good_tutorial

General text guide

2013-03-19T13:53:57Z

Program(s)+version: Photoshop CS4
Involves: No tools whatsoever - it's all about looking at text and composition
Translatable: Yes
Steps: Zero
Difficulty: Basic

Janie's Guide to Text - Part One

Hello everyone, welcome to my ~~guide to~~ monologue on text. Part One.

In my Guide to Negative Space, I mention that negative space is my favourite thing about icon making. But I'm happiest when my negative space lends itself to text use.

In this guide I will be talking about how I use text. I will talk about composition, making text fit with its background and, of course, my favourite fonts. The first part of this guide has to do with the ways of looking at text. How can we use it? In a later part, I will also spend time on how to make what you have in your head happen and share my favourite fonts with you. If this is something you are interested in, come on in!

"What do you want from me?!" the text yells
We often hear that a picture is worth a thousand words. And I suppose we wouldn't love icon-making as much as we do if we didn't believe that was at least partly true. But using text on graphics is a way to enhance the graphic. Text itself is a graphic form, just look at the countless examples of breathtaking word art.

Text on icons takes many different forms. Sometimes it works and sometimes it very much doesn't. In those last cases, you're probably pulling out your hair, screaming "WHYYY??!!!" at yourself or your graphics programme. Or, if you're not as dramatic as I am, you're just annoyed at the text not coming together the way you had in mind.

Needless to say, I've been annoyed. A lot. I still frequently agonise over the tilt of a letter or the spacing of a barely legible word. What helps is thinking about the different kinds of text on icons. Or rather, about the aim of the text and the placing of it.

If this sounds terribly odd or highly pretentious, please consider the next few categories. They are by no means definitive or exclusive, but it helps me compartmentalise the ways I use text.

First it might help to decide if you want your text to be the main feature of your icon or if you want to use it in a different way. Compare it to the Oscar categories for Best actress in a leading role and Best actress in a supporting role. You might want to use your text to support something in your icon, but it can still end up stealing the show. Please keep that in mind when reading the rest of this.

Fifty-fifty

firstillusion, queensjoy

The easiest category to define is the one where the balance of text and image is pretty much fifty-fifty. We all know them: wonderful images we know just how to crop, but they're not going to fill up our 100x100 pixel canvas. One way of giving those cut-off images extra impact is adding text.

Granted, this way of using text often looks quite choppy. Which is understandable, because image and text are fighting for dominance and the viewer doesn't immediately know what they are supposed to look at first.

Text as the star of your icon

Text that fills up most of your space

firstillusion, _puchula_, _puchula_, fuuurs, appleindecay
erzsebet, surrexi, krazykate2, sarisafari, sky_magenta

This category is deceptive, because at first you think the text is filling up most of the space and therefore 'disrupts' the image. What it often does, however, is draw your eye to a detail of the image. The text is definitely the star of the icon, but there is a reason the maker didn't make it a text-only icon.

Now we move closer to the area where star and supporting actress are starting to come together. It might no longer be glaringly obvious who gets paid the big bucks this movie, but what is clear is that the leading lady would not be able to shine without that one supporting actor highlighting all her awesomeness. Okay, lame, but some icons clearly benefit from text.

I have two categories that have to do with sight-lines/looking lines. I'm not sure if this is a proper English word, but I'm using it anyway. I know I tend to ramble about how the best icons are the ones that make the eye travel across it. Sometimes this is a rough diagonal, horizontal or vertical line, but sometimes it has a zigzag or another shape.

Well, text can be used to enhance those looking lines. Or it can be used to deliberately cut through the image's looking line and become a new focal point where the eye rests (for a moment). Because this might sound confusing, let me give you some examples:

Text enhancing the looking line

In the first icon, the text creates a diagonal from Amy and Rory and the car in the bottom left corner up to the text in the top right corner. Without the text, the eye would've been stuck on Amy and Rory because the sky has no other texture. Without text, it would not necessarily be a bad icon, but the text definitely makes it a far more interesting one.

The same goes for the second icon. Amelia is standing at the bottom of the stairs, which we can see moving to the top of the image diagonally. By putting text to the left of Amelia in a light colour (the same colour as the bannister) the looking line is elongated across the icon. Without the text, the bottom left corner would've been dark and the looking line would've stopped at Amelia.

Some more examples from people who show it much better than I do:

00mina, appleindecay, naginis, fuuurs
naginis, raiindust, juanxyo, fuuurs

See how all the looking lines except for the second one are diagonal? If the maker of the fourth icon had placed the text higher and on one horizontal line, the curve of Cher's body wouldn't have been followed. By placing the text alone that curve, the text follows the image. Conversely, the line on the image of the sixth icon runs from the top right corner to the left bottom corner. But by adding the text, raiindust has created a looking line for us starting at Ron's face and going down the icon to the text. Something similar goes for the first icon. Without the text, we would've seen Amy's pale face on the right side and then a whole lot of hair. Which is extremely pretty, but not as interesting as it is now.

Text to create a 'zigzag' looking line
It can also be the other way around. Maybe you don't want one straight line for the eye to travel across, maybe you want several points of focus. Text can help create that curve:

jam_min, absolutelybatty, firstillusion, lessrest
rabidline, solar_contrast, feel_the_fire, quiddity_

In these icons, the text provides either a new focal point out of the looking line, creating some sort of zig-zag (icons 2-4) or it forms a resting point for the eye, a sort of anchor. With the three standing figures, the sixth icon seems quite vertical, but the text has been placed on a horizontal line, creating a whole new image.

You might think I am talking out of my shapely behind or that I'm thinking waaaaaay too much about lines that aren't even there, but this is just the way I look at text icons.

Because the lines between 'star' and 'supporting role' are so blurry, I'll move away from the looking lines and give you a few more categories to consider:

The OTP text
Text can also stand out like above, but be more integrated in the icon:

firstillusion, firstillusion, firstillusion, absolutelybatty, blue_emotion
blue_emotion, calrissian, fuuurs firstillusion, firstillusion

Text in a supporting role

The text is visible, but it's not immediately your point of focus

firstillusion, firstillusion, firstillusion, firstillusion
firstillusion, erzsebet, firstillusion, appleindecay

Text is practically part of the image itself/becomes a graphic element on its own:

fuuurs, pamkips, calrissian, pamkips
sesperia, appleindecay, pamkips, pamkips

And text as a (purely) decorative element

systematichabit, firstillusion, pamkips, firstillusion, fuuurs
firstillusion, pamkips, fuuurs, pamkips, fuuurs

The text in these last icons isn't necessarily trying to convey words or even a message to us. It's clearly different from the previous category, because although the letters can still make up a rather important part of the icon, we are hardly ever supposed to be reading something into it. Tiny text, like in icons 1, 2 and 4, can be written by the icon maker to match the subject or the atmosphere of the icon, but it's not longer large enough for the words they chose to have impact on the viewer. For some makers, this way of using text helps them with their horror vacui (fear of the empty). It can definitely be an interesting way to spice up your icon.

Summing it up
Generally, I ask myself two questions about the text I want to use on an icon:

1) Do I want my text to steal the show?
2) Do I want my text to enhance the looking line or create a new shape?

Based on those two main questions, I can fill in all the details that are specific to the image.

Comparing these different ways text can be used has helped me use it more effectively. I also spend less time on it, because I generally know which effect I'm going for. That doesn't mean text is always planned or ends up the way you wanted it. These categories aren't anything except my own method of understanding text. Who is to say they make sense to anyone else?

And to illustrate that point, here are some examples of text that defy most of what I just shared with you, but where the text is so very special:

val_valerie, 00mina, appleindecay, renneroo_icons
previously princeton_girl93, delorentoes, erzsebet, sarisafari

Questions? Concerns? Thoughts? I'd love to hear them! :)

Basic guide to negative space

2013-02-10T10:39:24Z

Program(s)+version: Photoshop CS4 (for Mac)
Involves: scale tool, smudge tool, clone tool
Translatable: it's mostly a guide that deals with composition, so yes
Steps: -
Difficulty: I'm not explaining the basic basics of how to crop an image etc., so Medium to be safe. :)

This is a reworked version of a guide I posted to illusoir_icons after I had gotten questions about how I create my negative space icons. In this post, I talk about choosing your image, my take on the composition of negative space icons, and four different ways to create negative space when there is none. I use lots of example icons, so maybe there's a bit of inspiration as well?

We generally use the term 'negative space' if a considerable part of the icon isn't taken up by the subject of your icon. That means that most close crops won't be thought of as negative space, but there is no exact definition of what negative space icons are.

For example, it's quite obvious that these five icons qualify as negative space icons:

appleindecay, marylou_gr, thewonderlife, blue_emotion, regis

But what about these five, are they negative space icons?

lemonpunch, longerthanwedo, lessrest, llenka, tinebrella

I personally don't think it really matters what you call them, but I put these icons here to remind ourselves of the fact that there are so many ways to use the space in an icon. Composition is something most people don't actively think about when they start making icons, but if you take the time to consider it, it can make a huge difference to the way you icon and the impact your icons have on your viewer.

Okay, back to what we came here for: the basics.

Choosing the image
Theoretically, almost any image can be turned into a negative space icon. There are icon makers that are experts at cutting subjects out of busy backgrounds and putting them into new canvases with lots of negative space.

I am not one of those people. (But if you want to become one of those people, check out this incredible layer mask guide by lil_art or this useful tutorial by liquidlights.)

The kind of images I like to use for negative space icons are particularly images with clean-ish backgrounds and outside shots. Like these:

- -

- -

- -

- -

- -

- -

- -

Composition
Because the majority of shots in television shows and movies are medium shots (showing subjects from the waist up), the composition of a negative space icon will usually place your subject at the bottom of the icon (see above examples). I personally like to place the subject of the icon not dead centre, but more to the side. I have a couple of reasons for this:

→ I like using text sometimes and when I do, I like to create a diagonal line for the eye to travel across the icon:

→ I am very fond of repetition; putting the image in the same icon more than once. Again, I like creating 'looking lines' in my icons:

Any kind of crop is, of course, possible, but this one (just off-centre) is my favourite, because it gives me a lot of room to play with textures.

What if there's not enough image for the composition I want?
Well, then we get to the technical part of extending our background.

There are several ways to extend your background. I'm going to list the four methods I use:

1) The scale tool.
This works well if your background is already a solid (or almost solid) colour. What I usually do is prepare my base ( 1) Image > Auto Contrast 2) duplicate my base twice and change the blending mode of one to screen and the other to soft light, then play with the opacities), then copy those layers and merge them so I have all the effects in one layer, because you need a single layer for the scale tool to work properly.

I'm using this image of Doctor Who's Rory as an example, because the background of this image is already a solid white:

What I do once I've copy + merged the layers I had, is make sure this new layer is in a canvas that already has the height I want the new background to have. In this case that's a 300x300 canvas. (I made the background layer blue so you can see it better.) I then select a tiny strip along the top of the image with the standard box selection tool:

Then I go to Edit > Transform > Scale (this may be at another place in a different graphics programme). The selected area now turns into a box-y thing with points you can pull:

And that's how you extend your background quickly and easily. The bad news is that most often, images won't have this kind of solid background, so you'll have blur out any uneven patches or brush over them. Which is some of the other methods of extending your backgrounds tend to come in handier. But if you have such a solid background, go for it!

2) The smudge tool.
The smudge tool (that little pointing finger icon in your Tool palette) is a fabulous device. You can adjust the strength of the smudge and the blending mode, but because I tend to use the smudge tool for the 'hard' work, I hardly ever use those options. I use the smudge tool to create a basis on which I can then work with texture or other ways to make the background look like it actually belongs with the image.

Again, the smudge tool works best if you want to extend a background that is already relatively smooth, or backgrounds like walls that have a distinguishing feature that is difficult to erase. Think the door or window frames or a straight horizon. You can usually smudge lines quite easily, but you have to have steady hand. Some examples where I used the smudge tool to elongate vertical lines.

3) The clone tool.
Your best friend and your worst nightmare.

The clone tool (that little stamp icon in your Tool palette) is a wonderful, wonderful instrument. It will let you copy any part of your image and paint it on any other part of your image. But it requires precision and not all images have enough even patches to make it work. For a more elaborate explanation of how the clone tool works, please read this tutorial I wrote a while ago.

An example: below are two images. One where the clone tool worked well - and one where it required patience, patience and some more patience.

Image 1
> >

You can see where I cloned parts from - and where I cloned them to. There is some repetition visible in the larger image, but most of it isn't very obvious (at least not in a distracting way) in the smaller version.

Image 2
> > >

In this icon, the background was rather busy and the parts I could clone from were very small. So the repetition in the image was very visible. I tried to mask this with brushwork and textures, but it was far from ideal. Still, if you have a busy background or a background with a pattern and you don't want to alter the entire background, the clone tool sometimes is your best option.

For more technical and detailed information about creating backgrounds, read this amazing tutorial by shoqolad. For more tips on how to be creative when creating or adjusting a background, there's this fabulously inspiring tutorial by absolutelybatty.

4) Using textures
Like I said earlier, some people are geniuses when it comes to cutting out/masking subjects and creating whole new backgrounds. My favourite example is lemonpunch:

You can either use textures to replace the entire background, like lemonpunch - or parts of it, which is what I usually do. I tend to work from the background that is already there. This is ideal for when you've smudged your background, but it looks messy. Textures can easily cover this and by using layer masks, you can adjust almost anything. The trick is to find textures that work with your subject, the lighting of the original image and the original colour scheme.

If I work on a 100x100 canvas and I need to fill up space that has to be a certain way (tree tops for example) but which is difficult to do well, one trick that I have is that I cut a piece from the original image (or the image before I resized it) and work that into the image, like here:

The edge of the forest at the top (1) and the sand at the bottom(2).

Other times, I brush over the parts of the background that are distracting with a colour that is close to the dominant background colour, but a few shades lighter, then mask out my subject. Then I use textures to give the background depth:

Textures are so essential to extending backgrounds to me - and people are often surprised at how much work it can be. Negative space icons sometimes have a reputation of being the 'easy way out', but I think that if you only look at the icons + original images I posted at the top of this post, you can see what work can go into creating negative space. You don't have to, of course, but that very precise work can land you a spectacular negative space icon - and I personally would be nowhere without textures to help me create that negative space.

Speaking of textures...
They are, in my opinion, essential to interesting negative space. There are lots of cool, funny icons that have a solid background colour, but if you ask me, the best negative space icons have depth to them - and nothing creates that depth like textures do!

For example:

_omnivore_, ofraindrops animotus, fan_cifully, memonechan, a_thari

What about text?
Text is good. I love text. Especially on negative space icons. But I promised someone/people/appleindecay and ohgollygeedamn a specific text guide, so I'm saving this topic for a later date.

That said, I don't want to go without saying this:

→ Balance is the key to a good icon. Make sure your negative space doesn't become to busy, especially when you are using wild textures. Spend some time on a negative space icon and be critical of your composition: where can you erase/mask certain parts of textures so that the viewers eye doesn't get distracted from its path through the icon?

→ Sharpening is very important when it comes to negative space icons. Often times, your subject will be small and face and clothing lines especially tend to get pixel-y quite quickly. So be careful there. I always copy + merge my layers a few times during my process, so I have a 'back-up' layer if I royally miss up the sharpening and can work from that layer. Although I sharpen and blur at several stages during the making of an icon, if it's a difficult one, I tend to end with a copy + merged layer set to the High Pass filter at soft light - and I smooth pixel-y edges with a soft, small blur brush.

→ Vertical lines can work wonders for your negative space icons:

xeyra, prettyquotable, fancifull, _puchula_, chouchoune,
poshing, sky_magenta, ohfreckle, xeyra, juanxyo

→ + my personal favourite: repeating part of your icon as a line in your icon (not just in negative space icons though):

→ Negative space is a thing to experiment with, especially when it comes to sharpening, "how large does my subject have to be to still be visible and have ~meaning?" and using new textures to bring colours into your negative space. Text too, is easier to experiment with in icons that have rooms to spare, obviously.

→ Negative space is beautiful and I wish there was a club I could belong to. "Hello, my name is Janie and I'm a negative space addict."

So there you have it. I have no idea if that was helpful to anyone, but I suppose that's everything I have to say on the matter. Well, not really, because I could talk about this for hours, so if you want to discuss something negative space-related with me, please just leave a comment!

Questions? Concerns? Thoughts? I'd love to hear them! :)

How to soften things up

2012-09-06T10:42:49Z

Program(s)+version: Photoshop CS5
Involves: Curves, Blur tool, texture use, blobs, Blending Modes, Smudge tool, Filters.
Translatable: I think so.
Steps: ~13
Difficulty: Medium

I know, my icons don't look very soft - I prefer a bold contrast and a vivid coloring, not to mention a sharp quality that is probably the opposite of softness... but nevermind.
I know the theory. Sort of.
I didn't want things to sound too abstract, so I made an icon using all the techniques I know of and use to soften things up.
We're going from: to:

JUST AN AUTO-CURVES LAYER
>
First step is always Auto-curves. I'm becoming so freakingly lazy that if Auto-Curves don't work and the cap isn't just plain gorgeous, I throw away the icon without a second thought.

BEST FRIEND #1: THE BLUR TOOL
>
The Blur tool comes in handy when you're not working with a HQ cap, and especially if you want to make a medium/close crop.
The advantage of using the Blur tool instead of the Gaussian & Surface Blur Filters is that you can blur just the grainy/pixelly areas, leaving the rest of the picture untouched. Of course you can also use Gaussian/Surface Blur and then mask the layer, but it usually takes longer.
However, here is a general advice: use Gaussian/Surface Blur when the whole picture is grainy/pixelly, and use the Blur tool when there are only small/limited areas that need to be softened.
To blur, I usually choose a soft round brush and as for the opacity I'd say between 35% and 65%, depending on the cap and on how smooth I want things to be. On, and it also depends on the area I'm blurring: when working with backgrounds I usually raise the opacity of the Blur tool a lot - even 100%, if needed.
Back to my picture, I created a new layer and picked the Blur tool with the Sample All Layers feature checked. I zoomed in a little and blurred what needed to be blurred:
> >

BEST FRIEND #2: TEXTURES/SCREEN COMBO
>
I picked this texture by gallicka:

And I set it to Screen, then lowered the opacity to 27%.
Setting textures like that to Screen really helps to make things look soft, because they lessen the contrast and they also add some colours to your picture.
General advices:
- don't pick overly bright textures, and if you do, lower the opacity a lot: you want to soften the contrast, not to pimp up the highlights and say goodbye to shadows;
- in choosing the texture, try to figure out which kind of hues you want to add to the icon: don't pick a red texture if you're going for a cold coloring;
- pick a b/w texture if you want to work on the lighting/contrast and leave the colours unthouched;
- pay attention to the highlights/shadows scheme of the texture you pick: setting it to Screen will change the lighting of your icon and you have to be careful if you're working with close crops and the likes.
Btw, you may want to snag the set I picked this texture from. It's just stunning and I'm using it on everything.

BEST FRIEND #3: WHITE BLOBS/NORMAL COMBO
>
This is how this layer looks against a black background:

I used a soft round brush, painted some white spots here and there and set the layer to Normal, 25 % opacity. This is another useful trick to soften the contrast, but only in limited areas. More often than not, we're talking about hair.
I suggest you not to paint light blobs on people's faces, because it changes the lighting in strange ways and make the subject look weird.

BEST FRIEND #4: FAKE SOLID BACKGROUND
>
Sometimes it helps to replace the background. In this case, the background didn't look very soft, so I just painted over it. I used my favourite brush and wasn't very accurate on purpose: I wanted the edges of November to be a tad blurry, because sharp, defined edges don't look very soft obviously. Don't worry about the colour you pick, I usually pick a random colour from my image and once I'm done with the brush work I just go to Image > Adjustments > Hue/Saturation and play with the three sliders (Hue, Saturation, Brightness) till I'm satisfied with it.

BEST FRIEND #5: RANDOM BLOBS/NORMAL COMBO
> >
White isn't the only option available for blobs. I often pick a colour from my image with the Eyedropper tool and choose a lighter shade of the same hue to paint my blobs.
&
For both layers I used a pale beige, but on the first one I used Motion Blur and its effect on Normal was so subtle that I didn't lower the opacity. The second one is set to Normal as well, but at 50% opacity.

BEST FRIEND #6: RANDOM BLOBS/SOFT LIGHT COMBO
> >
Blobs can be useful to brighten/darken specific areas, and also to enhance the colors of some details, for example. I pick the color of the blobs accordingly.
&
I used those two layer to make small changes to her hair, and I set them both to Soft Light - at 100% and 67% opacity, respectively.

RANDOM BLOBS/NORMAL COMBO IS BACK
>
Nothing much to add, just some more painting on her hair to soften the contrast.

(Normal, 49% opacity)

A CURVES LAYER, WHAT IS HE DOING HERE?
>
Yes, Curves layers are male, deal with it. As well as Vibrance is definitely a woman. And Channel Mixer is- Nevermind.
So, Curves. Just because you want your icon to look soft doesn't mean that contrast is an optional. However, softness is about building the coloring and contrast gradually, in my opinion. This is why each step is very subtle and doesn't do much :D

HAVE YOU JUST SHARPENED THE ICON?
> >
Yes, I have. Because in my opinion soft doesn't necessarily mean blurry. I used Paint Daubs, because this is how I've been shapening my icons lately.
However, I also masked this layer a bit:
> >

BEST FRIEND #7: THE SMUDGE TOOL
>
The Smudge tool is very useful to even out the background and to soften the edges, especially.
> >
Any brush will do, I usually pick a soft round brush at 60-100% opacity.

CONTRAST & COLORING, NOTHING INTERESTING
> >
I used anther Auto-Curves layer to check the contrast and then a Color Balance layer to add some warm hues back. This tut isn't about the coloring, so I'm not going to elaborate this step further. c:

FINAL TOUCH
>
Just some more blurring.
> >

BONUS: ANOTHER EXAMPLE
> > > >
First of all I used the Blur tool and the Smudge tool on her hat, because it was very pixelly. Then I used a couple of textures set to Screen to brighten the picture, soften the contrast and enhance the warm hues of the cap. I'm quite sure I pasted this texture by innocent_lexys into my canvas twice and move it around a bit, like this:
&
I may have also smudged/modified the second one a bit, from the look of it.
And finally, I used Auto-Curves to check the contrast.

Tips and Tricks for Black and White Icons

2012-07-22T20:38:48Z

Program(s)+version: Photoshop CS5
Involves: Curves, levels, brightness/contrast, textures.
Translatable: The concepts should be translatable to any program with curves, levels, and/or brightness and contrast adjustment layers.
Steps: It's a guide.
Difficulty: Easy.

This is a quick guide on how I do black and white icons. I talk a little bit about cap selection (the whys and hows of the caps I pick), my general process for black and white icons, what textures I tend towards when I'm doing a b&w icon, and some random tips and tricks/notes. This is by no means a complete guide, and there's definitely not a 'right' way to do B&W icons (this is just how I happen to think about it).

If you have any questions, please let me know! I'd be more than happy to answer them. :)

Cap selection.

When it comes to picking screencaps for b&w icons, I tend towards two different types of screencaps: simple screencaps, or screencaps where I can remove the background so that my subject won't get lost in it.

Simple caps.

B&W icons work best when it is easy to find and focus on the subject of the icon; without colour to pull the subject out from the background, subjects can get lost if the icon is too busy or crowded. Thus, simple caps work best. What constitutes a simple cap?

(Click for full size.)

I'd consider all three of these "simple" caps because the camera has either closely cropped the subject for me, the subject is positioned so that a close crop would be good, or the background of the screencap is relatively plain (and therefore won't distract from the subject).

Background removal.

The other alternative, if you have a cap you think would make a spectacular B&W icon but for that busy background, is to cut your subject away from the background. Some good tutorials to look at for this can be found at good_tutorial (under this tag). The plus side of removing the background for an icon that you know will be b&w is that blending the new background and your subject together is easier, because your colours don't have to match! :)

Shadows.

I'm a sucker for a lot of contrast in B&W icons. While more matte tones work for some icons, I still tend towards really dark blacks and light whites for my icons because I like the aesthetic. While the shadows can be achieved through a combination of curves and/or levels, soft light layers, and brightness/contrast adjustment layers, I've found that if you start with a cap that already has the shadows there, the icon will come out much better. For instance:

(Click for full size.)

General Process.

Cropping.

The first thing I do when I start iconning (b&w or colour) is crop my image, because it's probably my least favourite part and I don't want to spend a ton of time colouring an icon only to find that there isn't a crop that I like. Usually, I just paste my image onto a 100x100 canvas and free transform until I'm happy with its placement (when I worked in PSP, I just used the crop tool, but I found it easier to work with than PS's crop tool).

For b&w, the main thing I've found to keep in mind is make sure the focus stays on your subject. Close crops where there are lots of shadows, centered negative space crops, etc. Without colour, it's easy to lose your subject if your background or textures are busy.

"Colouring."

I pretty much have two kinds of black and white icons: the ones where I sit down and plan for them to be b&w, and the ones where they end up b&w after being coloured normally. For each 'type' of colouring I'll give an example, to show how the process works.

For icons that I know are going to be b&w, the first thing I do is put down a black and white gradient map (found here in CS5), followed either by two curves layers or several screen/soft light duplicates of the base, depending on what looks better. If I use curves, I use the first layer to brighten the icon up and the second to introduce contrast back into the icon (usually, this second curve is 'S' shaped). From there, everything I do is strictly to increase the brightness and/or contrast as apporpriate - which I do using curves, levels, and occassionally a brightness/contrast layer. This obviously differs depending on the image.

As an example, for a simple icon that I intended to be black and white, this was my process:

The other b&w icons I do are as often happy accidents as attempts to make an icon specifically b&w, lol. Either the icon looks as good or better in b&w as it does in colour, or my attempts at colouring an image are either overwhelming the subject or just plain not good, and so they end up b&w instead. For instance, these icons are all b&w, but had colour versions already:

My process for getting these b&w isn't so different from my regular process. Once I've finished colouring the image, sometimes I'll check it with a b&w gradient. If I like what I see, I might save the colour version and then make the b&w alt. by tweaking the contrast and/or textures as necessary. For instance:

Tools.

These are approximations of the settings that I tend to use when making an image b&w. Obviously, it depends on the icon, and in all cases my main goal is really just to increase the brightness and contrast of the image. Usually, I'll use these tools in combination, not individually.

For more about how these tools actually work, see this guide to colour normalization that I wrote a while back, or browse through some of the tutorials I listed at the bottom of that post. :)

Curves.

>>

Levels.

>>

Brightness/Contrast.

>>

Layer blend modes.

I use lots of soft light layers, with screen layers underneath to keep things bright. I find the contrast gets too high if the layer is set to hard light, but if I use multiple soft light layers at various opacities, I've got better control over how much contrast an image will end up with. Soft light darkens the shadows beautifully.

>>

Textures.

Depending on your icon, textures might make all the difference in making your icon look like a finished product. These are some textures that I find myself coming back to again and again for my b&w icons, and what blend mode I usually use them on. Essentially, the trick is to make sure the texture complements the subject, not overwhelms it, because otherwise your subject might just fade into the background or get lost in the texture.

Below I've given some examples of different types of textures that I like to use, and how they look on an icon. I should note, though, that I'm big on using a whole ton of textures for colouring and lighting, so I've only listed the most obvious texture, not necessarily all of the textures on that particular icon.

Grunge/Splatter/etc..
All of these textures are by lookslikerain @ soaked

screen | soft + hard light | screen | screen

Lights/gradients.

hard light | normal (background removed) | soft light | soft light
midnight_road | midnight_road | soaked | lumsx

Everything else.

"Complex" B&W. (aka duplication)

Pretty much throughout this entire guide, I've been emphasising simple composition to make sure that the subject doesn't get lost in b&w. Obviously, I'm far from an expert, and all of this is pretty much just what I've concluded from my own experiments. And, as such, there are always exceptions to things that I've said. For instance, sometimes b&w can work with complex icons better than colour would:

In both of these cases, I've just duplicated the same base image - but, if these icons were colour, the colours would completely overwhelm the icon. The original icons are the black and white ones.

Light blobs.

So, these aren't really textures or gradients, which is why I'm not mentioning these up a little bit higher. Lighting is really important in a black and white icon because it is really what makes an icon stand out. Usually, if I'm not happy with how the lighting in an icon looks, I'll create a new layer, set it to soft light, and paint over various areas with a soft brush to brighten or darken those spots (light gray to brighten, dark gray to darken).

For instance:

+ =
(The "light blobs" layer is actually set to soft light, 88%. I painted over the blank layer with a largish brush (between 20 and 35 px, depending on the section), as soft as it could be.)

Other tutorials

These are some other resources that you might find handy when you're looking at b&w:

Tips & Tricks for Black and White by justmyb0nes
A Guide to Lighting by arctic_flower (colour icons, but the ideas can be used in b&w too)
good_tutorial has several icon tutorials that feature black and white that might be helpful (tag).

Tips & Tricks for Black and White

2012-04-29T14:54:57Z

Program(s)+version: Photoshop CS5 and CS6
Involves: gradient, color balance, exposure, brightness/contrast, textures, paint brush
Translatable: I don't know
Steps: /
Difficulty: medium

So I tried to divided it in three parts:
- Caps
- Tools
- White black & white
I hope I didn't do too many mistakes. English isn't my native language so don't hesitate to poke me on mistakes or sentences with none sense.
I never read any technical tutorial or open the help page of PS so I don't know the technical names of the tools either. I'll do my best though.
I think I answered everyone. If not, all my apologizes and poke me!

All icons were made by me and are snaggable.

CAPS
I am extremely picky when it comes to my caps. I only use HQ caps or blu-ray caps. For black and white, I also use grainy ones because it can give an effect on the icon without using a texture.
I don't use black and white as an "escape" when the coloring going wrong on a cap (exception goes for iconbattle). I choose black and white to focus on something. To me, it is a way to not be distracted by colors in order to focus on whatever you want to focus on. So, I choose my caps for different reasons:
- I want strong feelings
- -

- I want to emphasize something (bodies,something specific like what defines a character,...)
- -

- I want to make very strong contrasts
- -

You'll tell me: how that's color is distracting? It may be a perosnnal view but for example if you look at those icons:

-

On the first one, you see the colors and the light but you kinda forget the character. You see him just in the end. If you take the black and white one, you look directly at him because of the negative black space and the contrast with the light on his face.

I was asked "Can a cap be a beautiful colorful icon and a bad black and white icon?"
I'd say yes. For examples:
- -

Why those you'll ask. The first one because it has nothing but the colors. There's no emotion visible. The negative space is really extreme and if you'd do it in black and white, it'd just be a black (or white) blank space without a particular meaning. The icon won't be interesting at all.
For the second one, the character epitome is the red dress. She's a seducer, a femme fatale so making it in black and white might be pretty but what's the point? She would just be a woman in adress but in color, you directly think she's a femme fatale.
The third one is the typical example of a waste of a cap. In color, it would be all yellow and magnificent. In black and white: you can barely see what's going on.

So in short: no landscape, no blank character unless there's a meaning in making it black and white. For example, if you have a devasted landscape, black and white fits but if it a beautiful happy view of the sea, no black and white.
Use the color (or for our matter, black and white) to add meaning to your icons.

TOOLS
- Gradient Map: Black and White
Layer -> New Adjustment Layer -> Exposure
I never use the desaturated tool. I always find that the result is very poor. With a gradient, you will have a gorgeous "real" black and white effect. And if you want the effect to be a bit colored, looking washed out, you just have to play with the opacity of the gradient.
The gradient should always be on top. In other words, imagine a sandwich: on the top the gradient, then layers and last the base.

As you have probably already guessed, I never see the icon in color, always in black and white. In the end of the icon, I delete the gradient because I'm curious and want to see what it look like in color but most of the time the result is awful.

- Exposure: Depth
Layer -> New Adjustment Layer -> Exposure
Once the gradient is made, I create an exposure layer to give depth to the icon. The exposure will make it look darker and increase the shadows a lot. This is how I got my very dark and deep black that many of you asked for.
Play with the gamma at will. It is a very good way to give good contrast to an icon.

- Brigthness/contrast: Light
Layer -> New Adjustment Layer -> Brightness/contrast
After the exposure layer, the icon looks really dark. I always add a brightness/contrast layer to put some light. I always use the light and the light only, not the constrast.
Be careful to not make the brightness too bright because it will look ugly if you lost he shades of grey in your icon. Always be careful about the balance of your icon: too many contrast and light will kill it.

- Color balance: Contrasts
Layer -> New Adjustment Layer -> Color balance
I know it can seems surprising to use a color balance when you're working on a black and white icon that you see in blackand white and not in colour. But it is the most efficient way (for me) to work on the balance of the icon. Play with the settings without caring on how it looks like in color. Don't forget we are working under the gradient so you ar ein black and white. It will allow you to add some shades of grey or increase some darkness.
I can't tell you which color does what because it depends on the cap. Just play around! Don't hesistate to go to the shadows and the highlights too, don't limit yourself to the midtones.

- Smudge: Details
13th tool on the left sidebar
I always use the smudge tool to clean the cap a bit. Erase details that can be distracting. For example,here:

The original pic is here. I erased the right part of the desk, the cross and the candlesticks so that we will only focus on Spike.
I use the smudge tool. I prefer it to the eraser tool and the brush.

- Textures: Effects
I rarely use textures pre-made textures on my icons. The few times I do it is for challenges because I don't have the choice. I'd rather use my own light texture made with this little trick probably everyone and their mothers know by now.
Let's work with examples. Those 2 icons are a mix of both of those techniques

-

I used the textures by lookslikerain (here). I used them to have a special effect on the icon. It looks scratchy. I set them to "screen" to give them the good effect.
Note that setting an icon to soft light is also a good idea to make the icon looking darker. Setting to "multiply" or "color burn" can be a good option but there's a lot of work to do on the icon afterwards most of the time. It all depends on the cap, the texture and the effect you want.
For the light on them I used a trick I learnt from wickedgrdn long time ago. I open a new layer, paint a spot or a line with a white brush and then use the Gaussian blur (Filter -> Blur -> Gaussian Blur) to make the light as I want it to be.
Then always duplicate the base, move it above the texture you just create and set it to soft light. That way, you won't lose any contrast and you will improve the light.

WHITE BLACK & WHITE ICONS
I was asked how I do "white" black and white icons like
- -
The first is a mix of the cap itself and textures. The cap itself has a very light background so with a lot of light, it goes white. Then, I added a texture many times to make the whole character looking grey. The big light spots where already on the cap.
The second on is the result of many copies of the base all set to screen. (yes, that simple)
The third is simply because the background of the cap was very light (light blue) so I used a brghtness layer and it became white.

The reason why the "screen" option works well is because I always use very HQ caps. I never use the screen option if the cap isn't HQ. If it is LQ or grainy (blu ry caps can be grainy like BSG for example), don't try, you'll lose quality (unless you want it to be grainy of course).

THE END
I didn't say it all along but all the steps I showed here are layers between the base and the gradient. (remember: the sandwich)

Soft and glowing effect using gradients

2012-01-08T09:16:11Z

Program(s)+version: Photoshop CS4 (for Mac)
Involves: Blending modes, gradients, curves, brightness/contrast, selective colouring, vibrance
Translatable: Partly.
Steps: 12
Difficulty: Medium, because I'm going to assume you already know your way around most of your programme.

Both ohgollygeedamn and likealight asked me about this colouring at the Ask the Maker Meme 2.0, so here it is!

With Photoshop CS4, we'll be doing this:

--> &

1. I started with this cap, which I duplicated once and set to screen at 62% to lighten up the image quickly. I duplicated the base again and dragged it to the top. I set this layer to soft light at 41% to add some contrast. After this, I duplicated the screen layer and played around with the opacity until I liked the balance. In this case, the second screen layer ended up at 34%.

Step 1 is what I still consider 'preparing my base.' I think that most of my icon making process has to do with reaching a right balance between light and dark and softness and sharpness. During my entire process, I keep going back to my base, which in my case means that my layer palette usually includes lots of layers that are copy+merged from the ones below it. I keep creating new bases, as it were.

2. So these layers were copied and merged into my new base. On top of that, I created a new gradient fill layer in black and white (Layer > New Fill Layer > Gradient), set to soft light. Gradient layers quickly add any colour combination you want to your icon. If you then play with the blending modes, you get a multitude of effects to choose from.

My favourite is the soft light effect, because it adds the colours in a way that gives contrast to the icon - particularly when it comes to quickly adding light in a way that might work perfectly.

This gradient is absolutely nothing special: it's just the standard black and white preset and even the angle is straight-forward: 90 degrees. When I use gradients, I sometimes play around for ages until I have the exact angle. Sometimes, however, it looks good straight away. This was one of those times.

I did, however, end up slightly masking out a part over the right side of Amy's head near the top, because the gradients were lighting up her skin and hair too much there. (If you don't know about masking, I recommend this great tutorial.) I imagine I did this after I did step 3, because that's what happens when you duplicate your layers: proportions can get distorted and you either have to lower the opacity of the layer in its entirety, or you have to mask out parts (slightly) that are bothering you.

3. So step 3 was duplicating those gradient layers. I like duplicating my layers to see if the effect intensifies and becomes prettier or starts ruining the icon. The first duplicated layer was set to 47% opacity and the second duplicated layer didn't make it past 4%. This is what I mean by going back and forth between layers: it probably started out at a higher opacity, but got tuned down once I started focusing on colour and vibrancy.

4. Because I wanted to pack a bit more punch in terms of contrast, I used a curves layer (Layer > New Adjustment Layer > Curves>) to light up the entire image just a little in the RGB section. See a screencap of how minimal the change is here. In itself it seems hardly noticeable, but in the process of making this icon, I must have, at some point, felt it necessary.

5. If this next step seems contradictory to the previous one, that's because it is. Kind of. I never said I was logical! ;)

On top of this curves layer, I created a new Brightness/Contrast layer (Layer > New Adjustment Layer > Brightness/Contrast) where I dropped the Brightness to -12 while upping the contrast to 19.

If you wanted to know how I icon, this is one way. I use several methods to work on the contrast and light in my icons. I tend to try out all the options I know until I'm satisfied with the result.

6. But now, finally, for the colouring part. This was done with selective colouring, mostly. The first selective colour layer (Layer > New Adjustment Layer > Selective Colour), which is set to an opacity of 78%, has nothing altered except for some tweaks in the Neutrals to give Amy's light skin a softer, yellowy glowy feel:
Yellows: +21

I hesitated between going for the combination yellow+red or yellow+magenta for a long time, so earlier versions of the icon have Amy with red lips rather than the pink/magenta hue they have in the icon I ended up posting.

7. When it's specific colours I want to bring out, I tend to use a new selective colour layer for each colour. I'm not sure if that is an efficient process, but it works for me, so the second selective colour layer only has a tweak in the Reds and Magentas section:
Reds:
Yellow: -47
Magentas:
Magenta: +17

8. As you can see, the icon has lost some of it's yellow hues, which makes sense, because we've decreased the yellow in the Reds of this image, but needs to be remedied nonetheless. By far the easiest way of doing that is by using a Vibrance layer (Layer > New Adjustment Layer > Vibrance). Some people call this cheating. Personally, I find I need to watch out with Vibrance because the increase of yellow often disturbs the colour balance I've created in my icon. (Vibrance is almost always my last step.)

The Vibrance layer has an upped vibrancy of 100 and is set to 100% as well, so it's pretty straight up. But it works quite well here.

This is also the part where I guess the masking work of steps 2 and 3 happened. I can imagine myself duplicating the gradient layers at this point to see what they would do to my colouring, then playing with the opacity. This is also why that last duplicate gradient ended up at a 4% opacity. My layer palette looked like this.

9. Now for the part where I had most fun: cropping! This icon was made for a close crop challenges, and since I am always trying to improve on my close crops, I resized and moves this image around forever in a new canvas, ending up with all of these close crops:

But because I was kind of bummed that all that lush green couldn't make it into a close crop icon, I decided to do a zoomed out crop as well.

10. I resized the image to 100 pixels width and put it in a new 100x100 canvas. Behind it I put a new Colour Fill layer (Layer > New Fill Layer > Solid Color) and I used the eyedropper tool to select a colour from the lightest parts of the icon to serve as a background. I tend to use the eyedropper and then go a few shades lighter to make the colour as neutral as possible. (This icon has #fcf3c6 as its base background colour.)

11. However, I thought the background looked a little too neutral, so I used textures to give it some depth. I've used this texture by lemonpunch set to 100% normal blending mode, but pulled down to the lower left corner, like this. I duplicated that texture and set it to color burn at 100%. Then I took this texture by lemonpunch, set it to Overlay at 52% and masked most of it except for the upper right corner. I wanted the texture added to the yellow background, not to the image of Amy herself.

12. The finishing touch was a Vibrance layer on top of the layer containing Amy with the Vibrance upped to 54 and the Saturation to 2:

My layer palette of this icon looks like this.

Questions? Concerns? Thoughts? I'd love to hear them!

Multiple Animations in One Image

2011-10-30T05:09:01Z

Program/Version : Photoshop CS4
Involves : Adding text to animated GIF
Difficulty : Moderate to Difficult (basic GIF making and Photoshop skill required)
Translatable : To Photoshop CS3 & CS5, yes
Steps : 5 (not including making GIF from scratch)

*I posted this at icon_tutorial about 2 months ago, so I'm sorry if you've seen this before. But this one is improved version :)

Tumblr Graphics Tutorial :
multiple animations with continuous effect
in one image

In this tutorial, I'm going to show you how to put together multiple animations in one image with continuous effect (one starts after another).

Preview :

Step 1 : Preparing individual animations
As you can see in the preview, there are three separate animations combined in one gif. For the first step, we'll need to prepare each animation before putting them together.

I chose this video to gif, the part where Andrej says "I'm young, fresh, and unisex!" at 0:27. I'm going to make 3 animations, each for every phrase he said. I'm not going to explain basic gif making/gif coloring in this tutorial, if you need any reference please see basic steps at my tumblr gif tutorial or other basic gif making tutorials in this comm.

This is a cap of my base gifs :

Each consists of 7, 10 and 14 frames.

Step 2 : Placing text on animations
I needed text to go with the gifs, so I used type tool (the letter T on your tool bar) on each gifs. For a fade in effect, I used this manual setting for the text layers' opacity (look for 'opacity' in your layers window) :

Frame 1 : not visible
Frame 2 : 30% opacity
Frame 3 : 60% opacity
other frames : 100% opacity

Here is my result on my second animations :

(consists of 10 frames)

Finish it with Animation options > Flatten Frames into Layers so you have your text permanently flattened in every layer.

Step 3 : Copy and pasting animations into one canvas
Now that I had all 3 gifs complete with text, I wanted to put them all together in one canvas. When you're putting together animations, work with any gif you want to be in the bottom or right side of the canvas.

I expanded the canvas by doubling the height of my bottom gif with canvas size (Image > Canvas Size), but set the anchor so my original gif stays at the bottom.

On my second gif, I chose Animation option > Select All Frames > Copy Frames. Back to my third gif, I pasted it before the first frame, and when prompted with paste method option, I chose "paste before selection". I repeated the same step and pasted my first gif. So I have this; my first, second, third animation together in one canvas :

Step 4 : Creating static frames
As you can see in above image, my animation is in order, one starts after another, but I need to fill the empty spots so that it shows static frames. I made a table that I hope will make this easier to understand, imagine that this is an animation bar :

What you need to do is to know which frame is which (you can change your layers' name to help distinguish them). Go to your animation bar, select every frame of your first animation, and then, go toggle the visibility of the first frame of your second and third animation (see cap). Continue according to the table above. Here's what I got :

Step 5 : Finishing for tumblr
As you can see above the animation itself is complete, but it doesn't fit to tumblr's 500kb limit. There are some options; make the animated part smaller, delete some frames, or get rid of the colors.

Here's what I've done :
- added coloring
- resized my animation into 250px width (from 349)
- created a black frame (500 x 500)
- added some feather effect to the sides
- Save for Web & Devices > Option > Optimized to file size > 500KB > OK

Result :

and another version in b/w using gradient map layer (bigger animation) :

*Please note that this animation has 31 frames in total, but not many movements and or bright colors in it so I could maintain the quality, but this might not work well with busier animation.

**Another note : this style of animation were more popular back then when tumblr hasn't upgraded their photoset. But still, it's fun to learn how it's done :)

An awesome example from tumblr :

(via fuckyeahmisfits, click to reblog)

I hope this is useful and I would love to see what you make using this technique!

GIF TUTORIAL FOR BEGINNERS

2011-06-30T21:25:49Z

Program(s): VLC Player (not required), PhotoShop CS4
Involves: Screen-capping and making and editing animated gifs
Translatable: Only from PS CS3 - PS CS5 i'm pretty sure.
Steps: 5-6. All steps are lengthy and discussed in detail. I have a more detailed list of what the tutorial will cover underneath.
Difficulty: (Hopefully) Easy to Medium. As long as you have the patience to read. I wrote this for beginners so I don't skimp on the detail.

This tutorial is more of a beginners how-to to making animated gifs in photoshop. These steps are by no means the ultimate correct way of doing anything. I literally taught myself how to do this through experimenting so...it's the best I could do. I hope it helps.

What this tutorial will cover:

• How *I* screencap and import caps into photoshop (briefly)
• Creating the actual gif/animation
• Tips and Tricks to reduce the file size of the gif
• Adding colorizations/text to the gif
• Saving the gif for the web

Before you do anything, you are going to need caps. I don't know how to import .mov or .flv files into my photoshop. I've tried before but always get errors. You can use whatever method you like for getting caps. Personally, I use VLC player and manually cap the scenes I want to gif.

I find the scene (usually a few seconds long), and hold down the Command+Otion+A keys (Alt+Control+A on PC) as it plays. This creates the screencaps (everytime I click A) and puts them on my desktop. You can customize these shortcuts/settings in your vlc preferences.

Under Preferences } Video there will be a Video snapshots section. This is what mine looks like:

I set the file extension to .png and select the sequential numbering box. You NEED to turn sequential numbering on. You will see why later on.

Okay. So for the gif in this tutorial i'm collecting caps from a scene in the movie Clueless. When making a gif for tumblr, the smaller number of caps you can work with the better. Remember it's a trade-off. If you want to gif a longer scene, you'll have to compromise your image size, color and/or quality on the final product since tumblr has a 500kb limit. So try to make it work with a small number of screencaps if you can. Or you can do multiple gifs of one scene and post them all in one text post. Whatever works. (This doesn't apply on livejournal obviously).

So I let the scene play on my vlc player and I hold down the Command+Option keys and pressed the 'A' key over and over again. You'll probably manage to get a cap every half second this way. (As far as I know, there's no automatic capping tool in vlc. I hear KMPlayer is a good program for capping? Don't know if it's compatible with mac though). All the caps are now appearing on my desktop in the order I capped them. Once i'm done, I create a new sub-folder, highlight the caps and drag them into the folder. Because we turned sequential numbering on, this keeps our caps in order for us. So if your desktop gets crazy with screencaps like mine does sometimes, the sequential numbering will save you a lot of time.

Adding the caps to their own folder will allow me to preview the gif and weed out the caps I don't want or see if I need to go back and cap something I missed.

*Please note that i'm on a MAC computer. I'm not sure if there's a similar method to do this on a PC. But this part isn't necessary to making the actual gif so don't worry if you can't. Like I said before, it's just a way for me to preview the gif and my caps before going into photoshop. Any way you can manage to do this on a PC will work fine.*

Once inside the new folder, I put the view on column view and highlight the first file.

MAKE SURE you click the spacebar once to open up the the pop-out preview (it won't work right unless you use this) and then just hold down your down arrow on the keyboard. This will run through all the caps you took and basically will look identical to the gif you'll be making.

• Okay. Now that you've seen the caps in order and picked out the ones you really need, we'll move into photoshop. If you're making a gif that's 500px wide and you want to make it in color, I suggest having no more than 8/9 screencaps.

Next step is to highlight all your caps and open them in photoshop. All your caps will open in one tab. Don't do any cropping or editing right now. Save it for later on. Now we're going to layer them on top of each other in ONE file. This might be a long way to go about it, but it's the only way I know how and once you do it a lot it takes basically no time. Click the little double arrows on the side of your picture and a list of every file opened in that tab will be in a drop-down menu.

EDIT: Courtesy of carelesshands : If you go file > scripts > load files into stack, then select all your caps, it basically skips having to drag them all on top of each other and you just go straight to "make frames from layers"

Find the file with the smallest number in the filename (in this case it's ...00091). This is where sequential numbering is KEY. It would be practically impossible to do this without it. Drag that file out into a new window. Then go back to the tab with the rest of the files and click the double arrows again, this time selecting the next number in line (...00092).

Click on that LAYER (should be called Background. by default) and while holding down the SHIFT key drag it on top of the first file you pulled out (the ...00091 file) (this is just like you drag layers from a .psd file to another document). *Holding down the shift key will place it directly in the center, making the caps line up perfectly.

Now repeat this process until all the caps have been placed on top of each other in the new document. You'll just keep pulling over the next numbered cap in line until you're done. (...00093, ...00094, etc.)

• Now with all the caps in one file you can close the other tab/window. Double click the background layer on the new document to unlock it. Now you can do editing such as cropping and re-sizing. DON'T ADD ANY COLORING at this point. Just crop and resize. In this instance, i'm cropping the black lines from the sides of the caps and i'm cropping some unnecessary parts of the caps from the top and bottom. Any pieces you can part with will help your file-size in the end.

After cropping, I resized the gif to 500px wide. I like to sharpen each individual layer so I'm going to go ahead and do that. You don't have to if you don't want to. I just click the first cap in the layer palette and go to filter } sharpen. After that, you can click each layer in the palette and click Command+F (Control+F on PC) and this will perform that same sharpen filter on each layer you click. Command+F is the shortcut to perform the last used filter. This will save you time from having to constantly go click back to the filter menu and it applies to all filters.

Okay, now i've done all the editing I need to at this point. I WON'T BE COLORING UNTIL THE END.

• Now I want to start making the actual gif. Open your Animation window. Window } Animation.

When mine opens up it usually looks like this:

Just click the bottom right button and it should jump you to this window:

Now, with the top layer in your palette selected, click the this button in the middle bar and select Make Frames From Layers.

Now your animation window should have multiple frames in it.

Highlight all frames (either by clicking and holding down the shift key or clicking the same button you clicked to make the frames and selecting Select All Frames). By highlighting them all, you are applying any changes to all the frames at one time.

Now select the options to make your gif run 'Forever' and then select how many seconds you want each frame to run (click anywhere it says 10secs and a drop down will appear).

I usually choose .02 seconds, but .01 works better on this particular gif. You can de-select them at this point. I want to go ahead and make the last frame a little slower, so i'm gonna select just that one and change it to .02 seconds (NOTE: oops i actually forgot to do this when i wrote the tutorial but i usually do. Just pretend I did). If you want to put in a custom number for it you can just click Other... in the drop down and type it in (sometimes I do .03). Okay. Now we've got the timing of the gif worked out. You can always keep messing around with it if you want to change other stuff around. Just preview the gif by clicking the play button and it will run in the window where your caps are. When you stop the preview, always make sure you click back to the first cap in your layer palette.

Before I color the gif, i'm going to see what I can do to get the file size down a little more. In this particular gif the background doesn't move hardly at all...so I think I can get away with this next step and it won't look terribly noticeable. With the top layer selected click Command+J (Control+J on PC). This will duplicate the layer (Layer 8 in my palette). Now click the duplicated layer (Layer 8 copy in my palette) and select all the parts that will not move. What I did was use the free-hand lasso tool and select the girl's figure, leaving some extra room for her movements.

Now Select } Inverse. This will select everything BUT the girl (aka the parts that won't move) and click Command+J again. This will put the still parts in a new layer. Delete the duplicated layer (Layer 8 copy in my palette) you made before. You don't need it anymore. Now the still parts are placed over every frame in the animation and the girl is still visible underneath. This actually reduces the file size quite a lot. I know some people don't like it or don't think it looks right and most of the time you get strange movements and you can tell. But you gotta do what you gotta do sometimes to make the size go down. Just be careful with what you're selecting, keep testing the gif to spot anything too strange, and don't even bother trying to do this with backgrounds that move a lot. You will be able to tell and it will not look good.

Now it's time to edit color! Open the .psd you want to use in a new window. Make sure the top layer is selected in your palette (should be the still parts you just created). Now go to your .psd. I usually group all the coloring layers together (highlight all layers then press Command/Control+G) to keep things neat but you don't have to. Drag the (grouped) layers on top of your caps. This should apply to all the caps underneath (which is why the top layer needs to be selected. If you have another layer selected, the coloring will only affect that layer and whatever is underneath it). If you want to add text to your gif this is also a way to do that. Basically, anything you create on top of the top layer in the animation will apply to every frame underneath as well.

Now test your gif out by clicking the play button. If you are satisfied, let's get to saving!

Go to Save As } Save for Web and Devices. A new dialogue box will open. As you can see, these are all my settings:

I like to use Pattern because it adds an overall grain to the gif. It puts the file size up slightly but I like the way it looks better. You can adjust any of these setting though in order to achieve the look you want. Your colors are going to affect file size a lot too. I was lucky with this gif that I got to keep my colors at 256. That almost never happens. Note: Darker gifs with a lot of black will always have a smaller KB than brightly colored gifs. Less colors = smaller file. The same goes for Black and White gifs. B&W will always be smaller.

You can also reduce the IMAGE size in this dialogue box. I've had to go from 500px wide to 400px before to get the file size down. I like to keep it at 500px wide though if I can help it. I think it looks nicer on a blog that way.

Once you get your file size under 500kb you can save it! After that, just upload to tumblr and you're done!

Hope this was helpful!

Using layer masks in Photoshop: a basic guide

2011-06-14T13:23:19Z

Program(s)+version: Photoshop CS5
Involves: Layer masks, lasso tool
Translatable: should work in all programs/versions that have layers and allow to create layer masks
Steps: about 6 for the example, the rest is a guide more than a step-by-step tutorial
Difficulty: Beginner

A little guide on how to do layer masks and an example on how to change the background of an image using them.

Using Layer Masks in Photoshop

- A Basic Guide -

Since layer masks are a very essential part of my icon and/or graphic making process, here's a small roundabout on what they do and how you can use them e.g. for getting rid of the background in an image.

What layer masks do

Using layer masks is the undestructive way of "erasing" pixels from your image.

With the eraser in Photoshop, you literally delete parts of your image, meaning you lose them for good. With a layer mask, you just "hide" the unwanted parts of your image; you make them invisible. The advantage of this is that you can always go back and make corrections to your mask if you accidentally made a mistake but didn't notice until the very end. For example, you cut a person out from it's background and after you're almost done, you notice that the person somehow lost it's nose in the process and you want to get it back. If you used the eraser, you'll have a big problem now, you either have to live with your poor noseless soul (sorry, it's a really crappy example, I know) or do all the hard work again. With a layer mask on the other hand, you can just go back and make those parts of the image magically reappear again.

Basicly, this is just the long-winded way of me saying that the eraser is evil.
Layer masks are the good guys, and they can be used for a lot of things once you've understood the basic principle of them.

How to create a layer mask and how it all works

To create a layer mask, you just have to select the layer containing your image and then go to this little button here, right underneath the layer palette:

This will create a layer mask for this layer, you can see this by having a look at the layer palette, there should have appeared a white filled rectangle right next to the thumbnail of your image:

If you click the white rectangle, the "mask", you can simply paint on it using a brush. When you paint on it with black, these areas of the image will become invisible. Painting with white on it reveals the image again. Painting with grey will make the image gradually more transparent, depending on the shade of grey you use.

It's not that difficult after all, is it?

Instead of using a brush to paint in the mask, you can also make a selection on it and fill this selection with black or white or grey. I will use this method in the following example.

An example:
How to use a layer mask to change the background behind a person

Note: You can use this technique on larger images as well as on icons. I am using a mid-sized graphic here for better visuality.

I have an image with a person and I want to put that person onto a different background, a texture for example. To do this, I create a layer mask as described above, then I use the lasso tool

to make a selection around the person:

It will be easier if you zoom in a bit to do this, although it doesn't have to be perfect at this point.

When I'm done with the selection, I invert it by pressing Ctrl + Shift + i on the keyboard, so the marching ants will now cover the background instead of the person.
You could have made the selection around the background from the start, of course, but I find it a lot easier to do it this way around.

Now, making sure that I have selected the layer mask-thumbnail in the layer palette (and not the image thumbnail), I fill my selection with black and the background will vanish.

This is how the image and the layer mask are supposed to look like:

Now I can drag the texture that I want to use as a background underneath my image to see it how they blend together - for demonstration purposes I will just use an ugly dark purple color fill layer instead of a texture here; you should obviously aim for something prettier than that:

You can see that my selection was not very good to begin with, you can still see some light blue lines where the sky from the original background shines through. So that means that there's still some work to do.

To finetune the mask, I zoom in a lot (to about 200-400% or more) and use a very small brush (1-3 pixel, sometimes more; it depends on the image size that you are working on) to paint in black over these light blue lines to make them disappear. Again, you have to paint in the mask here, not onto the image.

When I think I'm done, I zoom out to 100% to see how it looks:

It's a lot better than before, but I still don't like the hard edges on the right side of the armour, especially on the right shoulder, it still looks like an obvious cut-out to me. So, I go and zoom in again and do the same as before, but this time I will use a larger and less hard brush to carefully paint over these parts with black to - hopefully - blend them in a little better:

And finally, it looks good enough to me:

Adding a quick coloring and a more beautiful background (texture here by spooky_window) might give us this, for example:

Of course, there's a lot more that you can do with layer masks, these are just the very basics and a rather simple practical example. You might want to experiment a little for yourself, painting on the mask with different kinds of brushes, changing the opacity of it, using harder or softer brushes, using a gradient... well, and whatever else you can think of.

Additional info:
If you're photographically interested and want to dig really deep into the realms of masking and see what great things can be achieved with them, I recommend a read of this very amazing tutorial by Tony Kuyper. You most definitely won't want or need this for creating something as small as icons or simple fanart, but if you work with larger graphics and especially with photographs, this can actually be a lot of fun to play around with.

Well, and that's it. I hope I didn't forget anything important and as usual, if something I said doesn't make sense or you have further questions, just ask. :)

Icon tutorial: colour enhancement in 7 easy steps

2011-06-12T14:48:21Z

Program(s)+version: Photoshop CS4
Involves: Auto Contrast, Curves, Smudge tool, brush work, texture use, Vibrance
Translatable: Not entirely, because it involves Variations. But you can work around them with Hue/Saturation.
Steps: 7
Difficulty: Easy, trust me

With Photoshop CS4, we'll be re-creating this icon:

>

shrimpy_19 and ofraindrops asked me about this icon and I happen to love it - so here's a tutorial and an explanation of my thought process, as requested by absolutelybatty.

I'm warning you: it's rather long. ;)

This icon was created for a challenge at icondare. The challenge was 'repetition'. I was a trial maker at the time, so I wanted to do something people wouldn't expect from me. I had to, because the members at the community had decided they weren't sure whether or not my work was good enough. So I felt I had to show them something new, something they hadn't seen from me.

absolutelybatty asked me about my thought process for icons like this and I'm glad she did. I think my biggest challenge to improving as an icon maker is expanding. I think I have some decent skills, but I need to get more creative and I need to push myself to not follow my standard procedures or simply do the things that come most natural to me. Some people excel at close crops. I don't. It's not that I completely suck at them but they don't come natural to me at all (which is why I'm working on them now). When I watch a film or television show or when I see screencaps or other pictures, my mind goes: "Ooh, that would make a great icon if I extended the background that way" or: "Look at that light! Can you imagine how great that would look if I cropped it in a corner and had all this negative space there?"

You get the idea.

Anyway, to improve as a maker, I need to think outside of my own box so to speak and being a trial maker at icondare helped in that respect because I had to figure out something to impress the members who didn't think I was good enough to be a full member of their community.

So I wanted to create a repetitive element that wasn't 'standard' but that also built on something that I think I do best: negative space. Basically, I wanted to diminish the repetitive aspect of the icon as much as I could. ;)

This screencap from The Fall by letsey_x seemed perfect for what I wanted to achieve.

Step 1
I cropped it so that the characters were just a little off-centre so there would be room for the repetition. I try to use the rule of thirds in all my icons as much as I can - or at least keep it in mind where I can.

Step 2
I extended the canvas upwards (Image > Canvas size) with the background colour a blue I took from the original cap with the eyedropper tool:

My base now looks like this:

Step 3
I copied this base and set it to auto contrast (Image > Auto Contrast):

Auto Contrast is useful to me because it gives an indication of the contrast already present in your image. If you see a big difference in your image before or after Auto Contrast, your original image had low contrast and (part of) that is now already fixed. If there's not a big difference, you know that fiddling with the contrast is not something to worry about in your main process, because the image is already nicely contrasted. The good thing about Auto Contrast is also that you can always fade it if you think it's too much contrast. Right after you have clicked on Auto Contrast, go to Edit > Fade to lower the amount of contrast.

Step 4
Then I used the smudge tool to smudge the background so that the entire background was blue:

When the background has a more or less solid colour, I set my brush (a round one with soft edges, the size depends on how large the canvas is and how much room there is between the edge of the original image and the people/lines on it) to 100% opacity. When there is texture or different colours in the background, I set it to a lower opacity and have to do much more work to get a gradual feel to the background. For this all-blue background, the smudging was easy.

Step 5
The image needed to be a lot brighter, so I created a new Curves layer (Layer > New Adjustment Layer > Curves) with the following settings in the RGB channel:
Lower point - output: 66 | input: 73
Upper point - output: 188 | input: 156

I just pull on the line to create a curve, but if you want to experiment with actual numbers, just click on the black arrow on the bottom to get those boxes where you can put your numbers. See here. If you want to learn more about Curves, you should check out this fabulous guide by talipuu at mirrorskies.

The icon now looks like this:

Step 6
You can see that the colours are already very similar to the end result. The only thing that changed between this and the end result is adding this texture by lemonpunch and setting it to overlay. This adds the texture but in a way that also lightens the image up. I don't remember the exact percentage, but you can see in the end result that it was lowered from 100%

(It could be that there was a Vibrance layer involved at the end as well. I do that often when I've uploaded an icon and the colours look different online than they do on my MacBook, stupid screen calibration. If your program does not have Vibrance layers, you can get a very similar result with a Hue/Saturation layer.)

Step 7
I went back to the base to smooth some of the lines with a (very small) soft round brush at a low opacity and then I merged the image and moved it into a new document to use that new document as my canvas to create the repetition. I'm sure you can image how the cropping of the repetitive elements went: I cropped them and then played around with different sizes on the original icon to see how small they needed to be to add to the composition rather than distract from the larger crop.

The end result is this:

I was inspired by this textured feel by some of the icons I had seen from the absolutely amazing talipuu at mirrorskies. I had no idea how they achieved their look (and still don't) but was determined to find a way to give my icon a textured feel somehow. For some reason I came across that texture in one of my folders and just decided to give it a go. It was the first texture I picked and it was a straight hit. I wish I could say there was more thought put to it, but sometimes, you just have luck and an icon turns out even better than you hoped.

So you see, this is really a much simpler icon than it looks.

I hope this was helpful to any of you. As always, if you have a questions, don't hesitate to ask. :)

MOD POST - "nominate" awesome tutorials/guides!

2011-06-12T12:44:37Z

Hi good_tutorial!

I guess you have all noticed that things have been very quite here lately, sadly not many people write actual tutorials these days. With this in mind I thought it was time to find a way to remedy this and bring some new blood into the community so to speak.

Now I want you to comment to this post with tutorials/guides you think would be a great addition to this community. Please make sure you check out the members list in the community profile page to see who's already members here. Also please make sure the tutorials you "nominate" follow the spirit of good_tutorial, tutorials with explanations, not just lists of numbers or .PSD files.

I will then try to contact the makers you "nominate" and see if they would like to join here and share their tutorial(s).

You can post/"nominate" as many tutorials as you like :)

Comments will be screened.

icon q&a post: coloring, cropping, general

2011-06-05T13:56:32Z

Program(s)+version: Photoshop CS5
Involves: coloring, cropping, certain tools to achieve both of those
Translatable: for all Photoshop CS versions. Some settings in certain tools might not be translatable for other graphic software.
Steps: No steps, just general information
Difficulty: Easy (but with basic knowledge of Photoshop)

The questions asked in this post were the ones that were asked in my Ask The Maker thread here at icon_talk.

First of all, I have to say, how I color my icons isn't really all that special. I use the same method on each and every one of my icons. Different settings, yes, but mostly the same method. I'm not creative enough to do a different one for every single icon I make.
My method changes though. It isn't the same one as when I first started making icons, and it isn't the same one as the gazillion methods I had after that. Once I find a good one, I stick with it. Until I get bored or accidentally find something else.

A while ago I did a tutorial here which is basically still the method I'm using now.

There were a couple icons though that tutorials for were requested, so I'll be explaining what I did differently to them.

(the steps below aren't the only things I did to those icons. I used the same method for every one of these as I did in the tutorial linked above. The steps mentioned are things I added or changed from that method)
Asked by hopeitallaway:
to

I remember this cap giving me trouble, but I really wanted to do something with it because the forest is pretty and pretty people running in it is pretty and... well, you get the point. But, just look at it, it's extremely dark, right? And not extremely colorful either.
I always put a yellow Soft Light layer on an icon (see above tutorial), which is usually at 20% opacity, but in this case I put it on around 60% because the colors really needed some yellow to liven up. That wasn't enough though, so there was a Hue/Saturation layer added with the Master Saturation +20, which makes it look a whole lot better. (I'm experimenting right now though and duplicating the yellow Soft Light layer and lowering that opacity isn't looking too bad either!)
Anyway, from that point on you can just fiddle with some Color Balance layers or something else to add a little more yellows and reds and there you have a colorful icon! I just came out with something like this. (I still hate that his face is kinda blurry though, but meh, it's about the coloring right now)

to

Damn you, Merlin caps! Always with the dark and uncolorful things you do! (they do have colorful scenes though, but seems they don't in this case :P) Anyway, there really isn't anything special about this icon's coloring. What I like is the crop though. I tell you, when you get the chance to icon a pretty cap with two people's face on it, try to crop it in a way you see the half of one person's face, and a half of the other person's face. If it looks good, you have a crop! Most times it doesn't look good though, either because the faces are too close together (and the cap's quality isn't that good) or too far away or something, but in this case it worked. Especially with Merlin being all busy scheming in the foreground (in the dark) and a noticably lighter nonsuspecting Morgana in the background. The contrast between the dark and the light skins of the two is the reason I couldn't get this icon to be really colorful. Because when you focus on Merlin looking good, Morgana looks wonky, and if you do it the other way around, Merlin'll look wonky. There were ways to fix that but I'm lazy and I thought it looked good enough as it is. (no one'll be using my icons anymore after this, I swear :P)

Asked by nalazhar:

to

Fun thing about The Vampire Diaries caps - when they're dark, they're dark. But when they're outside, they're COLORS COLORS COLORS (or at least, potential for colorful thinigs!) and nice quality at that. They were that less in season one though, but since season two it's improved. Anyway, even though there are some colorful things in the image, it's still quite dark so that's why there were lot of screen layers involved (2 or so). A simple Hue/Saturation layer with +20 also did the trick and I think that's all the 'special' things I did for this. Remaking it now I ended up with this.
This also goes for this icon.

Asked by lover_of_narnia:

to

Man, Legend of the Seeker caps are a DREAM. A dream, I tell you! They have amazingly choreographed fight scenes, slow motion hairporn, pretty people, amazingly gorgeous scenery, everything is colorful and all caps for it are AMAZING quality. I get mine from orange_corner where they have S2 in 1920x1080px so you know, you can't go wrong with that. This is why I can't stop iconning this show, people. One - I love the show to death. Two - everything I said above. Seriously, the original cap for this icon was this. It's in movement so it's not as sharp but because of the size of the image it's enough.

Now, back to the icon. It's one of my personal favorites, to be honest. The colors worked really well in my opinion, there's negative space that is actually grass and there's Kahlan in her pretty white dress but it's just her silhouette and not her face. So, basically, lots of Hue/Saturation, yellow highlights and red shadows in the Color Balance, enough Levels to make it contrasted enough, an extra added light texture set on Lighten to not have such dark blacks, and there you go!

I realize this was more of a LOTS cap lovefest than an actual tutorial, so I do apologize. But really, most of the work is done by the cap itself. Believe me.

to

Say what you want about Lumen, but she made for some great shots. This one has to be my favorite though. It's her lying down, her hair ends all curly, and the cap is colorful enough to work with. This is also one of those cases in which the cropping can make such a huge difference.

Basically, the same as the above tutorial. Make the colors that are already there pop by Hue/Saturation, Color Balance, enough contrast, and an added light texture. Here are lots of yellows and reds to work with and that's always fun.

II. CROPPING

In case you're still there, and I really hope you are, here's the cropping part which a few people asked questions about.

wandererjulia asked: this could be a very stupid question but how do you crop your icons?
do you work in a great canvas or you prefer to put the cap inside a 100x100 canvas and resize the cap until you're satisfied with the crop?

I'm lazy as hell and I just use the crop tool. It's the easiest and it works just as fine as the other way. I change my Crop Tool settings from the default though, like this:

Crop Guide Overlay - Changed it to Rule Of Thirds, which gives you something like this, which is absolutely perfect for center crops and the like.
Opacity - Needs to be dark, obviously, but not too dark so that you can still see what you're cutting off of the original image.
Resolution - (which is the option you see when you're not in the middle of cropping something) Always set to 72. It's what all pixel images work with and it also avoids issues of incorrect font sizes when you're trying to work with text.

If you don't like the crop tool because you can't really play with the crop after it's cropped (like with the Free Transform tool), do not worry because there's always this option.

exp0se asked: I love how you achieve such deep reds in your icons. I'd love some pointers on how you get your icons to be so wonderfully contrasted and rich in colors like red. You're one of the best I've seen at making icons look so vibrant and pretty. :)

Doesn't have anything to do with cropping but it's the only coloring-related question I got so I'm answering it here anyway.

By the way, hi there hon! I love you! :D Ahem. So basically, it's really just always a Color Balance layer at the end (in which I heighten the Red in Shadows and heighten the Yellow in Highlights), those really do the trick. After that one there's always a Levels layer (unless the contrast is already high enough), and change this setting. It usually averages between 10-30 and sometimes higher.

A general cropping guide would be wonderful! Maybe with some of this included?
A. Do you automatically see the crop in your mind when you pick out images, or do you just chose images at random and crop them with no idea or plan behind it?
B. Do you decide on one crop right way, or do you crop the image in different ways before deciding on one.
C. Do you color the image before you crop or the other way around.
D. Do you use the crop tool, the rectangular marque tool/free transform or maybe both? If it's just one, why do you prefer that one over the other?
E. What are your thoughts on the different types of cropping, negative space, close crop, obscure etc. And what are some of your tips on each one, what to look for in images, how to crop to make it look balanced, how to not make it seem awkward, that sort of thing. And maybe you could include some example icon for each category and talk a little bit about how and why you decided to crop each in that particular way?

A. I usually just save pretty caps. Or the ones that look croppable, like half of a person's face behind another person's face or the close-up or the far off ones. Caps are never predictable though. Sometims I look at a cap and think that it's very croppable, but then I try it and it never works out. Or the crop works but the coloring doesn't. And when even black and white doesn't work on it, I just give up. And sometimes caps can surprise you. When you can make a boring cap interesting, you know you've got a good crop.

B. It depends. Sometimes I know right away what I want. Then I do it and get on with it. Sometimes, like above, I have a pretty cap but I'm having trouble on how to crop it, I do try a couple different things. Which is where the History panel comes in handy, hehe.

C. I crop first, then color. If you're using a full-sized cap or even something 200x200, you can't test your 100x100 textures yet in full quality. And even then there's always a chance when cropping it down to the usual 100x100, it might not look as good anymore as they did before. I think it's just a personal preference.

D. The crop tool, as explained in wandererjulia's question.

E. I personally think that, if YOU think a crop's good, then it's good. Here's a couple I'm personally not a fan of though:

1.

The case of: I really want to make a negative space icon of this cap but there's already a background and I can't just smudge it out. Solution: get your Rectangular Marquee tool and extend it that way. Color it and done. Problem is: you can still see where they obviously extended it. What's the trouble in actually trying to hide that somewhat? You can go a long way with some textures and such. I've done such a thing in this one for example:

I mean, it could've been better. Much better. But it's not as noticable anymore, no? That's just what textures and a bit of smudging can do.

2.

Cropping off at the mouth so the rest of their head is visible? I'm not a fan of it. It can look good sometimes, every crop can depending on the cap, but in this case? No.

3.

Sometimes you have a cap but you don't really know how to crop it, so you just crop the whole face and be done with it. I know I've been guilty of this in the past, but it just looks really lame and bland. It just doesn't come off as interesting at all. Even when you have the most amazing coloring and texture work in the world, the crop makes the difference. It also depends on how close or far away the subject in your icon is.
Also, in the second icon, apart from the reasons I mentioned above, it's also way too close and I always do find it important to keep some of the person's hair in the icon. It makes a huge difference sometimes.

And now, the crops I do actually like! Ya hey!

CENTER CROPS
Some love them, some don't like them. I'm on Team number one. I love love love them.

CUT OFF SIDES OF THE FACE

MOTION CROPS
Whether it be moving hair, or a person running, or anything. It's a captured moment of something happening, it's interesting on itself. Look out for blurry hands or anything though, those never look good.

NEGATIVE SPACE
Because who doesn't love it, really?

CLOSE CROP
For this kind of crop, be sure to still have a piece of their hair and to not cut off their chin.

CUT OFF AT THE EYES
This one is a personal favorite, I have to admit. It's much preferred to cutting off at the eyebrows. Or worse, in the middle of their forehead. examples

OBSCURE CROPS

'NORMAL' CROPS
Sometimes the coloring really does make the icon instead of the crop. :P

Well, I hope this was of use for at least someone, because it really took a long time to write. :)

Cutting Out Images Tutorial

2011-05-13T03:00:55Z

Program(s)+version: Photoshop CS4.
Involves: Different techniques for cutting things out and making selections.
Translatable: Should translate to all CS versions of PhotoShop.
Steps: 3-5 for each tool.
Difficulty: Beginner to Intermediate

HOW TO CUT THINGS OUT IN PHOTOSHOP

I see people sometimes asking how to cut things out in PhotoShop and I know to some it seems like it would be obvious (LOL eraser tool duh) but I thought i'd make a ~~monster-sized~~ tutorial showing some different (and more effective, imo) ways to cut things out.

+++

PREP:

I usually go ahead and edit/color the image I want to cut out, since i'll probably be using one of the background colors as the new background after I cut it out. You can do whatever you want though. If the image is really dark and you can't really see the outer edges of the figure you're cutting out, I suggest creating a curves layer (that you will later delete) to overly brighten the image so you can see what you are doing.

Now, once you've done whatever editing you want to do, make sure you turn your background layer into an editable layer. You can unlock it by double clicking it and pressing OK on the dialogue box. Or you can just create a duplicate later (Layer } New } Layer via copy OR Command+J [It's better to use this tool instead of copy/pasting because by duplicating you aren't putting anything on your clipboard aka using memory]) and hide the background layer.

Now once the layer is unlocked/editable, you can begin the cutting out process. Here are a few different techniques. *Note: I won't even be going over the Magic Wand Tool since I think it's a horrible way to make selections. Moving on...

ONE: LASSO TOOLS

I don't usually use these tools if I can help it, since it's harder to get the exact cuts I want. But the magnetic lasso tool is pretty cool and I always use the polygonal lasso on geometric shapes.

Free-Hand Lasso: I won't even go into the free hand lasso, since it's pretty self-explanatory and not the best method if you want a clean cutout.

Polygonal Lasso: This isn't going to work the best on this particular picture, since it's a person and the curves are all organic. But keep in mind that this tool works great on more geometric, straight edge objects.

Magnetic Lasso: You could actually get a decent cut with this one. First, I would set the feather either 0 or 1. Not really certain on the other settings, but mine are highlighted below.

Now, simply click and release once and drag the mouse along the outer edge of the object. Anchor points will automatically be placed down as you drag.

And if you want to make sure an anchor stays put, just click and then keep dragging and it will stay in the place you clicked. I think with this method you will have to do some cleaning up afterwards since it's hard to be exact.

When you are done, double click to make the selection and you'll get the marching ants.

From this point, you can either inverse the selection (Select } Inverse OR Shift+Command+I) and then delete the background. OR if you want to keep your background for future use, just make the selection a new layer (Layer } New } Layer via copy OR Command+J).

+++

TWO: ERASER TOOL

I used to just do this all the time because I didn't know a better way. And there's nothing wrong with this method at all other than the fact that you are deleting information and that can sometimes be problematic if you accidentally erase something you decide you want back later on. Now, this method is full-proof for some people, but it's hard to get clean lines with this. You have to have a really steady hand and a lot of patience. To make things easier on myself, I usually take the polygonal lasso tool and cut out as much of the picture as I can.

After that, I zoom in and select my eraser tool. I usually use a pretty small brush with the hardness set to around mid-point. The less hardness, the more of a feathering effect you will get. The more the hardness is up, the more sharp the edge will be.

Then I just erase around her figure. This is pretty hard and tedious to do with a mouse. Fortunately for me, I have a bamboo tablet. So it's not too bad and it's fairly quick.

+++

THREE: MASKING

I was really scared of masking before I learned what it was. I actually find more use for it in editing portions of a selection rather than making selections. Here, i'll show you!

First, make sure your foreground and background colors are at default (black foreground, white background).

Then click on your brush tool and use similar brush size and hardness as you did with the eraser. (smaller radius [i'm using 12px], hardness around 50% or higher) You want a nice firm yet soft edge. Now click your mask tool button.

Now with both your brush tool and mask tool buttons selected, paint out the selection you want. In this case, i'm selecting Amy's hat. Make sure that black is your foreground color when you paint. It should paint a RED color overlayed at about 80% opacity by default. If you accidentally paint something you don't want, just flip the white to your foreground color and paint on what you want to take away. The white will remove the red color, thus excluding that part of the selection.

Once you are finished, click your mask tool again (or press the letter Q) and you will get marching ants around your selection.

Now you can edit the selection however you want. With an adjustment layer (Color Balance, Selective Color, Gradient Map) maybe? Whatever you want and it only affects what you painted red.

I'm not going to do a selection of her entire body since you basically get the idea from the hat.

+++

FOUR: PEN TOOL

I looooove the pen tool. Mainly because I like the firm smoothness of the edge. And it's nearly impossible to make a shitty cut-out with this tool (IF you use it right, that is). I'll try to explain it as best I can. But if you master this tool, it will be your best friend as far as making cutouts go. The beauty of the pen tool is you can click and drag and create anchor points, but they don't have to just be straight lines. You can make beautiful smooth curves.

First, open your PATHS (Window } Paths) and then select the pen tool (P).

*I suggest practicing on a blank canvas beforehand to get the hang of this*

Now make your first point. Anywhere is fine. Hold the mouse down with one click, then release and click to create a second point point a little bit further away (DON'T UN-CLICK). Now (without un-clicking) drag the mouse. This will give you some handle bars and will start curving the line you just made. Experiment and drag the mouse until the line of the pen matches the line of the object. Once it does, you can release the mouse.

Now, at most points you can just keep going and click to make your next point. But depending on how far you had to drag and how long your handlebars have gotten, you might want to hold the OPTION key down and click the last anchor point you made to delete the handlebar. This keeps the line from going all wonky (especially when you are going around sharper corners [like the points on the hat]).

I'm sorry if i'm just confusing the mess out of you. But if you click around a little bit and make some practice lines first, you will see what i'm talking about. As always, if you make a line you don't like and want to re-do it, just press Command+Z (go back a move) and try again.

Once your selection is done it will just be the pen tool line (no marching ants yet). To actually create the selection ants, you need to go to your PATHS window that you opened earlier and click the 'load path as selection' button:

Clicking this button will create the selection. (I don't know what the default is, but in my PS this button selects the INVERSE of everything I selected. So just go to Select } Inverse and it will flip). From here, you can duplicate (Command+J) the selection and it will go to a new layer and VOILA, you are done. You can, as always, erase any part of the selection you don't want at this point. I find that this tool gives me the best consistency of line and looks the nicest. It takes some practice though.

Okay, now that i've thoroughly bored you to death i'm finally finished! I hope this was kinda sorta helpful. :)

Coloring and blending tutorial

2011-03-28T22:24:08Z

How to go from this and this to

Program(s)+version: Photoshop CS3 (but most likely works with PS 7 and above too)
Involves: Curves, color balance, levels, layer masks, textures
Translatable: Mostly translatable; has a few selective coloring steps that are completely skippable
Steps: 57 (oh good lord)
Difficulty: Moderate/difficult -- I tried to make it clear where things are and how to find them, but the tut still assumes some basic knowledge of PS.

First, I crop and resize the first screencap:

I duplicate the layer (Ctrl + J) and set it to Screen, 50% opacity.

To give it some contrast, I duplicate the layer again and set it to Soft Light, 70% opacity.

Next, I crop and resize the second screencap, and paste it on top:

Obviously, there's some stuff I want to get rid of:

So I use a layer mask, using a black brush to paint over the parts I don't want:

-->

The second cap is really dark, so I duplicate the layer twice, setting the first to Screen, 100% opacity and the second to Screen, 20% opacity.

To create some contrast, I duplicate the layer one more time and set it to Soft Light, 50% opacity.

I want to add a yellow-ish tint to part of the icon, so I use a Layer > New adjustment layer > Color Balance layer:

Midtones: -19 -5 -14
Shadows: -18 -11 -20
Highlights: +15 +19 +9
Make sure "Preserve Luminosity" is checked.

But I'm only really interested in adding that yellow tint to <-- this part of the icon, so I use Layer Mask again.

-->

To brighten up the colors, I use a Layer > New adjustment layer > Curves layer:

RGB - Input: 122, Output: 143
Red - Input: 117, Output: 131
Green - Input: 125, Output: 131
Blue - Input: 255, Output: 255

Again, I'm only really interested in applying this to that bottom-right half, so I use another Layer mask:

Now I start filling up the surrounding white space. I first go back to the first screencap, crop out a piece of blue sky with clouds, and paste it on top of the icon:

-->

I use another Layer mask to get rid of the parts covering Juliet screencap #2:

I copy and paste the piece of the first creencap again, moving it to the left to cover up the white space there too.

-->

I use a Layer mask again to get rid of the parts covering Juliet:

There's still a sliver of white space in the top-left corner, so I copy the little piece of sky/tree, paste it, and then smudge the blue parts using a soft smudge brush at approximately 36 px diameter:

--> -->

Again, I use a Layer mask to erase the parts covering Juliet:

I want to cover up some of the white space here:

So I copy a little bit of sky again and paste it there:

-->

I use a Layer mask to remove the sides and corners:

Finally, I copy a bit more of the sky and paste it on top:

-->

I used another Layer mask to remove parts of it to make it look more blended:

Now that the white space is covered, I can work on getting the coloring where I want it. I first use a Layer > New adjustment layer > Levels layer:

-->

To brighten up the top-ish half of the icon (here: ), I use another Layer > New adjustment layer > Curves layer:

RGB - Input: 116, Output: 134
Red - Input: 122, Output: 132
Green - Input: 124, Output: 130
Blue - Input: 255, Output: 255

... with a Layer mask:

Then I use Layer > New adjustment layer > Color Balance to brighten up the icon:

Midtones: +32 +26 +17
Shadows: +7 +8 +5
Highlights: +20 +14 +12
Make sure "Preserve Luminosity" is checked.

Then I paste the following texture, setting it to Soft Light, 50% opacity.

-->

To brighten the icon a tad more, I fill a new layer with #DCC18A, setting it to Soft Light, 20% opacity.

-->

I create another fill layer with #FFB594, also setting it to Soft Light, 20% opacity.

-->

I then paste the following texture, setting it to Soft Light, 30% opacity.

-->

It looks a bit plain, and I want to add something artsy to it. So I paste the following texture, setting it to Multiply, 100% opacity.

-->

I'm only really interested in keeping this part of the texture: , minus the stuff closer to the center too. So I use another Layer mask:

I want to make everything brighter, so I use a Layer > New adjustment layer > Curves layer again:

RGB - Input: 123, Output: 135
Red - Input: 127, Output: 135
Green - Input: 124, Output: 130
Blue - Input: 255, Output: 255

I then use a Layer > New adjustment layer > Selective Color layer, set at 70% opacity to kind of enhance the reds/yellows. This layer doesn't actually make much of a difference though, so you can skip it.

Red: -13 +9 +9 +0
Yellow: -17 +12 +11 +0

I use another Layer > New adjustment layer > Color Balance layer to brighten the colors some more. (This is also where I realize, too late, that all my coloring steps are redundant. XD)

Midtones: +21 +19 +7
Shadows: +25 +24 +21
Highlights: +23 +19 +19
Make sure "Preserve Luminosity" is checked.

Then I merge everything down (Ctrl + Shift + E) and sharpen it.

I want a little more... diversity in texture, I guess, so I added a little gray streak, set at Normal, 40% opacity.

--> -->

Next, I want to illuminate the icon with a light texture, so I go to one of my favorite light textures, rotate it, blur it with Gaussian Blur, 5 px, and set it to Screen, 100% opacity.

--> --> --> -->

I don't want to obscure Juliet, and I want to make the light look like it's kind of swirling around her, so I use another Layer mask:

-->

I rotate the texture again and paste it, setting it to Screen, 100% opacity. This gives the icon some light in the bottom-left corner.

-->

I use another Layer mask:

-->

I then duplicate this layer, setting it to Screen, 30% opacity.

I use the light texture again, cropping and rotating it like so:

I paste it onto the icon and set it to Screen, 30% opacity.

-->

The icon is looking a bit washed out with all the light textures, so I paste the following texture and set it to Soft Light, 50% opacity.

-->

I add another Layer > New adjustment layer > Levels layer:

I only really want the layer to be applied to this part of the icon:

So I use another Layer mask, filling it in with black and using a white fuzzy brush to paint over the spot where I want the levels layer to show up:

-->

There's a white/washed-out part here that I don't really like:

So I paste the following texture, setting it to Hard Light, 100% opacity, and using a layer mask to erase most of the texture except for the part I want:

--> -->

Now I use my last icon-brightening texture, setting it to Soft Light, 80% opacity.

-->

I use a Layer mask to remove the texture from the right half of the icon, where things are starting to look too magenta:

Here I use Layer > New adjustment layer > Selective Color again, but again, it's optional. I just want to enhance the blue in the top-right corner of the icon:

Cyans: +100 +100 +100 +100

I duplicate this layer twice, setting the first to 100% opacity and the second to 50% opacity.

I use a little white blob, set at Screen, 30% opacity, to add some light near Juliet's forehead.

-->

In the corner of the icon, it looks a bit bumpy/rough:

So I merge everything down (Ctrl + Shift + E), duplicate the layer, and smudge that part of the icon.

So that's it!

I hope this was helpful and/or not too long and boring! Please let me know if there are any problems or mistakes, or if anything is unclear. Thank you! :D

Icon Tutuorial: Extending Background, Colouring & Texture Use

2011-03-02T22:42:12Z

This to
Program(s)+version: Photoshop CS5
Involves: Extending Background, Curves, Levels, Selective Colour, Colour Balance, Hue/Saturation, Filters & Texture Use
Translatable: To programs/versions that have selective color available.
Steps: 6 Steps | 39 Layers
Difficulty: Beginner - Intermediate

Step 01. Cropping your image

Note: I do Step 1 & 2 in various degrees in almost every icon that I work with. I find that if you work on making a base smoother and softer (no matter the quality of the image) then the overall icon is much easier to work with.

• I usually work with my images at 250px by 250px, just because I like the space that it gives me to play with, and when finishing an icon I find it much easier to work with that size and avoid over-sharpening anything.

• First, resize the image (Image > Image Size) to Width: 534 px, Height: 300 px. Now we are going to copy the image, and paste it into a new file.

Ctrl + A (Select All) > Ctrl + C (Copy) > Ctrl + N (New) > Image Size: Width: 350px, Height: 350 px > Ctrl + V (Paste).

• We now move the image we’ve pasted into the center of the blank canvas. We want Howard Bamboo (the central character) to be in the center of the image, and the other two characters to the left and right of him. Once that’s done, we now need to extend the background.

• There are various ways in which you can extend the background. Because of the detail in the top half of the canvas, I’m going use a copy, paste and flip method.

Using the rectangular marquee tool, select the amount of the image you want to copy, like I have here. Then Copy (Ctrl + C) and Paste (Ctrl + V) the selection.

We then need to transform the image so that the yellow semi-circle is flipped and reflected, creating a circle. To this, we go Edit > Transform > Flip Vertical. Use the Move Tool to push the image upward, until the flipped image looks like this.

There is still a little bit of empty space at the top of our canvas, so we simply Paste (Ctrl + V), and use the Move Tool to push the image upward again, until the image looks like this.

• Now we resize the image to 250px. Image > Image Size > Width: 250 px, Height: 250 px.

• Lastly, we need to merge all the layers, so our base is one whole image. To do this, make sure you right click on the image that says ‘Layer 1’ in your Layers toolbar. After this, you need to go Layer > Merge Visible (Ctrl + Shift + E). Now you have one complete image.

Step 02. Preparing your base

Layer 1: Base Layer. Remains untouched.
I always leave my first layer untouched. If I stuff something up and need to go back to the start, it’s nice to not have to crop again, especially if you’ve got a really interesting/unique crop going. It’s also handy to have it on hand if you want to use your base to add an extra element to the colours of an icon later on.

Layer 2: Duplicate the base layer. Filter > Blur > Smart Blur > Radius: 3.0, Threshold: 3.0. Set this layer to 50% opacity.

This can be interchanged with Surface Blur if available. I like the effect Surface Blur/Smart Blur has on softening the features of an icon, makes it smoother in general, and sometimes (especially for lower quality images or darker images that seem pixelated) that’s a real bonus.

Layer 3: Duplicate of the base layer, and drag it to the top. Filter > Blur > Guassian Blur > Radius: 0.5. Set this layer to 50% opacity.

Again, we’re trying to blur out any of the flaws and faults within the icon to make the base something nicer to work with. This will soften the edges of your image further.

Layer 4: Duplicate the base layer, and drag it to the top. Filter > Blur > Smart Blur > Radius: 3, Threshold: 3. Set this layer to 50% opacity.

Layer 5: Ctrl + A (Select all) the image. Ctrl + Shift + C (Copy Merged) the image and Ctrl + V (Paste) it as a new layer. Set this layer to screen > 100% opacity.

When you copy a merged image, you are copying the sum total of the previous layers together, as opposed to just one layer. This screen layer’s opacity could be anywhere from 10% to 100%, depending on how dark the image you started out with originally was, or how bright and vibrant you would like your icon to be.

Layer 6: Layer > New Adjustment Layer > Curves > RGB > Output: 147, Input: 117.

In this step, I’ve used curves simply to lighten up the image. Later I’ll use it to add some colour, but while I’m preparing the base, I just want to use the RGB layer to add more lightness to the icon.

Layer 7: Layer > New Adjustment Layer > Levels > RGB > 0 | 1.10 | 245.

For the same reason as Layer 6, Levels have been used here to help lighten the image. I generally use this setting, but if the image is especially dark, I’ll lower 245 to 235 to give it that little bit more brightness.

Layer 8: Layer > New Adjustment Layer > Brightness/Contrast > Brightness: -10 | Contrast: +10. Set this layer to 100% opacity.

This layer is used to add some contrast and depth back into the image. In this image that we’re working with, there’s originally not a lot of contrast to the image, so we leave the opacity at 100%. But if there was more contrast in the original image, lowering the opacity often works to add just enough depth to the image.

Layer 9: Duplicate Layer 8. Set this layer to Soft Light, 40% opacity.

This layer is used to add the same kind of contrast to the image, and also helps to distinguish the various colours in an image, and add contrast to them as well.

Step 03. Colouring your image

Layer 10: Layer > New Adjustment Layer > Curves > RGB > Output: 134, Input: 115

This layer with brighten the entire image – I mainly wanted to add a little more lightness into the foreground to highlight the characters.

Layer 11: Layer > New Adjustment Layer > Hue/Saturation > Master: Saturation +15

This layer adds a little bit of saturation to the colour in the image. In this particular image, there is a green overlay that I wanted to highlight, so I used Saturation though out the process to keep making it prominent. I also recommend using Vibrance if you have a version which has this tool.

Layer 12: Layer > New Adjustment Layer > Curves:
RGB Point 1: Output: 70, Input: 53
Red Point 1: Output: 187, Input: 197
Green Point 1: Output: 107, Input: 85
Green Point 2: Output: 195, Input: 159
Blue Point 1: Output: 83, Input: 101
Blue Point 2: Output: 199, Input: 173

Set layer to Multiply at 50% Opacity.

The image is lacking in contrast, so this curves layer works to achieve two results. The first is to add colour into the image, the second is to add contrast to this colour, and darken the image without taking away the work of the previous lightening layers.

Layer 13: Layer > New Adjustment Layer > Levels > RGB 0 | 1.10 | 235

We do loose a touch of the brightness by adding the previous layer, and this Levels layer helps to lighten the image and create more contrast between foreground and background.

Layer 14: Layer > New Adjustment Layer > Colour Balance

Midtones:
-20 | +15 | -10
Shadows:
-15 | -5 | -5

Set layer to 60% opacity.

With this layer, we are really trying to bring out the greens and yellows of the image, by highlighting them (using cyan, green and yellow hints) and taking away from the reds. However it’s the colours felt too strong at 100% opacity, so we drew it down to 60 to keep the highlights without overwhelming the icon.

Layer 15: Duplicate Layer 14. Set this layer to Soft Light, 30% opacity.

We do want to maintain a nice contrast however, and by duplicating the layer below, we achieve a nice distinction in contrast as well as continuing to highlight the yellows and greens in the image.

Layer 16: Layer > New Adjustment Layer > Selective Colour

Reds –
Red: +5
Magenta: - 10
Yellow: + 10
Black: + 15

Yellows –
Red: +5
Magenta: - 10
Yellow: + 40
Black: -10

Greens –
Red: +5
Magenta: +100
Yellow: - 40
Black: - 50

Cyans –
Red: +100
Magenta: - 100
Yellow: -100
Black: +100

Neutrals –
Red: +5
Magenta: -6
Yellow: - 16
Black: +5

Blacks –
Black: + 10

With this selective colour layer, in addition to highlighting the green, we want to draw away some of the yellow, and highlight some cyan in parts of the image as well. In doing so, we have a mixture of greens, cyans and yellows in the foreground and background to work to compliment each other.

Layer 17: Layer > New Adjustment Layer > Curves

RGB Point 1: Output: 194, Input: 179
Red Point 1: Output: 77, Input: 92
Red Point 2: Output: 191, Input: 182
Green Point 1: Output: 71, Input: 65
Green Point 2: Output: 142, Input: 121
Blue Point 1: Output: 124, Input: 99
Blue Point 2: Output: 176, Input: 205

Set layer to 50% opacity.

I wanted to soften the image colours a little, so with this layer, I added a touch of yellow to the skin tone to make it softer. It was important to lower the opacity however, because otherwise it would have overwhelmed the image. I also wanted to continue to highlight the greens and yellows of the background/foreground.

Layer 18: Duplicate layer 17. Set layer to Soft Light at 30% opacity.

As you probably know now, as much as I like softness, I enjoy contrast. And this layer adds back the contrast without detracting from the yellows and greens.

Layer 19: Layer > New Adjustment Layer > Gradient Map: Set to Hard Light at 15% opacity.

This layer further develops the contrast, but also removes a touch of the colour. This is okay though, because the intensity will be achieved again through the textures we are about to use.

Step 04. Texturizing your image

Layer 20: Texture by elli. Set layer to Screen.

I love this texture for two things: it’s blue tinge, and it’s textured feel. I think it offsets this image well, because it works with the colours that already exist, and highlights the blue rather than black parts.

Layer 21: Texture by drankmywar. Set layer to Soft Light at 20% opacity.

There are two problems that need to be rectified with the image. The first is that, as it is 100x100px in size, it is too small for our canvas. To fix this, we need to ‘Free Transform’ the image, and drag it out to fill the canvas.

Edit > Free Transform (Ctrl + T): Always remember to hit apply when the program asks you to Apply the Transform.

The second is the colour. Although the yellow highlights aren’t bad, right now I’m more interested in concentrating on the blue highlights, and therefore I want to invert the colours of the texture so it is blue, not yellow. To do this, I hit Ctrl+I, and the colours magically change.

TIP: This is something I play around with when it comes to most textures. Sometimes they have some amazing colours when inverted that just highlight an image in a manner that cannot be achieved otherwise.

Layer 22: Texture by drankmywar. Set layer to Screen at 20%.

With this layer, I want to highlight the green of Howard’s shirt, so I need the majority of the green to be vertical from the bottom up, rather than horizontal. To achieve this: Edit > Transform > Flip Image 90 degrees clockwise – and then Ctrl + T (Free Transform), and drag the image out so it covers the whole canvas. Now the green highlights the right part of the image.

Layer 23: Texture by drankmywar.

With this layer, I want to highlight parts of the blue at the top (where there is negative space), and parts at the bottom where there is green and yellow. To achieve this, I need to transform the image (Ctrl + T), however I don’t want to cover the whole image. I drag it so it covers the image vertically, however horizontally, it only takes up two thirds: See here.

I set this layer to Screen at 20%, so the green and yellow can be lightened, but the light texture doesn’t overwhelm the image.

Layer 24: Duplicate Layer 23. Edit > Free Transform >Flip Horizontal.

I now want to highlight the colours on the opposite side of the image. To do this, I duplicate layer 23, flip it horizontally, and then drag it across so it reaches the opposite edges to layer 23: like this. I keep the layer set at Screen on 20% opacity, as I want to keep the same kind of highlights on both sides.

Layer 25: Texture by elli.

Again, this texture is one that I love because of it’s textured feel, but to work with this image, we need to invert the colours to highlight the greens, rather than pinks. Ctrl+I the image, and then set the layer to Soft Light at 15% Opacity.

Layers 26: Texture by drankmywar.

This is where it gets fun! I want to highlight the edges of my image with the yellow section of the texture. So the first thing I need to do is stretch the texture so the yellow is large enough to cover a significant amount of the 250px edge. I don’t want to show you this part step by step, because this is the most guessed part of the icon. It begins with Ctrl+ T though, and ends with me shifting the texture towards the right side of the icon.

Once it’s covering enough for my liking (and I’ve hidden the red completely) , I set the layer to Screen at 100% opacity.

Layer 27: Duplicate Layer 28. Edit > Transform > Flip Horizontal.

Now I want to highlight the left side of my image, so I drag the texture across my icon, so it’s now only covering the left edge, and only the yellow is visible. I leave this layer on Screen at 100% opacity.

Layer 28: Duplicate Layer 27. Edit > Transform > Flip 90 degrees clockwise.

Here is where things get tricky. Firstly, you need to drag the image upwards so the yellow is covering the majority of the blue negative space around Ken (the male figure to the left of centre). The red of the texture is now visible, so we need to get rid of that. Use the Rectangular Marquee Tool (The little box of dotted lines in your toolbar) to select all of the visible red, and then hit delete. You’ll delete into some of the yellow, but that’s okay.

Duplicate Layer 28, and Edit > Transform > Flip Vertical. More red will appear, as will an empty space between both yellow textures. Use the Rectangular Marquee Tool to again select the red that is showing, and delete it. Then move the top yellow layer down until it meets with the bottom yellow layer.

Merge both layers together (Ctrl+ E), and then set this layer to Screen at 35% opacity. It may appear like a lot of work, but this one small layer breaks up the blue in the negative space (without overwhelming it) and also joins it to the green hues of the whole image.

Layer 29: Duplicate Layer 28. Drag this layer upwards into the top right hand corner. Set this layer to Screen at 50% opacity.

Layer 30: Duplicate Layer 29. Edit > Transform > Flip Horizontal. Drag layer into the top left hand corner. Set this layer to Screen at 30% opacity.

Layer 31: Duplicate Layer 20. Drag this layer to the top of your layers.

Drag texture sidewards until a semi-white strip appears on the left edge of the image. Set this layer to Soft Light at 40% Opacity.

After adding all the light layers to the image, we need something to bring back the contrast. This layer adds to the contrast of the image without detracting from the colour we’ve added to it.

Layer 32: Duplicate Layer 25. Drag this layer to the top of your layers. Set this layer to Soft Light at 50% Opacity.

Now we’ve got a little too much green in the foreground, and I want to erase some of the texture off the characters without taking away the strength of the background. Select your Eraser Brush, and set the Opacity to anything between 50% - 75%. Now erase the texture from the three characters once – you don’t need to be too careful, but you just want to erase enough to bring back some of their natural skin tone.

Layer 33: Layer > New Adjustment Layer > Gradient Map > Set layer to Soft Light at 30% Opacity.

This is to do two things. The first is to add my beloved contrast back into the darker areas of the image. The second is the tone down the colours, because of the next layer we are going to be adding.

Layer 34: Layer > New Adjustment Layer > Hue/Saturation > Master: Saturation +30 > Set this layer to 80% Opacity.

There’s a fine line between eye-burning brightness and saturation, and just right brightness and saturation, and without the prior and following layers, not to mention lowering the opacity, this layer could have been the first option. So when you are playing around with Saturation/Vibrancy, remember that sometimes, less is more.

Layer 35: Duplicate Layer 31. Drag this layer to the top of your layers. Set this layer to Soft Light at 50% Opacity.

This layer adds to the vibrancy of the colours without making it too intense. However I feel as though I need to keep the full strength of the texture in the background, but erase some of the foreground to distinguish between them more. This time, I set my Eraser Brush opacity to 50%, and erase over the three characters just once. It takes away just enough of the contrast.

Layer 36: Texture by elli.

Drag this texture so the white is centered, and then invert the image (Ctrl + I). Set this layer to Soft Light at 20% Opacity. Again, we just want to distinguish between foreground and background, and also add a little more colour back into the foreground.

Layer 37: Duplicate Layer 31. Drag this layer to the top of your layers. Set this layer to Soft Light at 25% opacity.

Another final touch to the contrast for me, as it adds that little bit more intensity.

Step 05. Adding the finishing touches

Now that I’ve coloured and textured the image, I’m ready to finish the icon.

These next couple of steps are always involved when finishing off an icon for me. Like preparing an icon base, I think that when you’re finishing an icon, it needs to look finished, rather than haphazardly thrown together.

Before we do anything though, I always re-size the image to 200 px by 200 px. Just because I’m neurotic, and I think the blur/sharpening combo works best on that size image before it gets resized to 100px.

We therefore need to do this:

Image > Image Size (Alt + Ctrl + I) > Width 200 (pixels), Height 200 (pixels).

Layer 38: Ctrl + A (Select All) > Ctrl + Shift + C (Copy merged) > Ctrl + V (Paste) > Filter > Blur > Guassian Blur > Radius 1.0 > Set this layer to 30% opacity.

Again, I use Guassian Blur to soften the image before sharpening the edges. I think the sharpening effect just looks a little bit softer if this is done, as it takes away that absolute sharp edge that can sometimes occur in an icon.

Layer 39: Ctrl + A (Select All) > Ctrl + Shift + C (Copy merged) > Ctrl + V (Paste) > Filter > Sharpen > Sharpen. Set this layer to 75% opacity.

Depending on the image you began with, the sharpen layer will be anything between 10% & 100% opacity. And while you’re image may look a touch too sharp now – remember we still have to resize the image to 100 px by 100 px, which will also take away a little of the sharpness.

You have to be cautious here, because sometimes sharpening can lead to rough, harsh edges. It’s always good to have a little bit of leeway on each side.

We’re finally ready to resize the image: Image > Image Size (Alt + Ctrl + I) > Width 100 (pixels), Height 100 (pixels).

Step 06. Saving your icon.

File > Save As (Ctrl + Shift + S) > Add a file name, and click down your format options to png.

PLEASE NOTE: ALWAYS SAVE YOUR ICON IN PNG FORMAT. I cannot stress this enough, hence the attack of the CAPS. Saving in jpeg decreases the quality of the icon dramatically. Saving as png maintains the higher quality of the icon, and makes it nicer to look at.

Once you’ve saved your icon, you’re done! (Which is the most *headdesking* thing to say, but it’s true!)

--

Endnote: I didn't realise how intense this tutorial would be when I began to write it - but my texture use was a little insane in there, so please don't hesitate to ask me questions if any of my explanations aren't very fluid. I hope someone somewhere finds it useful though, and that it encourages some of you to work with textures a little more, because they are so much fun!

As always, comments are love. Please let me know if you found this useful, or alternatively, hard to follow!

icon tutorial: blending

2011-02-27T20:15:32Z

Program(s)+version: Photoshop CS4
Involves: textures, blending modes; blending several images
Translatable: yes
Steps: 20
Difficulty: Moderate

Learn how to make this icon:

This icon is a collage of several caps, blended together.
These are the caps we're going to need: 1 2 3

Made with Photoshop CS4. It's only using basic blending modes, so it should be translatable to all programs.

1. Preparing the images

For this kind of icons I prepare the images I want to blend separately, for various reasons:

Even if the caps are from the same scene and have similar coloring, they still vary in nuances. It's easier to color each image and adjust adjustment layers for this single picture than trying to get everything right for three images.

If something doesn't work out the way I want later, I always have the source image ready. Replacing one image is a lot easier than fussing with several layers just to get back to a certain point.

So, the overall hue for these is yellow. I like yellow, so I'll go with that and just enhance the colors.
Most of the time I'm too lazy to make drastic changes in colors. Drastic changes also often result in quality loss, and that's the least thing you need for blending images.

First image

For the base image I cropped the image to 100x100.
I want to blend several images, so I'll need a lot of blank space to add the other images, so I cropped it with a lot of negative space around Joe.
Don't worry that the cap isn't "big enough" and that you'll get a stripe of black at the top. This is part of the space we'll need later for adding the other images. For blending images this part needs to be as dark as possible, so always make sure that your background color is set to black before you start cropping, and the stripe will appear in black.

After cropping I simply duplicated the layer and set it to Screen 100%.

+ -->

Save your image as .png.

second image

This is the image I want to place on top, so I cropped it just like the first one, just with the black part at the bottom.
I used the whole cap and not just parts of it, so of course you can crop it differently and just place it on top later. However, I like the smooth lines that appear at at the bottom and top of a cap if you it this ways, it makes blending a bit easier and the blending lines less harsh.

Again, duplicate you layer and set it to screen 100%.

+ -->

I want to stick with the yellow color, but also add a little something to make it a bit more interesting. For that I like to use textures, so I chose a texture by deny1984 that has yellow and a different color in it.
Paste it as a new layer.

For coloring with textures the only way is to experiment with blending modes. There's no magic or right solution, all I do is go through several modes until I find something that looks good.
Here I settled on Linear Burn, because it adds color without adding too much contrast (like Color Burn or Overlay would do).
I intend to do some additional coloring after blending everything later, so I keep this fairly tame and lowered the opacity of the texture to 75%

+ -->

It turned out a bit dark, so I'm adding a Brightness layer (Layer > New Adjustment Layer > Brightness/Contrast) and increased the brightness until I like it, here 24.

-->

We're done with this image, save it as .png.

third image

The first two images are going to be placed on top of each other, so the middle space of the icon is already occupied and additional images can only be added at the sides.
The cap is a partial face shot, so the ways to crop it with lots of space are limited. Here I chose to put it on the left side of the icon, so I need space on the right side.

The coloring is exactly the same as in the second image, so I'm not repeating the process. Just repeat the steps from the second image.

The only difference is that I rotated the texture 90° clockwise (Edit > transform > Rotate 90°CW), because the purple part on his face looked weird. It should look like this.

This is your result:

Flatten your layers (Layer > Flatten Image).
Now grab the Smudge Tool from your tool box and set it to a soft round brush, around 9 px. Smooth the right side until it looks like this:

We're done with this image, save it as .png.

2. Blending

Go back and open your first two prepared images. Grab the second with the Move Tool and drag it onto first one, until it appears as a new layer.

Now comes the blending part. I do that by setting the top image to Lighten.

Lighten works like this: In your top image everything that's lighter than the color in the image beneath it will show up, everything that's darker will disappear.

That means:
- Joe's face in the second image will show, because it's lighter than the black stripe in the first image.
- The black stripe in the second image will disappear, because it's darker than the image of Joe in the first one.

You result should look like this:

As you can see, a tiny part of Joe's neck (light) is overlapping his hair (dark) in the first one and shows. Add a Layer Mask (Layer > Layer Mask > Reveal All) to the second image and paint over that part with a soft round brush in black, and it will disappear.
You could also erase that part, but layers asks are much more convenient because the part you painted over isn't actually gone. You can always go back and make it visible again by painting over it in white and it will be there again. The general rule for layer masks is: white parts are revealed, black parts are hidden.

-->

Not bad, but the coloring is kind of pale. To add a more dramatic color, take the same texture you used before, paste it as new layer and set it again to Linear Burn. It looks good like this, so leave the opacity at 100%.

+ -->

The only thing that doesn't look good is that Joe's face in the first image is now very dark. Add a layer mask to texture and paint over this in a medium gray. You see it'll appear lighter.

--> You can't just use black on layer masks, but also grays. That way you'll only reveal what's underneath is partly. The darker the color you paint with, the more will be revealed.

-->

Now open the third image and drag in onto your canvas as a new layer. Set it to Lighten 100%.
There's a small black gap left to his neck, so move it a bit to the left and down until it looks better.

-->

By moving the image you'll lose the right part of it and a small stripe will show at the left where the layers under it aren't covered. Pick the Smudge Tool and smooth the black and yellow parts to the right until everything is covered.

-->

Note: this part of the blending wouldn't work without the texture. The texture on Linear Burn makes the lower left corner very dark, without it the yellow parts of the first image would be too light and Joe's face in the third image would be barely visible.
Click on the eye next to the layer with your texture to make it disappear, and you'll see what I mean.

The upper left corner is very dark. I want it to look exactly like the yellow parts in the lower half. To achieve that I'll actually use these parts of the icon:

Copy/merge everything as a new layer (Ctrl+Alt+Shift+E). Move it up until only the lower half of the image is visible.
The yellow part is on the right, but is supposed to be on the left, so flip it horizontal (Edit > Transform > Flip Horizontal)

To make it blend better, set it to Lighten 100%. Now parts of the image on top are covered. To fix it, add a Layer Mask to the flipped layer and paint on the right side with black until only the yellow part to the left is visible. Smudge in gray over the left side of Joe's face and hair, that way the yellow will be revealed at a low opacity without actually hiding his face.

Looking at it, Joe's profile at the bottom is kind of floating without showing his body. To fix it I will use a Soft Light layer with only the parts of the first image and the texture.
The reason why I don't simply duplicate the layer with the first image:
- it is very light at the sides, and setting it to Soft Light would cover Joe's face on the second image and make it too bright.
- merging it with the texture doesn't only make it darker, but also more colorful. This way I can easily enhance the colors.

In your layer palette, make everything but the first image and the texture invisible by clicking on the little eyes next to the layers. Select the layer with the texture and copy/merge everything as a new layer (Ctrl+Alt+Shift+E).
Your result will be this:

Drag it on top of your layers and set it to Soft Light 100%. Make all layers visible again by clicking in the now empty boxes next to the layers, so the eyes will show again.

The black part of the new Soft Light layer makes the upper part to dark, so erase it with a large soft round brush.

I noticed that the bar in the second image doesn't look good, so I smoothed it out with the Smudge Tool:

If you like the result, you're done. Flatten the image and sharpen if needed.

If you like light effects like I do, add a new layer under the soft light layer and paint some white lines with a hard round brush (~ 9px). Use Gaussian Blur (Filter > Blur > Gaussian Blur) on it until it looks good.
Set against a black background mine would look like this:

Leave this layer at Normal and lower the opacity until you like the effect. The Soft Light layer on top will make it automatically look like a light effect. Here I left it at 96%.

This is your result:

Your layer palette should look like this for the blending part:

I hope this wasn't too long-winded and complicated and actually helped a bit.
Any questions, just ask.

General Icon Making Tips and Photoshop Instruction

2011-02-27T19:37:02Z

Program(s): Photoshop (I use CS3 but certain parts will be translatable for earlier versions)
Involves: General icon making tips; including colouring, background extension, sharpening + more.
Translatable: Photoshop specific but certain parts are general enough for any program.
Steps: N/A
Difficulty: Beginner and upwards

Before anything else I want to stress that I always, ALWAYS, make the following adjustments on individual layers. See this cap for an example of what I mean. That way you can always go back and make changes to your 'adjustment layer'.

So first things first, we need an image! On the left you can see the original screencap of Lydia (with lurking!Josie) from Southland.

When I’m browsing thorough a folder of caps I am looking for pictures with potential. Two key features are: light and colour. I loved this particular cap because of the strong lightplay of highlights and shadow on Lydia’s face. Working with a flat image that has no immediate light source can be hard work.

Though this cap currently has a murky-green background I could see the potential to change this into a bright, pretty blue, using a few layers. I also loved how Lydia’s shirt was similar to the background colour, but seperated from the wall by her black jacket. This would allow her to stand out but also make a naturally easy colour-block style icon.

See example: right-hand screencap.

I find around 90% of the time that the original screencap I have chosen is currently too dark for the style of icon I like to make. A quick and easy way to brighten up your image is to use the layer style ‘Screen’.

Duplicate your background/base layer (a quick way to do this is to hold ctrl and press 'J'), and then in your layer palette look for your ‘layer style’ drop-down menu. Scroll down and click ‘Screen’. Then simply adjust the opacity of your layer until suitable.

Above you can see the first stage of my icon to the completed icon. As this is more of a guide to my process rather than a step-by-step guide how to make this exact icon I'm going to try and keep the explanation more generalised (bare with, I'm not good at this remember!)

So firstly I needed to blend across the plain background to cover up all the white and that icky yellow/green door by Lydia's shoulder. I do this easily using the 'Smudge' tool, choosing any of the standard Hard Mechanical round brushes and adjusting the hardness to around 80%. I have the strength at 70-90%. Then simply drag the background across, filling the space.

It is better to have your brush with a higher hardness because it decreases the radius of the cirle, meaning you are less likely to accidentally blur the edges of your subject which you are smudging around.

Once you've got a full background you might think it looks a bit rough; you can always adjust the softness of your brush (to say 30%) and the strength to 30-40% then sweep across the background and blend it together more softly. Just be careful you don't get too close to your subject or you may be smudging their edges without even realising (I have done this lots of times!)
Once it's a nice matte surface we can continue.

Next we go through our brightening-up process. As I said above a great easy way of doing this is with the 'screen' layer. A fancier way of doing it (but no less easy) is to use the 'curves' tool.

IMAGE > ADJUSTMENTS > CURVES or CTRL + M

I'm no professional so my explanation here isn't going to be gospel. Basically though the chart represents the light levels in the chosen layer. The chart reads so that the shadows are on the left, midtones in the middle, and highlights on the right. As you can see from this chart the light level is very dark, so I drag the line upwards over the empty area to brighten it. On icon you can see how the layer is immediately brighter and the highlights more prominent.

Here's a little introduction to the various colouring tools that I use regularly.

SELECTIVE COLOURING

(is your best friend)

Selective colouring is one of the most useful tools in Photoshop, and if you have it in other programs then that's awesome because it will definitely help.

With selective colouring you can pinpoint each colour within your image and adjust it accordingly. I always have my selective colouring set to 'relative' as it is more subtle than 'absolute', and will adjust with more detail.

COLOUR BALANCE

Colour Balance is less meticulous than selective colours but is a great way of increasing and adjusting certain colour levels. Maybe your image is too red? Drag the marker the other way to increase the blue and balance the colours out. It's easy and really useful, just play around and see what you get!

HUE / SATURATION

I always feel like a cheating cheater from cheatsville when I use hue/saturation, but there is honestly no easier or quicker way to add a little extra oomph to your icon. Just don't over do it or it'll end up looking like a psychedelic trip-out (unless that's what you were going for, in which case you just carry on there, you fab hippy.)

TEXTURES

Adding textures is a great way to add a little something extra to your icon. I often add textures throughout the process of making my icon, particularly colour/gradient ones, but texture can also be very useful right at the end. My current "thing" is adding a painted/grungey style.

THIS IS WHERE LAYER STYLES IS REALLY AWESOME.

The texture I have used in my example works wonders on 'screen', adding that white "grain" on the right-hand side. If you feel by using the screen style you have brightened up your icon too much simply duplicate the texture layer, set it to 'soft light' and adjust the opacity accordingly.

Sometimes I pick a texture and then go through every single layer style, just trying them all and see if any of them come up with an awesome, and unplanned, effect. Sometimes it's just all trial and error, so just have fun with it!

LIGHT

Light textures are excellent in highlighting certain areas of your icon and directing focus. I often use them to emphasis the light direction, or to add shine. Light textures usually work best using the 'screen' style (for highlight) or 'soft light' (for darkening/highlighting).

If you don't wish to use a ready-made texture you can always use your brush! Just choose a medium to large size round soft-edge brush (hardness at 0%) and dot white around in the areas you wish. Set to screen and adjust your opacity as necessary. Simples!

BLUR

You're probably looking at the blur picture going "wtf, why would I want to do that, you crazy lady?" but bare with, I will explain.

At the end of every icon I make this is what I do: I apply all the layers I have (IMAGE > APPLY IMAGE...) and set the blending to 'Normal.' I then apply one of two blur filters to the layer: Lens or Gaussian. It's entirely up to you, they will pretty much have exactly the same effect. Apply the filter as you see fit (for Lens I usually don't go over radius 20 and for gaussian no more than 1.5. Now set your layer style to 'Soft Light'. See how it instantly increases the contrast and accentuates the highlights? Adjust the opacity until you're happy.

SHARPENING

Sharpening and I have never been the best of friends, but with that whole blurring thing above you're going to need it.

I sharpen my icons in two ways: 'Smart Sharpen' and/or normal 'Sharpen'. Above the blur layer apply your image again and then go to FILTER > SHARPEN >. Choose 'Smart Sharpen' from the menu.

Now as you can see (and Lydia, quiet you, I'm almost done!) there's a whole bunch of options. What you can see in the example is what I would typically choose. All I can say is mess around with the settings and see what you are happy with. Some images will need greater sharpening than others.

After this I then apply my layer again (I know I know) and then on this one I do one more sharpen, FILTER > SHARPEN > SHARPEN. I then adjust the opacity of this layer (usually quite low) and VOILA, you've got your edges and details back.

Okay so I'm going to stop talking now, I admire you if you got the end! And if you did I hope you don't feel like suing me for the minutes of your life you just lost.

There are several things that I haven't gone over here but that's because each icon is different and Photoshop has an endless array of facilities. But as a basic tips & tricks tutorial I hope it was of some use to you.

Comments and feedback = ♥!

Icon Tutuorial: Coloring

2011-02-27T00:08:42Z

From this to
Program(s)+version: Photoshop CS5
Involves: Coloring an icon using: curves, levels, selective color, color balance, vibrance & filters available in Photoshop CS5
Translatable: To programs/versions that have selective color and vibrance available.
Steps: 4 Steps | 23 Layers
Difficulty: Easy

Step 01: Re-sizing the image
Note: I do this step in variations in almost every icon that I work with. I find that if you work on making a base smoother and softer (no matter the quality of the image) then the overall icon is much easier to work with.

01. Cropping your image
• I usually work with my images at 250px by 250px, just because I like the space that it gives me to play with, and when finishing an icon I find it much easier to work with that size and avoid over-sharpening anything.
• I cropped this image (courtesy of letsey_x) to 300px by 300px.
• I then resized the image to 250px by 250px.
• Now we have a beautiful, if not a little bit larger than usual base to play with.

Step 02: Preparing Your Base

Layer 1: Base Layer. Remains untouched.
I always leave my first layer untouched. If I stuff something up and need to go back to the start, it’s nice to not have to crop again, especially if you’ve got a really interesting/unique crop going. It’s also handy to have it on hand if you want to use your base to add an extra element to the colours of an icon later on.

Layer 2: Duplicate the base layer. Filter > Blur > Surface Blur > Radius: 3, Threshold: 3. Set this layer to 20% opacity.

This can be interchanged with Smart Blur if necessary. I like the effect Surface Blur/Smart Blur has on softening the features of an icon, makes it smoother in general, and sometimes (especially for lower quality images or darker images that seem pixelated) that’s a real bonus.

Layer 3: Duplicate of the base layer. Filter > Blur > Gaussian Blur > Radius: 1.0. Set this layer to 50% opacity.

Again, we’re trying to blur out any of the flaws and faults within the icon to make the base something nicer to work with. This will soften the edges of your image further.

Layer 4: Duplicate the base layer. Filter > Blur > Surface Blur > Radius: 3, Threshold: 3. Set this layer to 25% opacity.

Layer 5: Ctrl + A (Select all) the image. Ctrl + Shift + C (Copy Merged) the image and Ctrl + V (Paste) it as a new layer. Set this layer to screen > 20% opacity.

When you copy a merged image, you are copying the sum total of the previous layers together, as opposed to just one layer. This screen layer’s opacity could be anywhere from 10% to 100%, depending on how dark the image you started out with originally was, or how bright and vibrant you would like your icon to be.

Layer 6: Duplicate the base layer. Drag to top. Set to 10% opacity.

This step isn’t something I do often in my base preparation, but I believe I wanted this icon to have a little bit of a darker, more ominous feel to it. By dragging the base layer to the top, I added a touch of dark back into the icon.

Layer 7: Layer > New Adjustment Layer > Curves > RGB > Output: 135, Input: 122.

In this step, I’ve used curves simply to lighten up the image. Later I’ll use it to add some colour, but while I’m preparing the base, I just want to use the RGB layer to add more lightness to the icon.

Layer 8: Layer > New Adjustment Layer > Levels > RGB > 0 | 1.10 | 245.

For the same reason as Layer 7, Levels have been used here to help lighten the image. I generally use this setting, but if the image is especially dark, I’ll lower 245 to 235 to give it that little bit more brightness.

Layer 9: Layer > New Adjustment Layer > Brightness/Contrast > Brightness: -10 | Contrast: +10. Set this layer to 50% opacity.

This layer is used to add some contrast and depth back into the image. Not too much, but just enough to highlight the difference between foreground and background.

Layer 10: Duplicate Layer 9. Set this layer to Soft Light, 50% opacity.

This layer is used to add the same kind of contrast to the image, and also helps to distinguish the various colours in an image, and add contrast to them as well.

Step 03: Colouring the Image

Layer 11: Layer > New Adjustment Later > Vibrance > Vibrance: +50

Vibrance will heighten the colours of your image, and add a wonderful depth to them. You do however, need to take notice of the colours already present in your image. The image I’m working with had a lovely range of colour already within it, which made the vibrance layer work a lot more, than something like a duller image, in which you may need to add colour to the image before adding a vibrance layer.

Note: Vibrance as an adjustment layer is available on the most recent versions of photoshop, but not all. I absolutely LOVE and abuse at times the effect it can give – but if you don’t have it available, try playing around with Hue/Saturation. You need to be more tender and gentle with each layer, and definitely play around with the amount of Hue/Saturation you use, as well as the opacity – but can achieve similar results.

Layer 12: Layer > New Adjustment Layer > Curves > RGB: Output 144, Input 116.

We need to brighten the image a little more, so we add another curves layer that only plays with the RGB together, rather than the individual colours in the image.

Layer 13: Layer > New Adjustment Layer > Colour Balance >

Midtones:
+ 5 | -5 | +5
Shadows:
+9 | -5 | -10

When we alter the Midtones, we’re playing with the reds, magenta and blue hues of the image, adding depth to them. When we alter the Shadows, we’re playing further with the reds, but adding a touch of green and deepening the yellow colour. The midtones are a little softer than the shadows, and when you play with the shadows the colour effect is often a lot deeper than that of the midtones.

Layer 14: Layer > Duplicate > Duplicate Layer 13 > Set this layer to Soft Light at 50% opacity.

In this layer we’re taking the colour we added from Layer 13, and using Soft Light to enhance its vibrance and depth.

Layer 15: Layer > New Adjustment Later > Selective Colour >

Reds –
Magenta: + 10
Yellow: + 20
Black: + 35

Yellows –
Magenta: - 10
Yellow: + 25
Black: -10

Blues –
Yellow: + 19
Black: + 19

Magentas –
Black: + 24

Neutrals –
Yellow: - 10
Black: +5

Blacks –
Cyan: + 5
Magenta: + 5
Yellow: + 5
Black: + 15

Depending on the colours of your image, you may want to approach Selective Colour in several ways. For this image, because there are dominant blues and yellows, I’ve used selective colouring to highlight them, as well as draw out some of the reds & magenta's that are in the image. If you had an image in which you wanted to accentuate the greens and cyan's, you would play with those various colour tones within selective colouring.

Layer 16: Layer > New Adjustment Layer > Vibrance > Vibrance: 30

Now we want our colours to pop, and this vibrance layer helps do this, without overwhelming the image quality.

Layer 17: Layer > New Adjustment Layer > Curves > RGB: Output 132, Input 120.

I do actually use curves to play with the colours of an icon, but in this case I’m using it to add some more brightness and light to the image. Not too much though, as I don’t want to overwhelm the icon, or take away from the natural darkness that I like so much in the image.

Layer 18: Ctrl + A (Select All) > Ctrl + Shift + C (Copy merged) > Ctrl + V (Paste) > Filter > Artistic > Paint Daubs: Brush Size 1, Sharpness 0 > Set layer to 10% opacity.

Paint daubs is a magical device that can do lots of things. In this instance (and in most others) I’ve used it to create a blurred layer which softens the edges of the image, and gives it a glow of sorts. By lowering the opacity to 10%, it’s only very minor, but still enough to give that little bit of something back to the image.

Layer 19: Layer > New Adjustment Layer > Selective Colour:

Yellows –
Cyan: + 5
Yellow: + 30
Black: -5

Neutrals –
Yellow: - 15
Black: + 10

This selective colour layer plays strongly with the yellows. I wanted to highlight the yellow of the hair, as well as the more neutral yellows in the skin to contrast against the depth of the blue in the image.

Layer 20: Return to Layer 1 > Layer > Duplicate > Duplicate Layer 1 > Drag layer to the top > Set this layer to Screen at 30% opacity.

Like I said, sometimes having a base layer untouched can be a good thing. I added this layer to give a little bit more brightness to the image. It also adds a little bit of the original colour into the image as well, without overwhelming it completely.

Step 04: Adding the finishing touches

Now that I’ve coloured the image, I’m ready to finish the icon. Normally there would be an additional step in here, about adding textures – however I felt this particular image didn’t warrant any texture addition – because of the mixture of colours already within the icon.

These next couple of steps are always involved when finishing off an icon. Like preparing an icon base, I think that when you’re finishing an icon, it needs to look finished, rather than haphazardly thrown together and completed.

Before we do anything though, I always re-size the image to 200 px by 200 px. Just because I’m neurotic, and I think the blur/sharpening combo works best on that size image before it gets resized to 100px.

We therefore need to do this:

Image > Image Size (Alt + Ctrl + I) > Width 200 (pixels), Height 200 (pixels).

Layer 21: Ctrl + A (Select All) > Ctrl + Shift + C (Copy merged) > Ctrl + V (Paste) > Filter > Blur > Gaussian Blur > Radius 1.0 > Set this layer to 30% opacity.

Again, I use Gaussian Blur to soften the image before sharpening the edges. I think the sharpening effect just looks a little bit softer if this is done, as it takes away that absolute sharp edge that can sometimes occur in an icon.

Layer 22: Ctrl + A (Select All) > Ctrl + Shift + C (Copy merged) > Ctrl + V (Paste) > Filter > Sharpen > Sharpen > Set this layer to 50% opacity.

Depending on the image you began with, the sharpen layer will be anything between 10% & 100% opacity. And while you’re image may look a touch too sharp now – remember we still have to resize the image to 100 px by 100 px, which will also take away a little of the sharpness.

You have to be cautious here, because sometimes sharpening can lead to rough, harsh edges. It’s always good to have a little bit of leeway on each side.

Layer 23: Layer > New Adjustment Layer > Vibrance > Vibrance +50

I added this final layer to give a touch more bite and colour to the image. Generally I’ll be happy without this, but I wanted to highlight the colours a lot in this icon, which is why this layer is there.

We’re finally ready to resize the image: Image > Image Size (Alt + Ctrl + I) > Width 100 (pixels), Height 100 (pixels).

Step 05: Saving your icon.

File > Save As (Ctrl + Shift + S) > Add a file name, and click down your format options to png.

PLEASE NOTE: ALWAYS SAVE YOUR ICON IN PNG FORMAT. I cannot stress this enough, hence the attack of the CAPSLOCK. Saving in jpeg can make the quality of the icon decrease dramatically when it is re-uploaded. Saving as png maintains the higher quality of the icon, and ensures the icon is nicer to look at.

Once you’ve saved your icon, you’re done! (Which is the most *headdesking* thing to say, but it’s true!)

--

Endnote: I hope you've found this tutorial helpful in at least one way, or gained a little trick or two from it to use on your own icon.

As always, comments are love. Please let me know if you found this useful, or alternatively, hard to follow! Questions are always welcome as well :)

How To: Create Backgrounds

2010-07-21T03:29:58Z

Program(s)+version: Photoshop 7.0 (but it should translate over to all other versions of Photoshop, and is probably, at least in part, applicable to other programs)
Involves: How to create backgrounds (using the original screencap, textures, clouds, etc), using textures on icons, layer masks, the smudge tool, the Blur filters, general tips, and a little tiny bit of coloring using selective coloring, filters, and blending modes.
Translatable: Mostly, I think.
Steps: 4 parts with a variation of steps
Difficulty: Moderate difficulty, assumes a bit beyond basic knowledge of Photoshop (including: layer masks, blending modes, and a few other tools)

This tutorial is cut and pasted from a larger 'Ask the Icon Maker' post. I was invited to post it here, but I didn't think it would be appropriate to post the entire thing here, just the actual tutorial bits (warning: extremely colloquial). So if it seems a tad disjointed and/or overly personal, I apologize.

In my 'Ask the Icon Maker' post, I was asked: I would like to know how you do backgrounds/Ok, so I have been crushing on this icon for a long time. I was wondering how you made it look so epic with the clouds.

LAYER MASKS! THE BEST TOOL PHOTOSHOP EVER GAVE US! \O/ Oh wait. I'm getting ahead of myself. >.<

Okay, I'm going to try and split this part into coherent sections.

1. Part One: CLOUDS
>> Okay. So clouds. I like clouds. Putting clouds into an icon is pretty. But there's a way to do it and a way not to do it. You can't just...stick an image of clouds onto a screencap. You just can't.

First of all, I never, ever just put clouds onto a transparent background and call it a day. No siree bob. You want your clouds to look like part of the image, even though they are not. So, first off, I recommend actually using a shot from outside. I mean, you can use inside shots, but the lighting is really, really wrong for that. Anyway, then you want to prepare your "sky". How do you do that? Smudge it! Use the smudge tool and prepare a sky.

The top half of the icon has been smudged up to create a sort of fake "sky". That's where your cloud image is going to go. So then the next bit is to find an image of clouds. I usually just google 'clouds' or 'storm clouds' if I want epic stuff. I have a couple of cloud images that I have saved because they were truly awesome, and lucky for you all the cloud stock used in this icon I actually have saved! Ugh. Too big for tinypic. Sorry I can't share. D:

I took the full stock image and then I resized it to 300x400. I want my clouds to cover the whole icon to start off with because it is easier (and it looks better in the end) to integrate with the rest of my base if I do it that way. And then, I think, I change the blending mode to multiply. That way the fake sky and the real sky blend together and take on some of the same colors.

Okay, then comes LAYER MASKS! \O/ You can get excited now, lol. Alright, so you need to add a layer mask to the sky layer, and then MASK THE CLOUDS OFF THE FACES OF THE PEOPLE BECAUSE CLOUDS OVER TOP THEIR FACES LOOKS DUMB, YES? If you read my earlier tutorial, you know that I like to use soft, round brushes for this kind of thing (and by that I mean everything). Don't be afraid to let a little sky overlap the ground, but make sure it is completely off anything human (skin, clothes, etc).

And then I colored. Which basically involved a lot of soft light. Hm. But, when you are coloring, you should make sure to keep a light source in mind. In this one, the sky has the most light in the upper right corner, so I tried to keep my light effects in the same sort of range.

Old Icon	New Icon

Pretty darn close! I think I chose where to put the light blobs a little differently. More light behind in the old version; more light in front in the new version.

So basically, to answer both questions before we move along: To do backgrounds well remember one thing: INTEGRATE. Backgrounds aren't a separate part of your icon, and you shouldn't treat them as such. The goal is to make it a natural part of the icon. So use blend modes and layer masks to do that. :) As for epic: first of all, I point fingers at the image I used of the clouds. It looked pretty epic to begin with. Then, when I color, I tend toward the dramatic anyway because I am all about soft light layers (of various types), lights, and shadows. So I get bright colors, high contrast, and a variance of light and shade that looks...epic, I guess? :D

2. Part Two: PRETTY LIGHTS ETC

Okay. Pretty lights. Since this was all about backgrounds, I figured this sort of background was just a legit as the rest. Backgrounds like these are pretty simple because they always come directly from the screencap itself. I would suggest you use screencaps in which the backgrounds aren't too cluttered with various things. Try natural clutter like trees, or just simply walls and lights. Then use the smudge tool and go wild. If you are afraid you'll mess up, duplicate your base, do you smudging on the duplicate layer, and the add a layer mask and mask away what you don't want from the original base.

Usually I try and stick with a pattern because that means we don't have to blur and fix things so much later. But, for the sake of explaining things, let's just assume I went wild crazy and ended up with this.

So...that'll need to be fixed. Yeah. So this is where the handy dandy amazing Blur Tools come in! I'm not talking just the simple Blur either. Actually, I hardly ever use plain old Blur. We're talking things like Gaussian Blur, Radial Blur, and Motion Blur. All can be found under the Filter menu, and all are insanely fun to work with. So for the base in question, I just went through and tried out the blurs.

Gaussian blur is typically my go to for things like this, but in this case, it looked a little silly. And Motion Blur will be taken care of in the next section, which left me with Radial Blur! Which is the much overlooked runt of the Blur family, but on occasion has served me well! So. I used a little Radial Blur and the masked it away from Spock leaving me with this.

Which, idk, looks kind of nifty, right? I do use a lot of plain, negative space backgrounds (which we'll make fun of in the next section), but when I first started with Xena icons, I liked this swirly light background. It's fun. It spices up a boring icon. And...COLORS. Which is the most important part of backgrounds like these. If there aren't any pretty colors, then I don't bother.

I colored the cap (some variations to spice up the colors, lots of soft light, a little more blur, and bam), and I got this in the end. Which probably would end up in the scrap pile tbh because...the balance is off. :-/ The light side of the icon and the dark side.

So I copied bits from either side, dragged them to the opposite side, and then fiddled around a bit until I got something that was decent. Except, omg, this still sucks. Erm, you see the problem? THE FACE. Okay, I know we're focusing on background here, but...I'm making this icon as I go through this. So it is completely new, and IT NEEDS FIXING. SO the face. Too dark. Too blue. Ugly. MUST FIX.

So a couple of tan multiply layers (masked off everything but Spock's face) and a Diffuse Glow layer (also masked off everything except the face) later...

Which, okay, I'll probably still wander back and try and fix it up some more because I'm a crazy person, but for now I'll stop. *stares at icon; eye twitches*

BACK ON TOPIC. Swirling light backgrounds in summary go like this: SMUDGE TOOL IS YOUR FRIEND. Make sure you have lots of lights, and don't be afraid to experiment with the Blur Tools. And that's it!

3. Part Three: BACKGROUND EXTENSIONS

Yeah, yeah, I know. This was probably the section everyone was interested in anyway, right? I probably should have done it first because, honestly, this is the section that everything is sort of based off of. You have to use bits and piece of this to do Part One and Part Two. BUT WHATEVER. I NEVER GO IN THE RIGHT ORDER OF THINGS ANYWAY.

So background extensions. Um. Let's see. Where to start. I guess we should just start with the basics, right? So.

Part A

Negative space background extensions! These are pretty easy because all you have to do is extend, make it look "flawless", and then color.

I'm just going to do this quick and dirty. No images. You grab a screencap. I recommend one that has a pretty clear background to extend. You can use more complicated things (and sometimes I do), but they make it harder in the end and you'll spend a lot more time making your base pretty (actually, tbh, that's what I always spend the most time on which is why my coloring methods tend to be very simple. I spend half and hour trying to make the base ready. Ahaha.).

Okay, so you have your screencap. I tend to select the bit I want to use, copy, put it on a new canvas, resize so that the width is 300 pixels, and then put it on a 300x300 canvas. And then it is time to do the actual extending.

I do this one of two ways. The first way is probably the easiest. You select the bit of the background that is going to get extended, and then hit Ctrl+t (for those of you who don't know, that's the keyboard shortcut for the Free Transform Tool). Then you just drag upwards until you've fill all the space. 98% of the time, once you've done this you end up with a very subtle stretch line where you started stretching. Sometimes you get a bit of streaking too. So I use Gaussian Blur and layer masks to fix the background up, smooth it out, and give it that 'flawless' look that you can see in both my recent Disney 20in20 sets (especially the Lion King 2 one). Then you just color, sharpen, and resize. Pretty easy.

There's also the second way which you've already experimented with. This way is also better if you're dealing with slightly more complex backgrounds (ones that you can't just Free Transform). This one involves the Smudge Tool again. The smudge tool and I go way back, man. You use the smudge tool to get rid of anything in the background you don't want and also to extend what you do want up to fill the empty space. As with method one, you're going to have some imperfections that can easily be taken care of with Gaussian Blur and layer masks. This is the method I used to make THIS icon do what I wanted. I got rid of some of the wall, and extended the green part up and over to the left to create a full background.

And that's the easy way to do background extensions. So let's get into the harder bits.

Part B

Using the image/screencap itself to extend your background. These are two ways to do this, but the second way will be included in Part C. So, the first way.

The first way is this: Xena or Kahlan and Richard or Cara.
This way is way more complicated (though it won't sound it), and it can be used in infinite amount of ways (and I have used it in many ways). I don't just use it to create backgrounds, but also to create space/to add something else to the icon (as you can see in the example using Kahlan and Richard).

The easiest way to to go about it is just to create some space/to add something. That would be the icon of Xena. There, I just duplicated it, and put it behind her. There's an element of imperfection to it that I actually like. It is the one time I recommend not aspiring to the ultimate goal of "integration". I don't blur. I don't make to part of the rest of the background behind Xena. There's very clearly a line there.

The important part of using this method in this way is that the image you use behind the original base (which, for me, is always part of the original image to begin with) needs to be interesting. Not necessarily eye catching in a way that draws you from the subject but not so boring as a block of plain color (although, I do occasionally use that method as well. It has its place.). In this case, I used a bigger version of the trees/flowers around Xena because the colors and lights were nice.

The second way is Cara. Basically, you go through the motions of the first way, only you actually integrate. To do this, I use layer masks (to get rid of the line) and the blur tool to make it more natural. Once again, it is very important to use something that works with the rest of the base. So I used the trees behind Cara, only I rotated the image a bit so it wasn't exactly the same. We'll do a quick walk through of that then:

So I took the original screencap and resized it. That would be the bottom half of the base here. Then I ran into some problems in that the screencap didn't have anything big enough for me to just crop, resize, and use for the top half. So I had to take two parts of the original screencap and put them together using a layer mask (to get rid of Richard's head...). So the top half is actually two parts of the original screencap put together.

Okay. So. Layer Masks. It might be easier if I just show you the layer palette...here. Both layer masks have been Gaussian Blurred to make the edges softer and to blend everything together.

And then I color...(and ignore the fact that I can still see some of the line on my 300x300 canvas because I know I can't see it in the 100x100...). Coloring was a lot of soft light again. Some variations to add green to the background (masked off Kahlan and Richard, of course), and then some selective coloring because I hate red (Under Red, I took the little Yellow marker and got rid of all the yellow in the reds in the icon. Meaning that the reds all took on a blue/purple quality). Then a little Diffuse glow to make things pop. Almost all these layers had layer masks for various reasons.

The important thing to remember when you're doing this is, again, integrate. Even though you are using different parts of the same screencap, they aren't going to look nice together unless you pay attention to details. For instance, when I used a layer mask, I got rid of the bright white spot of sky by Kahlan's head to allow better integration with the green I was adding in. If you focus on that sort of thing, your icon will come out looking better.

Part C

Motion Blur! Fun times! Okay, I already did a tutorial for this here, and, yes, the Castiel icon used in the example at the beginning is made the same way. The thing with using the motion blur tool is...there's no way to hide it. If you've used it, I (and everyone else) can tell. There's no "natural" or "flawless" in icons like these. You can still integrate, of course! (And, actually, I did a better job on Castiel than I did on Buffy in the tutorial, lol.) But, ultimately, this is the least natural of the lot. And also the one hardest to do exactly right because each image is totally different, and you have to be careful or lines/parts of the original screencap duplicate will show up.

However, this is pretty fun and because you can change the direction and intensity (explained in the tutorial) of the blur, you can do all kinds of fun things. Just don't overdo it. Too much of this in one post, imo, looks ridiculous. I try and limit my use of it especially when I'm making it really obvious (like here).

And, yeah, there you have it! I think I've covered all the basics of extending backgrounds (at least the way I extend backgrounds). So time to move onto the last (and my least favorite) part!

4. Part Four: TEXTURES

Okay. Textures. :-/ Or adding a completely new background to an image (Ares go boom). Or getting rid of the entire background of an image (Want your bad romance). Etc.

Basically this is an exercise in patience and me trying not to yell at my photoshop program for various reasons.

So for Clopin (the first icon example), I used the lasso tool to get rid of the background. But I only do that with Disney icons because A) they've usually got black lines around them that make them easy to cut out and B) because the lasso tool hates me. I'm not kidding. I'll be almost done and it will suddenly select everything and I'll have to start all over again.

So for everything else there's layer masks. :D

So, um, for the icon of Santana, I started with the full screencap, and then, below that, I put a solid color layer (using the eye cropper, i chose a color that came from the screencap itself). Then I proceeded to mask away parts. Now usually I cheat a little bit. Or, you know, a lot.

If you'll take a look at the Ares example, you'll see that I use images that had the same sort of color scheme. I blended together Ares (who was dark and had flame shadows) with...an explosion that happens minutes later. So Ares, who was already dark, got blended with bits that were dark too. That meant that I could get away with not having a super smooth cut out of Ares. There are bits of the original Ares image in there that shouldn't be, but you can't tell.

It's the same sort of thing with Santana as well. Because I made the final icon so small, I could have a few mistakes. If you look at the 300x300 canvas:

You can see the imperfections. In around her legs, her hair, etc. There's color that overlaps and various places that aren't 100%. But you can't really see that when the final icon is so small.

So that's the trick really. I hate doing icons like this because I need place where I can lapse. I don't have the patience to stick to cutting out an image for an hour and a half. The one of Santana took almost an hour or so to get right as it was (and let's not even talk about the coloring because I worked even harder on that and I basically hate this icon, omg.).

Um. If you don't mind me being scatter brained, I'd like to return to the Clopin example because I am supposed to be talking about textures here...Um, the texture used in the icon is made by...chambertin (who seems to no longer exist?), I believe. And it had to be resized and recolored before I could use it to begin with. Which almost always happens when I used textures like this. Then I reinforce what I said way back when we were talking about clouds, by putting the texture on a blending mode that isn't normal. It was probably multiply again, but it could have been overlay. Not sure now. It has been way too long since I made it. Anyway, then you mask away the bits that cover Clopin. Then, idk, layer masks sometimes don't do the trick, even after they've been blurred. So I, a lot of times, take the base image, drag a duplicate of it to the top, and set that to soft light. Not for coloring purposes, but to soften edges and make things look more seamless.

So this section in summary: I hate the lasso tool (but use it for Disney icons), layer masks are, again, god, and to use textures in icons (as backgrounds at least), you have to work the same way you do with clouds: integrate!

And, yeah, I guess that's it. :D

A Guide to Colour Normalization

2010-03-24T21:16:03Z

Program(s) version: PSP X2, but the principles should apply in any program.
Involves: Colour Normalization of blue and red-tinted caps, as well as basic colouring tips for natural caps. Brief guides on the colour balance, curves, and levels tools.
Translatable: Yes.
Steps: None and many, all at once! This is a guide, not an A to B type tutorial.
Difficulty: Absolute Beginner. I've geared the guide towards people who are very unfamiliar with their program, but there should hopefully be some tips here that are useful for more experienced users as well. (: I assume a basic understanding of how to create layers, and how layers work.

This guide is very image heavy!
This guide covers, in this order:
-A brief introduction to how colours relate to each other, shadows/midtones/highlights, and the tools I'm using in this guide (colour balance, levels, and curves) for beginners
-Correcting blue screencaps using colour balance, levels, or curves.
-Correcting red screencaps using colour balance, levels, or curves.
-General colouring tips (i.e. odds and ends)

What it is not meant to be:
-an exhaustive list of colouring techniques
-the end point for an image colouring - these are just techniques that I use to get an image to a point where I can then apply other colour adjustments without the images looking horrific

This tutorial is intended to be a guide, not a list of "do this, then this". As such, I'm not going to list any of the exact numbers that I used to achieve the colours - rather, the focus of this tutorial/guide is on understanding the tools, and how you can use them to normalize the colours of whatever picture you are working with. Obviously, this isn't going to be an exhaustive list of various techniques; there is always going to be more than one way to achieve what you want. These are simply the ways that I find myself using most often.

If you've got any questions or comments, don't hesitate to reply to this post and I'd be more than happy to answer your question! If sections of this guide are unclear, it's probably because I don't realize it, so if you find something confusing let me know so I can hopefully reword it to make it clearer.

Preamble.

When I first started iconing, one of the things I found most intimidating was trying to figure out how to get rid of overwhelming colours (usually blue, in screencaps) - however, once I figured out how the colours related to one another, I found it a lot easier to get the results I wanted.

The tools I tend to use the most for colour normalization are the colour balance, curves, and levels tools. While there are plenty of tutorials out there that teach people how to use these tools (and I've listed some of them at the bottom of this guide), I'll include a brief primer for people who have never used them before or have used them, but don't really know how they work (just that they do).

While the directions are geared towards people who have PSP X2, the ideas behind each of the tools remains the same across programs (to the best of my knowledge). You'll probably find the tools under different menus, but the rest should still be applicable.

Colours in General

Most of figuring out how to normalize the colours in a picture relies on being able to figure out what colour is the predominant colour, and what the opposite of that colour is. If I'm stumped, I usually just refer to the colour wheel in my program:

Obviously, this is labelled very roughly, but the 'important' colours and relationships are labelled. When you're working with colours (and specifically, trying to get rid of or add certain colours) it's useful to know which colour is the 'opposite' of the colour you're working with. So, as you can see by my nifty labelled colour wheel above, the opposite of Red is Cyan, of Blue is Yellow, and of Green is Magenta.

So, logically, if you're trying to get rid of a lot of blue in your image, adding yellow will go a long way to reducing the blueness of your picture.

Shadows, Midtones, and Highlights

The best explanation I can give of Shadows, Midtones, and Highlights is that they refer to the darkness of the grays in the image, assuming that you have gray-scaled the image. The greater the range from shadows to highlights, the greater the contrast of your image.

When you're changing the colours of an image, it is important to understand the differences between Shadows, Midtones, and Highlights of the images so you know what aspect of the image you're adding colour to or removing colour from. If you imagine your image gray-scaled, you'll be able to guesstimate what areas of your image are the shadows, midtones, or highlights, but as a general rule of thumb the darkest areas are the shadows, the brightest/whitest areas of the image are the highlights, and the areas in between are the midtones.

Colour Balance

The colour balance tool can be found under Layers > New Adjustment Layer > Color Balance. It will open up and look something like this:

The three radio buttons at the bottom (Shadows, Midtones, and Highlights) and the sliders themselves are the parts of the box that you'll find yourself using the most. I rarely uncheck 'Preserve Luminescence'. Usually I start my adjustments with the shadows, but this isn't an exact science. Start your adjustments wherever you feel the most comfortable (or possibly, would like to see the most change).

Moving the sliders either left or right adjusts the amount of that colour in the image. So, moving the midtones Red slider right (towards the 'Red' end) increases the amount of red in the midtones of the image. Moving it left increases the amount of cyan in the midtones. Adjusting the same slider under the Shadows radio button increases the amount of red or cyan in the shadows of the image.

Levels

The levels tool can be found in two places in PSP X2. The first instance of the levels tool will only adjust the layer that you have selected and not the entire image that you are working on - it can be found under Adjust > Brightness and Contrast > Levels. This is useful if you are working with multiple layers of the same image, and you only want to adjust the colouring of (for instance), the layer that you've set to screen.

The second instance of the levels tool, and the one that I use more often, creates a new adjustment layer, which will then adjust the colouring of all of the layers below it. It can be found by going Layers > New Adjustment Layer > Levels. The tool looks something like this:

Under the Channels drop down box you have four options: RGB, Red, Green, or Blue. RGB adjusts the entire image as a whole (or, the Red, Green, and Blue channels all at once). The Red channel adjusts only the reds in the image, the Green the greens, and the Blue the blues.

I primarily use the RGB channel to adjust the contrast of the image I'm working on. The other three channels come more in handy when it comes to actual colour correction.

Each of the three diamonds below the diagram adjusts a different aspect of the image; the diamond furthest to the left (the black one) adjusts the shadows, the middle diamond adjusts the midtones, and the white diamond furthest to the right adjusts highlights. So, by dragging the white diamond further to the centre, a person is increasing the highlights (or reds, greens, blues) of the image. By dragging the black diamond further towards the centre, a person is increasing the shadows (or cyans, magentas, yellows - the 'opposites' of red, green, and blue) of the image. By dragging the midtones (gray) diamond left, a person is increasing the highlights (red, green, blue) of the image; by dragging the diamond right, a person is increasing the shadows (cyan, magenta, yellow) of the image.

Curves

As with the levels tool, the curves tool can be found in two places in PSP X2. The first place (Adjust > Brightness and Contrast > Curves) only alters the layer that you have selected, and not the entire image, but otherwise functions the same as the curves adjustment layer.

The curves tool that I use in this tutorial is the curves adjustment layer - Layers > New Adjustment Layer > Curves. This creates a new layer which adjusts all of the layers below it.

The curves tool opens up in a box like so:

As with the levels tool, there are four channels you can adjust using curves: RGB, Red, Green, and Blue. The RGB channel is used to adjust the brightness and contrast of the image, whereas each of the channels adjusts (surprise XD ) the red/cyan, green/magenta, or blue/yellow of the image, depending on which one you have selected.

For this section, I'll use the red channel for my examples, but the idea is the same for the other channels as well.

When you adjust curves, you add a point to the diagonal line in the middle and pull the point in one direction or another. Depending on the place where you add the point (to the left or right of the line) you will be adjusting the highlights, midtones, or shadows. The picture on the right isn't an exact example of where the shadows, midtones, or highlights will be on the image - they fall along a gradient, not a solid block of dark/medium/light gray - but it should give you an idea of what point on the line you want to adjust in order to adjust the shadows (or midtones or highlights).

Dragging the point you have created either up or down adds either red or cyan to your image (up for red, down for cyan). The further away from the centre line you pull your point, the greater the curve will be and the more colour you will be adding to your image. Adding more points to the line and pulling them in one direction or another further adjusts your image.

Blue and Green Screencaps.

When it comes to blue screencaps, my first goal is to get rid of the blue - and the way to do that is by adding the opposite of blue, yellow. Adjusting green screencaps is similar to adjusting blue screencaps - by adding the opposite of green, magenta, you can get rid of the green.

Colour Balance

When it comes to adjusting screencaps using colour balance layers, I tend to focus on all three channels - Shadows, Midtones, and Highlights. I don't just play with the Yellow/Blue spectrum, either. Don't be afraid to touch the sliders! You can always push them back to 0 if you don't like the result.

	This is my base image. Since it's dark, I duplicate the base twice and set both duplicates to screen.
	First things first, I start with the shadows. This (click) is what the image looks like when the yellow Shadows slider has been pulled all of the way to the left (-100). Ick. Reducing that even by half is a huge improvement, and for this image, I don't break -40. Whatever your screencap, you're going to want to toy with this until the image has lost some of it's blue-hued skin.
	Repeating the above with the Midtones and the Highlights gets me something that's a whole lot closer to natural all ready! The key, here, is to allow yourself to experiment; no two pictures are going to be the same, which is why punching in exact numbers isn't going to get you very far when you're trying a tutorial out on another image. Keep in mind, though, that you want to keep your adjustments (relatively) small. I very rarely find myself making a colour adjustment that goes beyond /-50 on any of the sliders.
	Now, my goal is to add some more fleshy tones into the image. I go up to the red sliders next, since generally speaking most people I know aren't either Green or Magenta. ;) I usually leave those sliders for last and make the smallest adjustments there. On the red slider, I make slight adjustments towards the right (or red) side of the slider. For this picture, none of the adjustments went above 20, but this might be in part because adding all of the yellow before removed some of the cyan cast of the image.
	Last but not least, I play with the Green/Magenta slider. Now, in the shadows and the highlights, I touched the slider towards the Green end of the spectrum ever so slightly (no more than 10) in order to brighten up the image. Pushing the slider towards the magenta end of the image makes things darker, but also has the added advantage of adding something slightly more reddish to the tone of the image (versus adding green, which only tends to make people look sickly). Since I didn't want to make the overall image darker I added green in the shadows and highlights, because this image mostly consists of midtones. My midtones adjustments have had the greatest effects, and when I push the midtones slider towards the green end of the spectrum, River looks terrible. Pushing it towards the magenta side, however, has a more positive effect. Obviously, this is going to vary depending on the contrast of your image, because the greater an image's contrast, the greater the range of tones (shadows, midtones, and highlights) your overall image has.

Curves

As with the colour balance tool, I don't just focus on the Blue channel when I'm using curves to correct the colour of an image - I focus on all four. Don't be afraid to experiment; as with the colour balance tool, you can always reset everything or get rid of your points if you don't like the result.

	I'll start with the same base image as before, but instead of using a screen layer to brighten the image up, I'll use my curves layer, so I don't have a ton of layers floating around at the end of everything (once I've applied other colouring techniques to the image). I create a new curves adjustment layer and under RGB create a new point near the middle of the black line, which I then drag upwards to increase the highlights in the entire image. The curve ends up pretty steep, because this is a dark image. (Cap)
	Desired brightness achieved, I then move down to the Blue channel. Again, I create a point near the centre of the black line, but instead of pulling up, this time I pull the point down. This adds more yellow to the image. You don't need to make a huge adjustment (as in, your point shouldn't be in the bottom right corner, or anything). (Cap)
	The image is still pretty blah looking, so I turn to the other channels at this point. I want to make River's face look a little more flesh-toned, so I add two points on the red curve. The first is near the shadows and the second is near the highlights, and both I pull up slightly. The reason I add two points rather than one in the middle is because this gives me a little more control over how dramatic the curve is - I found when I was just adding one point, the red became too overwhelming. (Cap)
	I'm pretty satisfied with where this image is at, but I decide to brighten up the image a bit using the Green curves channel. I add a new point near the bottom of the line (in the shadows) and push it up very slightly, and it brightens the image considerably.(Cap)

Levels

Last but not least is levels! To shake things up, I'll be working on a picture of Jayne this time.

	As with the picture of River, this cap is rather dark. I'll use my levels adjustment layer to brighten things up, rather than a screen layer (though that would work just as well). I take the white diamond (the one furthest right) and push it to the left. With this image, it goes past the centre since this picture's rather dark. (Cap)
	I scroll down to the Blue channel and this time, only adjust the gray diamond. Most of the blue in this image is in the midtones and the highlights, so I push the midtones right to add more yellow. I don't touch anything else.
	Now I move to the red channel. A lot of the blue in this image also seems to be cyan (so, lighter/brighter blue) and so I opt to add more red to the image. This time I adjust the highlights diamond (the white one) by pushing it further left in order to add more red to the image; the midpoints diamond also moves. While the background still remains rather blue/cyan, Jayne's face looks relatively human. I could definitely stop here, if I so chose.
	In order to brighten things up and add a little contrast, I go over to the Greens channel. This time I move the shadows diamond closer to the centre (so, right), which adds more magenta to the image. Then, in order to counter the magenta, I move the midpoints diamond to the left, which adds a little green to brighten things up. The change isn't big, but it does add a little more contrast (especially to larger images). (Cap)

Red Screencaps.

The ideas behind adjusting Red screencaps really aren't that different from adjusting Blue screencaps, except you're working in the opposite direction; rather than adding yellow and red, you're adding blue and cyan! Even so, I've written up a guide for each tool (colour balance, curves, levels), just for the sake of thoroughness. (:

Colour Balance

	So we begin with a lovely picture of Inara! Obviously, this image is incredibly dark so I duplicate the base twice and set both duplicates to screen.
	Now we have a very red, but brighter, picture of Inara. In a new colour balance layer, I increase the amount of cyan in the shadows, midtones, and highlights. The midtones and highlights weren't adjust a lot (no more than -15 for either of them), but the shadows needed a bit of help so that was a bigger adjustment.
	Now onto the other sliders. I increase the amount of blue in the shadows, but when I move to the midtones I decide that adding more blue would probably be detrimental to the whole 'normalization' idea. Instead, I increase the yellows very slightly, so as to add a more natural tone to Inara's skin. Lastly, I go to the highlights and increase the blue just a tad so the yellow and red doesn't look completely overwhelming.
	Last but not least, I adjust the Magenta/Green slider. I don't touch the shadows, since I'm satisfied with where the shadows of this image are, but I want to brighten the picture up slightly so I increase the greens in the midtones and the highlights just slightly.

Curves

	Now we'll be working with a very red image of the lovely Inara. (: First things first, it's time to lighten this image up. Instead of using several screen layers, I open up my curves adjustment layer in order to cut down on the number of layers in the image over all. (Cap)
	Now that the image is bright enough (for now) I hop down to the Red channel and create a new point near the middle. I start to drag it downwards, but the image gets to be too dark for me so I move the point further to the right and continue pulling down. Eventually, that results in a point that is near the top right of the box. (Cap)
	The picture looks better already! In order to warm the image up a bit, since adding cyan cools it down a lot, I go down to the Blue channel and create a new point in the darker-midtones/shadows area and pull down slightly to add a little yellow. It's not a major change, but it does warm things up a little. Then, to increase the brightness of the highlights, I create another point near the top of the line and pull it up, adding a little more blue and brightening things slightly. Overall, it's still a very minor change, but I'm satisfied. (Cap)

Levels

Last up is levels!

	Again, since this is a dark image, I'll use levels to brighten it before adjusting the colouring. I drag the highlights (white) diamond almost halfway across the bottom bar to the left, brightening the image up considerably.
	Now to get rid of the red in the image. I move down to the Red channel and pull the midtones (gray) diamond to the right in order to add a great deal more cyan to the image. I also moved the shadows (black) diamond slightly right, but this also has the added effect of darkening the image. (Cap)
	In order to brighten the image somewhat, I go down to the Blues channel and pull the midtones diamond slightly right. This brightens the image and takes away some of the yellowish aspects that still remain in Inara's skintone. At this point, I'm satisfied with the image (aside from its darkness, which can be fixed through later colouring).

General Colouring Tips.

Obviously, this section is going to be heavily influenced by my own personal tastes when it comes to icons and icon-making - there's no real 'right' way to colour an image. This section is geared towards achieving a more 'natural' colouring, or general tips and tricks to get the sort of colours you're going for.

These are tools beyond what I discussed at the beginning and together, these do not create a complete list of all of the tools to use when colouring your images; there are a ton of tools in your graphics program that can be useful for this. These are simply the ideas I find myself using most often, and have noticed cropping up in tutorials most often. Remember: experiment!

Colour Fill Layers

These are the simplest way to alter the colouring of your images. When it comes to colouring images, I tend to use a combination of multiply, burn and soft light layers. I may at some point put together a group of tables to illustrate various combinations of colours and blend modes, but since this guide is both long enough and already ridiculously image-heavy, that will have to wait.

In PSP X2, the layer blend modes can be found in the layer palette. (Cap)

Here's an example of a colouring achieved solely with colour layers and duplications of the base (so, without adjustment layers):

>
After duplicating the base three times and setting the duplicates to screen in order to brighten the image up, I added a new flood fill layer with a light cyan colour that I then set to Burn in order to darken the shadows of the image (and add a blue tinge to them). I then added a new flood fill layer with brown, set to multiply at a low opacity, in order to add some browns to the image. Next, I added a pink layer set to soft light at 50% opacity, partially to brighten the image up and partially to add some pinky-reddy tones to the image (to return things to a less corpse-like flesh tone). Next, I duplicated the base and pulled it to the top of the image, which I then set to soft light and set to around 25% in order to add some contrast to the image. Last but not least, I added a white-black gradient (white on the top) that I set to soft light (around 60% opacity), in order to create a light source coming from the top of the image.

Brightness and Contrast

Personally, I prefer icons that are bright and have a decent amount of contrast, which makes the colouring pop. For example:

>
>
>
I tend to use Brightness/Contrast layers to increase the brightness of an image without having to use another screen layer (which will wash out the shadows as well as brightening the image) as well as increase the contrast. The Brightness/Contrast tool in PSP XII can be found under Layers > New Adjustment Layer > Brightness/Contrast.

There are several other ways of increasing the contrast:

1. Using curves. To use curves to create contrast, create a new curves adjustment layer and create an 'S' curve in the RGB section. This makes the shadows darker and the highlights brighter, and viola! Instant contrast. You can do something similar using levels, by moving the black and white diamond of the RGB channel towards the centre.

2. Soft light layers. Duplicating your base or copy-merging your entire image and pasting it as a new layer, and then setting this layer to 'Soft Light' will increase the contrast of your image. As well as those two methods, another I occasionally use is duplicating the base and, after setting the image to soft light, adjusting the Hue/Saturation of the duplicated base either to brighten the image, or take the colouring out of the duplicated base. Here's a comparison table:

Base
Base + Coloring
Duplicate Base, Pull to Top Set to Softlight
Copy-Merge Image Paste as New Layer Set to Softlight
Duplicate Base, Set to Softlight Adjust>Hue and Saturation>Hue/Saturation/Lightness Grayscale (Cap)
Duplicate Base, Set to Softlight Adjust>Hue and Saturation>Hue/Saturation/Lightness Increase Brightness (Cap)

Channel Mixer

Since this guide is already quite lengthy, if you're interested in learning more about the channel mixer I would suggest that you read this tutorial, written by yours truly, which has a couple of paragraphs on the use of the channel mixer in the middle. Another good tutorial that explains how the channel mixer works is this one.

Useful Tutorials.

These are some tutorials that I've found incredibly useful over the years when it came to understanding how various PSP (and by extension, PS and possibly GIMP) tools work. In large part, these tutorials have helped me when it comes to writing this guide, because a lot of the above comes from a combination of playing around and reading.

The concepts of all of these tutorials are translatable, even if they are originally written for another program.

A series of tutorials by bethyj, explaining basics of icon making, fonts, text techniques, blend modes, texture use, and all sorts of good things. Written for PSP 9, but useful even if you don't use that program.
Understanding Histograms, Levels and Curves by nomadicwriter
Photoshop's Channel Mixer by pixelempress
Beginner's Guide to Color Balance by cardcaptur
How To Use Levels (A Guide) by effulgency
A colouring tutorial written by myself, that includes a brief explanation of the channel mixer.
Selective Colouring: What is it and how does it work? by going_x_crazy - this tutorial obviously isn't translatable for users of PSP or GIMP. However, if you are interested in translating tutorials that use selective colouring for your program, it might be an idea to read this, just to learn what's really going on, so you can replicate it.

*To the writers of these tutorials - if (for whatever reason) you don't want your tutorial listed here, just let me know and I'll take it out. (:
**Caps were all taken from Leave Me the White.

extending backgrounds

2010-01-08T22:15:48Z

This is a little tutorial on how to extend the background of an image. I'll try to explain 3 different techniques I use. I tried to explain everything as detailed as possible. If there are any questions left feel free to ask away!
I'll start with what I feel like is the easiest way to extend the background of an image and then get to the more difficult techniques. But this isn't all that hard.
The coloring of the icons is not included.

Program(s)+version: Photoshop CS 2
Involves: smudge tool, rectangular marquee tool, clone stamp, layer mask
Translatable: should be translatable to programs that have the above mentioned tools and work with layers
Steps: up to 6
Difficulty: medium

extending backgrounds
using the smudge tool
using the rectangular marquee tool
using the clone stamp

To make the tutorial a bit easier to read I decided to not include every detail in the text. The more detailed explanations (things like adding new layers or where the tools are located) are linked in the tutorial. You can find them at the bottom of the page or by just clicking the appropriate links in the tutorial text.

USING THE SMUDGE TOOL

		I'll start with this icon. It's fairly easy since there's only that corner of the wall in the background.
01		This is my base. A base like this one is pretty easy to work with. If it wasn't for the window above Peter's head or the corner of the wall above Olivia's head you could just smudge the background out over the white part of the canvas.
		Of course that's what you can do here, too, by just smudging over the window and the wall corner. But the result you get looks a bit more unrealistic than the alternative (at least in my opinion) with everyting that's going on in the background of the image behind Olivia and Peter. It feels like that wall corner needs to be there in order to make the perspective look right.
02		Anyway, in order to extend that corner all way up to the top I like to use the smudge tool. (Where do I find the smudge tool?) If you look closely at the image there's this black box-y thing right above Olivia's head and next to the corner. There's only two pixels (in height) of just the wall corner between Olivia's head and the box. So this little part is what I want to smudge all way up to the top. I use a square brush for this. (How do I get a square brush?) In this case it's easier to work with a square brush because it gives you a straight line at the bottom. (Where do I choose a brush?) So with the square smudge brush (strength set to 100% Where do I adjust the strength?) I drag the little bit of wall corner all way up to the top of the canvas. Make sure you smudge in a straight line, you don't want that corner to look all bendy. This is a little tricky. You'll need a pretty steady hand for this. If the line gets bendy you can just smudge over it or you can try to keep smudging in little steps.</a>
03		For the rest of the background I just kept smudging the wall up until the entrie white part of the canvas was covered.
04		In order to smoothen things out a bit I used my favorite brush (made by inxsomniax) set it to a smaller size (about 10-20 pixels Where do I adjust the size?) and lower strength (about 20-30%) and smudged happily around until it looked right-ish.
05		Then you can go ahead and apply your coloring. Sometimes the background starts looking pixelated. To smoothen things out again, just use the same technique again. First you'll have to either merge all layers down or copy merged and paste as a new layer. (How do I do this?)
01		This is my result after smudging the background.

USING THE RECTANGULAR MARQUEE TOOL

		The second one is a little bit more difficult with all the light reflections in the background.
01		This base is a little harder... if you want it to be. Again, you could easily just smudge out the background all around Sam and Dean and have it all blank... which is exactly what I did at first. But I felt like that didn't really work with the image. So I decided to use the light reflections from the windows to fill the background.
02		First I did what I did with the Fringe icon. I used a square brush to smudge the background all the way up to the top. Now while there is the pattern from the reflection of the windows on the bottom left half of the image, there's just a straight vertical line in the top half.
03		I used the rectangular marquee tool and selected the window pattern. (Where do I find this tool and what does it do?) Then I copied that part and pasted it in as a new layer. (How do I do this?) I draged this new layer all the way to the top of the canvas. I added a layer mask to hide the bright part at the bottom of that layer and make it blend in with the background of the base. (How do I add a layer mask and how does it work?) This is my result.
04		I used the same technique for the reflection behind Sam's head. I selected the area I wanted to duplicate, copied it and pasted it on top of everything as a new layer. Again I used a layer mask, this time to get rid of Sam's head. In order to complete the reflection (one little square is missing since it was covered with Sam's head) I added a new layer. (How do I add a new layer?") I used the eyedropper to get the color of the reflection (How do I do this?) and then used a small round brush to draw the rest of the reflection on the new layer. (How does this work?)

USING THE CLONE STAMP

		The alternative to the marquee tool is the clone stamp.
01		If you have an image with a background similar to this, I feel like it looks better to not just fill the top of the canvas with the wall's color but to have all those little blinky lights go all the way up to the top, too.
02		First I use my square brush again to smudge the background all the way up to the top. I do this in order to get the basic color I want in the background. Then I look at the lights and what else is in the background and decide what should go all the way up to the top.
03		I use the clone stamp set to a small round brush and click on the part I want to duplicate while pressing the "Alt" key. (Where can I find the clone stamp?) By doing this you copy the part of the image you're clicking on. When you start painting with the clone stamp it will copy the part of the image you just Alt-clicked on. So extended the lights on the right side all the way up to the top by clicking (i.e. copying) in little steps. I could have used the rectangular marquee tool here, too, but if you want to copy small parts like I did here it's just easier to use the clone stamp.
04		I did the same with the lights and the wall behind Elizabeth and probably used the smudge tool to soften everything a little in the end - just like I did in the first example before.

Where is the smudge tool located?
You can find the smudge tool in your little tools window. You can switch between the blur, sharpen and smudge tool by right-clicking on the symbol and then choose the tool you need.

How do I get a square brush?
You can easily create a square brush in Photoshop. All you need to do is open a new canvas (File >> New) and fill it with black using the paint bucket from your tools palette. Then you can go to Edit >> Define Brush. A little window will pop up. Put in a name for the new brush. Click "ok".

How do I choose a brush I just created?
When you're working with the smudge or paint tool, just click on the little downward arrow in the top bar, right next to the brush you're currently using. A little box will appear in which you can find all your brushes. Here you can also adjust the size of the brush and the hardness.

How can I adjust the size of a brush?
You can adjust the size of a brush in that little drop down box in the top bar.

How can I adjust the strength of a brush?
You can adjust the strength in the top bar.

How can I merge or copy merge all layers?
To merge your layer go to your layer palette and right-click on the top layer. Choose merge visible.
To copy merge all your layers got to select and click "all". Then go to edit and click "copy merged". Then go to edit and click "paste". The keyboard shortcut for this is Ctrl+A (select all), Ctrl+Shift+C (copy merged) and Ctrl+V (paste).
I prefer the second method. This way you keep all your coloring layers and can always go back to change something or look at how you did something.

Where can I find the rectangular marquee tool and how does it work?
You can find the rectangular marquee tool in your tools palette. With this tool you can select any rectangular part of your image. For example in order to copy that part as I am about to do here.

How do I copy and paste?
After you selected the part you want to copy go to edit and click "copy". Then go to edit again and click "paste". The keyboard shortcut for this is Ctrl+C and Ctrl+V.

How can I apply a layer mask and how do they work?
Go to your layer palette and select the layer you want to add a layer mask to. Then click the third little button from the left at the bottom of the palette. A white square will now appear next to your layer. This is the layer mask. Make sure the layer mask is selected (just click on it to make sure). Now you can start drawing on the layer mask. If you draw on it with black everything you draw will disappear. If you use grey instead the layer will start fading. Using layer masks is a lot easier than using the eraser tool because you can always draw on the mask with a white brush to correct mistakes since drawing on it with white will make the layer visible again. This is what my layer palette looks in the end.

How can I add a new blank layer?
You can add a new layer by going to the layer menu at the top and then click "new" and then "layer".

Where can I find the eyedropper tool and what does it do?
You can find the eyedropper in your tools palette. By clicking on it your arrow will become a little pipette. With this you can choose a color from your image by clicking on it.

How do I draw on a layer
Just choose one of the round default brushes and adjust the size and hardness.

Where can I find the clone stamp?
The clone stamp is located on your tools palette.

Please let me know if you have any questions, if I messed something up or if any images/links don't work!

Comments are appreciated.

Happy birthday good_tutorial

2010-01-01T22:50:23Z

Today good_tutorial is 1 year old!

Happy Birthday good_tutorial!

I would like to take this opportunity to thank you all for making this community a success, and I would especially like to thank our members here, the tutorial writers, for sharing their wonderful talent and knowledge with all of us. It's very much appreciated and I hope all of you will continue to write tutorials and share them here this coming year, as you did the last :)

GRAPHIC TUTORIALS: BLEND MODES, BRUSHES, TEXTURES

2009-12-17T17:23:21Z

I didn't realize I never posted the last part in my tutorial series, my apologies!

Program(s): Paint Shop Pro 9
Involves: Basic tips
Translatable: It's quite specific to PSP, but is translatable to most other programs.
Steps: 0 - It's just a list of tips.
Difficulty: Beginner

The main purpose of textures and brushes are to enhance your image, not completely distract from it. You want to attain a certain effect, not compromise the integrity of your base image. Because as I’ve continually said before you add textures, brushes or text you need a quality base image. When using certain textures or brushes however there are times when you create the base to match or fit them. Even on that occasion you still need to concentrate on the quality of that image, don’t let the brush or texture distract you.

NOTE: These do's, don'ts and tips are written based on the assumption you have read the previous tutorials. These tutorials are to be used as steps. By this point you should be somewhat more comfortable in the basics and tools of your Paint Shop Pro.

BLEND MODES: Normal. Darken. Lighten. Hue. Hue (Legacy). Saturation. Saturation (Legacy). Color. Color (Legacy). Lumincance. Luminance (Legacy). Multiply. Screen. Dissolve. Overlay. Hardlight. Softlight. Difference. Dodge. Burn. Exclusion. You can set a layer (except your background layer) to any one of those. But what do they do? Sure many of them are fairly self explanitory, but when do you know which to choose? Here I'm going to try and explain when and with what I'd use each 'Blend Mode'.
NORMAL Normal is normal. There is no special effect, you cannot see through it. You can erase parts of it or stamp a brush onto it, but it is what it is.	+
DARKEN/LIGHTEN These two do the same thing, only at opposite ends of the spectrum. It sounds simple enough, but when you use a brush and set it to Darken the darker parts of that brush is what you will see the most of over your base image. Set it to lighten and the lightest parts of the brush will be seen over your base image. I use lighten to add light textures (though I mostly use screen - see below). Light texture by masterjinn @ deviantart.	+ / DARKEN / LIGHTEN
HUE/HUE(LEGACY) COLOR/COLOR(LEGACY) These are all really very similar. Basically, if you make a new layer and fill it with a color or brush and set it to any of these blend modes the background with be tinted with the color of your new layer. The difference between the regular and legacy is the contrast.	+ / HUE / HUE (LEGACY) / SATURATION / SATURATION (LEGACY)
SATURATION/SAT.(LEGACY) When you make a new layer and fill it with a color or texture and set it to this it can make the colors in your base image either duller or brighter (higher saturation of colors). Again, the difference between regular saturation and the one with legacy is basically just the contrast. I've used different tones on the layers to show the different variations without having to make four different icons.	+ / SATURATION / SATURATION (LEGACY)
LUMINANCE/LUM.(LEGAGY) I never use this blend mode. There may very well be a brilliant way to use it, but I haven’t been patient enough to experiment with it. Basically what it seems to do is make the colors in your icon base become as saturated as possible, usual to a degree and contrast that is undesirable. Even when I change the opacity percentage I’m not very happy with the effect achieved. It doesn’t even matter what color you use, only what brightness. I've used different tones on the layers to show the different variations without having to make four different icons.	+ / LUMINANCE / LUMINANCE (LEGACY)
MULTIPLY This is one of the blend modes I use most and find most useful for adding textures and brushes. When you set a brush to multiply all the darkest areas will stay normal and the lighter areas will be see through. Though only the black areas will stay, anything lighter will be come slightly see through, but the darker it is the darker it makes the base image below it. First brush by threeeyespei. Second brush by deleted user. Third brush by kekoah.	+ = + = + =
SCREEN This is the opposite of Multiply. I didn't put them together because I use them so much I felt they deserved seperate descriptions. When you set a layer to screen all the lightest parts will show up, the darkest parts will be harder to see. I mainly use this blend mode for light textures. First brush by innocent_lexys. Second brush by deleted user. Third brush by kekoah.	+ = + = + =
DISSOLVE I've never found a use for this blend mode. I can't see what purpose it could serve for graphic making. No matter the opacity you set it at it will look like a bunch of dots.	100% AND 80% OPACITY 40% AND 10% OPACITY
OVERLAY / DODGE Overlay and Dodge are not exactly the same, but they are very similar. They both achieve the similar effect of brightening the base image. It's just that Dodge does it to a much higher degree. So for the most part you don't want to use dodge at 100% opacity. If you want it to be a little more subtle I suggest you set the layer to overlay. I sometimes use overlay to achieve a brighter more contrasted icon. Try duplicating the base image and setting it to overlay, see what you get. Granted it won't work with every base, and you should be sure to fiddle with the opacity.	+ BASE IMAGE + NEW LAYER / OVERLAY W/ NEW LAYER AT 100% vs OVERLAY USING DUPLICATED BASE / DODGE AT 100% OPACITY vs DODGE AT 50% OPACITY
HARDLIGHT / SOFTLIGHT I use these blend modes quite a bit. I tend to duplicate the bases and set them to either softlight or hardlight. It helps to give just a little more contrast and depth to an icon. They both do the same thing only Hardlight, as it's name implies... does it harder, to a slightly harsher degree.	/ / BASE / BASE DUPLICATED & SET TO SOFTLIGHT / SET TO HARDLIGHT
DIFFERENCE / EXCLUSION Again these achieve similar effects, just to different degrees. Depending on the color and brightness it can look a bit like a negative image.	NEW LAYER / DIFFERENCE / EXCLUSION NEW LAYER / DIFFERENCE / EXCLUSION
BURN This is basically the exact opposite of Dodge. In stead of brightening it will bring out the darker tones in your base image. The darker the brush or new layer, the darker the base image will be.	LIGHTER COLOR SET TO BURN DARKER COLOR SET TO BURN / THEN SET TO 50% OPACITY

But what is a brush? Well, brushes I tend to liken to a stamp. If you can add it to your PSP brushes, enabling you to change the size and color then it's a brush. You wouldn't want to do that with a texture, as you'd mostly want to use them as-is. Brushes can be any size, as you can always resize it for the icons. So perhaps the bigger the better, as resizing to make something smaller is best. Resizing to make something bigger will make it look grainy and of poor quality. How do you make a brush in Paint Shop Pro? I explain it here.
BRUSH DON'Ts You need to make the brush a part of the image. That means creating shadow, or perhaps erasing part of it with the eraser brush set to a lower percentage hardness. Slapping on a brush doesn't make it a good icon. Where you place the brush can make or break your icon. A perfectly good base can be wasted with poor brush placement. It's not just about placement either, the wrong color or size is just as disastrous. To the right are examples of poor brush placement. Below this are the DO'S, how that particular brush should be used with each base. Also, remember LAYERS ARE YOUR FRIEND. Always create a new rastor layer for EVERY brush you use. That way they are each seperate and can be moved around and edit seperately!	ICON 01 - ICON 02 ICON 03 - ICON 04
BRUSH DO'S - ICON 01 For this particular base slapping a flower brush on doesn't quite work. The people and grass already make it a bit busy. So let's use a technique we used for text: use a texture. I used the same texture and shadow technique as I did before. Notice how in the second icon the flower brush isn't overpowering the image. However the color of the flower brush still ties it in with the image and makes the icon one cohesive unit. Again I added tiny text for a little more detail, but made sure to erase the part of the tiny text brush that was over the flower.	DON'T ------ DO
BRUSH DO'S - ICON 02 Placing a brush directly over your subject is a bad idea 98% of the time. I don't say 100% because there are always artistic exceptions. Until you've reached a level where you can tell the difference between artistic and just bad placement however, I suggest you offset your brush or place it in the less busiest part of your icon. To correct this brush don't I just moved it over off the center of the subject. It's still partially on the subject, yes; but it's over a less busy area and the white shows up well against the dark blue. I added a shadow to the brush as well to really make it stand out. Yet again I've added tiny text. I like using it, I can't help it! If you don't like the way it looks then don't use it.	DON'T ------ DO
BRUSH DO'S - ICON 03 Like I said in Text: Tips and Suggestions, don't overuse tiny text! Slapping a tiny text brush on an icon to fill space can help, but not if it looks like an essay length brush like this icon don't. If I've said it once I've said it a dozen times: LESS IS MORE. If your subject is centered then put a small line of tiny text centered over top like I've done to the right. Maybe add some real text as well, tiny text doesn't always look good alone. If you can't figure out what text or other brushes to use with the tiny text maybe the icon would just be better off plain.	DON'T ------ DO
BRUSH DO'S - ICON 04 One of the biggest issues I see with icons today is poor brush use. Please understand, putting a brush on an icon will not and cannot make it look better if you don't do it right! A bad base image cannot be made better with a random brush. And a good base image can be ruined by a bad choice in brush, I've seen it happen many more times then I would like. You may look at this brush DON't and think "Why would anyone do that?" I ask myself the same thing when I actually see people use brushes like this. What I would do to correct this particular poor brush use is move it, make the brush a different size and use it much like I used the texture in Text Meets Texture. I gave the same larger brush a shadow (at 36% opacity) and added some text (Georgia, size 6) and tiny text. The new final product is better, not great, but better.	DON'T ------ DO
EXAMPLES Here are a few more examples of icons with brushes. I tend not to use brushes very much. I mostly use tiny text, my Yankee symbol brush or any brush that will help text stand out. That doesn't mean other brushes wouldn't look good, but like I said I feel less is more.

Now what is a texture, what's the difference between them and brushes? A texture is, obviously, something you use to add texture. It's not a simple stamp brush. It has a certain texture or design. It may be done in black and white or have a brilliant color scheme. They can be as simple as a light texture, added to give the icon some dabs of color and light. Or maybe a light grey gunge texture to give the icon an aged look. Or they can be as complex as a scrapbook texture. They could also be stock images. Also, you tend to set the texture layers to something other than 'Normal' to achieve an effect. You'd set a light texture to 'Screen' or 'Lighten'. A grey texture you'd want to set to 'Multiply'. Whereas a brush layer you'd most likely want to leave set to 'Normal'. They can be 100x100 pixels or the size of a desktop wallpaper. As I've said before I prefer most of my textures to be much larger than icon sized. That way I can move it around and choose just the right part of the texture to use. I prefer having one very large texture than 10 small ones, it just seems much more versatile to me.
TEXTURE DON'Ts Just like brushes you need to make the texture a part of the image. Textures can be a bit trickier. Not only are there many different types, but because of that they have many more uses. Experiment with different types, check out different tutorials. There may be many ways to use them, but that doesn't mean there's less room for error. If anything there's more. To the right are examples of poor texture use. Below this are the DO'S, how that particular texture type should be used with each base.
TEXTURE DO'S - ICON 01 Light textures can take away just as much as they can add to your icon. They're often and easily extremely missused. Take my DON'T example, I see this far too often. Not only is the light texture covering the face, but you can't make most of it out because it's being used with an image with such a bright background. How do I fix this? Simple. First... CHOOSE A DIFFERENT LIGHT TEXTURE! One that doesn't cover the face. I used THESE light textures by ohstarfish over at tenandathird. So I chose one that fit better. Still, how do we make the light texture more visible? For this I added a grey texture by ohfreckle. I moved it around so that the big scratches didn't cover the icon, just the grey areas. I added it as a new layer and set that layer to MULTIPLY. The light texture I set to SCREEN above the grey texture.. NOTE: I played with the Hue/Saturation/Lightness a bit, but I put that layer OVER the grey texture layer and UNDER the light texture layer.	DON'T ------ DO
TEXTURE DO'S - ICON 02 A grunge texture can, like any other texture, work well if you choose the right texture for the right icon. This icon for example has the wrong texture. It's far too dark and distracting. It might work if it were paired with text that might explain why the subject is being covered. However to correct this I just chose a different texture. Texture on first icon by kekoah, texture in second icon is by ewanism. Adding a grey grungy texture can help tone down the contrast and color of an image. This can make it easier to use text or a lighter brush on top and have it visible without having to give it a shadow.	DON'T ------ DO W/OUT TEXTURE -- W/ TEXTURE
TEXTURE DO'S - ICON 03 Setting a texture to 'Multiply' is not always a good idea. It works well when you have a grey texture and a subject on a light background. However it doesn't work well with more colorfully designed textures. Even a grey texture set to Multiply on a subject with a dark background won't always look very good. Take our third DON'T example. Setting this scrapbook texture to Multiply has only succeeded in making Draco Malfoy look green. To fix this one I set the texture to 'Normal' and erased enough so that only half of Draco's face was showing. I gave the texture a bit of a shadow as well. This isn't the only option you'd have, but this is how I tend to use textures. Go here for three other texture tutorials I made back in 2006. I'm not as fond of the icons anymore but the techniques are still relevant.	DON'T ------ DO Scrapbook textures by tragic_icons.
TEXTURE DO'S - ICON 04 Stock images can be very useful textures. For this particular DON'T and DO I used this cloud stock image by noisettee. When placed as a new layer and set to Multiply you get the DON'T, but if you just move the texture a bit and set it to Screen you get the DO. A small change, but a huge difference.	DON'T ------ DO
EXAMPLES Here are examples of various ways I've used textures. I'm not sure how to teach you how to use a texture, it was such a trial and error process for me. Plus with the sheer number and varieties of textures the possibilities are so vast. It's hard to know where to begin. However HERE are three other texture tutorials I made back in '06; three different ways to use textures.

TUTORIAL SERIES:
INTRODUCTION
A GOOD BASE
BEGINNERS GUIDE TO FONTS
TEXT TECHNIQUES
>>>BLEND MODES; BRUSHES & TEXTURES: THE DO'S AND DON'TS<<<

The Ultimate Screencapping Guide for Mac Users

2009-11-24T23:40:23Z

Program(s)+version: UnRar, HJSplit, VLC Player, Automator, NameMangler, EasyBatch, ViewIt
Involves: Learning about HD files' image quality and size, how to cap, batch conversion and renaming, sorting, uploading your caps to a server and adding those caps using Coppermine for sharing in a gallery as well as uploading zip files for sharing.
Translatable: No, Mac only.
Steps: A lot.
Difficulty: Moderate

There always seems to be fairly well done tutorials for various Windows programs and how to cap out there over the years, but when I switched over to Mac over a year ago it took me a while to find all the programs I use now because there was no guide like this to simply follow. I had to learn through trial and error and searching and searching for the program I wanted for each task. There was no step by step how-to. Now that I have a set method, I figured it would only seem reasonable to share.
So for anyone interested in switching to Mac, new to Macs, or for the occasional dual Mac-PC user, this is THE guide to all things screencapping. I'll try not to get too technical, I want this to be pretty easy to comprehend.

Table of Contents
Applications You'll Need
High Definition
High Definition image quality and file size
High Definition File Size Breakdown In Comparison to Standard Definition
How to Cap in VLC
Autocapping and manual
Autocapping
Manual Capping
Batch Conversion and Renaming
Batch Conversion
Batch Renaming
Batch Conversion and Rename Using EasyBatch
Sorting Your Caps
Using ViewIt
Uploading caps to a server and Coppermine
Uploading zip files
Megaupload
Mediafire

APPLICATIONS YOU'LL NEED

UnRarX | If you have rar'd files.
HJSplit | Java version for Macs, if you have them split into .001, etc files
VLC | The screencapping program, also a standard media playing program that can play anything.
Automator | Standard application on all Macs. Once you figure out how this works, you can have a LOT of fun with this app, it can do about anything.
NameMangler | A powerful and light renaming tool.
EasyBatch | A shareware alternative to Automator and NameMangler.
ViewIt | Image viewing and sorting program.
FireFTP | An FTP client integrated into Firefox, for uploading files to a webserver.

HIGH DEFINITION*
High Definition image quality and file size
Files sizes are key in relation to the image quality in high definition files. The key point being that high definition files (HD) are large. Depending on size and how they were ripped and encoded, they can be massive, other times they can be not so massive. But usually they will be pretty huge, especially in relation to SD (Standard Definition). Now I'm gonna lay out the basics in resolutions for you.

624x352 = Standard Definition, (File formats: .avi, .mp4, .mkv, etc) - [Example] These files are actually compressed HD files most of the time. Sometimes they are straight up SD from the source, which is even lower quality than a compressed HD file. The quality of these files are severely reduced in relation to the HD files. The image size can vary depending on how it is compressed, but for tv this is the most common size. Nothing to fret much since these aren't the files I'd recommend capping from.
1280x720 = High Definition , (File formats: .mkv) - [Example] These are the most common HD files out and you can find most popular shows and freshman shows in HD, or anything on a major network or a cable network. (Note: Any caps made at "1248x704" are not HD. Simple math will point out that 1248x704 divided by 2 is 624x352-the SD resolution, only doubled. Doing so does not give you HD caps. You will get blurry caps at a high resolution, yes, but worse quality. 720p files will always be 720 pixels. If you doubt that a cap is 720p cap is not truly 720p, then you are likely correct to assume it's not true HD, and rather more upscaled images. Which is annoying, frankly. And pointless.)
1920x1080 = High Definition, (File formats: .ts, .mkv) - [Example] Unlike SD files being scaled down from 720p files, 720p are not scaled down from 1080p. These are from a separate source. This is the best quality and can be found for the most popular types of shows like Lost, Heroes and Smallville as well as most popular movies.

* You will need a fairly fast computer to play and screencap 720p, and especially 1080p/i files. At least 2.4Ghz or higher processor and 2GB RAM. I can't fairly assess what a minimum requirement would be but this is what I have and it works splendidly.

High Definition File Size Breakdown In Comparison to Standard Definition
Just a reference point so you know you're getting the right file and what to look for. Sizes for the half hour shows (that range from 20-30 minutes) will vary slightly.

720p

SD	HD
175MB	558MB
233MB	745MB
350MB	1.09GB
550MB	1.49GB
700MB	2.19GB

Movies can generally be anywhere from 4.67GB to upwards of 24GB depending on how long the movie itself is, whether it is a 720p or a 1080p/i, how the video and audio was encoded and how fast the bitrate is, among other things. Any 1080i file of tv is varied from 2 to 5GBm depending, again, on the length and how it was processed. Now that you know the basics of HD files, let's get on with how to cap them.

HOW TO CAP IN VLC
Autocapping and manual
VLC is the only freeware on Macs that I like, or have found, to be a comprehensive and powerful media player that you can not only view a myriad of files formats with but it's also the perfect (in my eyes) screencapping program. I've never found a file it couldn't play. And as long as you have a fairly fast computer, things should run pretty smoothly from here on out. If you're not sure about how fast your computer is, click on the apple in the menu bar and hit 'About this Mac' and if you'd like to know more specifications, hit 'More Info'.

Autocapping
This is a feature I only generally use with DVDs or I'm really lazy (in which I will still set the capture function ridiculously high so that it grabs enough frames). Because VLC only lets you manually cap files at source size, but you can upscale resolution if you wish for them to be larger, or leave them be. I upscale them to my screen size (1440x900) although the caps themselves will need to be measured, or the aspect ratio will go wonky. Also, you will need a version BEFORE 1.0.0 release. The new version seems to have messed up the settings or names for what is needed to cap and I tried figuring it out with no luck. I'm running on 9.8.

DVD Source Sizes
The fixed source size for PAL is 720x576 for NTSC: 720x480 or 853x480 (older things in full screen can be 640x480). Widescreen is the norm these days but you will come across full screen on some shows, like Dark Angel, It's Always Sunny in Philadelphia (although as of S5 it is in widescreen, though I've never come across any HD files of this, sadly), and The Wire to name a few of the top of my head.

Open VLC, it should look like this:

Now hit ⌘+, which will open Preferences (I like shortcuts). You may still go to VLC>Preferences if you wish. And it will look like this:

Then hit 'Advanced':

Now go to the Video tab>Output Modules:

And choose from 'Default', which is what it needs to be for manual capping to 'Image Video Output':

Then go to Image File:

Choose PNG as the file format to save in (we'll convert them to jpg later--capping straight into jpg loses quality). Set the size you'd like the caps to have. If it's a DVD and you're trying to upscale it, play a video and put it on full screen and take a manual cap.

if you're not sure what hotkey it is, go to Interface>Hotkey Settings and scroll until you see 'Take a video snapshot'. Mine is 's.' but you can make yours whatever.

And if can't find that cap you took, go back to Video and see the destination folder (any manual caps, you can choose a destination, but autocaps will always save straight to the Macintosh HD folder).
Now this image is a reference point to how big you want it. Open the cap in Preview or Photoshop, whichever image editing software you have and set the image size to whatever you'd like. Make sure the aspect ratio is set on, so when you input this upscaled ratio into VLC, the caps come out correctly.

Now go back to Output Modules>Image file and insert the new size of the caps. Set it to png and have the recording ratio be whatever you like. It depends on how action oriented a show or movie is but the more caps I want, I lower the number. For me, if I want a cap for about every half a second (hey, some good moments can last a split second, I don't wanna miss 'em!) I would set it at 10, every second about 15, if I want it every other second 18 and then if I want it every few seconds, I'd set it between 18-23. But the norm for me can be around 19 for standard slow stuff, action-y, I can get it down to 10. You play around with it and see what you like. And you may put whatever filename you prefer. i like to put the show or movie title a dash and then my username so each file is shown as made by me.

Now to figure out what title you need to cap, drag the DVD disc that is on the desktop to VLC's yellow block. Now when you've got the DVD menu up, hit play on whatever you want to cap and go to Playback>Title and see what title it is at. You need to know this so it caps what you want.

When you know what title it is, make sure to check No DVD Menus and set the title to whichever it is you want and whatever chapter you want (I'm assuming starting at Chapter 1--the beginning of anything).

Once you hit start, lower the volume and let it do it's thing. These files will save to the Macintosh HD folder. Remember to change the output back to Default if you're going to be manually capping!

Manual Capping

Set the filename and directory where you want the caps to save. Set the files to save as png.

Drag the file to VLC (or open File>Open File>Browse) and start capping using your set hotkey. The images will save the directory of your choosing.

BATCH CONVERSION AND RENAMING
Now there are two different programs you can use to batch convert. I personally bought a license for EasyBatch because it's a bit more straightforward and you can convert to .jpg and rename at the same time (or do one of the other, as well as resize, add a watermark, etc) but for those with the need for freeware, Macs come with Automator.

Batch Conversion

You will choose these four selections from the list.
Find Finder Items
Get Selected Items
Copy Finder Items
Change Types of Images

Now choose the current destination of the .png caps you want to convert jpg, set the settings like in the image 'Whose extension is equal to png' in the 'Find Finder Items' section, then the 'Get Selected Items' will get them. 'Copy Finder Items' is there to make sure if something goes wrong, the original files are still there. 'Change Types of Images' is the conversion part. Choose the destination to save them to as well as changing the file format to .jpg.

Hit 'Run' at the top and the workflow will start, and you will have your converted files in the destination you chose. If all went well, you can delete the original png files.

Batch Renaming
If you accidentally forget to change a filename setting in VLC or if you want to just change the file names, add in your username, whatever.

Open NameMangler and drag the files you want to change the name of, how many digits you want (my standard is 4 because this has precedent when adding them in a Coppermine gallery). This is also handy if you cap a file and VLC crashes or you need to finish it later and VLC starts caps back at 1 again, you can start a number to continue sequentially. Like if the last one ended at 1147, you could start the next one at 1148. Very handy, eh? Keep the step value at 1, that keeps the files one digit after another (if you did 2 it would go 1, 3, 5---which is...useless to me). Hit the checkmark and the files will rename!

Batch Conversion and Rename Using EasyBatch
If you don't mind buying the license, I highly recommend this program because it cuts out the use of two programs and puts it in one if you need

You can just choose the images from the directory, set however you want to resize, rename and convert to and the directory to save them to. I always check 'Move original files to Trash' so when the conversion and/or renaming is done, the originals go to the trash. Hit Start and the process will run and your files are changed however you set the program up to change. So now that your files are successfully capped, converted and renamed, it's ready to sort through them.

SORTING YOUR CAPS
I'm a big stickler for sorting caps. Doesn't matter if things are autocapped or manually capped, I always, always, always sort through them and weed out the bad ones. I feel it isn't far to make an excess of caps and dump half the work on the people that want to use them and make them sort through ~4000+ caps~. Whenever I see someone showing off how many they made, and they make a superfluous amount I think it's rude. As an example, I get a little trigger happy with my hotkey on some shows, like Fringe (so much pretty) and White Collar (faaaaaaces!). Most times I end up with around 3000-4000 caps. But I always sort through them, and then it's scaled down to about 1,200 to 1,700. All the scenes are there but the excessive amount has been eliminated.

Using ViewIt
ViewIt is an optional shareware program but you don't need to buy a license. It will have a 30 second delay but I always just hit Later aka Never.

Hit File>Add and open the folder of images you need to sort. Make sure they're in sequential order by going to Images>Sort by>Name. Now just go through them with the arrow and delete key. Arrow over the ones you want to keep and delete the ones you want to get rid of. I sort out images that are near duplicates, anything blurry and other things I accidentally capped I didn't need. It's all up to you what to keep or get rid of.

When you're doing going through the files, go to Tools>Choose Folder to choose the destination you want to save this new set of sorted images to.

Now hit Tools>Finder Previews and Create. The newly set of sorted images will save to the directory you chose. When finished, you may delete the original unsorted folder.

Whew. Now we're about done.

UPLOADING CAPS TO A SERVER AND COPPERMINE
This isn't exclusively Mac related and telltale covers it well here way better than I could.

UPLOADING ZIP FILES
It's all personal preference, I suppose as to where you upload and why but these are the two sites I like. You can upload zips to your host at your website if you get one, but we've had problems and had to move to many times over this so we don't upload zips to the server anymore, so we just use free uploading sites. To make a zip, you will click on an folder in finder and hit Compress or it might be 'Make Alias' if it's an older Mac.

Megaupload
I love MU because it is the most popular, common uploading site. I upload single zips of things over 100MB here, mostly.

Mediafire
I love Mediafire because while it only goes up to 100MB per file for free users, you can have unlimited parallel downloads and the speeds are good. So if you'd like to split them into separate folders of 200MB or less if the folder is more than 200MB so they can be zipped and uploaded onto Mediafire, then that's just the same. Personal preference and all that.

And now you've got your caps made and sorted and uploaded so you can share with the masses!

Please do let me know if you find anything confusing, or if you have more questions or problems and hopefully I can answer them!

An easy-to-follow, step-by-step guide on how to make smooth, good quality animated icons

2009-10-12T18:17:12Z

This is my first tutorial. I decided to make an easy-to-follow tut on how to make animated icons because I can't count the amount of times I'd searched around for tutorials on how to make animated images when I wanted to learn how, only to discover nothing made any sense or I got lost along the way because the instructions were so confusing.

Today, we'll be making this icon:

Programmes: VirtualDub (for screencapping), Adobe Photoshop CS - CS2 and Adobe ImageReady.
Involves: How to screencap, how to import frames and images, how to crop, colour adjustment layers, tweening, how to optimise animated icons to fit 40K, how to set timing for icons.
Translatable: Probably not. :-(
Steps: For screencapping: 4 steps. For making the icon: 15 steps.
Difficulty: Actually really easy once you know how to do it. But I will go with Medium - Hard.
WARNING: THIS IS VERY IMAGE-HEAVY.

SCREENCAPPING.

If you don't have VirtualDub, you can download it here. VirtualDub isn't the screencapping programme you *have* to use - if you have a screencapping programme already, then just use that. I choose to use VDub because it's very easy to use.

In order to have smooth animated images, you will need to screencap your footage every three frames. This is how many frames I screencap, anyway, and it works really well for me. (If you're not going to use VDub, then set your programme to capture every three frames and save those frames to their own folder, then skip the rest of the VDub part of the tutorial.)

Step one:

Open VirtualDub. Open your file.

For today's tutorial, I will be using the kiss sequence in the House episode, Under My Skin (5x23). Now, when you open your file, you may get a pop-up message like this:

Don't panic! Ignore it. Just click 'OK'. Then your footage should open in two windows, like this:

Step two:

You will need to scroll through the footage to find the point you want. When I am capping to make a number of animated icons, I use the scroll ball on my mouse to go through the footage slowly. But in this case, I know exactly where I want to go, so I have moved the frame cursor to the spot I want. When you have found the spot you want to start capping from, click on the mark-in button:

Then I move the frame cursor to the spot I want to stop capping and click the mark-out button:

As you can see, the segment to be capped is marked in pale blue on the timeline. That is what you will be capping.

Step three:

Now you will need to select the number of frames to cap, the file name, file name suffix and the folder in which to save the caps to.

First, create a folder on your computer for the screencaps to be saved into before you do anything. Then go up to File. Now, this next step may vary according to the version of VDub you have. That's not a problem - in older versions, you will probably need to go to File > Export > Image sequence. In my version (VirtualDub-MPEG 1.6.19 build 24568/release by Avery Lee), you will need to go to File > Save image sequence.

A pop-up wlll then appear:

In filename prefix, type in the name you choose for your files. (I chose 'bothsidesnow' as my file prefix.)
In filename suffix, including extension, type in .jpg
In minimum number of digits in name, type in 3

Then you will need to select the folder in which to save your caps. I created the folder name 'BSN' (acronym for 'Both Sides Now'), thus I selected this folder. Click on the browse folder button next to the directory to hold images bar and a pop-up will appear for you to select the folder you created earlier:

Keep your output format on JPEG at 100%. Then hit OK.

VirtualDub should now start capping your footage every three frames. A pop-up like this shoulder appear the moment you hit 'ok':

This is the progress of the capping; it will show you how many caps it has taken out of the complete number of caps, how big the file will be, and how much time is estimated for VDub to cap everything. Once VDub has finished capping, this pop-up will automatically disappear. You can close VDub now, as we no longer need it.

Now if you go to the folder you'd made earlier and selected to have the images saved in, all the caps should be there. Bravo! We're about a quarter way there now!

Step four (optional):

Okay, now depending on the pixel dimension of the gif (we'll talk about that a bit later), you can have up to 50 frames in a gif. If you noticed in the pop-up I had when VDub was capping my footage choice (see the last image), you will notice I had 492 images to be saved to the file. WAAAAY too many. If you tried to import all of those into ImageReady, you'd be waiting forever and your programme could likely crash. SO, if you've found that you've capped way too many images, don't worry! Go to the folder you've saved your images in. Create a folder within that folder. Now have a browse through them and choose a particular segment of image sequences, like what I've done here:

Move selected caps to the folder you just created and viola! You have a much more manageable amount of screencaps to work with. Repeat the above couple of steps if you plan on making more icons from the caps you've made, if you have a large number of them.

RIGHT. So, we've dealt with screencapping. Now we can finally move onto the fun part!

IMPORTING, IMAGEREADY AND PREPARING/COLOURING YOUR ICON.

This really is the fun part. There are SO many things you can do with animated icons. Try not to be daunted - it looks complicated because there are a fair number of steps but it really isn't complicated at all. Hopefully, my explanation of what to do will be easy to understand.

Step one:

The very first thing you need to do is open up ImageReady. You can open up Photoshop, too, if you like, though we won't be needing that just yet. Bear in mind that I am using an older version of PS and ImageReady, so where applications are on mine may not be the same as yours. Knowing your ImageReady and PS programmes would be handy in this case, but if you're not too familiar with either of them, don't worry. Just do some browsing around and you'll find everything I'm talking about.

Now, go up to File once you've got ImageReady open and go down to Import. A side-box will pop out, like this:

Click on 'folder as frames'. Another pop-up box will appear, a list of all your folders. We're going to be importing our screencap folder, so browse for this and when you find it, select it. Click 'OK'. A progress bar will then appear, which will tell you ImageReady is importing your folder.

Now you should have something like this:

The large box with your image in it is the first frame of your image files. Underneath that is an oblong box labelled 'Animation'. This is all your image files in frames. If you press play on the Animation box (circled in red in the image below), you will be able to watch all your frames as an animation.

Play it a few times to make sure everything is okay. It should play smoothly.

Step two:

Now, what we need to do is import these frames into Photoshop. So, go to the beginning of the animations in the Amination box and select the second frame by clicking on it. Not the first - the second.

Now scroll to the end of the animation. Press the shift button and keep it held down while clicking on the very last frame. This will select all of the frames except for the first one.

These frames need to be deleted from the animation box. (Don't worry, we'll have them all back in the animation box a little later.) Making sure all of the frames, except for the first one, are still highlighted, go down to the bin icon at the bottom of the animation box and click it. This will delete the selected frames, so you will be left with only the first frame.

DO NOT delete anything in the layers box. The frames have been deleted from the Animation box but not from the layers. If you delete the layers, you will lose them and and you'll have to start all over again. :-(

Step three:

NOW. Go up to File and select edit in photoshop. A progress bar will pop up, like this:

Depending on how many frames you have and how fast/slow your computer is, this progress can take a couple of minutes. Usually for me, it zips right across to Photoshop within a few seconds. But larger amounts of images take longer to import.

Okay. Now switch over to Photoshop. You should see something like this - your first frame in a large image and all your layers in the layer box, all of them deselected except for the very first layer:

Step four:

The very first thing you need to do is crop, if you wish to crop at all. I don't want the 'Global' watermark in my icon and I also want to focus a lot more on House and Cuddy's faces. So, I take the Selection tool and I select which area of the image I want to crop. I then go up to Image in the tool bar and go down to crop and select it. This crops every single frame in my icon automatically.

Before cropping:

After cropping:

Step five:

Next, we need to resize. This is where things can be a bit tricky. A .gif is an image that uses an RGB (a colour model that has the primary colours red, blue and green combined in various forms to create a vast array of different secondary colours) palette of up to 256 colours. The higher the file size, the more colours it uses. The lower the file size, the less colours it will use. Livejournal allows gif icons of no more than 40K, which is pretty damn small. Thus, you need to get your icon down to 40K without losing too much quality or colour. The bigger your image, the bigger the file size is going to be. The smaller the image, the smaller the file size, which means the more colours you get to keep in the image.

Confused? Lol, I hope not. But the point in my blabbering is: in order for me to have a good quality animated icon that retains a good number of colours, I need to make the image fairly small. So, go up to Image in the tool bar and select 'image size. I decided to go with 65 x 39 px. This is what I end up with:

Step six:

Now we need to create the canvas around the image so that it has a background. Go to your layer box and, with the only visible layer highlighted, select a new layer. Move that layer down to the bottom so it is now your very first layer.

The canvas needs to be enlarged so we can actually see it and use it. Go up to 'Image' in the tool bar and select canvas size. In the numerical boxes, type in 100 in both of them. Hit 'OK'. The canvas will then appear as 100x100 sized with the image of House and Cuddy still at 65x39px in the middle of the empty canvas.

You're about halfway there now! I know this is a lot of steps but trust me, once you've gotten the hang of this, you'll be making animated icons in ten minutes.

Step six:

Now we need to colour and brighten the images so they actually look nice instead of dull and lifeless. Go to your layer box and scroll to the very top layer. Right-click on it (this will make the layer appear visible but that's okay) and click on select layer transparency. This will create a "marching ants" selection border around the image House and Cuddy are in. What this does is prepares the image to have a new layer to be created with a layer mask, which will enable all the images to be brightened and coloured without affecting the rest of the canvas. Er, does that make sense?

With the image of House and Cuddy still outlined, go down to the create adjustment layer button at the bottom of your layer box (the little black and white circle button that looks a bit like a yin yang sign) and click on it. A pop-up box will appear; go up to curves and click that. A new curves adjustment layer will appear in the form of a layer mask at the top of all the images.

I adjusted the curves RGB channel to Input: 49; Output: 119. This brightened House and Cuddy significantly and makes it a lot clearer. I repeated step six again, except with selective colouring this time, to make the colours in the image stand out more. And to give the colours one final little punch, I repeated step six again except with hue & saturation to bring out a bit more contrast. So, now the top of my layer box and my image looks like this:

See how much clearer and better my image is? This same colouring will automatically apply to all the layers, so I don't need to repeat any of step six anymore.

Step seven:

Now we need to deal with the background and with text and other things to make the icon look pretty. Hide the very top image layer you made visible while you were adding colour adjustments in your layer box (don't hide the colour adjustments themselves, though) amd go back down to the very bottom of your layers and select the empty layer we created earlier. What you choose to put in your background is entirely up to you, but I chose to go with this background:

by worship_elle

I pasted it in the empty layer and got this result:

Step eight (optional):

I want to make a small border around the image, to make it sound out against the background more. So, just like in step six, I right-click on the bottom most image and select layer transparency. Then I click new layer and move that beneath the first image layer but above the layer with the background in it. Just in case that sounds confusing, here is a screenshot:

To create the border, go up to to edit in the tool bar and select stroke. A pop-up box will appear like this:

Select the 'outside' radio button and choose what pixel width you want your border to be. I chose 2px and used #7C704B as the colour for my border. Hit OK and deselect the selection border around your image. It looks nice if the border is a nice contrast to the background colour but compliments to the image, too. This is what I got:

See how it makes the image stand out against the background a little more? This step is entirely optional, of course, but I thought I would include it just in case people were interested in knowing how to create borders around the gif image in their icon. :-)

Step nine:

NOW. I swear we'll be almost done soon. I know this seems like an awful lot but I wanted to explain each step thoroughly so that everything makes sense. But now, we have to do something with the background behind the image. The adjustment layers will affect that part of the background. The best type of background to choose is one that is either the same as the background itself, or a colour that compliments the background. Scroll to the very top of your layers. Right-click on the top-most image (NOT the adjustment layers) and select layer transparency. Go back down to the create adjustment layer button at the bottom of your layer box (again, the b&w one that looks a bit like a yin yang symbol) and select solid colour. Move that layer to the very top of your layers, above the other adjustment layers.

Select the colour you want for the background fill. I chose #9E9064, a lighter colour than the border but still compliments the background. I copied this layer and dragged the copy all the way back down so that is between the border layer and the first image in the layers, as well. This is so I can make a smooth and clean fadeout later when I take everything back over into ImageReady.

Now I want to add text. With the text tool select, click on the icon and put in whatever text you wish to use. I chose the text, "I always wanna kiss you" (something House says to Cuddy just before they kiss in this scene, SQUEE!), and I set the font to Trajan at size 8px, with the colour #766125.

See how the very top image layer is visible. Click on that to make it invisible. Leave all the other layers exactly as they are. So, you should have: your text layer, colour adjustment layers, border layer, background layers and the very first frame of the icon visible. The rest should be selected to invisible.

Yes? Everything good to go? OKAY, AT LAST. We can return our icon back to ImageReady! Go to File > Edit in ImageReady and everything should import back to IR successfully.

OPTIMISING, TIMING AND SAVING YOUR ICON.

PHEW. WE ARE ALMOST THERE.

Step ten:

Now, when your layers return to ImageReady, you should see something like this:

You should see only one frame in the Animation box. That's good if you do. That means everything imported properly. On the top right corner of your Animation box, there is a little tab. Click on this and a box of options will appear. Select make frames from layers and every single layer in your layer box will appear in the Animation box.

Step eleven:

What we need to do now is delete every frame in the Animation box that isn't House and Cuddy. Press the ctrl button and hold it down, and select with your mouse the background frame, layer frame, colour adjustment frame, every single frame that doesn't have House and Cuddy in it. Once all of those are selected, hit delete and you should be looking at something like this:

FROM THIS:

TO THIS:

Now go to your layer box and make every layer that does not have House and Cuddy in it except for the very two top layers visible, so you should end up with something like this:

Step twelve:

WE ARE ALMOST THERE. DON'T GIVE UP YET! We need to adjust the timing of each frame. This is a step you must do, otherwise your animated icon will end up very slow. Beneath each frame in the Animation box is a little tab that has 0 sec on it. Adjusting each frame individually is a real pain, so click on the very first frame, scroll to the end of the Animation timeline and hold down the shift button. While holding that down, click on the very last frame in the timeline. This will highlight every frame in the box. Now click on a tab that has 0 sec on it and a box of options will pop up.

Click on Other... and a box will appear that will be titled Set Frame Delay. Set delay for all frames to: enter 0.06. The reason for this is because 0.06 is a good time in which the animation runs at normal speed. Any greater number than this will start to make the animation go slower. And any value less than 0.06 seconds will make it go faster and look like something out of The Benny Hill Show. This of course depends on how fast you want it to run. If you want your animation to go slower, then enter a value closer to 0.08 - it will run slower but still smoothly. Any greater value than that, however, will make the animation run staccato rather than smoothly.

Hit 'OK'. All the frames will automatically change to 0.06 second delay. Make sure your icon is set to loop forever (bottom left-hand corner - there is a drop-down box that has 'one', 'forever' and 'other' to choose from. Make sure 'forever' is chosen, or it will not loop properly) and now you can preview your icon! Hit 'play' in the Animation box and watch your icon a few times to make sure it runs smoothly.

My icon currently has 85 frames, which is waaay too many. This means I need to delete some frames to bring the number down, as too many frames will make the file size of the icon much too big for LJ. After watching my icon loop a few times, I worked out which bits I no longer wanted. I decided for now to get rid of the first 25 frames. I did this by selecting the very first frame, holding the shift button down and selecting the 25 frame. This highlights all 25 frames I wish to get rid of. I then click the trash can button on the bottom of the Animation box and delete those frames. So, now I am left with 62 frames. Still too many, but I will worry about that in a moment.

Step thirteen:

We now need to add in the text layer and the colour layers we'd edited in photoshop earlier. Go to the very end of your timeline. Select the very last frame and then go down to the icon at the bottom of the Animation frame that looks like a piece of paper. Click that and the frame you selected will be duplicated.

Click on the duplicated frame. Now go to your layer box and scroll right up to the top. The duplicated frame should be the very last House and Cuddy layer. Make that layer invisible and in its place select the colour layer at the top to be visible. Here is a screenshot to show you what I mean:

Step fourteen:

The next step is to fade the last couple of frames into the background colour. This step is called tweening. (This step is optional if you don't wish to fade anything. If you prefer your icon to go straight from the image to the text, skip this part.)

With the very last frame selected (the one that now has no image of House and Cuddy in it and has a solid background), find the button on the bottom of the Animation box that has three tiny circles on it. Click this and a box will pop up, asking you to enter the number of frames to add. Type in 2 and then hit OK.

The result should look something like this:

We need to do one more frame delay adjustment, this time only on the very last frame, the one without House and Cuddy, and with the solid colour. Click on the drop down box beneath the last frame that should have 0.06 sec on it and select 2.0 (or 2 seconds). The text in the little box beneath that frame should now read 2 sec.

And now, we're going to make the text layer visible on that very last frame. Keeping that last frame selected, go up to the top of your layers and change the text layer to visible.

Step fifteen:

SO CLOSE TO THE END NOW. What we have to do now is to get the icon at 40K or below without losing too much quality. How to find out what size your icon is, go to the canvas where your icon is and at the top, you will see a few tabs: original, optimised, 2-Up and 4-Up. Select the optimise tab. At the bottom of that screen, image information will appear: the percentage size of the canvas (should be 100%), and two long, thin boxes. The most important thing to take note of is the file size.

This is where this box is going to come in very, very handy:

See the settings I have it on? Set yours to the exact same: 0% web snap, 40% dither, transparency selected, no matte, again no dither and add metadata selected. These settings I have allow me to include up to 50 frames in a gif, as they free up more space that I can use for my icon size. I won't be getting 50 frames out of this particular gif but I have done with others. 'Dithering' is where you'll find your file size will go up; the higher the dithering, the higher the file size. Dithering is a feature that spreads colours out more, especially on graphics that are already low in colour. You ever seen .gifs that seem particularly "spotty" or grainy? That's dithering. And while it's a good feature where your colours are limited, it eats up your file space like whoa. I choose to go with 40% or less because that's a good range of dithering that spreads my colours evenly and doesn't take up too much file space. Anything between 0% - 40% is a good range. So, don't feel you *have* to stick to 40% dithering. Decrease to 20% or even 0% if your file size is on the large size.

On the top right corner in the box I just showed you, there is a little tab. Click on this and a box of options will appear. Select Optimise to file size... and another box will pop up. In this box, you should see a heading Desired file size and a small box next to that. In that box, type 40 and hit OK.

My image is currently at 73.53K - much too big for LJ. So, I put in the value of 40K to optimise the icon to that file size. This is result I get:

So, I've got it down to 39.31K. But look at how horrendous the icon looks. Turns out that with 65 frames, I'll only get 17 colours in the icon, hence why it looks so grainy, washed out and basically really bloody awful. This means I need to delete more frames. The best way to do it, if you're not sure how many frames to delete, is to delete slowly. Meaning, only delete a couple of frames at a time so you can see how it looks as you decrease the number of frames. Each time you delete, hit the 'optimise to file size option. I did exactly this a number of times and turned the dithering down to 0% until I finally get down to this:

MUCH better quality. So, I now have 33 frames and my icon is down to 36.16K. Which means I can optimise it again to get an even better quality. Thus, I optimise and viola! My icon is now at 39.1K and has 83 colours. Perfect. All I have to do now is save and then I am done! The best thing to do first, however, is to preview it in your web browser. On your tool palette, you should see a little symbol that represents your browser. My browser is Opera, but you can select the default browser to be Explorer (whyyyy anyone would still be using that, though, I have no clue) or Firefox. Click this button and your icon will appear in your browser so you can view how it behaves.

Does it look good? Does it run smoothly? Are you happy with it? AWESOME. Now you can save your icon! Go to File > Save optimised as and save your icon wherever you wish to save it.

And now you're DONE. WHOOO!

This is my final result:

Other results I got with the same process:

(These are free to take - just credit my art comm, bold_as_brass.)

I hope this tutorial was easy to understand and easy to follow. And please, if you end up making any animated icons using my tut, post them in the comments so I can see! I'd love to see what results people get. :-)