No announcement yet.

Tutorial Thread: Graphics from Start to Finish

  • Filter
  • Time
  • Show
Clear All
new posts

  • Tutorial Thread: Graphics from Start to Finish

    Did you write a tutorial in which you explain step for step how you made a graphic? Please post it in here.

    Any size or form is welcome in this thread. So post your tutorials for wallpapers, icons, wallscrolls, Tumblr-graphics, blends...whatever! It's up to you! If you want to write a tutorial with a lot of pictures or if you prefer writing it all out. As long it's made by you.

    Don't be afraid to write a tutorial. Even if you are new to fanart or this forum, afraid your skills aren't advanced, or afraid your English isn't great, all are welcome. Most fanartists learn a lot from seeing/reading how others work. Something that seems simple and basic to you might be a revolutionary new trick to another.

    Thank you to Nina for creating this thread.

    Lydia made the punch!

  • #2
    Tutorial Thread: graphics from start to finish.

    FanArt From Start to Finish

    First I pick my two photos and arrange them on the black canvas. Next I erase away parts of them on a layer mask.

    Next, I want to add some depth and interest to the negative space and I really like the pipes in the right pic and the little red blops in the left pic, so I duplicate those layers and place them in various places, using rotate, to give everyting a new direction. There are five or six layers of this extra stuff, making ten layers for the whole piece so far.

    Next, I am going to add a gradient layer because I find it easier to see where I am going once I have my color scheme more or less set. This one was pretty easy to color. I didn't have to lay around with the settings at all; I just chose the eigth gradient in my Mix N' Match set and set it to reverse and really loved the colors. So that's it.

    Now I want to add a stock image of some candles. I size it down and apply the median filter at 17, add a layer mask and erase areas of it. Additonally, I also adjust the color using the Hue/Saturation selection; I just tone down the red as well as bring down the Lightness.

    Now I notice that the piping on the right picture looks a little harsh comparitvely to the rest of the piece. SO I am going to slect that portion of the pic and apply some median at 4.

    I also am going to add a little bit of the Tweaker filter by XeroGraphics to the pictures of Matt and Audrey. It doesn't change anything drastically, it just brightens it up slightly. So now I am taking another stock photo--this time of light streaks--and placing it over all the image layer. It's very red so I am going to tone it down like I did with the candle picture. I am also going to use Selective Color, selecting Yellow from the drop down menu, and moving the bottom level (Black) all the way to the left. I will do this three times and it get rid of all of the yellow in the streaks.

    Next, I am going to size down the light streak picture a bit and desaturate it a little bit more and then set the opacity to 44% and also Auto Contrast it. I am also going use the burn tool and run along the edges.

    Now for a little brushwork. I am going to use Brush "132" in my set Randomness 3 and then on my brushes palette I select the options: Shape Dynamics and Scattering. Below are the settings I've chosen for Shape Dynamics.

    Then I run the brush along the left side, using white as my color ad the layer opacity set at 22%. And I add text. Now, for the text, I am using two fonts, Corleone Due and Charlesworth. I use the Edit > Transform >> Skew function and slant the text. I want to have large spacing between the letters so I set the spacing at various intervals in the thousands on my text palette. And I use some drop shadow and various gradients on them. It just takes a bit of fiddling around to get the effects I want.

    And then I am done! I flatten my layer, duplicate it, add some smart sharpen and then go to Edit >> Fade to Soft Light at 34%.

    follow daydreaming on twitter / livejournal / tumblr / facebook for instant site updates!


    • #3
      Justin's Tutorials

      Full icon Tutorial #1

      Go from


      First I duplicate the layer and set it to screen. This lightens up the icon alot

      Then I add color balance with these settings

      -30, 0, +23
      -28, 0, +58
      -42, 0, +42

      And I now have this

      Now for a gradient map

      I add this () gradient from Misplaced. I set that layer to screen with an oppacity of 54% and this is my final result


      • #4
        Full wallpaper tutorial

        Im going to show you how too make this wall

        Full sized

        First I choose a picture from Glamorous SMG. I chose a BTVS season one promo picture of Buffy and Giles, placed it on my canvas and useing a soft round brush I erased the egde

        Full Sized

        Im now going to add textures all from Black Lagoon

        I used these three textures

        Texture One || Texture Two || Texture Three

        all on normal and erased using a soft round brush, I now have this

        Full Sized

        Im now going to add gradients to my piece, all of these were from Black Lagoon, I don't remember exactually which ones but they were both on Soft Light @ 100.

        Full Sized

        Im now going to add filters to my piece, I used both smart blur (Radius 3, Threshold 5) and Paint Daubs (Brush Size 2, Shaprness 3) and this is now what I got

        Full Sized

        Time for text. I used simple Georgia text saying 'Buffy And Giles'<BR>
        and arial saying 'I'd be no where without you'

        and my final piece

        Final Wallpaper


        • #5
          We'll be making this wallpaper: Alyson Hannigan

          1: Create your base and fill it with black #000000
          2: Choose your image, i chose this one of Alyson Hannigan
          3: Drag the picture onto the base and put it where you think looks best
          4: Go to basic brushes (They come with pS) and choose a brush with smooth edges. I used a brush with smooth edges and at 200 px.
          5: Gently delete the edges of your picture. here's what i got so far: Preview 1
          6: Go to filter->Blur->Smart blur and use these settings: Radius: 3,0 - Treshold: 15,0
          7: Now take the blur tool at strength 40% and blur all the rough edges. This is what we got so far: Preview 2
          8: Take this TEXTURE (Made by Sayjay @ Radiance) and place it above your images and delete the parts that's covering her face. Preview 3.
          9: Take this TEXTURE (made by me) place it above the other texture and set it to linear burn and delete the parts over her face: Preview 4
          10: Take this TEXTURE (Made by me) place it at the top and put it to soft light 50%. Preview 5
          11: Take this TEXTURE (made by webmistress @ Black Lagoon) and put it to overlay and delete the edges. Preview 6
          12: Take this TEXTURE (Made by Nikki @ Misplaced moments) and put it to screen 50% and delete the parts you don't want Preview 7
          13: Create a new layer on top of the others and press ctrl+shift+alt+e and put that layer to soft light 60%
          14: Add some text and you're done. Final
          Dodge Razor history


          • #6
            Tutorial Thread: graphics from start to finish.

            How to make a pop-art style wallpaper

            It was hinted that some people wanted a tutorial on how to make pop-art styled wallpapers, like Untitled#3 and More Time. Though they weren't exactly made the same way, the process is similar and I'll take you through both of them - at the same time!

            What we'll be making today is this piece however...

            For these you don't necessarily have to have a tablet, but it definitely helps! There is potential - OK, almost required - drawing involved. Though don't let that scare you, it really is quite easy.

            First of, you have to choose which image or images you're going to use. You have to choose an image that you really want to work with because once you have, there's no going back. Unlike when you work with caps or stills, you can't simply change mid-way through, so be sure you have an idea of what you want to express - or at least that you're sure which face you want to be looking at for a few hours!

            For this piece I chose to work on this picture of Kristen Stewart and Robert Pattinson from the set of New Moon. First I open up a new canvas (1440x900px) then I copy the picture onto the canvas - it becomes Layer 1. I then duplicate Layer 1 and re-size it (Layer 1 copy) so that it's just over 26% larger and tilt it a bit. Like this.

            If the image is a bit blurry, don't worry. As long as you yourself can make out the gist, it's OK. Remember: you won't actually use that (pixlated) thing in the end!

            What I do now is put the opacity of Layer 1 copy to about 65%. This way you'll see where lines needs to be drawn and where you've already made them. This is where the fun begins! There are several ways to make lines of course, the most painstaking is to just do it all by hand. But, if you're like me and a bit shaky at times - or if you don't have a tablet - paths is by far the easiest.

            What you do now is start to trace all the lines you wish you have. It's important to remember that this is not like when you use this tool for cutting out images. You don't necessarily have to close the sub-path, in fact, it's almost beneficial if you don't. Now, another thing to have in mind is that the beginning and end of each "stroke" will be thin whilst the middle will be fatter. Lay your strokes so that that works to your advantage. (Of course there will be opportunity to correct this later on.) To end a sub-path without closing you press Ctrl and click somewhere not on the path/anchor point. To edit an earlier path you click on it whilst holding in Ctrl.

            I start by tracing Bella, like this. If you notice that one of your earlier anchor points are off simply hold in Ctrl and click that anchor point, now you can re-position it, change it's angle or pretty much whatever you want. If you want to delete the anchor pont, click it without holding Ctrl and if you want to add one, click the path (where there isn't an anchor point!). Here I fixed the bottom of Bella's shirt:
            Before, after.

            I start with a sort of basic path for the bodies. I leave out the faces and most of the hair, mostly because I want to see how this first batch of path/s work out, but also because I hate having too much in the same layer.

            Now, pick the brush tool, go to "Calligraphic Brushes" and pick the first one (that's 7px), then change the settings like this. The brush size is a matter of taste and size of the image of course, but the shape dynamics are very important! See here.

            Create a new layer (Layer 2), then go back to the pen tool, right-click and choose "Stroke path". Make sure you click the "simulate pressure" before you click OK! Now you can delete the path and see how it looks... For this image, I think it looks OK, everything except Bella's hair (the reason why you should wait!), so I'll mask that away before I continue...

            Now comes face and hair. You do the exact same thing, only now you really have to be careful. Faces especially are hard to capture if a line is a bit out of place. This picture really isn't a good example for how faces look though...

            So I begin tracing the faces. This really is all improvisation though, as with all art nothing is set in stone. When I feel like I'm more or less done I create a new layer (Layer 3) and stroke this path too with the same brush settings.

            So how to go about the hair? Well, it's up to you. I try to follow some of the main directions of the hair, to mark out certain locks, the texture of it and to make it look less flat. Here especially, you have to think about how the simulated strokes are going to look! Now that I'm done with that, it looks something like this.

            Now that I'm going to start with colouring I'm going to need to keep track of all the layers more easily, so I make two groups called "outline" and "colour" like so.

            I start with the faces, mostly because they're the most fun! I pick a light beige colour (#e8d7bd), but that really doesn't matter because I'll most likely end up changing it anyway. That is the most important reason as to why all colours should be in different layers - because otherwise changing the colours afterwards will be difficult if not impossible.

            It is also now that the whole mouse vs tablet comes in. Because if you have a tablet it's very easy to use the brush tool and simply draw all the colour-fields. If you have a mouse however, I would recommend that you use the pen tool.

            Quick pen-tool instruction:

            So, instead of drawing up vector paths, you chose to make vector shapes.What you do then is simply to start marking out his face, like so.

            Then you do that instead of using the brush to colour for the rest of this tutorial

            The Brush
            Now, for this piece I'll draw with a brush - I feel it lends me more freedom - but the principle is the same for both, the way you go about it is really only a small difference.

            Now I choose a round brush, drag the spacing down to 1% (as little as it goes, otherwise the edges will looks odd) and fix the shape dynamics like this. The size of the brush doesn't really matter, it's all to your preference.

            I start by filling his face... This is a bit like paining by numbers - but not quite! Since this is the first layer, going over the edges isn't bad per say, it's actually kind of good. Because other layers will come on top. Furthermore, if you constantly keep to the edges there's risk of gaps between the layers - which actually is bad.

            Now my piece looks like this. After this I could go either way: continue on his face, continue with his body or start on Kristen. Since her face is so...interlinked...with his, I start on her. I choose a bit more flushed colour for her, for now. Since her layer is above his, the edge between her and him actually is important now.

            I continue to map out the major areas of the piece. When choosing colours I tend to go for a "middle-colour". Ergo, not the lightest, nor the darkest in the pallet of that area. One problem that arises with working with a brush is that sometimes you miss a spot - so be very careful of that! I end up with this:

            Now that all the major fields are done I can go two ways: either the more typical Lichtenstein style, or a softer more vectoresque style. I want to go with the vectoresque, but for your benefit, I'll briefly head into the Lichtenstein style. Before I start: I'm a big fan of order - especially when it comes to these kinds of pieces. Therefore, I have ordered my layers into groups of what they contain, like this. Just a tip

            Now, for this extreme pop-arty style, you're going to need certain patterns. They're not hard to make, but if you're lazy, just steal mine!

            If we shall begin with Edward's face as an example then? Take the layer that is his face and lock all the pixels like this (If you've used vectors, this isn't necessary). Then go Edit>Fill>Color and chose an extreme red (like #ff0000). Then you select his face by holding in Ctrl and clicking on that layer. Then you create a new Pattern-layer (since you've made a selection it should only show up where his face is. You then choose a fitting dotted pattern (I chose Nr 3 here, and put that to scale: 50%). Once that's done, you put that layer to screen which allows the red to be seen through the black dots!

            Then you do that for the rest of all skin areas. Now the thing about pop-art is that one just cannot be afraid of strong colours! Since this part of the tutorial is mostly me joking around, I just changed the colours of all layers to something more extreme and ended up with this.

            Now, it still isn't quite pop art the way it looks best. The larger black shadows are still amiss. Now, since I'll most likely end up using these kinds of black areas in the final version as well, I'll actually put a little time into them

            The thing to think about here is what you want to shadow. There are two primary reasons to shadow (in this kind of piece); to make it look less flat and to pull the focus to certain areas. So, to do this I'll use the same method as I did colouring - only now I'll use black and put the layers in my "outline" group.

            A positive thing about using a tablet while doing this is that it's much easier to get varied shadows - especially in the hair like this. It's possible to do with a vector of course (I did it before I even got a tablet), only takes a little more time. So for this piece I mainly focused on their hair, but added a few shadows to their faces and bodies as well. Now the outline looks like this. Putting the colours on again I now have this.

            Now, I'll get into more detail about how to go about the kind of piece I actually want to end up with! Back to the faces!

            Let's start with Rob shall we? What I do first is pick a colour that is just a bit darker than the skin tone I already have and start mapping out the shadows on his face. Now it looks something like this. I then pick a lighter colour and map out the highlights. The more different shades the smoother the piece will look, but since this style is of the un-smooth kind I keep to max three shades per "field". Don't be too picky about the colours at this point, they'll be changed later. The important thing is to get the basics there. And to have fun of course, this is easily the part of the process that is most "artistic". The only colour-filed I don't add two shades to is Pattinson's pants, since they're so dark. When I'm done with the basic mapping I have this:

            Now, I'm seriously hating some of the colours, especially the skin colours and his hair (actually, I hate his hair period). I play around a bit with the colours on his hair until I land on something that I think keeps the reddish tint he's supposed to have, but still looks OK. At any-time you want to change the colours of a layer like this you have to remember to lock the pixels - otherwise the entire canvas will be filled! Also, I strongly recommend fill since if you brush over there's a large risk of missing bits. Again, this problem you don't have with vectors - the colour-changing is just a tad easier then!

            Finding a good skin tone is even harder! It's can't be too red nor too yellow - contrary to popular belief our skin tone is actually much redder than we think, although we're far from pink!

            When I've changed the colours so that they're OK (for now) I start to think about the background. It's not always necessary to have one, as seen in Untitled#3, but it does add a little something. Since the picture of them is from the scene in Volterra, I thought I'd maybe add red flags in the background, as well as some stone steps next to Kristen. I start by using these two flags and position them again and again to make several flags. I then use the same method - albeit less detailed - as with the main image to make the flags.

            I then use a different on set image for the steps, as seen here. I also add some tilted stripes to the right, just to escape the boredom that is grey. This is the background I end up with - which will have to do.

            Because I think the outline is a bit weak on the main image, I simply duplicate the outline group.

            To make the thought bubble, I start by putting some round rings using the circle shape tool. Then for the bubble itself make my own path - again, the same way I've been doing the black lines for the entire thing. My bubble looks like this. To fill it with white, I take the wand select tool, then expand that selection with 1px (Select>Modify>Expand). Then I fill a new layer with white. I use the font Pupcat to write my text.

            The fun thing about these kinds of pieces are the many things you can do with the text. In this piece I had a more simple approach, but comic books are text heaven and there's tons of inspiration to find there!

            Now, despite having picked all the colours myself - I don't like them. So I add a curves layer. that end up looking something along these lines... Then I add this gradient map from Salt'n Burn and put it on Soft Light, 52% opacity.

            At this point I'm so sick of the stone-steps to the left that I take them away and add some stripes there too, instead.

            I'm starting to feel quite done with this now, just a few minor things... I make a new layer then I stamp (not merge!) all visible layers onto that. This layer I then down-size to 90% of it's original size and fill a new layer with black behind it: thus giving the piece a border without cutting away from it

            Add my signature, and I'm done!

            1440, 1280

            I hope this tutorial was somewhat coherent, or better yet, helpful! If you have any questions, just ask!

            Images - Everglow
            Stocks - Wikipedia
            Gradient - Salt'n Burn


            • #7
              How to make a comic-book style wallpaper

              I wrote this tutorial somewhere around 2006-2007... However, people did seem to like it so I'm adding it here as well. It goes through the process of making pieces like my Domestic Bliss, The Class Room Profession: parts 1, 2 & 3, The Best Part and Amateurs.

              We'll be making this:

              The resources I will be using are from Oxoniensis Art, Pretty as a Picture and Belladonna.

              First I open up a canvas with a white background. Then I add about 12 screencaps (all from Pretty as a Picture). Now, what I do is I roughly re-size and crop them. However, and this is important, I don't really crop them, I use the rectangle tool to make a vector mask which I will later on change to fit the structure better. For now, I just want to get an overlook of how the piece is going to work, if the piece is going to work!

              Note: I suggest you really think of what is supposed to be said, by whom, in each frame. Make sure you've got enough room for all the text.

              Now you neaten the images up, making the rows etc. fit together. Preferably by using the rectangle tool and vector masks (you can use regular layer masks but vector masks will make things easier further ahead).

              Now, hide the background layer. Then make a new layer (Ctrl+Shift+Alt+N), then press Ctrl+Shift+Alt+E to stamp the visible layer into the new one. (Now you can un-hide the background layer.) Duplicate (Ctrl+J) this layer and set it to screen. Which percentage you have to use on this layer all depends on your caps, but about 70-80% is usually good.

              Duplicate both the original and the screened layer, drag the two duplicates to the top and merge them (Ctrl+E), like this. Then duplicate this layer. With the now duplicated layer (I know, it's messy) you go Filter>Noise>Reduce Noise. I don't know what I would do without this filter! My settings are:
              Strength: 10
              Preserve Details: 8
              Reduce Color Noise: 12
              Sharpen Details: 51
              We can name this layer "Noise-less".

              Next step is to sharpen it up a bit. Duplicate Noise-less and go Filter>Other>High Pass and set it to 0,5 px. Set this layer to Overlay 100%. Not a great difference, but a difference none the less.

              Now, Rory's eyes in the picture in the middle is bothering me, they look a bit funny. Therefore I use the Burn tool and the Sharpen tool to make the eyes a bit darker. This is a good time for small touch-ups in the images, if you feel that any needs to be done.

              Now, go and hide the background layer again and repeat the Ctrl+Shift+Alt+N and Ctrl+Shift+Alt+E. (Then un-hide the background layer). We're naming this layer Base, because it's the base for the effects. Now, duplicate the Base and select one "box" (you can do this either with the Marquee tool or by holding in Ctrl and clicking on the vector mask belonging to the image) then go Filter>Artistic>Cutout. The settings differ from every image, it's therefore important that you do each separately, like this. The settings are usually between:
              Number of Levels: 7-8
              Edge Simplicity: 1-3
              Edge fidelity: 2-3

              Now add a layer mask (the little square button below the layers with the circle in it) and mask out "disturbing" parts (faces and delicate details etc.) and set the layer's opacity to 80%.

              The next step is to yet again duplicate the Base, drag it on top just for now, and go Filter>Artistic>Poster Edges. I usually set my settings somewhat like this:
              Edge Thickness: 2
              Edge Intensity: 1
              Posterazation: 6
              Now drag this layer under the cutout layer and mask out the things you don't want (black spots in faces and the like).

              Here comes the most wierd one, hope you can follow me (and you should know, this differs from every piece, all depending on the images, so you may have to experiment a bit). Duplicate the Base, drag the duplicate on top and go Filter>Sketch>Photocopy. The settings I used:
              Detail: 5
              Darkness: 3
              When you're done, take the same layer and go Image>Adjustments>Levels, settings:
              Input Levels: 195; 1,72; 214
              Don't think we're done... Now go Filter>Blur>Gaussian Blur and set it to 0,5px. Set this layer to Multiply, about 70%. Again, mask out the things you don't want.

              Again, this is touch-up. Go through your filtered layers, mask out or smoothe an edge. Just, try to be content with the image you have before you, because now we only have text and colour left.

              Select all of your frames/boxes, whatever you want to call them, like this. If you used vector masks, select one mask using the Ctrl method I said earlier, when you want to select the second mask press Ctrl+Shift. Now create a new layer, then go Edit>Stroke. I used the settings:
              5 px; #FFFFFF (white); Inside.
              While still selected, create another layer and go Edit>Stroke. Now I used the settings:
              2 px; #000000 (black); Inside.

              It has now come to the speech bubbles. These can be fun, or they can be a pain. If you had planned when you picked the caps and re-sized them, the text won't be such a problem. To make a speech bubble you pick the Custom Shape Tool and select the shape of a speech bubble. Make sure you have the square with the four corner-squares, but without the pencil, selected because this creates a shape layer using a vector mask. There's not much more to say about the bubbles. However, I do have some tips if you want them. Firstly, you want to have readable text, i.e. swirly text don't work. I usually use Comic Sans or Arial. Secondly, try to make it feel more like a comic book by using exclamation marks, bold/italic text and the like. Moreover, I love thought bubbles, but it can be hard to think of any good/funny ones. Finally, there are *no* rules in comic books! Experiment! Now, I also add my background. Make a new layer just below Base and fill it with a colour, pattern or texture, whatever you want.

              Here comes the gradients! Basically, with comics I try to be careful when I use gradients. I don't exaggerate and I try to keep things smooth. The gradients I used are from Belladonna and my own.

              This one is from Belladonna, I put this on Soft Light 80% opacity, just to lighten the piece a bit.

              This one is my own. I put this on Soft Light 70%. Mostly, I picked this one because I liked the colours.

              I made this quickly out of two colours and put it on Soft Light 60%. This was for getting some contrast and bringing out the black.

              Mine again, Soft Light 40%. Again, to bring out the black and this time white as well.

              Next, I add this paper stock from Oxoniensis. I use this to get a bit of texture, and put it on Multiply 30% opacity.

              Wow! Almost done. For the final touch and an even more comic book feel, add a big flashy head text, name the comic. For this piece, I used the font Loverboy. First I wrote the title ("First Meeting"), then I duplicated it, changed the colour of the text and nudged it a bit to the bottom right. This creates a shadow-y feel and a 3D feeling. And if you're dying to find out, the Gilmore Girls text is called "Poor Richard".

              Despite the ancientness of it all, I hope it helps


              • #8
                Beginner tutorial

                Following discussion that we needed some new tutorials to help beginners I have come up with the following tutorial based on a piece I did ages ago. It's a fairly simple piece with a few images and a few uncomplicated techniques that are often used in fanart. Hope you find it easy to follow:-

                This tutorial will show you how to resize and move images around the canvas, flip them, use layer masks, layer blend modes and gradient maps. Finally you will see how to add text.

                This is what we'll be making:-

                The images I used were from Free Images but I can't find them now as I've had them for a really long time but you can find them here (click for larger image):-
                Cloud, Steeple and Wing image

                The other two images were from Jenni Lou's wonderful Daydreaming site:-

                Anna image
                Dean images

                Tutorial for "Forgive"
                Open a new PS canvas (CTRL + N) in the size you want it. I used width 1600 pixels, height 900 pixels as that’s the size of my laptop screen, but you can make it any size you want. For the tutorial I would suggest doing the same size if you want to replicate it exactly though. I had my background colour set to black and so in “Background contents” I chose Background colour from the drop down menu. I then had my new canvas open with a full black background.

                I open the cloud image and pressed CTRL + A to select the whole image then Ctrl + C to copy it. I then went back to my “Untitled document” and ensuring the background layer was selected I pressed Ctrl + V to paste the image into my working doc. The cloud image is much larger than my working document and so I needed to resize it to a workable size.

                If you are comfortable with how to re-size an image you can ignore the steps under the spoiler tags

                In order to resize I ensured the layer was selected by clicking on Layer 2 in the layers palette, then clicked the Move tool:-

                A dotted line appears around the cloud image with little square “handles” at each corner and in the centre of the edges. In order to re-size you drag these corner handles toward the centre of the image. Beware though it is easy to pull the image out of shape by doing this. If you drag it downwards instead of diagonally you will make the image smaller in height but keep the width the same – change the aspect ratio of the cloud image. We don’t want to change the aspect ratio, we just want to make it smaller so thankfully PS has a handy tool that keeps it for us when we drag.

                Once you have your dotted line, click on any of the handles and you will see that the dotted line becomes a solid line and the options bar along the top changes. In the options bar there is a chain link symbol between the width and height % boxes. If you click that you will maintain the aspect ratio whilst dragging.

                You can also re-size by changing the % in these boxes, useful if you’re taking a large image like this down to much smaller.

                I resized my image to about 50% and it now looks like this:-

                I wanted the light to come from the opposite direction and so I flipped the image by using Edit>Transform>Flip Horizontal, ensuring the cloud layer was selected.

                Ok now I want to soften the edges. I never have hard edges in my work, although that can be an artistic element, I find it difficult to pull off so always soften away the edges. There are many ways to do this, but the way I favour is to use a layer mask.

                If you are familiar with layer masks and how they work you can skip the part under the spoiler tag

                Layer masks are a way of erasing parts of an image, but not permanently. If you make a mistake with a layer mask you can put it back the way it was easily. I like them because I create my art in an organic unplanned way and that leaves me room to change my mind as the piece progresses. I also change my mind a lot and make lots of mistakes so I couldn’t work without them!

                The way they work is, you paint black over the area you want to hide, but and here is the wonderful thing, if you want the image shown again you just need to paint over the black areas with white to bring it back to how it was.

                In order to use a layer mask you have to add the mask to the image layer by clicking the icon at the bottom of the layers palette. It’s the third one in and looks like a square with a circle in the middle:-

                Once it’s clicked a white rectangle will appear in the layer:-

                This white rectangle needs to be selected when you paint the black on or you are just painting onto the image. In the screenshot above you can see a border around the white triangle and that indicates that the mask is selected. Once that’s selected you simply select a brush, make sure the colour is black and paint away onto the image. If you want the effect to be paler (i.e. not as much erased) you can paint it with a lower opacity brush.

                The best way to learn with masks is to use them, because you can always return to the original image. I suggest that once you’ve added the layer mask then paint black somewhere on the image that you want to keep, like the blue areas of the sky and you’ll see it all disappear and the black of the background come through. If you then change your colour to white and paint over the same section in white it will restore the image as good as new. Then change the opacity of the brush to 50% and with black colour selected again do the same to see the different effect

                I masked away all the hard edges so that you can’t see where the cloud image ends and the background begins.

                Now take the steeple image and paste it into the document. Resize it to roughly 70% and set the layer blend mode to “Soft Light”. Move the image so that it looks like this:-

                You may need to mask away the edges a little if you can see hard edges around the steeple image.

                If you know all about layer blend modes you can ignore the next bit under spoiler tags

                So let’s talk a moment about layer blend modes. The different blend modes will cause the interaction between the top layer (steeple) and the one below it (cloud) to change. This blend mode “Soft light” is kind of see-through – you can see both images quite clearly but without affecting the brightness of the 2 images. Other layer blends affect the brightness, deepen the colours have stronger effects etc. The effects will also differ depending on the light and shadows, colours and composition of the images concerned. I’m sure there are better technical explanations for the way these work but I don’t know them.

                Take a moment to change the layer blend modes to all the ones available in the drop down box so that you can see the different effects. Now because I create organically, I tend to try lots of different blend modes before I settle on the one I like best and sometimes it changes as I go through. For this image I liked both Overlay and Soft light, but thought that Overlay was too strong. As you do your art you may find certain blend modes work better for you. I tend to use Soft Light, Multiply and Lighten a lot and others very rarely but it is a personal thing.

                OK so next I pasted in the image of Anna above the Steeple image. I resized it down to about 90% and moved her to just to the left of the steeple at the top of the piece. I masked away the background leaving just Anna behind. I wasn’t very accurate in removing the background so there are parts of Anna that you can’t see very well, like the edge of her hair and jacket but for this piece it doesn’t need to be precise. I also masked away the bottom half of her body so just chest up was showing.

                I set Anna’s layer to Screen. This is how it looks now:-

                OK now I want Anna to have wings, because she’s an angel.

                I pasted the wings image under Anna’s layer, resized it to about 50% and tilted the wing tips upwards so they looked more natural and moved it into place to the left of her shoulder. I then masked away most of the pale area around the wings and set it to soft light. To make the other wing I duplicated the wing layer by pressing Ctrl + J - that copies everything that I did to that layer as well. Once duplicated, I flipped the new wing layer horizontally and moved it into place by Anna’s right shoulder. Now we have this:-

                Now the problem I have is that Anna isn’t standing out as much as I’d like. I don’t want to lose the translucency of the clouds popping through, but I want to see her clearly. I duplicated the Anna layer (CTRL + J) and set this to Soft Light, which deepens the colour and brings some more detail into her jacket. She’s more obvious now.

                So I now add the final image of Dean looking up. I pasted him over all the layers, resized to about 70% and positioned him below and slightly to the right of Anna, basically so he’s looking at her. I masked away the edges and set the blend mode to Screen.

                I wasn’t very happy with the positioning of the steeple now, Dean basically covered most of it and so it was unclear what it was. I moved the steeple layer around a bit until I found a better composition. It ended up only slightly further right than where it was to the right of Dean’s head, but I changed the layer mode to Overlay to bring the silhouette effect out a bit more. This is what I end up with now that all the images are in place and I’m happy with the composition:-

                So now I’m going to look at the colouring. For me colouring is always a matter of trial and error. I rarely have a colour scheme in mind when I start, although I might know whether I want warm or cold colours.

                I use colours to highlight areas, to improve the balance of colours between the images, accentuate emotions and to create a more attractive image. The images in this piece are pretty balanced already. The background stock images are very saturated and the two main person images and fairly similar in brightness and contrast, however Dean’s face is a little more red than Anna’s So colour correction is not much of an issue, but by using a gradient map I can make this piece look more cohesive.

                Explanation on how to use Gradient Maps in the spoiler tags

                To add a gradient map to the image click the icon at the bottom of the layers palette that is a circle divided with one half white and one half black. This will give you a pop-up menu:-

                Choose Gradient Map and a new layer will appear which has the divided circle and a layer mask. You will immediately notice a change to your piece. If you colour selector is still black and white from using the masks earlier your piece will probably now be in black and white.

                Double click on the divided circle on the gradient layer and a new tool box will appear:-

                (How it actually looks will depend on the version of Photoshop you have but it will be something like this)

                Click on the drop down arrow to show the different gradients available. I can’t remember how many gradients come pre-installed with Photoshop but it’s not many. You can create your own gradients, but I would suggest downloading some more. I have lots, some from Daydreaming and Salt n Burn, but I’ve had them for years so I don’t know if they’re still available.

                For now though just click on the pre-installed red and green gradient. The piece will change colours too. This solid colouring is rarely suitable for what I do, so I change the blend mode to Soft Light. Now it will look like this:-

                You can see what a difference that makes. Each gradient will make the colouring look different so spend some time playing about with the different gradients and how they look with different blend modes. In addition to the different gradients and layer blend mode combinations, you can also reverse the gradient for different effects. There are many options even with a limited amount of gradients.

                In the end with this piece I used two different gradient map layers because I wanted a depth of colour that I couldn’t get with just one. I like the way the lighter region made Anna look a little ethereal and how that light reflected down on Dean’s face. With the other combinations, that contrast was not so evident but really here it’s a personal thing:-

                Now for text. I will happily admit I’m not good at it. I’m rubbish at font choice and its placement. Where you put your text should probably be incorporated into your thought process whilst you’re placing your images, and I rarely do that so nearly always my text placed where there’s room. Don’t do that, it will make your life easier if you think about text earlier than I do.

                I generally do the colouring before I add text so that I can blend the text in with the image and not have it stand out like a sore thumb. To add your text click on the text icon in the tool bar (the icon is a large T)

                How to add text under spoiler tags

                Once you click on the Text icon click somewhere on the piece and a flashing cursor will show where to type. Once you’ve type your text, click the tick in the options bar at the top to commit the text.

                From here you can change the font, size, the character spaces, add bold or italics, colours etc from the text palette. Font choice, size and colours are a personal thing so play around until you find something you like.

                I chose two simple fonts, picked colours from the main body of the piece, one light one dark and put one line in capitals. This is all to break the text up a bit, make it more interesting and stop it being to blocky. I also changed the blend modes. The top line of text was set to “Difference” and the bottom line of text was set to “Colour Burn”. I placed both text layers below the gradient layers.

                Before I finish, I look over the piece in closer detail to see if there are any edges I’ve missed, any places on the piece that look weird or dodgy. For this piece I notice that you could see a bit of tree from the steeple image around Dean’s nose which looked odd so I masked it away. That made Dean’s face lighter but I was ok with that. I also masked away some edges around the wings that were still knocking around. This step is optional, and I’m a bit of a perfectionist so by all means ignore it if you’re not as up tight as me!

                Here is the final piece:-

                Here is a link to the full size piece and to a copy of my PSD file if you want to see how I created it compared to your own version:-


                Hope you enjoyed the tutorial and it was all clear. If you have any questions, PM me and I’ll be happy to answer.

                Peter Capaldi is the 12th Doctor