Pro Techniques to Create Compelling Composites

[Music] [Julieanne Kost] Hello, everyone. How are you this morning? Thank you for joining me after the MAX Bash.

You guys are hardier than I am. I was not able I know myself better. I'm not going to a Bash and then getting up for an 8am. So I'm super excited to be here. My name is Julieanne Kost. I work for the Digital Imaging Team at Adobe. And as always, I'm very excited but also very nervous to be here. Thank you for coming. So I've got a lot to cover, so I'm just going to jump into it. So in this session, we're going to walk through my creative process, as well as my technical process. I'm sure that we all have very different ways of working. I'm primarily a photographer not a designer. This is the one big design show I do every year.

But what I have to say, I hope that it is helpful, informative, and you take something away from it. A quick bio. I am a lens-- I don't like to call myself really a photographer because then people ask questions. So I just say, "I'm a lens-based artist." And they go, "Oh." And they don't really know what that is, and so they don't ask any questions. So terrific. But I think that what I'm really trying to do with my artwork is work through my issues with a camera and the computer. So I'm not a photojournalist. I'm not trying to fool anyone with my images, which should be pretty obvious, all right? So right after college, I had a job, a day job at a medical imaging company where I had access to Photoshop after work. And that might not seem like a big thing. But this was in 1992, so it was, I couldn't afford the Quadra 800.

Yeah. And Photoshop. No way could I afford it. But I had a wonderful instructor. I was going back to school for photography, and he recommended me for a job at Adobe in '92, and I got it in technical support. So if you ever called technical support and you got the right answer? All right. Now actually, technical support was this perfect blend of my psychology major and then my love to teach. So anyway, enough about me. Let's talk about you and Photoshop and compositing. So the three images that you just saw are what we're going to walk through today. I'm going to have to do this very quickly if we can walk through them all. So yeah.

How many of you guys are doing composites and stuff just personally for yourself? This is great. This is my artwork. Okay. So one of the things that I love is that the thought, in all seriousness, that art can be very cathartic. And for me, this is my cathartic part of my job. That's why I love to do this class at MAX. It's much better or it's not much better. It's much better than therapy. Yes, it is. It's helped me discover things about myself, like that I'm an introvert, and it also helps me when I reflect back on the artwork because a lot of times you just get in the zone, and you get to point B, and all of a sudden you look at the artwork, and it can really reveal a lot about the relationships between events in your lives and things like that. And then also someone yelled out once that I should have realized that my color palette was somewhere between doom and gloom. So there you go. Okay, so people often ask me where I get my inspiration, so I just want to talk about that for a second. I read a lot. And it's not like profound reading, right? Last night, I actually finished American Dirt. That's an amazing book. You guys should all read it. Really good. But I find that I get more interesting ideas from books than I do from, say, watching TV or watching movies. And I think that's because the book, when you read the written word, it's really making your imagination go. You have to figure out who all those characters are, and you have to visualize the places that they're in. So I would just strongly encourage everyone to read when they have a little bit of downtime as opposed to watching TV. But that's just me. My motto is create more, consume less. So every time I find myself scrolling, especially when Firefly first started and it was like building it, I'd be like, "Oh, time for the phone." And then 20 minutes later, you're like, "Oh, that only took a few seconds." But now I wasted 20 minutes. So anyway, as I read, I do jot down anything that I think is interesting, anything that creates some strong visual in my mind. And you can do this electronically. You can do it old school and put it in a notebook. But that way, if I have I think, well, I want to make composite, and I don't know what to start with, I have this book of ideas. So I never really have to start with a blank page. I just go back to one of these quotes that made some visual impression on me, and then I try to recreate it. It doesn't mean that I have to finish the image based on that, but at least it gives me a starting point. This image that's up there right now, Hindsight, was from, something I read from Jessica Blau. And it occurs to her that the thing that she had lost may had never really been there to begin with. So you might not really see that in the image, but it gave me a start to create the image. Okay, and it's not just what I read. So if you don't read, that's fine. How about song lyrics? This was Regina Spektor. Two birds, I won't sing it for you. You don't have to worry. Two birds of a feather say that they're always going to stay together, but one's never going to let go of that wire. He says that he will, but he's just a liar. You can get a lot of good stuff through music. And maybe this would be good stuff through prompting with ChatGPT. I wrote my own review with it. Any of you out there who don't like to toot your own horn, just write it in ChatGPT. These are the things I did. I did them really well. Make it sound really good. Put in all that flowery big word stuff that my boss would like that I don't know how to do. All right. It can also be experiences. So this image I created during the pandemic when it seemed like some people were barely able to keep afloat.

And I do like to start my illustrations, like I said, with a concept, but I also like to watch the image just unfold.

So last example, I created this right after I was diagnosed with cancer in 2021, and everything is fine, so that's fantastic. But at the time, I didn't want anyone to know. So we were all working from home. It was during the pandemic.

Because I didn't want anyone to know, I just had to keep saying, "Oh, how are you doing? Everything's fine, everything's fine, everything's fine." So anyway, other people during the pandemic had way worse than I did. So but like I said, for me personally, art is just this amazing form of therapy that I love. And I always have self-assignments, multiple self-assignments or self-projects or side projects going on just for me because those side projects is totally an investment in you and your time, and it can help you learn new technologies and expand your creativity without the pressure that you might have in your job. Anyway, that's another talk. So regardless of whatever type of compositing that you're doing, I would say that these are the primary things that you need to consider. Most of you guys are designers, so you know this already. Probably inherently, it just comes to you. But color and tone is pretty obvious.

Contrast and density, so that means really like you want to have the same contrast, same density throughout the whole thing. You can't have one object that has a really dark black shadow and one object that has a lighter black shadow or a shadow that has more blue in it and then a shadow that has more yellow in it because the things were taken at a different time of day, right? The direction and quality of light. I don't know how many of you guys are photographers, but I'll tell you the easiest way to make a composite in the universe because I do a lot of landscape, more realistic composites as well. But you take a picture, and there's this amazing iceberg floating by, but there's all this other ice and stuff in the way. And it's like, that's not what we're focused on. We don't always see the camera. The camera sees everything in the scene. I'm just looking at the iceberg. That's what's important. That's what I want to capture. That's the story I want to tell. So I'll take the picture of the iceberg. And then that same day, a few minutes later, when the boat goes a little further or something, take another picture with a blank background that has the same direction of light and the same quality of light so that you can use that so you don't have to make up a new background, right? Okay, structure, texture and resolution, that's just you got to make sure, the ISO of the camera. What's the grain structure? What's the texture in the image? Is there a lot of grain, a little grain? You're going to have to maybe add some grain or add some texture in order to get them all to even out. Focus and depth of field. If you are the photographer, try to capture everything in focus. So much easier to blur things after the fact than to try to make them sharp.

If you want to watch for your reflections and your shadows and all of your images when you're compositing, don't forget to take out those reflections because it's funny sometimes. And then scale and perspective. And then odd elements, just things that just look weird. Now this is all in like an ideal world. My world is a little bit more whimsical. So I did want to create something. Usually if you've been to my classes before, they are a little bit darker. So I want to do something more whimsical here. Yeah. So if you want a formula, if you came here for a formula, how do I start making composites? Here's my little recipe. So I create a stage, right? I create a stage just like a theater has a stage or a movie has a set. I'm trying to create the background to put my elements in. Then I will add my primary subject, and then I will add my secondary subject or any supporting characters in my story. And then I will add-- Sometimes I like to add just a little, one or two little treasures. So if someone does look harder, they'll see something more than someone who's just glancing at it on their phone or something.

And then I'll just say that I then take out everything that doesn't belong, right? So if there's something to my image and it's not adding to the story, I'm going to remove it because I do think that less is more. So here, we're just going to walk through. So I start creating the environment, right? And I just wanted to have this foreground where you can imagine this car racing across. And I know. Who out there is like, "Oh, my God, the horizon's crooked?" I knew there would be people does it bug you? I know. I knew there would be some people you're like my dad.

All right. And he's awesome. All right. But then I started to play around with clouds to set the mood of the day. I'm ignoring, I'm going to totally ignore it, especially at this point. I don't care if the horizon is crooked. I'm just trying to set the stage. I can blend my colors later. I can resize things. I can change the perspective of things later. I'm just trying to get the mood, right? So I'm trying to set the mood, and I'll do this with multiple images. I'll just throw them all in there and see what works.

I try not to get bogged down at the details at this point. You can spend so much time on one little component and then realize that you don't even want to use that component later. So this is really just more the conceptual stage. And when I am out photographing, I'll take pictures of everything and anything. If there are good clouds, photograph the clouds. If you are going to photograph the clouds, try to get a little of the horizon in so you know where that would be so that when you make your composite, the clouds will look correct, right? And if you see something in the foreground or you see an object that you think you might want to use, shoot it from multiple perspectives, right? That's going to make your life so much easier. Walk around it if you can, shoot it high, shoot it low, all right, so that you have all those source images to work with. So this is the combination that I thought would work best because I thought maybe the giraffe would be tall, so I wanted the blank space for the head. And then it was time to add the car. So this is the car that I decided to add, and I used Adobe Camera Raw, but you can use Lightroom or Lightroom Classic, whatever you want to enhance the colors and the contrast and get rid of I don't know if it shows up there or not, but there's reflections on the car on the left. And, yeah, so I've tried to get rid of them on the right, just enough that I know that I'm going to be adding a little bit of a texture, some structure over this. So that is going to work just fine. And I'm doing this while I have it in Camera Raw. If you don't have the RAW file, then most of the time, I think if you're buying stock or something, all this stuff should already be removed. But for me, it's not. And this was just random. I was in a neighborhood, and I saw all of these cars were parked outside. So I'm like, "Stop the car," and just went out and took pictures of them. But then when I decide that is the car, I will actually open it in its own document. I don't bring it into the composite. Why not? Because I might decide later that I don't want it. So if I'm going to spend time to extract something from the background, I want to do it on the full res file that hasn't been transformed, especially if you're doing a layer mask instead of a vector mask, right, because your vector masks can scale, and they'll be beautiful. But your layer masks, if you scale them down and up and rotate and all those things, they start getting softer and softer, and you might have to redo it. So I do this in its own document, and I use the Pen tool, and I go around it. And you're thinking, "Oh, my gosh, you use the Pen tool." Yes. That is my favorite tool. I love the Pen tool. But all of these other tools really have gotten so much better. It's incredible. But in this case, it's a car, so they're pretty harsh lines. And then I also wanted-- I didn't want the whole car. I didn't want the bumpers and stuff. Okay.

And we can probably see that the light is coming from the upper right to the back of car. So then there were the wheels in front. So the driver, obviously, when he parked, turned them a little bit. And again, didn't bother me, but those same people that didn't like the horizon, it's probably driving them crazy right now. You're the same as my dad. I showed him the same image to my dad, and he's like, "The wheel." I'm like, "What about it?" He's like, "It's not right." I'm like, "I don't care." So I thought, "Okay, I will fix the distortion a little bit." The thing is, the wheel's hubcap has red round stripes around it. So when you take it at an angle, you're never going to distort them back. I mean, you can use liquefy. You can use perspective warp. You can use whatever you want. But if you really want them to look absolutely perfect, you're going to have to redraw them, right? Use the Pen tool, and then you're going to have to shade them and all of that stuff. But, yeah, I didn't care. But then I thought, "You know what? This would be the perfect time. I should just go to AI and just make a whole bunch of wheels." And you can. But it wasn't the wheel that was supposed to be on that car, so I just couldn't do it. Anyway, I had bigger problems. I had to stuff a giraffe in the car. So I was like, "Forget the wheel. I'm doing the giraffe." So luckily, I'd been to a little, a living desert park somewhere in Southern California where you can walk on a plank, and then you get to take pictures of the giraffe half their height, right? So the perspective is okay.

I thought it was fine because obviously, I don't want the back feet. I mean, I put them in the back of the car because I was pretending there would be room for the feet. But anyway, as long as the focus and the depth of field match the car, it's all good, right? So I've got my little primary subject, and I actually did use the-- I don't think I used object select. Maybe I did. Either that or select subject, and then you use select, and mask, and it was amazing. And I like the way that the head fits in the cloud, but the face isn't engaging with the viewer. So if this wasn't a square image, I could probably get away with it because if it was a horizontal image, then it would give him space to be looking off to something. And if you had to add type, that'd be great. But I didn't. So I wanted a more engaging giraffe. So I have another picture of the same giraffe. So you just do a little head swap, right? So you just go whoop, whoop, whoop. All right. So then there's focus and depth of field. Those all matched up, so that's fine. And then, of course, it's time for our secondary subject. So we've got the primary. That's the giraffe. We've got a secondary. That's our warthog. We've got to rotate them a little bit. But again, when you're trying to photograph components, make sure you get the whole body, right? Because I did try. I had some other ones, and I went into Firefly, and I'm or actually Generative Fill, and I was like, "Legs, legs. No, it didn't work." So I do try to include as much depth of field as possible. Like I said, it's easier to blur than to sharpen. So then just by putting the secondary element in the composite, it's going to encourage the viewer to see a relationship between them. I mean, they have to, right? First you see a giraffe, and it's like there's a giraffe in the car, okay. But then you start thinking, "Well, what's the relationship between the giraffe and the warthog, right?" So when you have these secondary elements, know that your viewer is always going to assume that there's a relationship, which is great because that's going to help engage the viewer, right, and help them and encourage them to ask questions about the image. And then I adjusted the lighting to both of them. So let's see. Is it between there and there? Can you see the lighting? I can't see from here. So you can see under the warthog, he gets a shadow, and I reversed the lighting on the giraffe so that it would match the lighting on the car, right? So you've got to pay attention to those direction of light. And then I added the little hat, and you probably can't see it, so there it is a little bigger, little hat. That's Russell Brown's hat from a previous event. So always take pictures of things that are weird, and you're not going to see them again. All right? But there's no shadow on the head, right? And the lighting's off, right? So the lighting's coming from the wrong direction on the hat. So just go into Camera Raw, or you can use an adjustment layer and just lighten up that side. I mean, you might get away with it because you'd be like, "Oh, well, he was in the shade of the giraffe, but no, I don't think so." So I just did that and then add the little shadow. And then we can add the shadow under the car. But it seems a little static at this point. So I thought, "Okay." I'll just add some water or some horizontal lines because in this one, I tried adding things like I spun the wheels, and it just looked silly. And I tried to kick up some dirt. And then I was like, "Well, it's not important to me." What's important is the giraffe and the warthog. So I just added some water and that horizontal line because it's helping move your eye across the image, then you can see that it's also like adding what looks like a little bit or imagined speed. So and then you had to fix the reflections in the wheels, right? So you got to just-- I just hid everything except for the background, merged all of those layers together and then just put it right there in the wheels, scaled it down, just put it in there. And then supporting elements, so this seems a little off balance to me, right? So I wanted something in the front of the car, so I thought, "Well, why not put a raven?" I tried adding the tail of the giraffe. That just looked weird. I tried putting the feet out. I tried putting the hooves of the-- It's all this stuff that just I didn't need it. You know what's going on, right? Plus, I like to leave a little bit up to the imagination, right? So I was even trying to put little wedding cans at the back of the car, but it's just less is more to me sometimes. So then I've always used my own photographs in the past, but I don't have a pink scarf, and I don't have one that's flowing in the wind, and I don't have a wind machine and all of those things. So I started exploring with AI because now I'm feeling like how is AI different from Stock imagery? And so it's not. Is it? I don't know. I don't know. So I'm fine with using it. I made a whole bunch of versions. By the way, in case I forget to say this later, when you get those variations, you get three variations that you can click through. On that layer, you can always right-click on that layer and say Convert to Layers, and it will break apart the variations into their own layers because sometimes you part of variation one and part of variation two, and you want to mask them together. So you can break apart those variations by just right-clicking or Control clicking on Mac and saying Convert to Layers. They're very special, but there's nothing special about those layers except that they're together. They're just pixel-based layers as variations. So when you break them out, they just become layers. So then once they become layers, they are just regular layers, so you might want to convert them into smart objects. Okay. All right. So that's the same thing. Warthog needed a little bit more hair, right? So you give them a little hair. That's also, I mean, gosh, that is what AI is just made for is just adding a little bit of hair. And the thing is if there wasn't any hair there in the beginning with, you can just paint some hair on a separate layer to give it an idea, right? Because the way that the AI is working is it's only working on the layers below it, right? It only makes the information. So if you're on this layer and you've got the warthog, just add some strands of hair on the warthog to help it know what you want and then run Generative Fill. And it'll be like, "Oh, you wanted hair there." That's great. And it'll do it for you. And then I like to add these little treasures. So I put a little pearl thing in the beak, and I put the little things on the hat. And anyway, repeating colors and shapes, you guys know this. You're designers, it can help to unify the image, and it can help to lead your eye around the image. And then making it pink on the left, and you've got the pink scarf on the right. And then also making the density of those shadows consistent is really important. All right, so then I just added a little bit of texture, a little structure. All right. And then we can-- Whoops. And then we can do a final curves adjustment layer, right? Just darken down those edges. Keep your eye within the frame. Cool. One last thing. I will just say I try not to stop at my first what I think is the done final image. I will go ahead and try to make a few other iterations. I had all sorts of stuff in here. I had a lighthouse in here. I know a lighthouse. Why would you have a lighthouse? I don't know. I thought it'd be cool. But anyway, I try not to stop at the first solution. All right. And then this is the globe. And this is the image we're going to walk through right now. So I'm going to quickly and hopefully-- Because I want to sit down to do this. Okay, so let's start with this. We're going to do this globe, and I'm going to start with my background image. So this is this little-- Oh, I forgot. I have to plug in one more thing, one more thing here. All right. Excellent. And gosh, why is there so much stuff up here? All right. Let's do face ID.

Usually by the third day of a conference, face ID doesn't recognize me. It's like, "Who are you? Who's the really tired lady?" Oh, that's me. Okay. So let's go ahead and start by adding the globe, right? So here, I've already got all the layers in here because I think you guys know how to take layers bring layers into a document. So I've got a globe up here. This is a shot in Ireland. This is actually something that they put on graves. And so I want to go ahead and mask this out. So I'm just going to grab the Marquee tool. Now I actually have my tool set up completely different if you were in my previous class, and I separate the rectangular and elliptical. And I redid the keyboard shortcut for elliptical to be O because that would be far more easy for me to remember because an ellipse looks like an O. In case you're wondering, I don't call it the elliptical. All right. So we've got the ellipse there. And we can just add our mask, except that I don't have-- There we go. Can you guys see the bottom of the Layers panel? Yeah? Okay. Good. All right. So there's my mask. I got my mask around my globe, and then we just grab our brush, which is the B key, right? And we get a little bit bigger of a brush, and we can come in here. And we can do this with a different opacity if you want to, but I'm just going to paint with black. And we can just paint this out, right? So...

So this is something-- I mean, I wouldn't have taken that picture otherwise, right, except that I was like, "Oh, that globe is really cool. I might want to use that someday." And then we'll just take off the bottom of it right there. Then we can add a little bit of glass back in just by using this picture right here that's just a texture. So again, anytime I see textures, it was so nice that there was rain here. I mean, it wasn't-- Yeah. It was nice that there was a little-- We don't have rain in California, so it was nice to see rain. And I took a ton of pictures of just rain and rain on windows and condensation because I never know when I'm going to use them. So I want this to blend, though, right? So we better pick a blend mode, maybe overlay or soft light, probably soft light. It might even be a little bit too much. So I tap the V key that gives me the Move tool. And that way, I can just tap 4 and that'll give me, like, 40% or maybe 5. I don't really like this right here. So let's just grab a mask. And with a Brush tool, we'll just paint that out as well, right? So I'm not painting out the globe. I'm just painting out these highlights that are going to appear because those highlights are going to make it look like there's reflections on the front of the globe, and it finishes off the fact that it is a globe. I might go back to this one just for a minute, tap the 2 key and just give me 2%, and just paint with black and get a little bit more of the actual glass out of there. Okay. So after that, we need to start adding our little raven. So here's our raven right here.

I'm going to move it up here, and I'm just going to pin it, right? So you can pin it here. I think this is a good place to pin it. Oh, not reset. Silly. Silly, silly, silly. You people make me nervous. All right. I'm going to pin it right here because look at that's pretty much a waste of space, right? Hardly anything ever shows up there. So I just put it up there instead. The other thing that I-- Well, yeah, okay. So there's the raven. So how are we going to select him? That is up to you. You could just remove the background, but I'm just going to select the subject. Because if I select the subject, that will enable me then to go into Select and Mask, right? So we can go to our little Select and come down to Select and Mask. And that brings us into Select and Mask. And I like to preview this, right, on the layers so that I can see what's behind it. And you know what? For right now, let's see how we're doing here at 100%. Let's see what we're looking at. Let's refine the hair there. He's got pretty short hair, but this right here is showing up. Let's see if Object Aware is any better.

Ding, ding, ding.

All right, so now I can go back and forth between-- Oh, that one's better. Okay. So I like that. So because I'm using a RAW file, this is probably all I would do. I would say let's output this to a Layer Mask at this point and click OK.

But why would I say that if I'm using a RAW file? Because if I'm using a RAW file, I can double-click on the RAW file, and that's going to bring it up in its own dialog. And then I could go over here to my masking, and I could create a new mask around any areas where you saw that white halo. And I could say because I've already got the selection right now all I need to do is take my brush and say, I don't know, paint around here, maybe paint around here a little bit, and then just take down maybe the whites and the highlights and the whites here. And so it doesn't matter that I paint it over here because this area is going to be masked. So I'm just painting along that edge there. And when we click OK, it'll update that. And so we shouldn't see as much of those halos. I probably didn't darken it down enough, but I could do it again. But I also know that I'm going to add a little bit of texture, so it'll be fine. Okay, so after the raven, let's zoom back out so we see where he is. We need to put the raven in the basket because why wouldn't you, right? So we're going to come over here. We're going to grab our object select, which is the W key. Now I can click on this, and it will select it. But a lot of times, I find that if I click and drag over it, I don't know why, but it gives me a better selection. See how it went all up in there in the basket area? I don't know why, but it does. Oh, the other thing you want to make sure of is that, well, make sure of-- Some people don't want or can't have their images, anything done in the cloud, in which case that's fine. You can do it on your device, and it is quicker. But you can also change this to cloud. In fact, if we go to our preferences and we go to our image processing, we can set our default right here. So now all of my defaults for select subject and remove background, as well as object select will be done in the Cloud. So I never have to change it anywhere else. So let's make that one more time because it will make a better selection if we do it in the cloud. Okay, so that's the basket. We select it, and we get our little mask. All right, then we quickly get off of that tool so we don't see the pink. All right, so we've got it, but obviously, we have all this stuff in the center as well. What do we do? Should we do the Lasso tool? Let's just do the Lasso tool and come right down here.

Don't sneeze. Don't sneeze with the Lasso tool or you'll lose your selection. Okay. So what do we do? We fill that with black. Great. And now we need to figure out where it needs to go. So we need to maybe move it up a little, move it over a little. How about that? See, I don't really like this part of the raven over here, so let's grab the Lasso again. This little thing looks weird. I'd rather have that just disappear there. All right. So we've got that. Let's zoom out. And oh, oh. See, I did it.

I clicked when I should have double-clicked. All right. So there we go. We've got that. We go back to the raven, and black is the foreground. So Option-Delete. We'll get rid of that. And now we sit in there. Okay. So we sit in there. The color is a little weird, but and I want to mask the bottom part of the basket. So depending, like a lot of times when you've made a precise selection around something, maybe like that car that I was talking about before, and then all of a sudden you're here, you might not want to ruin this mask, right? So you might want to group it or put it in a group, right, because this is the basket, and it has its own mask. And if I don't want to ruin that mask, I just add a group and then add a mask, right? So now I can go ahead and come in with a brush, a smaller brush at 100%, and maybe come along here and hide everything that's within that group, right? But then if someone says, "No, no, we don't like that." You can take that off, right? So with the-- Yeah, okay, you get it. All right. So let's add his hat because they should, ravens should always have an acorn. Okay, so with something like this, with the object select, if I tap W, again, you definitely want-- Oh, it's gotten better. Before it was picking both of them. It wouldn't just pick one.

Well, never mind. It's getting easy-peasy. All right, grab that Lasso tool...

Around the raven. And we've already got the acorn. All we need to do is fill it with black and then deselect that. Great. It's on his head. So we need to do some shadows, but I can do that later. And we probably need a duplicate of this. I don't know. It looks too light. So let's just do Command-J. That will give me a duplicate, and we can just set this to multiply. And then we can maybe just decrease the opacity there. There, I like that better. Okay, and I really like that color. So above the basket, I'm going to create-- Is it a basket? What did I call it basket? Above the basket, I'm going to create, in fact, above the group, I'm going to create just a Solid Color fill layer. Now if I hold down the Option key and I do the Solid Color fill layer, then that's going to give me the ability to say, "Yes, I want this in color" And click OK. And then I can pick a color from the acorn, choose OK. Oh, I should have clipped it while I was in there. It added it. You know why it added it into the group? So I had the group selected, and it always will add into the group. If you close the group and then add something, it will add it above the group. Okay. Just something that will win that Trivial Pursuit for you. Okay, so that looks pretty good. I'm going to go with that. Command-0, let's back out. We also have a blowfish, of course. We need to put him up there. See, I like the fact that the raven's the one that's supposed to fly, and therefore, he's in the water in a little nest, and then the blowfish is in the sky, because never mind. Okay. Right here. We can just grab this. Now we have the blowfish. Now we can say, "Terrific." I want to add a little mask to that. Now he's flying up there. Whenever I photograph objects, I try to photograph them. It was a blue background, but the thing was is that it was evenly lit blue background, so I didn't really have any shadows or anything. So try to shoot against something that either isn't brightly colored or is evenly lit or has-- So there's not a lot of shadows or things that you have to select around. Okay. And then we need a crown on him.

I guess it's a him. I don't know. So again, with a W. So even though I use the Pen tool all the time in my previous lives, these selections are getting so good. Oh, it didn't quite grab the top part there. Okay, now I want to get rid of this area here. So I'm going to hold down the Option key, and that'll get deselect. What did I just say? Just gave you a compliment, Photoshop. Said, "You were getting so good." Okay, Option. I don't want this area in there. Thank you. I don't want that area. You know what happens when that happens? It means that you're not being sincere. Photoshop knows. Oh, see? I thought it automatically went to add mode.

Maybe it doesn't. Maybe I had to hold down that. There we go. Okay. So I still have this, and that's going to bug me. So let's just delete that as well. Okay. Oh, kidding. Let's add that as well.

Okay. And we can mask it.

And I'm not on the right layer. Where's the crown? And then I will mask it. Right there. And then we need to add a shadow to that too, but I'll do that in a minute because it's also-- I have to get these little bumps on the head, right? So let's get this little dude. Bumps, they're spikes. They're probably poisonous. I don't know if I need that one. Let's try just with these two first.

Did I need that third one? I did. I needed that third one. Let's hide that. I want this one too.

Right there, right, because they are in front. So let's just fill that and deselect it. So now they're in front. Okay, so I need to add a layer effect. You know that you can just double-click on the Layer, and it will bring up Layer Effects. I never really realized that. I was like, "Why is it always coming up?" And I was like, "Oh, well, and I would always still pick it from the bottom of the Layers panel." Well, I guess the good thing about picking it from the bottom of the Layers panel is you can actually tell it exactly which one you want. So here, I just want a little drop shadow. That's fine. I can leave it as is, but then I need to separate the drop shadow. So when you have layer effects too, there's a lot of hot spots on the Layers panel. If you click on the Effects, you get a different menu than when you just click over here, right? So it's a lot shorter on the effects because you only get the effects options. So that's an easier way to maybe create a layer, right? And then usually what I do is I just Option click on the mask and wrong layer.

This one, Option-click to add a layer mask. And that way, it hides the whole drop shadow, and then I can just grab my brush, and I can paint with white, and that would just paint that in, right? Because otherwise, it gives you a shadow everywhere around there, around the whole crown. And I only need it at the bottom. Okay. Command-0. We could do the same thing. Let's go back to the basket.

Basket. 8:49. Okay, we're still good. Basket and double-click. And it's going to use the same settings that I used before. So let's just drag that down. What have I got it? I've got it on the wrong layer. No. It's because it's the group, right? So I put it on the layer, but what I really need to do is I need to put it on the group because the group has a mask on it. So I just drag the drop shadow up to the group, and it would add it there. And then I should probably do it to the acorn hat as well. Let's do it to this one. Just double-click. Where's our little Drop Shadow? That's a little bit too much for the hat. And just reposition that a little bit. Oh, do you see how they're both repositioning? That's because the Global Light is set on. All right. So if I turn that off, now I've really messed this up. Hold on. For some reason, it's not undoing. So let's just do that again. Drop Shadow. So just undo the Use Global Light. I mean, yeah, because I just want to scoot that just up a little bit. Okay. And probably decrease the opacity. All right. And then I'd want to turn those into layers as well like we did with the crown and stuff, but I'm not going to take the time to do it because I want to get to some more AI stuff. So here, there's the just-- I take pictures and then just blur them, right? So anytime I see some interesting earth tone colors, I'll go ahead and just blur it. This is a drop cloth. And then we can just set the mode to Color, right? And so we can add our little color in there. But then the raven needs an eye. So you know when birds blink, they get this really cool iridescent, right, so now we could zoom into there. And again, I would probably-- Because I've already got it in place, I would just hold down the Option key, and that gives me my little mask.

Come on. On the eye, Option-click. And then just a little small brush. Evil eye. All right, and then maybe on this color layer, I don't want it to hide. There is a little bit of color there, so let's paint with black. And maybe on the crown, I'll paint with 50% black. Let's load up the-- Where's my crown? Oh, you know what? That's so much better. Oh, sorry. I took off my glasses. Now I can see this. So here's my crown. If we Command-click on that layer, it'll go ahead and load up that layer as a selection, right? So now I can be on the drop cloth layer mask painting with 50% black, and I'll get just a little bit of that color back from the crown. So that's how I typically add things in. And then Command-0, and we've got our little condensation. And I've never ever used before the Divide blend mode, but there is now I can now say there is a reason for Divide. So there it is. Use the Move tool. Maybe tap 6. You get 60%. And then a paper texture, again, setting this to something like overlay or I don't know. Let's do overlay. We'll just leave it as is. Oh, I forgot the pole. There's also a little pole, so there's a little piece of rope right here.

See this little piece of rope that I wanted? I wanted him to look like he was one of those, like a horse on a merry-go-round. So yeah. Because why not? All right. So I start my selection, and then we can just hold down the Option key, right, with the Lasso tool, and that gives me the Polygonal Lasso tool. So we can just grab that, come up here, and then just release, and it will close the selection. So here's the rope. We add our little mask. Everything else goes away. If there was a little halo on this, we could always come in here and say, "Hey, let's just do a little Inner Glow, right?" But we don't want it White, and we don't want it Screen. Instead, we want it Multiplied, and we want it Black. And that would just darken down those little edges that we might see. We should probably have that either fade into the water. So let's go ahead and just add our little layer mask, get our little brush. Whoopsie. Where am I? On the rope. Wrong layer mask. Make sure we're painting with black and fade that in. Okay. Command-0. How's that look? Let's do another one. So Command-- We can get rid of this. We didn't use it. So let's delete the vector mask. Command-J, right, we can make a duplicate of that. Use the Move tool and just scoot that up. Boop. So it goes right up to the top of the globe. Okay, we're done. That's it. Okay. But that's it for this one. Now we got to do some more AI stuff, right? Okay, so let's go to that raven one. Does that make sense? Yeah. Are we good? Are we too fast, too slow? Mm-hm. Okay. So the raven, I needed to add a monk. This has been really hard for me. This whole, like, I know. You guys are designers. You don't care. You'll use you use Stock photos. You do use what the clients tell you to use, all that stuff. But I'm like, "I can't use anything but my own images." And then my boss said, "You have to." So I said, "Okay. I will." So a few things first. Let's go to Lightroom Classic. So I am making a ton of elements in Firefly, right? The elements. So some of the things you can do is on a gray background. That would help, right? Use the word complete. You want a complete human. You want a complete dog, not a part of a dog, not the head of the dog. You want the complete thing that you're looking for. That can help. It doesn't always work. It's like it's got a mind of its own. Yeah. So but I am creating a ton of elements because, I mean, where would I come across this? A skeleton of a dog. But the dog's face is really funny. So don't expect it to make everything in one go round or in one image, right? If I was going to put this dog with the mountains and this blah, blah, blah, blah, I would make them all separate and put them together using those exact techniques that we just walked through. I hate to say it, but none of the techniques has changed. I know that in the description of this is we're going to learn how to do Firefly with all these techniques. They're the same techniques. It's Select and Mask. It's adding a shadow. It's doing all these things. Where's the light coming from? Do I need to flip this? It's all of the things we've already talked about, all those things to consider, right? You want some wings. I mean, how easy is it to get wings now, right? But again, they don't always do them, the whole thing. But what do you have? Generative Expand. So if it does cut something off, try using Generative Expand to see if it will fill it in. If you want a vase and you know you're going to put it into a background that is gloomy or something like that, try making it maybe in the environment. You don't have to use the environment that Firefly creates for you. You can extract the vase and then go use it in your own images, okay? That one should look familiar from the float thing, right? So, I mean, it took me forever to get a lady is half her head submerged in the water and but you can do it with the prompting. It just takes a few iterations.

And then we need to talk about grain because I don't hear a lot of people talking about that. And being a photographer, I'm a little bit sensitive to it. So let me just actually-- Let me open this. I'm sorry if you were in my previous class on tips, but I think this is important. So first of all, if I was going to fill in this area and this area, I would first go to my-- The W key, which is the Object Selection tool, all right, and I would select this frog. And I'm sorry, but this is the cutest frog in the world. Look at the little belly. That is so cute. Okay. And there's no way I'm going to move this. And then this is the joke. Are you ready for it? I'm going to use Command-J to jump him to his own layer.

That was good. Don't worry. Sometimes the jokes are just for me. I'm not because-- So the little guy is now on his own layer, right? And that is important because if I want to fill in a background and I go over here to the Lasso tool-- Can I fill this in first? Yes, I can. No, Julieanne, just do it all at once. But it's going to be different. That's okay. All right. So let's say I want to fill this in. I'm going to go. Did you see? I selected some of that frog, right? You want to select some of the area that you want Generative Fill to blend with. Don't do the precise selection like the car. If I was going to say I want a new background with that car that I cut out for the first one that I was walking through the slides, I would say I would select inside the car. Maybe you still use the Pen tool, but then you're going to have to-- If the car is selected, you would need to contract the selection into the car so that the Generative Fill knows that you want to take the background into the car. Then it will distort the car. So that's why you want to jump something to its own layer first like the car, or in this case, the frog so that when I go to Generative Fill and I hit Generate, it's going to go ahead-- And did you notice? I did come down from the frog. Even though it was hidden, I'm just in the habit of just being above the layer that I want the Generative Fill to work on. So Generative Fill only considers the layers below it. All right. So now it's filled it in. There I've got one. There I've got two. There I've got three. Let's say I like that one the best. And by the way, this is what I meant by you can right-click and see. You can just say Convert to Layers. And now what I have is my three different layers, my three variations as layers within a group and then within the selection that I made. So I'm actually not sure why-- If I do Command-G here, did something change? Oh, no. Sorry. I did-- I'm on Command-Shift-G.

Nothing changed. I don't know why it groups them and groups them.

Yeah, I don't know why. Okay. Now something else to learn. All right, so let's say which one did I like? Let's say I like that one. Okay. So look what it did to the eye, right? So that's the original. That's why I jump it to a layer. See also, right here, it's expanded the frog up here, right? So I could just go to layer one, add a new layer, just a new blank layer, get the Clone Stamp tool or whatever tool you want to use, and just, hello, maybe use all layers.

Use all layers, current and below. All right. So Option-click again, and then just clone that down, right? Maybe come over here, clone that down. Come over here. That looks weird too.

What am I doing? How come I apply? Wow, it really extended the frog, didn't it? Holy moly. It's never done it quite that much before. I was getting myself confused. Well, that's not-- Oh, now I am-- Now I'm-- Oh, you guys. Oh, come on. Save me. What am I doing wrong? Yes. I've got the top layer wrong. Yes. Yes, yes, yes. Duh. Put it underneath, right? Okay. There we go. Sorry, Mr. Frog. All right. Thank you for saving me. Okay, but here's the problem. Do you see the difference between here and here? See the structure difference? Grain, no grain. Okay. So what do we do? We go to our Generative Fill layer, and then-- Oh, but when you expand them to layers, it's no longer a smart object. So we're going to right-click and say Convert to Smart Object, right? So now it's a Smart Object. So now when I add my filter, which you could also have just said Convert for Smart Filters, and I go to Camera Raw as a filter, and we zoom in here to 100%. And you don't have to view this in full screen. So we can scoot this over, hopefully, so we can see them both. I didn't do a very good job. Okay. let me cancel out of here. All right, let me cancel out of here. And let me scoot this over, right, because what's important is that I see the screen there. So let's zoom back in, scoot over to here, go down here to my effects.

Always a little different between Lightroom, Lightroom Classic, and Camera Raw. And then come down to my Grain, and then we can add some Grain in here, maybe quite a bit of Grain. But it doesn't quite look the same, right, because it's not rough enough. So if I go to the left, it's going to get it rougher. If I go to the right, it's going to be really clumpy. So almost like a reticulated image if you go to the left and more like real grain to the right. So I think that's a little bit too small. Okay, anyway, so you go in there, and you add your Grain. And then when we come over here, it's going to look much more like it's real. And then if you have more than one layer-- Oh, we didn't do this layer. But if you had more than one layer, you could just Option-drag your smart filter to the other layers, and that way you could get the same grain throughout the entire image. Okay. I'm not sure that made this image much better, but we're going to just go with it. All right, so let's close that one. The other thing that I would suggest is go ahead and make things in small bits. So I wanted this raven to have this sash going across it, but and then I wanted the sash to have this little ribbon that had a little diamond on it. And I thought, "Oh, okay, sash with ribbon with diamond and blah, blah, blah." No. No, no, no. Grab your Lasso tool and say, "Okay, I want it to be like the sash is going to come right down here, and then it's going to come maybe down over to here." And then it can come up, say, to here, maybe a little less right there. Okay, so you've got your sash. So now all you need to do is type in red ribbon sash. Red ribbon sash, I'm glad I spelled it right. And it will create that. So I'm not trying to ask it for too much. I'm trying to be very specific with what I want.

By the way, while I'm thinking of it and it's doing this, if you're using generative image-- Oh, well, it already did it. Okay. Never mind. Let's go to here. Oh, look. One little feather. Come on. Give me more than that. There we go. That's what I want. So you might need to do this a few times, but you're going to get what you want. And usually, if you've got hair, or feathers, or something like that, it will blend it over it. So now I've got this portion of it. So now I could come down and say, now here is where I want my red ribbon with diamond pendant.

Pendant? Pendant? I don't know. Maybe it'll know.

Maybe we need a spell check in there. Oh, that'd be awesome. Huh, spell check in the prompt? For me, that'd be so-- Then I wouldn't be embarrassed on the stage. Oh, look. Oh, that's see-- But it put the pendant in the sash, right? That's cool. That's not so cool.

Well, that's actually probably more what I asked for. Sometimes it'll put-- Look at here. So I've obviously done this before because I was scared I wouldn't get on the internet or something while I was here. So let's just hide this for a minute. So here-- Oh, yeah. Then they made that. That's what made me think of this was I had this little-- I had a photograph of this little emblem, and I thought, "Oh, I'll put that on the raven." And then I drew the ribbon on the either side with a Lasso tool, and I said, "Red ribbon," and that's what it did. So then I guess, "Oh, well, that's cool." Maybe I could make him some kind of, I don't know, add a little bit of bling to him since he is the king. So then here's one with a diamond. See, so before, it was doing it with a ribbon. So I mean, I hate to tell you, but you're not going to get the same thing twice. You know you never are. That's really fun when you demo, super fun because that's the thing. That's this whole mind shift we're going to have to get used to because we used to be able to do things. Everything in Photoshop was repeatable. You do it a million times, and you could always repeat it. Now it's like, this is crazy, man. All right. So there's pendant, and there's pendant. Yeah, so just give it a try and keep trying out different things, right? Then I tried to get a monk back here, like a monk in the fog. No way. I tried for the limit of my patience. So I just made it in Firefly. Don't forget, you don't have to make it in here. The other thing that people were getting confused of in Russell's Lab, which was really good because I say people, that means me. If you do generative image, if you do this, you make a selection, and then you click Generate Image, it doesn't care about your selection. It's going to generate the entire image of whatever you prompt. It doesn't care. So you don't use Generate Image for generating filled areas. You use generative image to generate an image. Okay, all right. So I actually went through-- I went through this with-- I just put in Labrador. I made 2,500 of these. I put in Labrador, and then I just changed the effects for each of them. And then I did it with three pairs, and I got a furry pair. It was awesome. But while we are in Generate Image, I will tell you, if you're using the materials and you come down to fur because you want something made of fur, also, put that in the prompt because otherwise my prompt was three pairs on the table, okay? And I said make it with fur, or I said use the fur materials effect, right, so the fur materials effect.

And it would put it on fur. So I'd have three pairs on fur. Yeah. It was driving me crazy. So I was like, "No." So anything that you add, I found it works better even if you're doing layered paper or you go to watercolor or art nouveau. Put the effect in the prompt as well. It's almost like you're doubling down on it then, and it really does come out much nicer. Yeah. Okay. So that's Generate Image. Yeah, the monk thing just did not work out. So I just generated an entire image of the monk. Where is our monk? Actually, do I have him already in here? Yeah, so here's the monk without the mask, right? So here's the monk on the Generative Fill. And that way, you know what else? If you do it with Firefly or if you do it on a new document that's 2kx2k, you'll get the full 2kx2ks worth of information. So you might have more image quality than if you just made a selection within an image and said make a monk. Okay. So yeah, so I just outlined it with a Pen tool, but I knew that I want the shadows. So it is-- Is it a smart object? It is not a smart object.

Okay. So smart objects-- How many times do you do this? I do this all the time. Oh, it's not a Smart Object. Command-J, disable the vector mask or delete the vector mask, Convert to Smart Object. Oops. You clicked in the wrong place. Convert to Smart Object and then move the mask and then grab that and then hit delete. Write an action. It's all recordable, and then you can do it in one click. If you're not a fan of keyboard shortcuts, actions are fantastic. So I have a bunch of actions here. But you know what you do? You put them in button mode, right? So now I have all of these actions, and I just place them. You can just place them over there, right? Because the thing is, if you're using this all day every day, you're going to need your muscle memory, right? And why should you have to come over here and wait for a tool to pop out just to change the tool? Or why should you have to come here and go to filter and then be like, "Oh, I have to go to blur gallery, and then, oh, I have to come over here?" I mean, if you're using all of these things all the time, I showed in my class my workspace is a little bit different. So my workspace is actually over here, right, because everything then is on the left-hand side. And I can get my tools, their options, the buttons, the layers, everything. And even my Properties panel is underneath my Layers panel so that if I add something like an adjustment layer, like curves, look where my cursor is already. So really, I mean, if you're in production and you're doing stuff all the time, write a little action that just does one step or multiple steps that you continuously find yourself doing. Okay, off the soapbox. Let's close this. Let's go back to wherever I was, which is probably Essentials. Okay, so yeah. So I've got this guy, and now he has a smart object. And now I can do Command-T, and now we can bring this down. And we've got him right there. And that looks good. And use the Move tool and maybe make them 50% and then zoom in a little bit. And then we can just use our other mask. See, so I cut them out, right? But then I want to add a layer mask here and then use my brush and then use black and then just slowly paint this out so that I can use the shadow from him, right? So that's why-- Sometimes I just add a little extra area, right? So now-- And okay, I don't like them at 50%. I'm going to bring it back up to 100% because-- Sorry, I can't get over that. All right, so because it's a smart object, I could definitely use an adjustment layer. But I could also come over here, and I could go to-- What am I looking for? I'm looking for curves. So that's under Adjustments. And then we go to Curves. And then we move this out of the way. And then I take my blacks up instead. See how I'm lightening it this way? So he's not transparent, right? And when I click OK, look what it becomes. It becomes a Smart Filter called Curves, right? So it's not just filters. As long as you have a smart object, you can add your image adjustments this way as well. I mean, you can do it with an adjustment layer. That's totally fine. But I've seen some of your layered panels. They're crazy, right? So this, I can just collapse really easily and be like, "Yeah, just hidden, just gone. It's a calm. I'm good." All right? So, let me just do one other thing. Let's see. For this image, just real quick, I had to help it. Okay, so that was before, and that's after.

I'm just going to open this image. Again, I'm sorry if you were in my other class, but this will be a good refresher, right? So because it's-- This is the AI that I'm doing. Instead of using the Crop tool and then expanding this, right, and filling it with Generative Expand because this is a huge file. This is 7,900 pixels. Oh, but I'm looking at the height. So 4,400. But my camera shoots 8,000 pixels on the long edge. So it would fill that with 2,000 pixels maximum and then have to interpolate it up. So you'd want to come in here and you can make an action to make selections like this that are only 2,000 pixels high, and then use your Generative Fill, right? So you'll want to do this in blocks if you have a really large high resolution image. And it's fine because once you build your one block, you'll pick your variation, and then it's best if you stick with that variation. So make it a hard decision. Sorry, I know you like flexibility, but make it your decision. And I'll just do this one other one fast because I want to show you the third step here, which is Generative Fill, Generate. So make your decision on which variation you like because it's hard to toggle on and off variations because if you do it in chunks, this next variation is going to be based on the one that you've already created because I had overlapped that selection, right? So here's the thing. Here's where I'm going to break my own rule though. Down here at the bottom, if I use that same thing where I just select a chunk and then another chunk, it's going to keep filling it with the land, right? And maybe you don't want the land. Maybe you need to put text down there, all right? So go along here and help it, right? So do Generative Fill and hit Generate and say, "I think it's going to be smart enough." It might add some more icebergs, but I think it's-- I'm telling it look, there's water, and there's ice. I don't want the land. Whatever I've selected, get rid of it. So it does, but it's low res. Well, fine. Now I can go in with my Marquee and grab this area that's 2,000 pixels, Generative Fill, Generate, and it's going to generate higher res water. But because that land is gone, it's not going to get confused by the land. Does that make sense? Yeah? Okay, all right. So we probably don't really need to do this. And then one last thing, selections really do matter. So we'll open this image here. And if I get my new Lasso tool, my Selection Brush tool, and we paint here, and I'm sure you guys have all seen this. But man, it was a lifesaver. I got a family of 14 that sent me their Christmas card, and they all wore T-shirts.

Yeah, and not the same color T-shirts. And they were like, "Could you just put us in suits and ties?" I'm like, "Yes." And I did, and I swear it worked. I mean, they'll know because it's them, that they're a little thinner, or wider, or something. But if you had a client, I was just dumbfounded at how quick it was. Okay, so selections matter. What do I mean by that? If I select her hand and if I fill in this area, they'll just give me-- I'll get a straight result. She'll just be standing there straight. But if I leave that hand in there, well, so she says, let's generate it and see. Oh, no. Cancel. Generate.

Select, please. Reselect. Phew. Reselect seems to be fickle sometimes, but denim shirt. Blue denim shirt. Do we care if it's blue? Sure. Blue denim shirt. Generate. If I hadn't left the hole where her arm the-- I don't know, hole and her hand there, it wouldn't take those as clues, right? And it will probably make her hair a little bit longer, which-- Oh, much longer. Well, as she's-- And who's in the background? Well, who's in my studio right now? Oh, my goodness. How awkward. Let's try another very-- Oh, well, now you're looking very stylish. And now you're-- Oh, look, and now you have a new necklace. Fantastic. But selections do matter, right? So make sure that you are accurate with your selections. That includes shadows. Make sure you get the shadows if you're selecting something. And just even a little bit, if you-- Don't quite get something in there and you don't notice it, it'll just make the selection again, and you'll be fine. What else? I was going to do something else with this portrait, and now I've forgotten what it is. And it's 9:15. Anyway, so I am out of time. Look, you guys, I hope you really enjoyed this. I hope you enjoy the rest of MAX. I hope it was worth coming to after the Bash. Have a great day, you guys.

[Music]

In-Person On-Demand Session

Pro Techniques to Create Compelling Composites - S6300

Sign in
ON DEMAND

Closed captions in English will be added in early November.

Share this page

Speakers

Featured Products

Session Resources

No resources available for this session

About the Session

Join Julieanne Kost, digital imaging evangelist at Adobe, to learn how to expedite your creative process using new and established features in Photoshop. Discover how Julieanne seamlessly merges disparate visual elements to create a single, surreal, composited scene. She’ll demonstrate key techniques to incorporate into your workflow that will enhance efficiency from initial concept to final scene. 

Topics include:

  • Using the latest Photoshop features powered by Adobe Firefly to spark creativity and streamline your workflows
  • Choosing and unifying elements to create successful compositions
  • Leveraging powerful new and established Photoshop features that will transform your compositing workflow 

Technical Level: Advanced

Category: Generative AI

Track: Graphic Design and Illustration

Audience: Art/Creative Director, Graphic Designer

This content is copyrighted by Adobe Inc. Any recording and posting of this content is strictly prohibited.


By accessing resources linked on this page ("Session Resources"), you agree that 1. Resources are Sample Files per our Terms of Use and 2. you will use Session Resources solely as directed by the applicable speaker.

Not sure which apps are best for you?

Take a minute. We’ll help you figure it out.