. Aiken cease down here. There's some very dark grey underneath their arm that should be black. It works the other way to If I make black my foreground color, I can't paint on white, but I can paint on dark Gray. It's awesome. It really is amazing. I mean, when I think over the years when I learned that tip in the amount of time that that's saved me now be careful, though, because when I first started doing this, I would live in my layer mask and be up all overlay painting. And then I'd go. So far, when I look at it in context on Lou, I think going a bit too far. So I want you to see that's how overlay mould work. But here's how I'd actually use it. Make sure you're still viewing the photograph in the context of the background but painting on the mask because that way every decision you make you can go. Okay, that was kind of bad. Let's try this way. That looks a little better because if you're only looking at the layer mass, people tend to go too far and say I still see gray Gray is okay. It's just some areas. For example. I wouldn't want to see gray around her eyes because that would mean her eyes would be sent me see through. So there. I want to make sure it was white, but around these edges, where there's little hairs, that's okay, that actually help sell it. That there you see little, tiny, kind of barely there hairs. So the other main tip, I would say, is, while it's good toe, look at the layer mask. Here's a good example. If I was only thinking gray, I might obsess over getting rid of this great thing. But then I realized that's part of her arm, and it's on that other mask anyway. So why would I spend all the time fixing it? Just like down here? I can see a bunch of gray and think, Oh, no, I should fix that. Well, no, because that's already there on the other one. So you have to always be thinking cumulatively and in context. Okay, so now let's say for the sake of today's demonstration that I'm happy with the way her hair is looking at all the edges. But remember, way back in the beginning, I deliberately used a kind of a weird setting of her to make it easier to see what I was doing. That doesn't look very realistic in terms of matching the environment. So the advantage of back when I said duplicate too smart objects or to make too smart objects is now I can double click on either one of them. It's alright. Let's try to make this a little more believable and try and match the environment. I wish there was a way to do this more in context, but you can hopefully at least kind of see somehow or another part of your photo had a bigger monitor. You could do that, and if it takes a couple of back and forth, that's fine, too. But you see now updates, and now right away that's a little better change the color temperature have changed. The exposure maybe needs just a little bit lower. Never. Now, what I used to say at this point is how doe I combine these two into one and then I realize why. I mean, I could but I don't need to because I have kind of like having the edit ability of two different layers because it before for some reason, had my head I should combine to make one layer mask on like this. Long is, they're both the same smart object, really doesn't matter, so you could select both and make him a smart object. But there is long as you know, if you go to move her, this is the only thing you have to be. Make sure that you don't just click on one layer, cause then that would be bad. So that would be the main reason why consider put them is a smart object or putting her into a group is now those two layers will move as if they're one. The reason we're seeing that is obviously my mask that wasn't that wide originally. So if that happened, I just have to go on repaint that one little part. So if you ever move around the point where you see something else that just means initially when I made the mask, it stopped at the initial boundaries of my photograph. Okay, but that's that procedure of going through and saying Put them on the photograph first, adjust with the selection in mind. First, treat the hard part separately. Body hair. You might end up doing free because of the nature of the person that you're trying to extract. Our goal is to make it look realistic at the end result. So the steps steps you take to get there are less important to me. And preserving the two layers with their own mask means that I can really continue to tweak it. So if I go back later and zoom in closer and go, I'm still not 100% sure of this. At least I know I can go to the Let's call it hair mask because that's really what it is, a layer mask for her hair and continue to tweak it. And the only thing I've ever seen where I had toe pay close attention to you got to make sure there's enough of an overlap. You don't see a little tiny hairline where you realize that two masks are kind of touching but not overlapping. So there's no rule to say Make do this much. But just make sure because otherwise you might see what looks like a little hairline right where the layer master kind of touching. So when I'm making my selection, I usually make sure that I'm saying, Okay, let me overlap this one by a fair bit and then overlap another one just to make sure that doesn't enter into things questions about that part at all. Or, you know, we have just a couple Dave s. Fredericks asked. Why is he using the quick mask on the woman and not and not deleting the background? That makes sense. Um, well, first of all, there's a word in there that I don't like, which is deleting, right. Of course, I don't want to lead anything, so that's the first thing. So I'm not exactly sure that quick mask is just a viewing to see whether my selecting looks good. Um, I could delete the background, but the problem is your selection has to be incredibly accurate because If you delete part of her hair, it's gone. So I think they meant that they might be thinking using a layer mask and said to lead in the background here. And the reason is because if you there is a tool called the Background Eraser, which will say, Let me just get rid of all that but it gets rid of it. So if you realize later I cut off a little part of her ear hoops, whereas with the mask the information is always still there. So that's why I always use a mask just in case and destructive and nondestructive. And because I have that in the Properties panel that second go around right and say now that I've already made the mask, let's go back here to this thing called mask edge and see if I need to do any further tweaking of it. So that's the other reason for using a mask instead of just deleting as it gives you that second chance to kind of tweak a little further cool. We have a question from relaxed now out of Pittsburgh, says I'm working on an image where I want to select a large number of droplets from a splashing water. I will. I will then composite this on top of another image. Any suggestions on how to do this election of water droplets and letting some of the background show through? Well, that's a good question, because normally our goal is to say, I don't want any of the background because it was shot with the purpose of taking it away. But my first reaction is I would probably use the same principle of doing this camera so I could adjust it to make the contrast easier to make a selection and then pull it back later. And in that case, I'd be okay with seeing gray in there because gray would mean part of the background is still kind of there. Any time you're trying to select anything that's, like, difficult, like a veil or something that semi see through. That's where you do want gray in your mask as you want to still see some of the original and some of the background and some of the new background you want it to be. You can't completely hide avail because it will be well hidden, so you need to make it in semi see through. And how much touch up would you do on her dilating on her arm and her face at this point? Well, that's sort of the next step. This is to say, Get her in here and make make it look good. And then the other part of compositing is the make it believable and realistic. And a big part of that is things like lighting and coloring and all that kind of stuff. So it's important to remember that whole concept of do that last or later so that don't be afraid to in camera converted a black and white if you need. If that's gonna help you make a better selection, that's okay, because eventually all change the color setting so it matches her environment and then build on that even further. But the biggest thing in compositing is getting a great selection. You could turn into a mask, so however you have to do that if that means you have toe darkened or lightened her completely temporarily, then you can put her back and then say only now I need to worry about coloring and all that kind of stuff. I just blown away by all the time that we have a couple questions here. But this is fantastic at learning this. Especially with that jump back to a CR so it makes a huge difference. Yeah, let me ask a question Day that I just occurred to me, Do you have presets that you use in a CR to do this kind of thing like a black and white mode? Or, you know, at one point I tried that, but I to me it was it's so dependent on the image on some photographs. Or like actually, I can use it, just as is so I as a rule, I don't use the preset in that case, because on some photographs, I'm thinking who I should make this black and white. Others I look actually worked pretty well the way it is. So you certainly could. Especially if you had say you were doing ah, whole basketball team where they were shot the same way on the same background. They have the same uniform. That might make good sense to do a preset for that temporary adjustment, cause you know, you're doing a Siri's. They're the same Cool. Tina. Yeah? What technique would you use for incorporating those shadows to make her more three dimensional to the background. Well, that's That's the next thing we're gonna talk about. Let's go back to the very first question we get in the Q and A when you're selecting either the subject or the background, Is there a benefit to one versus the other? I do this every day. I select objects and put them on white backgrounds. And if I have a really pure white back and I usually tend to select the background, um, I'm or yeah, then invert the mask and my more apt to get a fringe one way or the other. The only thing I would say is depends how you're selecting it. So a lot of the reason a lot of people would do used to do that. Select the background in verses because the magic wand go click and get the whole background. The problem of the magic one is it may not be as accurate that looks because it is based on color. So with the quick selection tool, I've tried both, and I've hardly ever seen a difference. So if you select the person, not the white or the white and then invert. Usually it's the same. Okay, so the my favorite expressions selections is end up with a good selection. So if it's easier sometimes to go select all the white and inverse, nothing wrong with that. But with the quick selection tool, something on a white background, it usually does much a pretty good job of finding it, cause it's like there's his arm. There's the white back. Where there's the product, there's the white background. So the quick selection tool. If it was me, if I was the president Adobe, I probably wouldn't be here. I'd be in some cushy office. But anyway, I would change it to the edge detection selection tool, because any time to me, any time, like magic or quick, that kind of sounds hokey like really a magic wand. I mean, it is pretty cool, but it really isn't magic. So the quick selection tool it looks for edges. So if you have something on a white background, either way it's gonna find those edges. So I would say Try it and you may find, oh, in this case because the lighting it did get a little bit of a friend, So maybe I'll try. I guess that's the main answer is if you try it one way, you see a fringe, try starting over the other way around and see if it makes a difference or not. Okay? And I think we're going to get into this next. It was talking about wanting that realistic shadow. I spent all my time eliminating that shadow. So when we get there, you could give me well and shadows air. An interesting thing, because here's here's sort of a side note I think is interesting is most compositing. You see looks like this waist up because it's easier trying to put someone with their feet and make it look realistic and worry about shadows. In this case, there wouldn't be a shadow because she's standing far enough from that wall that in the real world there wouldn't be a shadow. So, Yea, I want to make a shadow so seriously. That's the reason why a lot of people would you see a lot of compositing. It's done like this because it's easier. So if you do decide to do full length and we'll look at a couple than what I tend to do is try and take a reference photo with something that has shadows in it is that way. Now you can tell because, for example, is working on one where my first reaction was I should put in a shadow. But we've done a test shot for lighting purposes, and I looked at the test subject. They weren't casting any shadow because it was such an overcast day. So I was like, Why would I put a shadow in if there wasn't one? In reality, a lot of people jump to the assumption. Gotta add a shadow cause there's always shadows. No, there aren't in the real world, there are times where you don't see shadows. If you do, then use that as a frame of reference. Cool. So in this particular case, so let's talk, first of all, about the whole shading highlight shadow kind of thing. In this case, when I look at the photograph, there's no obvious light source to say, Well, everything I can tell the sun or the lights over here, And that actually helps because if you're not, if you don't can't see one, you know that makes it easier. But let's pretend that the light source was over on the left hand side, and I could even artificially enhance that. And I want to accentuate where there are shadows on her and to try and blend her in. And for the purpose of demonstration, I'm gonna go more extreme than I normally would. So you can see me doing something cause obviously, our main goal is to make it subtle, but going back to our discussion of working with flexibility. Here's what, Another reason why I use adjustment layers. I want to be able to darken enlightened certain parts of her in a very subtle way. But as soon as you say subtle, it's hard to see. So what I would do is add whatever we want. Let's just do levels adjustment layer on top of here, and I want to clip it because I don't want affect the background. I only one affect hers. Remember, That's what this little button does. It says Whatever I do next will only affect the layer below. So I'm gonna dark in her to the point of she spent way too much time on that tan machine and just looking bad because I want able to see what I'm doing. Needles say I'm not gonna leave it like that, but it's accentuating. Okay, there's already shadow here, and there's already shadow there, and I can kind of see what's happening. I take the mask for this, and I inverted, so it's completely hidden. So whenever you had adjustment, layer filled the mask with black. It means you're temporarily hiding effect that allows me to go to my brush tool, paint with white and say, Get out of overlay mode. First of all, one of the recurring themes is Photosynth has a long memory, So because I did overlay 10 minutes ago, it's still remembering that now this is a hard concept to live with it first, because it looks like I'm wrecking this photograph because it looks like it's ridiculously dark. But I'm not going to keep it that way. This is just to help me see what I'm doing. So I'm kind of going through going where they're some shadows. That kind of makes sense now. So live with. Be prepared for the fact it's gonna look weird and bad at first, but that's because I want to see what I'm doing. Once I'm starting to safe is a government. Okay, those that's where I want the darker areas to be. Now I'll go back toe levels go, but I don't want that dark. So you see how now it's much more realistic. But the challenge I think people give themselves is they move levels this much and then they try and mask and they're going to see it. Well, then see it over. Do something so you can see we're doing. And I would add another same idea for lightning things. So first I'd make a really light. So as I was painting all time painting with white and then lower it back. And the other thing is, I was painting with a soft edged brush. But sometimes it's not. When you look at it finally, your liking and still a little obvious. Remember, one of the other benefits of the way the mask works is this mass panel. Now I can feather it again. I'm making a very subtle change. It might need to be able to see it, but what's happening is now the mask is starting. Teoh blur a little more so I can soften the effect. It's not quite as harsh or obvious, so this theory could be applied to anything. So let's say for the sake of argument, you're trying to composite someone into a Nalley way where there was a red neon lights, he thought, Well, realistically should have a little bit of red kind of well than at a hue saturation. Justin Layer make it red really red or purple or green so you can see what you're doing and then pull it back to the point where it's really subtle. But it adds to the effect so that all other concept of this recurring theme of end up with in this case I want to end up with the kind of toning changes I want. But I want to make it so I'm not struggling and zooming in really far. I still want to look at it, kind of in context and go okay. If I zoom out this far, you can see it's a subtle difference, but it's adding to the effect. Okay, so that's the first thing. Is this whole kind of coloring thing to make sure that the colors match? Okay, um so here's my alleyway. I was gonna open that he said open it area and go back here. So this is a full length shots, and now we're adding a bit more of a challenge for ourselves. And I'm deliberately using one where it's even more of a challenge because they have no frame of reference. There's not a person standing there are. There's not something in the alleyway. So if this is gonna be kind of up to me to figure out so normally I would try, even if I have toe put a light stand in there to go. I know this light stand is this high. Then when I go to put the person in, it gives me at least a frame of reference to go. How big should I make them? So this is obviously a studio shot in the place her in here since I used my son's photos, it is only fair they used my daughter's photos for a part of it. So and again. Same thing here. I deliberately shot this to be dark. But trying to make a selection from this might be more challenging. So saying concept, I might temporarily change the exposure slider so that when I bring her in, it's gonna be easy to see what I'm doing. So this is the hardest part in this case. I don't have any frame of reference. So I'm just kind of kind of guest eMate that if she was standing on alleyway, she might be this big because I don't know if that's a huge window or a small window, and it's It's funny how perception works, because we did this during a compositing workshop. I was teaching in my studio, so she was there. Stephanie is very good at photo shop, so she's looking in my shoulder. That's too big. That's just that's too big. I'm like, This is not helping, but her person I was like, That looks purchase like, really, it doesn't to me and it's, you know, so that that's the challenging part of this. So if you can find something, if you have to physically put something in the photograph to use as a frame of reference to say, now I know how high that is, then that's the approach I would take now. From this point on, it's the same kind of thing, and I'm not gonna worry. This is we've already seen how to do the whole edge thing. So I'm gonna do this part really quickly and not worry so much about making it look ideal because that's really not what I want to talk about this point. So we're gonna pretend in this case that it looks really good, even though it clearly doesn't to make it a layer mask. And now that again, use your imaginations that I've done all that same thing with her hair and it all looks good on, I say. But I want to go back to that kind of darker feel that it had because that was kind of in a deliberate effort. Now the alleyway looks to light. Well, that's OK, because that's a camera smart object. So I could change that a little bit. And this is exactly kind of as I'd start doing it now the our lives too dark and you kind of do a little bit of that back and forth. But a certain point I'm like, that does look a little odd to have no shadow at all, even though I can't see a light source, and I might artificially add one to make it look like there was some light over here, One of things that I try to do if I'm shooting in studio or outside with the intent of putting someone a background is have a little shadow. See, there's little shadow there that was from a light over here, because I'm thinking even I want to use this beauty dish, the light, her face. If I'm imagining the light sources over there somewhere, then why not put a light over there somewhere? Because that will cast a fairly natural shadow because I got the light up nice and high, like the sun would be. So that, at least, is giving me an idea. So what I try do as much as possible instead of guessing what I used to do and I'm sure other people do. This, too, is they would take the entire layer of her fill with black and then transform and do with this cast shadow. But that's not That's not how it looks in the real world that hardly ever happens that you say, Look, the shadow is that ace skewed representation of exactly the same person. It's lines that kind of start out really sharp and then fade out into something I mean, occasionally you will see a very specific dark shadow, but not that often. So in this case, this is really what I want to emulate is what you're seeing here. So above the background, I would just take my paintbrush layer here and just paint what it might get my settings on here. Okay? I just have to temporarily put it above so I can see it. I'm just trying to emulate as best I can at first just with black paint. So now that I've done that, I can hide the mass. Put this back down. I would let me ask a question just for purpose of review if I wanted a way to be able to, let's say, experiment with a filter. How would I do that? I would make this layer smart. Because then my filter can also be smart because I have no idea how much, Tobler that instead of guessing Oh, I should blur exactly this much. I have no clue. So I'm going to just sort of say that that's kind of what I was thinking. But let's take this layer and convert to a smart object. So now I can try doing a bit of blur. Here's the other mistake I think most people make with shadows is they go there we go and they fall too far. But they go. They make it really blurry. Well, a real shadow starts out sharp, close to the person and then starts to get a little bit softer, but not usually a soft, as most people make them. So one of the things I tried to do was spend some time deliberately is gonna sound weird. But it worked for me deliberately. Take photos with shadows in them. So you start kind of studying. A real shadow goes much further than I would have thought, and it actually stays sharper a lot more than I thought. In my mind, I used to show techniques of doing drop showers and get really blurry really quickly. But in most cases, it doesn't actually look like that. So what I would do here is try this blur a little bit, But then again, closer to her, the shadow should not be a soft. That's the other reason for doing a smart filter is because of the layer mask, or that you say the filter mask because now I can take my brush. Make sure black is my foreground color in the paint. Say, uh 50%. And that way closer to her, the shadow is sharper and then it gets softer. And then once I've done all that, I would generally lower the opacity and or change the blend mode to multiply, cause that will try to put into effect anything that's underneath. Like if there was a rock there, it wouldn't cover it up. You kind of see the rock looking darker now. Also, remember, we're looking at this in really, really zoomed in view so I would zoom out. There's one thing that I realized I did that's throwing me off completely. And that's there should be a shadow right there between the front of her shoe and are hell Now, if I had not used a smart filter, I would have been in some difficulty at this point because I would have this blurry object and I can't paint exactly the same level of blurriness. But because this is smart, when I open it, it's gonna show you Well, here's just the object that you did so I can kind of remember. Okay, my brush was about this big. Unfortunately, can't really paint on. Here is I'm not seeing it in the context of my photo, but at least it's reminding me how big a brush that I use. Okay, so now that I know that, then I could make another new layer just pink, that little part in there. And then I really wanted to make it the same. I would convert this to a smart object and then option or all dragged the smart filter so that one is the same. So now I've got the start. I think it's still a little too intense. I may have to lower the opacity mawr, but that's kind of the thought processes to use whatever you can that's already there keeping in mind. Like I said before, there may be some situations where this could have been a covered alleyway where there was just no light coming in other than ambient light, that's not going to really make much of a shadow at all. Or if there is one, it would be so subtle. You barely even see the fact that there was a shadow there. I think one of the things that as photo shop users we tend to do most of time is overdue shadows because years ago we saw a technique that said, Here's how you make a cash shadow you duplicate and you make it blurry and you transform it that way and because it looked cool. But it really wasn't generally a realistic shadow. It was just a technique that kind of work. But then, with as we take more photographs, start going on that shadow that shadows don't really look like that in real world question. Could you use the ingredient tool on the mask to get that kind of Yeah, I mean the great until basically we're trying to on the mass for the filters introduced some different amounts, and you could either do a Grady into, say gradually or paint with the low opacity and do it that way. So yeah, any Anyway, what I'm trying to do is not make it even across the board because a real shadow, very rarely. I mean, sometimes if it's a really strong shadow, I've taken somewhere. The only time of day to do it was at noon, where the shadow was very hard. That's a little different, but day to day in most shadows don't look like that. And without kind of saying the same thing numerous times, the recurring theme here is even at this point where I think I'm getting close to being finished, I'm still not 100% convinced doesn't matter, because I can still double click camera and say, Let me adjust this further and sad faces will too bright or whatever it is. You can still do that because word taking advantage of these temporary nature of things to help us do our job. And then once we're saying now I think I'm close to the end of what I want to do, then I can pull it back sometimes because the only downside is when you double click and go to camera raw, you're not. You're no longer seeing it in context. And that's one of the few downsides to me is I want to change like the color temperature. I'm kind of guessing a little bit. So as much as I love using camera raw, there would be times where it might be just as easy to add an adjustment layer like photo filter, because this one you can say I would like to have this kind of bluish tone, but I just wanted on her. So again, clip. So now I have this ability to kind of adjust in exactly in the context of what I'm doing. So photo filter, hue, saturation, something that's lying uto change the coloring a little bit. This is another option. I tend to do as much as I can and camera. But as I said, the one downside to it is sometimes you it's hard because you're going into camera, moving a slider, going back and going. That's not quite it. This is It's a little more interactive because you're seeing the subject in the current background. So we have Eagle Eye out there. Snafu. Who asked? I noticed that when he made the selection in his example, he didn't have the auto enhance checked for the quick selection tool. Was there a reason for that or no, I click Well, here's the thing on my version of photo shop when I launched up every morning, it's always on, so never even think to check it right. So because I'm using its let whole combo cooking in someone else's kitchen that must have been turned off because I always leave it on. The only time I ever turned that off is if it's a huge file and I suddenly getting a progress bar going. Hold on on will be with you in a second. That's when I turn it off. So that was just me not remembering to check, cause I never turn it off. Perfect. Thank you. Else did. Thank you. They just want to make sure it wasn't intentional. Good cackle. Thank you. Know we're good to go. Okay, so here is just yet another example of the same thing, this whole putting, everything in context kind of thing. So here's my two photographs taken a photograph in the studio of Says Dare on one of my son's friends, and I actually gave him special permission to wear a Canada shirt, even though he's not Canadian, because it's the only soccer shirt I had. And again, I was trying to create a sort of, ah, deliberately darker look. So I did that first and then thought I gotta put him somewhere, so I later, probably a few weeks later, actually drove to my local soccer field, and this was a little challenging because the only time I could go was in the middle of the day. So I had to try and change my camera exposure to make it a little darker to begin with, even I could do part of camera. But I also thought, I want to create depth of field a little bit. So let me tell you, it's really tough to do depth of field when there's no one standing there. It's like, really difficult. So I had to, like, focus on a post and then hit my exposure a focused Emmanuel so it wouldn't refocus so that I could create the illusion of aperture setting for depth of field. But to me, if you can take a photo with fake depth of field fake in the sense it without someone saying there is gonna look way better and try and do it in photo shop photo shop, you can try and fake depth of field with some of the blurring filters, but it's always better to do this. So before I move on, let me show you this example. I don't actually have the original photograph, but my wife and I were in New Orleans and I had her stand over there so I could create depth of field and I kept my camera in the same places. OK, now move. And she moved out of the photo. So I got this background. That's depth of field. That's not Photoshopped blurry because you could get to it with wholesale blurring. But it's never gonna be exactly the same. So any time you have, you see a cool background, you have someone with you, do that tripod even better. But if not handhold and say, OK, let me get depth of field okayed manual focus. Now you move out of the way and then we'll get that result. Okay, so that's That's how I was doing this one here. So because I have no idea what I want, I'm just gonna first of all, just open that one because I just want to get it opened, and then as we saw before, So this is just a quick little review of it. I'm not going the whole thing just to give you an idea of kind of the approach of how quickly you can do it once you start to get the idea. This is how dark I want him eventually. But against could be pretty hard for the quick selection tools. So I'm gonna do something like this where? Yes, his face is blown out. But I'm not looking at that. I'm looking at his hair against the background because that's the challenging part. And then I can I decided I want him standing right here that at this point, same as before. Turn on auto, enhance in the quick selection tool and make your selection so again, through the whole thing. But here's example. One for the first time I did it, I completely missed the fact that it missed a couple of red parts because of the way that the shading and everything and looking edges so quick, little going into quick mask. Okay, I still have more work to do, etcetera. So Okay, I'm not going to the whole thing because we've already seen how to do it. But this is just to kind of remind you you get to a point where you're kind of all right. That's starting to look decent. I would only zoom in closer, of course, and make sure had every little piece. But then when I go to refine edge. I'm gonna do refine edge in the context of that background on layers. So clearly that's bad. And I need to go back and do it. But that's kind of the ideas. Eventually you do all this. You end up with something, looks much better than that and go back to camera to put it more back the way you want. And because this is a camera smart object, I can also change that. So it gives me both possibilities. All right, so a couple of little other things we need to talk about and then have some more question time at the end, and the 1st 1 is what happens when you do all this work And this one I deliberately shot on a green screen because as where does this sounds? Actually wanted to introduce a bit of a fringe so that I could show how you could try to deal with it. So this was actually taken on a green screen background. See that little part there that's a different shade of green. That's because in refined edge, when I said Decontaminate edge, that's what it part of what it does, it actually alters the original pixels, which is OK, but I want to make aware of that that they ever decide. Well, I'm gonna reuse this green screen somewhere else. It won't all be the same color, because that's what that decontaminate at did. Thing did. So I end up with this, and I'm like, OK, overall, I love the depth of field everything, but I'm seeing a bit of green here and there, some green reflection, that kind of stuff. Green screen for Target, by the way, is really interesting when you're doing large volume. I wouldn't shoot people on a green screen as a one off thing, but if you had to take like basketball players and put them on new background, I'd be all over green screen. Because Green Screen has the one benefit of you can automate it potentially and get at least close to the end result. MAWR automatically talk about that one second, but this is the one downside. The trick. The green screen is half the person as far away as humanly possible from the background and make sure the lighting is good so you don't get this kind of reflection. I kind of wanted to show you this, so I kind of made sure there was a bit of green going on. Okay, so for example, appear in her hair. I could definitely see some green going on. Now, this technique is really good for dealing with any kind of colored French, not just green. It's if you see a little bit of blue or a bit of something or whatever, something that's kind of spilling in a bit, and you want to make the edge look better. This only works if you've already got to this point where you have a layer with a mask so that effectively, I've cut her out already. Okay, so what I do is I add a new layer on top. Now, let me pause and say, This is one of these techniques that should have a little wait for it symbol because it about halfway through you're going to like, really, that's better, because it's gonna look way worse before it gets better. So warn you that in advance. So I've got the new layer I take my eye dropper. I got my paintbrush actually the easier way and hold down option Ault and I pick the color that her hair should be. And then I paint over on, check my settings, that sort of a low opacity and just do a little bit of painting and occasionally picking a different color like this that looks way better. And then on the white. I just do that and, as you can see in no time at all, I mean, I'm already to give this to my client ago. There you go. That's all fixed. Sure, So it's not quite finished yet. So we want to mix things up a little bit, and I'm only doing this because I've done it before the first time you're gonna be like So The trick is especially with hair every so often optional clicking to sample a slightly different color because most people's hair, of course, is not all the same color. And then we do a couple of things. First of all. Clearly, I don't want all this extra paint spilling off into the background like that. So I do this layer clipping mask, optional click between them. So now it still looks bad, but least the colors are only inside of her, not on the background, and I change the blend mode to color. Now the color is not exact yet, but as you can see to see the difference, how it it went from green to kind of being more like her hair color. So once I've done that now, I could start choosing other colors and painting and seeing it kind of in person. But that's a very simple way to get rid of any kind of fringe without spending forever. Actually, painting on it is just to use that that color blend mo because that way I can go in with my paintbrush, pick up this color paint down here and it looks like I'm painting away the green in this case. But what it really what I'm doing is just going over those areas and saying anywhere, My, I see a little bit of green now. The other option, if it's really bad, would be to use like a huge saturation adjustment layer and go to the greens and say lower the saturation, and sometimes you might have to do that. But this technique is I show I'm showing the context of green it often just a little tiny fringe around on outside. That's a some color you don't want. So this is just a simple way to do it. Okay, Okay. So way back when somewhere I said, don't worry about obsessing over Hera's much because there's a trick that can make your life simpler. And every time I show this people, I think many people think at first that I'm kidding, but I'm absolutely not. And this occurred to me a while back I was working on a technique and and I realized that I was zooming in dramatically and trying to get every little hair. And then this thought occurred to me, which I will share with you momentarily and I did it. At that point, I was working in an office with a bunch of photo shop experts and instructors. So one of the tests of Does this make sense or not is to ask them to come and look at the photograph. But what I did was I very quickly hit the layers panel so they couldn't look over there and see how I did it, because that would have given it away. And I just said, What do you think of this? And they're like, Wow, that's That's really good. That's like a really good extraction. How did you extract every little hair? And my answer was, I didn't I painted some back in. So instead of obsessing over getting every hair, don't just add some. It's so much easier. And I do this all the time now, and I get compliments all the time like, Well, you must be really good at masking Hair and Esco, Thank you, even though it's a bit of that. But then it's fooling the eye by adding more hair where I need it. And once you do it once, it's actually pretty simple. It's just like anything. The first time takes a couple of runs, so you add a new layer. What we're gonna do is take our pink brush tool, but I'm going to change it to ah, hard edge brush, and I'm gonna make it really tiny, like one or two or three pixels. Is this your hair brush tool? Exactly. It's a hair, and you can actually say I call about because you can actually save a preset and call it a hairbrush. Okay, so you make it really small. That's still too big, so it's do it type it in. Okay, so you end up with a little brush that looks like this. Now, if I just left it like that, it would be bad because I was having to be painting and I'd have all these little brush strokes. So what we need to do is change the behavior of how the brush is working now. You could theoretically do it just with your welcome tablet. But even there it takes too much effort, frankly, to figure out the right amount of pressure. So in the Brushes panel, where you can use all these settings, there is a setting called shape dynamics. It's set to pen pressure because I have a pen, but I can actually change it to something called Fade when you do it, says Fade, 25. And what makes this kind of interesting is the question is 25 what pixels? No percent? No, it's steps, whatever that means. So fade 25 steps. What's the step? I have no idea. That's why we take advantage of the preview. So it defaults to 25 look down the bottom. There's a little tiny thing now that's fading out. So now, over here. I do a brushstroke unlike while that faded out really fast. Almost too fast. So that's exactly how I do it is go. I don't know. So I'm gonna try. What about 65 then do a brush rogue. Okay, that's more what I had in mind. So I can't tell you Use this number because it will vary with what you're working on. But things I will tell you is keep the brush hard and small and then try different fate amounts. So once you're happy with it, Fair. Okay, this is the mount I want to use. Then this is how you use it. Over here, you can see there's already some hair, but it's not quite as obvious as I'd like. So on my new layer, I've got my brush tool with my new hairbrush. I option are all collect and say This is the color I want and I just brush and all I'm doing is adding in more hair, this little bits. But look at that hair does not look better than it did a second ago. And that's just me, following the contours of the same hair. But the way the brushed was working as it's fading out, so it looks more like hair, and it's gonna look obvious when I'm zoomed in this close. But what I, I would remind you, is no one's gonna look at it this close. So the recurring theme is as you move around ago and this part here, I'm not too convinced that looked a little cut off here. Optional click and just add a few strokes. Now this fade to me now looks a little bit too much on this side, so we'll go back and the fade more toe something a little lower and that you may have to do that occasionally. Kind of mix it up a little bit. But all I'm doing is trying to follow what's already there and just kind of add more to it so you wouldn't want to do is call. Well, you could, but whatever it takes for realism, but honestly, guys, this change the law things. I realized I was one those people where I was completely obsessed with Oh, that I missed a hair there and I spent an hour trying to get every little tiny little hair, and I did this and it took like, 10 minutes, and I was like, Wow, that kind of works Now again, this is zoomed in close for the purpose of demonstration. I made my brush a little thicker than I normally would, so it's a little on the thick side, but that's OK. You can see zoomed out when people look at it. Or maybe look at it this size. That's how they're gonna look at it. And now there's just looks like little fly away hairs that you would otherwise struggle to get. Now, once you've done that and you found this is kind of the size of brush that works well, you could if you wanted to go in and in here there's an option, says New Brush preset. So then you make a little preset, which is your fading hair brush. My feeling is, while that's OK, don't assume that the fate amount will always be 50. Sometimes that might be six. Sometimes it might be 80. It's gonna really depend on the resolution of the photo, how much the hair they have. As we joked well, before some people, I don't use it all on my photos because it may be a little bit down here overall. I mean, so remember when we thought about presets? The idea of a preset is to get you started, so there's nothing wrong with saving a brush preset for this. But be prepared to change it on the fly and go, Oh, down here. That faded now needs to be more or less. But honestly, when I started doing that, now I do it all the time on photos and it's cut out. Can't pick a number of, let's say, for sake of argument, half a hour to 45 minutes that I used to spend trying to get every little hair and paint in overlay mode and make my mass look better. Now I get as accurate as I can and sometimes with financial does a wonderful job. Other times I need a little something to fool the eye, because whether we realize it or not, when we look at a photograph that we think might be compositing, we look at hair because that's since it's the hardest thing to do. That's one of things you look for to say. Does this look realistic? That and lighting. So here's just have questions in a second, but I thought this was a really interesting story about the world of compositing. Most people know Joel Grimes from the world of compositing. He taught here Photoshopped week previously, and he does really good work. And he told me this story that I thought was really interesting about people's perception because he started posting his composites on flicker a long time ago before everyone knew about him and people wrong, adding, All these concrete is asking for feedback at first what you think of this house, the lighting look and people were adding feedback. After a while, he became very known for his compositing because that's sort of his style. And then one time on Flicker, he posted a photo he done on location that wasn't a composite, and fighting went, I don't know, Joel. This one doesn't look real at all because he was so known for compositing that in their brain they were looking at it going. It doesn't look really interesting because it actually is. You know, it was the location shot. This is the background. But because people in their head were so much thinking, well, Joel does compositing that they were looking at it through those eyes on. How funny is that? And that's why if you tell people, here's the composite I just did. Then they're gonna look at it and look to try and see the edges. If you don't just say, here's a photo I made because a lot of people use that term when they're saying and said, Take a photo. I made a photo. This is a variation on that. So if you include the word composite, then people will look at it that way and start to be more critical. If you just go look at this, then hopefully no one will say, I don't know. I mean, if they still do, then there's work to be done. But the point is, don't put that thought in people's head, you know, unless you're you know, you know you're doing it for a client where they're saying, Well, I want to be a bullfighter So if you can put me in that background, that would be awesome. Well, obviously that's not a real photo, so that that's a little bit different, but the more we can do to make it realistic and take away that perception so much the better. Yeah, I think one of the dead giveaways. Let's look at the photo you have up right now is that infinite depth of field. And so the suggestion you made to shoot background out of focus in cameras a really great suggestion, I think. Yeah. And that's one of the challenges is that backgrounds are You can take them anywhere on that. I'm glad you mentioned it. Tweak the thought in my head whenever I'm walking around and I look at something interesting, like that alleyway. I didn't take one photo. That alleyway. I took 15. And somewhere from this height, and somewhere from this heightened some of because I wanted different options, I wouldn't I'd hate to say, put this person on there doesn't really match a well. So if you're out shooting backgrounds for compositing, doing them in different levels do some depth of field stuff. So that gives you those options. Thank you. We have some other. Absolutely. You just touched on the fact that things that give away composites law are the hair. And then also the lighting people. When you talk about the alleyway, you said you'd probably add lighting to, you know, you know, memory mirror, what was on her face. Can you kind of show that we have a couple people asking about? I mean, there's there's various ways of doing that. But my first thought is always, of course, do it in the most adjustable way possible, so I would probably consider something like levels or curves, at least to begin with. In fact, let me just levels will be easier to show, I think, in this case, but it could be anything. But It's going to things where at first I'll lighten up the whole thing and then invert the mask. Take my paintbrush hold. This is where the little hairbrush would not be so terribly useful because it would take four days. That's that's one thing. Remember with brushes. Once you do one thing like that, it's going to stick with that brush. So for a while will be like this is the worst possible brush ever. So now I paint with white and I'm tryingto like lighten it so that I'm creating. At least there it's lighter on this side. You could take to whatever level and actually draw a shape with the polygon lasso tool. Then really blurred, so it looks more like a beam of light. But realistically that unless it's a window where you can see Oh, there actually is that already happening in an alleyway? It's more just the feeling of there's light over here somewhere. And by doing it this way, then I could still go in and feather it, even Mawr, or if I wanted to make it an overall more of an overall fact. I could lower the density of the mask, so there's like anything a phone. There's probably five ways to do it, but I would always start off with a way to at least let me edit it and say, Let me try a setting that I can then go back and change it if necessary. Cool and Roger asks, What's the difference between using the edge reduction option under the refined edge and using the different option under the layer menu? Could you? Well, different ages kind of. It just looks basically on a pixel basis where you can type in 12 or three pixels, and that that works as well. I like the percentage things I think it gives you. You can be a little more. You're doing it percentage wise instead of either one pixel or two pixels or three pixels. So there might be times where doing it minus 2% is better than minus 4%. So technically, it's probably very close to the same. But because I'm doing it also in a during the selection, I'm seeing it. Maura's part of the overall process. Cool. Thank you. And this is sort of ah, reiterating question. I know we've talked about this a lot over the last few days, but if for V case on this is a yes or no, I think if your image is not a raw images, this process still possible making a smart object well, and it takes a sleigh different approach. But if you have a J peg, you can still open that J peg and force it open in camera raw. They open with camera raw, then moving forward. It's quite similar. It just doesn't start that way. So with my raw files, that happens automatically because the way my A. C R. Is set up. But if if you don't have that, then you just have to take that Jay Peak and say open this in raw, and then it gives you that ability to do that back and forth. Kind of adjusting. Thank you. Great. We have a question from Bonnie Pfizer. Do you try to match the depth of field on both of both shots to use the same F stop on both? Well, if if I'm shooting for background to me that the eyes are mostly like does the depth of field look right from the background standpoint and then the person, it's less important to use it. I honestly I tried, like writing it down, saying make sure shoot the same way. I didn't really see a difference. It was Mawr. Does the background create the illusion of that? I shot this with depth of field, so you certainly could do that. But in, uh, in addition to that, my own question do you do any anything on the edges to soften them up, or like on the edge as opposed, Like cause you made this perfect selection in my to Crisp? Do you ever do anything? You know, my impression is years ago there was this sort of school of thought that whenever you're making a selection, always put a small amount of feathering on it, and that was sort of the way everyone did it. And I started looking at riel photos and going. But if it shot a certain way, people don't have that soft and agit that could quite easily look crisp because the sun was shining right here. So that a very defined edge. So I would say I wouldn't do that as a general rule. I would look at ITM or does this make sense? So the more time you could spend, like when you go into a photo shoot with someone walking around outside, take different photos than look at them and say What is the edge look like in real life? And if the because of the nature of the way the photos taken, you're like, Oh, maybe I do need to soften this, but like one of some people I know, it's like, Well, with hair, one of my solutions, I just really make it soft like feather it. So the hair blends in better, I'm like, but in real life, hair doesn't blend in its there, so we gotta always be comparing to same stories that drop shadow technique we used to use was cool, but then it didn't really emulate riel shadows. It was just a technique that people use in photo shop because they could. Right. So we'll study real photos and use that as your deciding factor is toe. What do I need to work on? Brilliant. Thanks. And there was a lot of those types of effects, right? David? Those early days, we do those things. It was like, Did they really looked really good? Yeah, exactly. Right, Because exactly. Hey, so Frank K from Georgetown, Indiana, ass. I understand taking so much time and effort for your paying customers for your personal stuff. Same same workflow, same decision, same time and without just putting it, putting it out there. It doesn't take me that long anymore. It used to when I was doing compositing. I think is gonna take me three hours. Now. Realistically, Aiken, I feel like I can do are realistic composite in 20 minutes, in part because I shot it properly to make sure it looks good. So I have in my own studio. I know if I put these lights here with these settings, that's gonna emulate things and make it easier to select. And with these methods of saying camera smart objects to do the hair set within the body, that's cut my time in half. So whether it's paying customer or not, I'd still take the same amount of time. It's not that much as it used to be. A lot of people shied away from compositing because it used to be well, it's gonna take a few hours, not necessarily. I mean, here's when it takes a few hours is when you take a photograph, someone in front of a brick wall and then you say I should put them on a different background. That's way harder. So don't if you're outside, you think, Hey, even have the remote his chance of doing composite this person. Oh, look, there's a white wall over there. Let's have you stand in front of that or something. Contrast ing, and that's I think the problem and people have asked this question all the time and say, Well, I have this photo and I want to composite them and I don't would ever want to be facetious, But sometimes the answer is you know what? Take a different photo because gonna take four hours to extract someone off that background, take a different photo and use a grey background or something. Generic war. Move them further away so you can create depth of field. So by the nature of your camera, the edges air more detectable by quick selection. A lot of compositing is how you take the photographs. The problem was always that people like, Well, this is what I got. So now I need to do a construction. Well, that's gonna be harder. I'll show you an example. What? I mean, I used to show this in the class. If this is the only photograph you have that's gonna be much harder contracted can even see her hair. So the only hope I would have would be doing that same over a just in camera raw. But frankly, it's still a lot of effort. And I didn't take this photo for that reason. But I always bring this one to show if someone said now put on different backgrounds like, you know what? She's actually on a nice background. That's why I took it that way. So don't make your life harder by saying, Let's just have you stand over here in front of all these trees, and then later on, I'll extract you. That's way too hard. Find a plane or background that where there's contrast edges. Think one of my other favorites. Freshens. Think like photo shop. The quick selection tool looks for edges, so you should, too, when you're shooting, create edges for the quick selection tool. Don't make though it's life more difficult by having a person with brown hair standing in front of a brown door because that's gonna be harder for Photoshopped. Eso think the way photo shop would have? How could I make photo shops life easier? I love that. Just what an amazing thing. Just putting in your mind. Think about shooting for a photo shop. As a matter of fact, Dave, Well, what if I do that? I realized no. One other folder that I want to show you, because this is where compositing and that kind of thing becomes fun is when you can just start shooting. With that in mind, here's a dancer who wanted something really cool, so that's just like you stand there. All stay and still. Now you move over there. It was very easy to composite that because there was no overlap. Here's my all time favorite project with compositing. I had this idea of having the same model be like superhero villain, but fighting herself now if I had done it where she was standing here and here and there was no interaction, it would have been OK. But this was much more interesting to me so that the end result was actually having Hello fighting. Now here's important lesson. I thought I would make my life easier by using green screen. It actually made it much worse because she set up wearing the world shiniest black pants that were like she was wearing a mirror green everywhere so it actually weighing more challenged. If I do it again, I would definitely not use green screen again. That the reason ideas I'm thinking those It was actually for a project from the company that Westcott that makes the green screen. But that kind of idea of shooting with that in mind and creating something that you just couldn't do normally in one shot to me, that's kind of where the fund exists, one that my daughter did many years ago. Same idea, same model twice but interacting with herself in different outfits. So that takes more pre planning. But again, something very different. That isn't just a standard portrait. And that's leading into what we're going to next. Yeah, so let's talk about it. Well, the course graphic is that one of the girl reading the book and everything flying around. And that was kind of an idea that I had that I want to create more unusual portrait's for someone, words like she's really into literature. So let's do something really fun that relates to her hobby that I couldn't easily do by myself with one shot. So I set that up and said, Okay, if I do it this way and I kind of thought through it and that's what we're gonna do in our photo shoot is sort of simulate the same thing. Build it again with some some props and things, but kind of go through the process of how I think like Photoshopped to say in the first shot there will be a C stand that's in the way, but I'm not gonna take every shot with that there. Some of it it won't be there. So when I pulled them together a separate layers. I take the pieces I want from each layer toe, build a portrait using layers. So I'm actually looking through my viewfinder going. That's a layer that's another layer, and I'm building it in my head so that when I get the photo shop, it's easier to put everything together and then create something, and we're gonna look at two examples. One will shoot there and build another one. I did the other day here, where I just took some photos in the alleyway behind the building with another example of how you don't have to be in a studio. You can just take photos outside, but use that same thought process to create something that's different and interesting.