One of the easiest uses of a layer mask is to make some text (or any other layer) fade out.

Just follow two easy steps:

Then just drag in the image to draw the gradient. You can drag exactly vertically or horizontally (remember, holding down the Ctrl key while you drag will help by constraining the gradient's angle to a multiple of 15 degrees), or you can drag in any direction you like.

In this case, I started with a layer that had the text and drop shadow already merged together, and then dragged diagonally, from the lower right toward the upper left of the text-and-shadow layer (Figure 9-1). You can see the gradient in the layer mask's thumbnail in the Layers dialog.

Figure 9-1. A gradient on the layer mask makes the text fade.

A layer mask can also give an image a fuzzy border, so that it fades into the background. This gives a nice effect for images included on web pages.

The procedure is straightforward. First choose your favorite selection tool. Turn on feathering (you'll probably want as much feathering as possible), and then make a selection where you want your border to be (Figure 9-2).

Figure 9-2. Make a selection where you want the border to be.

Now add a layer mask to the image layer: right-click on the layer in the Layers dialog and choose Add Layer Mask... (In some GIMP versions, you may not be able to create a layer mask unless the image has transparency. If Add Layer Mask... is grayed out, use Layer

Now you have an image with a fuzzy boundary. Everything outside the selection is transparent, so GIMP will show it with a checkerboard pattern.

To find out how the image will really look on a website, there are two additional steps. First, hide the selection with View

Figure 9-3. Make a new layer mask, initialized to "Selection."

Figure 9-4. Adding a white background layer and making the selection invisible gives you a much better idea of how the border will look.

If you're happy with the result, make the selection visible again (that way, you won't be surprised the next time you try to select something), and then cancel it (Select

What if you want more feathering than the 100 pixels that GIMP allows in its selection tools?

While you still have the selection, you can use Select

A better solution is to make sure the layer mask is selected (it will have a white border, as in Figure 9-5)—and then blur it with Filters

Figure 9-5. A selected layer mask has a white border.

Blurring the existing 100-pixel feathering by another 250 pixels gives a more gradual border, as shown in Figure 9-6.

Figure 9-6. Blur the layer mask to get a more gradual border.

But wait—what's wrong with Figure 9-6? Notice what happens at the edges. All this blurring has made the border extend beyond the edges of the image. If you put the image on a web page, you'll see squared-off edges, as in Figure 9-7.

Figure 9-7. Blurring beyond the size of the image creates squared-off edges.

This happens because the oval frame gets bigger when it's blurred, and the image isn't big enough to contain it. For a really fuzzy border, you have to leave quite a bit of room around your original selection. If you want a lot of fade at the boundary, your best bet is to avoid cropping much until the very end.

Of course, since the layer mask is just a black-and-white image itself, you can modify it in a variety of ways to make all sorts of borders, like the ones in Figure 9-8. Experiment! You may not like most of the effects, but you might also find the perfect look.

Figure 9-8. You can make all sorts of frames using layer masks.

Figure 9-11 shows all of GIMP's layer modes, using the tree from Chapter 4 against a background of light blue/cyan.

Figure 9-11. All of GIMP's layer modes

Be careful not to draw too many conclusions from this one diagram, though. The effects of each layer mode can vary considerably depending on the colors of the layers involved, as you'll see. Also, though it may not seem like it as you read through, all these modes have their uses—which are often not what you'd expect at all! But more on that later.

Dissolve and the two special drawing tool modes Behind and Color Erase were discussed in Chapter 4. Now I'll cover the other modes, what they do, and some ways to use them. You'll learn even more uses for the layer modes in Chapter 10.

B - F

(128, 128, 128) - (255, 0, 0) = (0, 128, 128)

Addition mode is straightforward: the two pixel values are added together:

If that makes any of the colors greater than 255, then that color is set to 255. Addition will nearly always make the image lighter.

Subtract is the opposite: the front layer is subtracted from the layer below it:

Anything that would become negative will be set to zero. This will make colors darker, naturally. You'll often end up with a lot of black or near-black from using Subtract mode.

Taking the difference of two numbers is the same as subtracting, isn't it? Usually that's true. But in GIMP,Difference mode is, well, different. Difference is a subtraction too: but after subtracting, if any color values are negative, GIMP uses the absolute value of that color instead of truncating it to zero. Think of it as showing the difference between two colors regardless of the direction of that difference.

This means that instead of a lot of dark areas, the regions that would have been black become lighter again—especially in the colors that would have been absent. You can see that in Figure 9-11: areas of the trunk that were dark blue in Subtract mode become magenta in Difference mode.

Difference sometimes creates strange, garish colors and unusual transitions between colors. It can be useful when you need that sort of effect. In Chapter 10, you'll see several other uses for Difference mode.

Multiply multiplies the value of the pixels in the two layers (bet you already guessed that). This generally makes the image darker.

Why should that be? Multiplying two numbers should make a bigger number, shouldn't it? And bigger numbers are lighter?

The trick is that the result is normalized to 255 (in other words, divided by 255). Otherwise, too much of the image would end up being too bright.

So, the equation for Multiply is

Divide is a little more complicated. As you learned in grade school, you can't divide by 0 (at least without getting infinity). To guard against division by 0, GIMP adds 1 to the front layer's pixel values. Like Multiply, Divide is normalized, though this time it's normalized to 256 (1 is added to the normalization because of the 1 that was added to the front layer's pixels).

Divide mode often results in nearly white pixels that look "burned out."

These four modes are closely related to Multiply and Divide. Burn and Overlay are similar to Multiply and make the image darker; Dodge and Screen are more like Divide, and make an image brighter.

Dodge divides the back layer by the inverse of the front layer (the inverse comes from subtracting the front layer from white, just like Colors

In Burn mode, it's the back layer that is inverted rather than the front layer; then after the division, the result is inverted again:

Burn ends up looking a lot like Multiply, though there are some differences. In Figure 9-11, you can see that they both got darker, but Burn took on the blue color of the background layer, while Multiply turned black. With another color combination, though, Burn might end up darker.

In Screen mode, the values of the visible pixels in the two layers are inverted, multiplied, and the product inverted again:

This sometimes results in a washed-out look and a bright image with muted colors.

Overlay is a combination of Multiply and Screen modes. The back layer is inverted, multiplied by twice the front layer, added to the original rear layer, and then multiplied by the original rear layer. (If you didn't follow that, don't worry about it.)

It usually doesn't darken nearly as much as straight Multiply, though in this case, it has taken on so much of the background color that you can scarcely see the tree any more.

Hard light is another combination of Multiply and Screen modes. Depending on the color of the rear layer, you can sometimes think of it as an opposite of Overlay mode.

The formula is complicated because it has two parts: one for dark colors, and another for bright ones.

Hard light can sometimes be a useful mode for combining two photographs when you want bright colors and sharp edges.

Soft light is not the opposite of Hard light, but it does tend to soften edges and dim colors. It's yet another mixture of Multiply and Screen modes, and its effect is most similar to Overlay mode. In fact, in some versions of GIMP, Soft light and Overlay modes are identical due to a bug.

If Rm is the result from Multiply mode, and Rs is the result from Screen mode, the equation for Soft light is

These two are easy: they do just what their names imply. For each color component of each pixel, GIMP will show you the darker (or lighter) of the two components. So Darken only will make the image darker, while Lighten only will make it lighter. So Lighten only would be

To get Darken only, just use Min instead of Max.

Grain extract is supposed to produce something akin to the "grain" you might see in photographic film. In reality, I find it's more useful for giving images an embossed and desaturated look.

Its equation is surprisingly simple:

Grain merge adds texture to an ordinary image, as if adding film grain or texture from an artist's canvas. The equation is the mirror of Grain extract:

You can use these two modes together to combine the texture of one image with the color of another, as you'll see in the next chapter.

These modes are straightforward once you understand what they do, but their names can be a bit confusing.

Hue mode takes the hue from the front layer (unless the front layer's saturation is 0), but takes saturation and value from the back layer. Saturation and Value modes, likewise, take the named value from the front layer and everything else from the back layer.

Color mode gives you the color of the front layer with the lightness of the back layer. It's like Hue, except that it takes saturation as well as hue from the front layer.

You'll see in the next chapter how you can use this mode to colorize an image, using a photograph as the back layer and a solid color as the front layer. It's as if you took a black-and-white version of the back layer and covered it with a colored filter—only it works better than simply using a translucent-colored layer in Normal mode, since that would darken the image unacceptably.

Saturation and Value modes aren't needed very often (though Saturation can be used to remove color from selected parts of an image), but Hue and Color modes can be quite useful for colorizing images, as you'll see in Chapter 10.

Here's a straightforward example of how to use layer modes to create interesting logos. Start by making some text in basic black on a white background (Figure 9-15).

Figure 9-15. Start with some black text.

For once, rather than keeping the text layer separate, in this exercise you should merge the two layers. Right-click on the text layer in the Layers dialog and select Merge Down(Figure 9-16).

Figure 9-16. Merge the text layer into the background.

Make a duplicate of the text layer, Gaussian Blur it (Filters

Figure 9-17. The duplicate text layer has been blurred and inverted.

For the next step, you need to replace the original white background on the bottom layer with a medium gray. (Remember your old friend medium gray?) Turn off visibility temporarily on the blurred and inverted layer so that you can see what you're doing on the lower layer. Then select a medium gray foreground color: in the color chooser, set H and S to 0 and V to 50, or type 808080 into the HTML Notation box.

In theory, you should be able to use GIMP's Select by Color tool to select all the white background, and then use Edit

Antialiasing smoothes the text by making pixels at the edge slightly transparent, blurring the interface of type and background. That means that along the edges of the letters, the color isn't exactly black or exactly white, but some in-between shade of gray. Select by Color will leave some of these areas unselected, and when you fill with gray, you'll end up with speckles left over (Figure 9-18).

Figure 9-18. Light speckles left behind when you try to use Select by Color on antialiased text

You could clean those up by hand, or by using operations like Select

So what can you do? Layer modes to the rescue! Make a new layer with the Layer Fill Type set to Foreground color, so you have a layer of solid gray. Use the down arrow in the Layers dialog to move that layer to the bottom of the stack, below your black-text-on-white-background layer.

Then set the text layer mode to Multiply, and magically, the white background turns to gray (Figure 9-19).

Figure 9-19. Multiply mode changes the white background to gray smoothly, without any unwanted speckles.

Why does that work so well? How does Multiply get rid of the speckles so cleanly?

It's a trick. No antialiasing speckles are harmed in the generation of this background. Instead, it's darkening the white to become the background color, without changing the black text (or its gray antialiasing) very much."Darken only" works as well as Multiply, and the trick works for colored backgrounds as well as gray ones. On a gray background like this one, you could also use Burn; but with a colored layer underneath, Burn will change the color of the black text. (As always, if you're not sure, try several layer modes—or all of them—and see what works best.)

Getting back to the 3-D letter project: turn visibility back on for the top layer (the one you blurred and inverted). Then set that layer to Difference(Figure 9-20).

Figure 9-20. Set the top (blurred, inverted text) layer to Difference mode.

Nice effect! But you can get an even nicer, more 3-D effect by using the Move tool to move the top (blurred, inverted) layer a few pixels in any direction.

Don't like the gray color? No problem. Make another solid-color layer, any color you like, put it on top, and set it to Overlay mode (Figure 9-21).

Figure 9-21. Add a color overlay.

Or set it to Darken only, which maintains a medium-gray background while turning the text your desired color (Figure 9-22). That might be useful if you want to select the text and paste it somewhere else. (If you need to paste the text without the background, try using a copy of the original text or blurred text to paste into the QuickMask or use as a layer mask. Don't forget that you need to click the Anchor button in the Layers dialog after pasting into either the QuickMask or a layer mask.)

Figure 9-22. Try Darken only mode for the top color layer.

I'm sure by now you realize that I've only given you a tiny idea of the useful effects you can get by combining modes. It would be an epic task to list all the possible combinations. So, make lots of layers and experiment for yourself!

Start by making a drop shadow from the Tux layer.

Now use the Perspective tool to drag the shadow off in the right direction (Figure 9-24). Try to match the shadow angle that you see on other objects in the same photo. Don't worry if the base of the shadow doesn't quite match the base of the pasted object; you can fix that separately.

Figure 9-24. Use the Perspective tool to drag two corners of the shadow box in the right direction.

Until now, the shadow layer has been on top of the Tux layer. Obviously that's not right. Now is a good time to move the shadow layer down in the layer stack. Then use the Move tool to move the shadow so that its base is in the right place (Figure 9-25).

Figure 9-25. With the shadow layer under the Tux layer, move the shadow to the right place.

That looks pretty good—but you're not quite done yet. Change the transparency of the shadow layer so that it's not completely black. Then you can see some of the texture and color underneath it (Figure 9-26).

Figure 9-26. Give the shadow layer some transparency, so you can see the colors and textures underneath.

Of course, sometimes you may need more work to make a shadow look perfect, especially if it's being cast onto a curved or irregular surface. But for most of your shadow needs, the Perspective tool combined with Blur and layer transparency will give you everything you need.

You probably remember your introduction to perspective in Chapter 4, when you drew the box for the tree planter. That was a reasonable introduction to single-point perspective (Figure 9-27).

Figure 9-27. A classic example of single-point perspective: three boxes drawn with one vanishing point

The "point" in question is the vanishing point. This is the place where parallel lines seem to converge in the distance. Imagine looking down the lane stripes on a long stretch of road. You'll note after some distance you can no longer distinguish them as separate stripes. If you can see far enough, the entire road disappears to a point in the horizon.

Since there are three boxes in Figure 9-27, you can see what happens when foreground objects are offset to one side or another. But that tells you something else: the vanishing point does not have to be in the center of the image.

Use any object (such as your hand, or the opposite page of the book) to block out two boxes. Now you have an example of a drawing where the vanishing point is offset to one side.

This is similar to the effect at work in the Tux image. But in Figure 9-26, the point where the shadows converge is way off to one side. Why is that, and how do you figure out where the vanishing point should be?

In Figure 9-27, all the horizontal lines are parallel to each other, and so are the vertical lines. The only lines that have a different angle are lines that recede "into the paper." Real images are often more complicated; we can do better.

How? By adding another vanishing point, as in Figure 9-28.

Figure 9-28. Perspective using two vanishing points. Note the guide layout is very different from single-point drawings.

Not coincidentally, the box has been drawn so that the angle from it to the left vanishing point is similar to the angle the shadows follow in Figure 9-26. What happens if you scale up part of the two-point drawing and paste it onto the Tux picture (as in Figure 9-29)?

Figure 9-29. Centered on Tux (the focus of the image), the two-point cube's perspective lines show the relationship of Tux's shadow to the hiker's.

An understanding of perspective will help not only with drawing, but any time you have to make an alteration to an existing image. Getting the angles, warps, and shapes "just right" can seem impossible using LARTM ("Looks About Right To Me"), but is easy when you know how to draw perspective lines.

GIMP doesn't let you set up guides that aren't exactly vertical or horizontal to help with the Perspective tool. But that's okay: by drawing lines in a separate layer, as shown here, you can still find your vanishing points, and then let them guide you when using the Perspective tool. Then make the layer invisible before you save.

In many cases (such as this one), the vanishing point is so distant, it's off the edge of the picture. That's perfectly normal. Sometimes old-world draftsmen (in the days before computers) would stick a pin into their drafting board way off the work, and line up their guides with a ruler.

Not so easy with a screen! But you can use the same trick you see in Figure 9-29. Make your guides smaller on a separate layer, and then scale them up to the proper size.

By using vanishing points with GIMP's Perspective tool and adding a realistic shadow, you can convincingly stick just about any object into even the most distorted background.

In Chapter 4, you learned how to create simple parametric brushes. But what if you want to create a brush of a different type?

The Clipboard Brush

The easiest way to make a brush out of an image is the Clipboard brush (new in GIMP 2.4). Simply copy anything in GIMP, then look at the list of brushes either in the Brushes dialog or in the Brush menu in the Toolbox. The first item will be a brush made of whatever you copied (Figure 9-36).

Figure 9-36. The Clipboard brush puts whatever you last copied as a brush in the Brush menu and Brushes dialog.

Simple Image Brushes

Any image can become a brush. But GIMP has two types of image brushes: color and grayscale.

A grayscale image brush works like most of GIMP's standard brushes: anywhere the brush is black, GIMP will paint in the current foreground color. Anywhere that's white will not affect the image. Be sure to use white rather than transparent for grayscale brushes: GIMP will refuse to save a grayscale brush with transparency.

Warning

If you try to save an image as a grayscale brush and you get an error message, check to make sure the image is really in grayscale mode, not RGB (it should say grayscale in the window's title bar). If not, use Image

Color image brushes keep their color when you paint with them (like the Vine brush discussed in Chapter 4). Most good color image brushes will be small and will have transparent backgrounds.

Hey—that describes Wilber! How would you go about making Wilber (Figure 9-37) into a GIMP brush?

Figure 9-37. Wilber, the GIMP mascot

Making an image into a brush is easy—just save it in the right place using the right image type (Figure 9-38). The right place is the folder check-marked Writable in Preferences

Tip

If you have trouble finding your GIMP profile folder in the Save as dialog, try clicking on Browse for other folders; then right-click in the file list and select Show Hidden Files.

Figure 9-38. Save as "GIMP brush" (.gbr) to the brushes directory in your GIMP profile.

The right type is GBR, for "GIMP brush." You can expand GIMP's Select File Type list and click on GIMP brush, or just save it to a file ending in .gbr and let GIMP figure out the type.

Once you click Save in the dialog, you'll be prompted for a Spacing and Description (Figure 9-39).

Figure 9-39. Options when saving a brush

Spacing is how far apart the images will be when you draw lines and curves with the brush (Figure 9-40). The Description is the name that will show up when you choose the brush from the Brushes dialog.

Figure 9-40. Spacing controls how far apart the images will be drawn.

After you've saved a brush, you need to Refresh brushes—the button in the lower right of the Brushes dialog—before GIMP will see the new brush. Of course, exiting GIMP and starting again also works. Then you can use your brush.

Although it's fun to draw silly lines made of Wilbers, a more practical use for simple custom brushes is as a stamp, or an image you need to use frequently. Suppose you have a signature that you like to add to your artwork. You could keep your signature in a file and open it every time you needed to use it, copy it, and then paste it into the new image. But if it's already loaded as a GIMP brush, it's always accessible through the Brushes dialog. Much easier, and you don't need to remember where you stored the file.

Animated Brushes

The real power of custom brushes is in animated brushes. These are made from images with multiple layers, just like the animated image you made at the end of Chapter 3. As you drag the mouse across the canvas, instead of seeing the same object repeated many times, you'll see slightly different objects. GIMP's Vine brush, and others such as the Pencil Sketch brushes, Felt Pen, Confetti, and Galaxy, all use animation to give you a random-looking, constantly changing line. The animated brushes have a red corner in the Brushes dialog, so you can tell them from other types of brushes.

To illustrate, I'll demonstrate how to make a simple animated brush: some cartoon grass.

Start with a small image. I'll use 50 × 50 pixels. It shouldn't have any background, just a completely transparent image. (If you start with an image that has a background, Add Alpha Channel followed by Clear can fix that.)

Now draw a blade of grass (Figure 9-41). Notice that I zoomed in the window to 400%: it's hard to draw in a 50 × 50 window at normal size.

Figure 9-41. A single blade of grass against a transparent background

Tip

Since GIMP doesn't offer any way to draw a line that starts wide and then tapers to a point, I drew each blade as a constant-width line, then zoomed in and used Lasso Select and Clear to taper their tips.

Make a new layer, and draw another blade of grass. Grass will look more realistic if each blade is a slightly different shade of green as well as a different shape.

Continue adding new layers until you're happy. I'll stop at six layers (Figure 9-42).

Figure 9-42. Six blades of grass, each a different color and shape and in its own layer

Now it's time to save the brush. Use the same folder you used for simple brushes—the brushes folder in your GIMP profile directory—but this time use the GIMP brush (animated) type, a .gih extension (Figure 9-43).

Figure 9-43. Save with a .gih extension: type "GIMP brush (animated)."

This brings up the Save as Brush Pipe dialog (Figure 9-44). Pipe is another term for an animated brush.

Figure 9-44. Options when saving an animated brush

Spacing is the distance between consecutive brush marks, just as it was with simple brushes. It's specified as a percentage of the brush size. You'll have to experiment with this setting to find the right one for your brush. For grass, 8 or 10 works fairly well.

Description is simply a text description you can give to your brush. Number of cells is the number of layers in your image. GIMP should initialize this for you.

The rest of the options—Cell size, Display as, Dimension, and Ranks—control how the frames of the animated brush are chosen from the image, and in what order. For a simple animated brush like this, leave Dimension at 1 and set the first Ranks column to the number of cells (6, in this case).

Note

For more complicated brushes, you can choose to divide each layer into a grid, and then use each grid cell as a frame in the brush. Dimension and the other Ranks fields tell GIMP how the grid is set up and in what order to choose the frames. These make brush design more complicated; you're probably best off ignoring them until you get comfortable with animated brush design. They're used for fancy effects like changing orientation according to the direction you're drawing.

The menu next to the Ranks entry controls the order in which cells are chosen. Random picks a random layer each time; Incremental uses the layers in order; angular and velocity let you create more advanced effects, such as changing the brush image according to the direction or speed the mouse is moving. The other options, pressure, xtilt, and ytilt, use those values from a drawing tablet if you have one. If you don't, you're better off sticking to Random or Incremental.

Figure 9-45 shows the effect of drawing with the grass brush.

Figure 9-45. Drawing with the animated grass brush

You've probably used some of GIMP's built-in patterns by now. But you can get a pattern from a photo too. For instance, what if you want a pattern of spots copied out of a cheetah photograph (Figure 9-46)?

Figure 9-46. Cheetah spots to use as a pattern

The easiest way to test a pattern is to use the Clipboard pattern, which works just like the Clipboard brush. Copy your pattern, and it will automatically appear as the first item in the Patterns dialog, or in the Pattern Fill checkbox you get in the bucket fill tool (Figure 9-47). Drag from the Clipboard pattern into an image window, or use Edit

Figure 9-47. The first pattern is a copy of whatever's in the clipboard.

Well, of course there's a catch. When you fill with a new pattern, you'll almost always see a tiling effect, as in Figure 9-48.

Figure 9-48. You can see seams at the edge of each tile.

What went wrong? There's a sharp boundary where the bottom edge of one copy of the pattern doesn't match the top edge of the next copy (and similarly with the left and right edges, though the effect is less obvious). The eye is very good at picking out boundaries. That won't do!

Fortunately, GIMP has a filter to help with creating patterns which can be found at Filters

After running Make Seamless on the cheetah pattern, copying it, and filling with it, you get something more like Figure 9-49.

Figure 9-49. After running Make Seamless, you can't see the boundaries as easily.

Running Make Seamless sounds easy, and it is. But it's not perfect—notice the repeating group of spots you can see in the cheetah pattern when it's tiled. Patterns with straight lines, such as bricks, may end up warped or blurred in unnatural-looking ways. Even fairly random patterns like cheetah spots can become blurry or blotchy. If I move the selection over just a few pixels in the original cheetah image, copy it and use the Make Seamless command, it comes out like Figure 9-50.

Figure 9-50. Moving the selection over slightly can make a big difference in how realistic a pattern looks, even with Make Seamless.

The key to choosing a good pattern is finding one that doesn't change much from one edge to the other. If it's brighter in one corner, if there's a big dark spot in the middle or a line along the bottom, if there's an obvious shape inside...it will be obvious where one tile ends and the next begins. Sometimes this is acceptable (and even some of GIMP's built-in patterns show obvious edges), but most of the time you won't want that.

There are also plug-ins you can download, such as Resynthesizer, and other techniques such as a high-pass filter and the Clone tool, that can go beyond Make Seamless to help make your patterns tileable.

After making brushes, I'm sure you can guess how to save your pattern once you're happy with it. Navigate in the Save as dialog to the brushes folder in your GIMP profile, and save it as type "GIMP pattern" (with a .pat extension). When you're prompted for a description, choose a name that makes sense. Then open the Patterns dialog and click the Refresh button in the lower right.

You can make your own gradients, too, by using the Gradient Editor.

The easiest way to learn the Gradient Editor is to edit an existing gradient. Select a gradient similar to your goal, make a copy of it by clicking the Duplicate button, then click Edit to get the Gradient Editor (Figure 9-51). You can also use the context menus of the Gradients dialog (right-click on any gradient in the list, and then select Edit to edit that gradient, or New to make a new one).

Figure 9-51. The Gradient Editor

At the top of the Gradient Editor is the name (for a new gradient, "Untitled"). You can change this to the name of your new gradient.

Below the name is an image of the gradient as it currently looks, and below that, an area for range selection sliders—the blue-and-white bar with black-and-white triangles.

In the range selection area, black triangles define segments. Each segment has two endpoint colors. White triangles represent the midpoints of ranges. The currently selected range is blue; the others are white.

In Figure 9-51, there are two segments. The first (leftmost) segment is active, since that's the one with the blue bar below it. Its left endpoint color is blue, and its right endpoint color is yellow. The second segment has a yellow left endpoint, and a blue right endpoint.

Each segment will fade from the color at one endpoint to the other. How fast it will fade is controlled by dragging the midpoint slider (the white triangle). Drag it to the right, and you'll see more blue; to the left, and you'll see more yellow. You can also drag the black endpoint sliders, or even drag the entire selected zone.

You can control many aspects of blending, and add or delete segments, by right-clicking in the gradient to get a context menu (Figure 9-52). The most important entries are Left Endpoint's Color and Right Endpoint's Color to choose the colors used in the current segment.

Figure 9-52. The Gradient Editor's context menu

When you've edited a gradient to your liking, the Save button adds the gradient to your GIMP profile. You don't need to choose a file name or file type for it: GIMP will handle that for you.

Most people won't ever need to create or modify a gradient. But for those who do, GIMP's Gradient Editor gives you plenty of power to do whatever you need. Experiment and see what it can do for you.