Skip to main content

Smart Filter Techniques

Lesson 29 from: Photoshop Finishing Touches

Dave Cross

Smart Filter Techniques

Lesson 29 from: Photoshop Finishing Touches

Dave Cross

buy this class

$00

$00
Sale Ends Soon!

starting under

$13/month*

Unlock this classplus 2200+ more >

Lesson Info

29. Smart Filter Techniques

Lessons

Class Trailer

Day 1

1

Course Intro

05:55
2

Layer Masks

15:37
3

Adjustment Layers

23:47
4

Clipping Masks

08:38
5

Intro to Groups & Smart Objects

23:44
6

Quick Mask

09:18
7

Defining & Creating a Brush

14:49

Lesson Info

Smart Filter Techniques

first thing I want to talk about in this segment. It's a little more on touching on from some of the advantages and ways we can use smart filters because we talked earlier about why smart filters air useful. The fact they're so edible won't show a couple of specific techniques that kind of build on that idea and show how you can do things in a way that's, I think, a little easier than it was in the past and gives you that ability to explore. So as an example, I have this photo I want to start to play with a little bit. So before I do anything else convert to smart object. And what I'm gonna show in this case is a technique that there's a lot of things that I like to show a photo shop that I feel should have a wait for it or it's not finished yet. I call on it because about halfway through we were like, not sure about that one right there, so this would be one of those because at first is gonna look like I just doomed doing a horrible job of like completely blowing something where you l...

ike, Why would you ever do that? Let's get a little closer so we can see. So I always thought, in photo shop, you know, there's thes blur filters like Ghazi and Blur, or some people call it galaxy and blur and things like that. But I was wished there was, like, soft, dreamy focus blur like the old days because what gasoline around the outside, their lands to create that kind of blurry effect. And there really isn't the filter that does that. But you can get to that and result and do it pretty easily, using smart filters so you can use just about any filter ice tend to still use ghazi and blur. But you could use some of the newer ones that do a bit more of, ah, a blur. But the key to this technique is that you have to be prepared for the fact that initially it's gonna look bad before it looks good. So I'm gonna blur it quite high, and I can't give you a number, but to the point where you don't want to see detail, you want us roughly tell. Okay, I only know that's lady an umbrella, cause I know that's what it looked like before Now the good part about this is because we're going using smart filters. We can try this and then go back and tweak the results so I would just say Blur a lot and then we can always tweak it. So you click. OK, I mentioned before that one of the advantages of this smart filter is that not only can we adjust the turn this on and off and double click to adjust the settings, we also have order right here. These other options by double click on that, let's look at her face instead. All right, try again. Here we go and that one of them is simply lowering the opacity. And as you do, look what happens right away. We're getting that kind of dreamy focus instead of out of focus. So it almost was adding, like this really soft texture to it, so that can create a really nice effect, because your taking away the blur, but it's still kind of blurry, and this works better than other methods. I've seen where someone would blur it and then try and fade out the opacity or something. This kind of does the same thing, but in one spot. And the advantage that, as I mentioned, is now that I've lowered that opacity setting for the filter, I can go back and say, Well, now, what would happen if I do blur it, even Mawr or a little less? You can see each case getting a very different results. Some of them will actually look almost harder. But at a certain point, you kind of go right about there are now I'm starting to get that softer focus effect that I was looking for. Now, once you've done that, if you look at it and say, Well, I like what has done overall, but there's something about her face I don't like is much or some detail area. Remember, this is the mask for that filter. So we go back to our brushes and get this stuff pretty regular. Let's make sure we have a brush gonna act normally here because I got some other weird settings on that should be okay here, and I basically I just want sir, why my shortcuts Air suddenly not working. It's interesting. Okay, so I just want Teoh paint and get this out of her face a little bit. If I paint completely black, it tends to look a little artificial, because now there's nothing there at all. So there's a couple of choice here. One would be lower the opacity of the paintbrush, considerably like 20% and just do a little bit. So it's really subtle. That's certainly an option. But the other thing that I could do leave it at 100%. So I can really see what I'm doing is what I want to do is make sure I'm seeing the area's very clearly that I want to say, Take that away, take the effect away. But then if I double click on this, it gets me back to the Properties panel. One of the properties is density. So now this is like lowering the opacity of the brush after the fact. And one of the things that I think is that in Excuse me, a very valuable principle to think about in photo shop, and that's a term that I use in almost every class. I'm surprised I haven't used it so far in this class, and that's I think about what I want to end up with sorry, being in this case, I want to end up with a very subtle masking of her face or the filters in there. But if I paint 10 or 20% it's almost like I'm going. I think I've got all our face, but I can't really tell by painting 100% black, even though it initially looks really artificial now, it means if I can see what I'm doing, then I can lower that density toe, pull it back to more reasonable number, and I use that philosophy and a lot of different things, and we're gonna take a sidestep room. I think this is important toe talk about this a little bit outside of even the world of smart filters. Turn this off for a second. And let's say that in this photo I want to try and change the color of her dress just ever so slightly. So let's start off by making a selection for a dress that will make life a little simpler. Probably I'm just using the quick selection tool here. The quick selection tool. It's one of the few tools you don't need to hold down the shift key to add to the selection. You do need to hold down option all to take away area. So I don't want her arms and everything else. I just want just her dress and missile. But there, Okay, so now if I want to change it, my goal is to end up with a subtle change. But if I just do that right away, I might not be able to tell whether really made a good selection. And master not so. What I may initially do is go overboard and deliberately change it toe like fluorescent purple or something, cause that way, now I can look at it and say, Yeah, okay, I can see a bit on her hand that shouldn't be there. And now it's easier for me to see. So that's philosophy is when I use all the time, let me show you doing it without, um making a selection first to show you that both possibilities exist. So the other way to do things if you don't it's hard to make a selection is you can just add your adjustment layer, change the Q and saturation dramatic, and court's gonna change it on everything at first, which looks really bad we take the layer mask and inverted, and then I take my paintbrush at 100% capacity with white. Wherever I paint, I'll be painting in the new hue and saturation adjustment layer that I just created. So this way I can do this without as much effort. The main point I'm trying to make here is if I had made a subtle little shift in the color of her dress, I'd probably be one of those people have my face about two inches away, my monitor going. I think that looks good, but I can't really tell. And I don't wanna spend all that time questioning does that looking good. And I want to it really fast. So this way can go in and paint all these areas. If I go on to her arms like I'm doing here, you can see it's very obvious. So then I can switch colors back to black and a mask it back out again. So by deliberately over doing it, it's much much easier to see what I'm doing to make sure I'm not making any glaring errors. And for now, we're gonna even though they're still Mr. But we're gonna pretend that I've done it perfectly well. Now that I've created that now I could go back. It's OK, let's put these back to kind of normal for a second. Because, remember, all I really want to do is change the color of her address. It's a small amount, so now that's a very subtle change. Can even see it changing a little bit. So it's a subtle change, but it was much easier to think of that end up with that approach. So when you're using filters or adjustment layers or anything else, part of the reason that I advocate so much about working non destructively is for this reason, this health would be more accurate because Aiken deliberately over do something. Make sure I'm getting things in just the right place and then pull it back again. And that applies to anything, as long as it's on a layer that has a mask so you can adjust on the fly like that. That's a very good approach to take, I think, and I know without question, it's over. The years save me time because I remember I used to be the person I zoom in tow like a ridiculous high zoom percentage and I was painting going. I think that's working on. I realized, Why am I working so hard? I should be able to tell that anytime you have any filter adjustment layer or effect that allows you to deliberately push it over, do it and then come back again. Why not? Why not take advantage of that? And that's how weaken start to experiment with a lot of different things. It also suggests if you try a filter and you're like, Whoa, this is just a bad filter that I'll never, ever use. And there are some of them that your first impression is like that. Try them as a smart filled because you might change your mind. For example. It's not what I use a lot, but I was trying to make a point here. For years, people used to joke about this filter called Find edges because there's no options when you choose it. It just does it, and this is typically what it does, and more often, not your liking. That's beautiful, right? Why would I ever ever want to do that? Because it's just really that, but because it's now a smart filter. I can't change the settings because there's no settings to change. But I can change these settings. And, for example, if I change the blend mode to something else and lower the opacity that's completely changed the effect of that filter What I use it maybe I don't know, but see how it's now. It doesn't look anything like the the find edges filter that were used to. So if you're ever in any filter and you're like, Oh, that's kind of wonky, wacky, why would I do that? Keep with it and then try some of these settings. Try different a blend mode. Try Lauren the opacity. Try whatever it is you know, some combination, and you may well find that it's like, Oh, that actually has some possibilities. On the other hand, you may be like No, still really bad Don't like it at all, and that's okay, too. But that's why I think we need to be prepared for experimenting and seeing what happens because part of what makes these filters so smart now is that it's changed that allows us to change the behavior of filters that otherwise honestly weren't really that useful because a lot of filters and Photoshopped for many years. I'd be like, Nope. Talking is that one can think of a reason why I want to, But now that I know I can go in and do this kind of thing, this is how Aiken, you know, get to the result that I want. Okay, now, we touched on something a little earlier, but I want to show it to you live so that there's no confusion about this because the great thing about smart filters is they come with a mask, as we saw. The downside is there's one mask, no matter how many smart filters you have. So, for example, maybe I want to do something like this, working in a blur and do maybe surface, blur something like, It's too much, just a little bit. And that looks a little over the top. And I'm gonna fix that in a second. But right now, the whole photograph has this surface blur applied. I really just want apply it to her skin tones, so I'm gonna take this mask inverted so it's filled with black. Take my paintbrush 100% opacity, foreman, color white and start painting and by doing it this way by applying a filter, filling the mass with black and then painting with white. It's almost like you could think of this like my paintbrush has just become the paint with the surface blur brush, because what I'm doing is revealing this filter wherever I paint with white. Now it's still too big, I mean to intense by the way I'm should verbalize us. I'm varying the pressure of my pen on my tablets, so I'm not constantly changing the size of my brush because I have pressure sensitivity turned on. So for some areas pulled down, the space bar lets you scroll to a different area. And I'm doing this with the full recognition that this is too much. But using that same philosophy, it's easier for me to be able to mask over her skin than it would be if I started with a really subtle amount of that surface blur. I just go through, pick all the spot's gonna pretend I did them all looks OK, maybe a bit more here that now that I've done that and come back and say Okay, now that I've mastered all, maybe that was a bit much. Let's pull it back this way. Lower the numbers a bit and I might also double click on here and pull back the opacity a bit enough. We turn that on and off. You can see there's a change. It's pretty subtle, but it's enough to get the effect that I want now. The problem happens, though, but now I want to sharpen everything else. Well, with this one smart filter, the problem be, if I added a sharpening filter, it would be mass. So now I would surface, blur and sharp in exactly the same areas, which is not what I want. I want the opposite. So the only way to do that effectively without and still preserving all the edit herbal nature is to duplicate the entire smart object and then kind of start again, meaning, take this and throw it away and then now do a sharpened filter. So let's do smart sharpen again. I'm gonna overdo it so you can see what's happening here. So now it's sharpened everything, including the parts that I just previously did that, um, surface blur on. So I don't really like the way it looks. I don't want to start over again. I mean, I could paint all over again, but really, all I want to do is say everything except her skin tones is what I want to work on. So I already have a mask here that I painted previously. That's her skin. Any time in photo shop, you see an element you can use. You can copy it just by dragging. But I want to tell it. Replace this one right here. So I hold down Option or Ault and I click on this mask and drag it right on top of this one is you want to replace it? Yes, I do. Then I have to invert it. So it's the other way around. So now, in one document, I have the this smart filter which is doing the blur on her, and the other one is doing this sharpening on the other area, each with an appropriate mask. And of course, once I do that, if I look mark closer, Mr Bid here, I could still continue to paint on those mass. But I see people all the time that every time they start a new layer with the mask, their thought process we don't have to start again. And they started completely blank mask and paint from scratch. And while you can do that, if you already have a mask that's pretty good. It might just be the opposite. Why not just duplicate it? Replaced the blank one that's there and then, if necessary, like in this case, all I've done is massed out her skin tones that I might want to mask the umbrella that I paint that as well. But it's still saving me from doing something from scratch every time on the classes I taught here a couple years ago, the 1st 1 of personally that was called Photoshopped productivity. It was all about ways of automating things. And one of the statements I made there and I make all the time is I think a lot of people come in with the wrong interpretation of automation, and they think of automating the entire thing. But to me, anything that can save me a step. So if I can duplicate a layer, copy the layer mask and tweak it, that's better than starting from scratch, because some of the work is done for me. So my one of my favorite expressions to always be thinking compared to the alternative so duplicating that layer mask and inverting it was much faster compared. The alternative be starting from scratch and painting again. You have to a little bit of tweaking. That's still better, then, starting from scratch every time. So if you can see it on there somewhere, you can use it because this is the same smart object duplicated twice. If I double click and go, well, here's the original one, and I'll just make some change so you can see I did something to it and save it whenever you duplicate a smart object in this exact clone of the other one. So that means both of these will have that same information on it because they're just complete copies of each other so that way, and have to ever worry that What if I do all those smartphones and then decide she's a little too light or too dark? Not a problem. Just come back to this document were to that and add curves, brightened everything up. It saved wait for it, and then when I come back here, it's brightened up, but it still has those same effect supply. Now the same theory applies. This is a file that I just was started out at as a smart object. But if I started completely from scratch again with one of my raw files on their off, I'll I guess I'll use this one. And since it's there long as I opened as an object, if I duplicated three times and do things that still the same smart object. So if I have done something to these and I come back and change a setting soon as I click, OK, they all change. Okay, now that's assuming you want it all to change. So let's visit this for a second. Let's say on this document that I was going to apply a filter. Let's do sharpened because we can see it pretty easily. So I've done a heavy duty sharpening to all this area, but that I want to make this sky look quite different. I want to make it a different color or something. Well, if I duplicated this and change the sky and camera raw, both of them would change because that's what happens when you duplicate a smart object. Whether you have filters on their or not. So if I ever want to make a separate smart object still tied to this original photo that I can edit separately, there's a separate command that we use. That's called New Smart Object by a copy or via copy. Would you like to say it? Now? I can throw this one away. Let's say on this one, I'm just looking at that sky there. I want to change it to a very different color just so we can see something was to Hugh and we're gonna change this. There we go. So here, take my quick selection tool at Layer Mask and now I've got what I wanted it. It's the same photograph, but now I have the bill to edit the two parts of it independent of each other. So the difference is when you duplicate a smart object, it's an exact clone copy. So anything you change to the original, it will change the other one, which, in the case of the girl with umbrella, make perfect sense because I wanted them both to be the same. If you're doing something like a landscape where you want the sky to be edited independently of the rest of the landscape. New smart object by a copy lets you edit them independent of each other. Still, it's the same file, so you're not having to worry about will they line up because it's exactly the same document you're just doing in a different way. So one of the things that we have to recognize that makes Photoshopped challenging is when people say, for example, in this case, well, should I duplicate the smart object? Or should I use new smart object by a copy? And the answer is while it depends, and it always depends because depends on what you're trying to do next. So part of the challenge of deciding something is you have to kind of think through it. Well, if I duplicated, it's an exact clone. Is that what I want or do I want to separate versions like an edit independently, and the answer will depend entirely on what you're trying to do next. But as long as you understand those two things Air four and the difference that means when we're using things like smart filters weaken on top of that stable on this layer, I want the smart filter toe act little bit differently because I'm gonna adjust the settings and camera raw differently or whatever it might be. Okay, so a couple of important lessons there one is think about the end result and work back from there. What I want to end up with and the other is, if you're including smart objects, there's only that one smart filter. So if you're trying toe have two different types of mass, you can't. I just there's it's just not possible. Okay, now there's another approach we could possibly take, and this is a new option. Show you this so you can see this one. It's a bit of a brain teaser A little bit, Um, but it's still it's kind of interesting. So let's do the same kind of thing to start off. I'm gonna do girls and blur the gamble. Lower the opacity, just a hair. But in this case, I'm also gonna paint on the mask. I said, I really don't want the filter in here in these areas, and I'm doing it really zoomed out far. Just kind of give you the idea. I would normally be much more accurate, of course, but for now we're just gonna go with this and then use this feather thing to kind of soften it up more like maybe that So now I said, Well, now that I've done that, I want to look this capacity just a bit more. I think there we go. But now I want to sharpen some of the details. So before I show you, option one would be to duplicate whole thing and then start again. The other choice, which is this is an optional thing, and some people like this better. Some people find this more confusing is whenever you have a smart object with a smart filter, it means you have that one mask. If I take this layer, which is already a smart object, and I right click and choose convert to Smart Object. Now I'm taking that special magical container and putting inside another container. So now it's like I'm starting again. So now I could do and do some other filter. Let's do something crazy and obnoxious. I don't want to that one, though. Let's do ad noise just so it's really obvious. Not that much, though, So now, in this one, I could paint the mass completely differently. cause I'm embedding one smart up inside another smart object. The only thing that throws most people off on this one is in order to get to the original, you have to just keep digging. So when you double click to look inside a smart object, there's the other Smartwatch. We want to get right back to the original photograph. I have to double click again. So this is like think I was used the When you take Tupperware containers, you can stack inside each other, but you can see through them or some people like Russian dolls. There's another one you can do the same kind of thing there's Technically, there's probably some limit as to how many times you can put a smart object inside. A smart object. I don't know what it is. I've never reached it. I did hear rumors of a guy in Texas once it had, like, 12 and his brain exploded because we just liked by too much information. But there probably is some limitation. Realistically, it's unusual to do more than a couple, but it sometimes is a really good solution because we have to think about it is whenever you convert something to a smart object. On some level, it's like you're kind of resetting, so you can try something new. But you still have access to all that information. Which reminds me I meant to answer. You ask the question before and I completely forgot to show you it, so I won't do it now. So I don't forget you talk about when we were doing text. Could I put texture inside the tech? So it's something I don't know why. That just reminded me that nothing to do with it, but show you now. Okay, so let's get someone text on here. Okay. So I need to have some text. And I wanted to looked like this. But I also want that text actually have texture. And the problem that we run into is one of the ways we could use. Texture is the texturizing filter. The problem is, if you go to the filter menu and find where it is now is in the filter gallery. This is the warning dialog box that comes up and it says you've got two choices. Rast arise or convert to smart object. I have to tell you that if you have an older version of photo shop. All you'll see is Rast arise. This is a dialog box that just magically appeared one day, and I was so happy because it used to always bother me when it would come up and say, This type player must be Rast arised before proceedings. People Okay, so they'd rast arise and then wonder why they couldn't edit their type anymore, because Rast arise means converted to pixels. So a much better choices convert to a smart object. Because now I can. There's my texturizing filter. Aiken texturizing to my heart's content. It's an edible filter because it's a smart filter. But what if I want to edit my text? Well, if I try to, all it's gonna do is add. New text is technically that text. The original text is inside that smart object container. So this way I just have to get used to the fact that to edit the contents of the smart, obviously, you have to look inside of it by double clicking. So I double click. There's my text, and now I could pick some other whatever it is to change it to that. You can see I've made some change if I save it now, when I come back here, there's my new type Still has the texture on. So that option of convert to a smart object is a way that you can preserve the edit ability of our type but still have a texture that's edible as a smart filter. So that's kind of option. One option to would be if you want to use an actual texture like one of these photos, and I would do it this way, place in photo shop and stretch it to fit this, enter and then make it clipping group. So now you are clipping mask in. See now the actual photographic texture is looking like you're only seeing inside the letters. Thank photo stops pretty smart in some areas because it's really interesting to me the way that all these things work together. So, for example, I don't really need this filter gallery with texture. But since I have it, if I change the blend mode to something now, it's gonna pick up that texture and the photograph. Texas kind of mixed together, maybe multiplies a little too intense. So maybe overlay or something else, or keep it at normal, but lower the opacity a little bit. We're getting a bit of that color, texture, something like that. But if I turn this off completely, I don't need it. You can take this layer and do layer style. Things like that still bevel in Boss, and it's smart enough to realize you've got bevel in a boss. But there's a texture in it, so it will still include that as part of it. So even though you're using that clipping mask idea when you add any kind of layer style like Bevel in Boss, you're still going to see the photo in there as well. And just to make a point. Not that I probably would in this case, I suppose. But if I did want to blend these two together different way, there's always that can always pull in the old blend if sliders that what happens if I take part of this layer, maybe the darker parts, and have them start to disappear so that the text the red underneath is going to start to show, you know, I mean set multiple times. There should never be a point or like and well, I guess I have to stick with that because there should always be the ability to take it some other step further and say, What else can we do? So a couple of different options and there's no easy answer. To say this one's better than that wanted Complete depends on what you're doing. I personally, with text specifically, tend to do the clipping mask unless there's a particular reason why it makes sense. If I'm duplicating it more than once that I would do it as a smart object, because if you edit the type in one, it'll update in the others.

Class Materials

Bonus Materials with Purchase

Tool Kit
Action Kit
Luminosity Action
How To Use Photoshop Actions
Starter Kit

Ratings and Reviews

karlafornia
 

I like Dave's teaching style: methodical, well-organized, VERY knowledgeable, interesting, relevant, and delivered with a really good sense of humor (he's a very snappy dresser, too!). Most of all, his lessons are most useful in teaching me how to save time processing my photos in a NON-destructive way and with a stream-lined workflow. This particular class is not only versed in technique, but I LOVE how he encourages creativity through experimentation and "playing" and pushing the envelop with the program. that is not as scary as it sounds because Dave is all about working with smart objects, smart filters and other such ways designed to save us from destroying our photos or work that has to be redone or scrapped because we went down a road of no return.

a Creativelive Student
 

Dave has a brilliant (as well as humorous) way of teaching and I always learn something new from him. I have purchased many of his previous classes and love every one of them! Thank you for another great course!

Student Work

RELATED ARTICLES

RELATED ARTICLES