For photos intended for web pages, it's usually best to keep the largest dimension of the image at 640 pixels or less (400 × 300 images are free on eBay!). If you have a lot of images, or if many of the people who will view your images have slow network connections, smaller is better.

The Resolution section in the Scale Image dialog allows you to specify the number of pixels per inch (or other units, if you choose). This doesn't change the size of the image; it's merely a note that gets added to the image so the image size can later be converted to physical units (such as inches). Most of the time, you won't need to look at these values.

The Interpolation Quality controls details of how the image is scaled. Linear, which is the default in at least some GIMP versions, does not produce the best quality, but it's fairly fast. Unless you're on a very slow machine, you'll probably want to change this to one of the higher-quality settings, such as Cubic or Lanczos. These settings will run a little slower, but they'll produce much better-looking images, especially if you ever scale images larger instead of smaller. I recommend changing this value permanently, in the Preferences dialog under Tool Options, so you don't have to think about it when you're using the Scale dialog.

Usually, when you click Scale in the dialog to make a digital photo smaller, and GIMP rescales the image, the window will resize to be tiny (Figure 2-3). Don't be fooled: the image probably isn't as small as GIMP is showing you.

Figure 2-3. The scaled image now looks too small!

What happened? Since the original image was too large to fit on the screen, GIMP showed it as "zoomed out"— in the example in Figure 2-3, it is shown at only 33% of actual size. After the image is scaled, GIMP keeps the same zoom setting (it may or may not make the actual window smaller, depending on your setting for Resize window on zoom in the ImageWindows preference category). So now, even though the image could easily fit on the screen, you're still seeing it at 33%. The solution is to zoom after rescaling. Use the Zoom menu in the image window, the menu under View

Scaling an image is easy, and it's an important operation when you're sharing images with other people. For many images, scaling may be the only operation you need. But if you plan other operations that may affect image size, such as rotation or cropping, do them first. Scaling should usually be the last thing you do when you're editing an image; that way, you have the greatest control over the image's exact final size.

The exception that proves the rule is a very slow machine. If you scale the image first, subsequent operations will go much faster. However, you'll find that this matters only with very slow machines—or very large images!

Most of the time, Image

You'll find it in the Toolbox next to the Rotate tool, or through the menu Tools

The Scale tool is a bit trickier to use than Image

Third, by default it doesn't maintain the layer's original aspect ratio. You may end up with a "funhouse mirror" version of your image where everything looks tall and skinny or short and fat. Keep aspect in the tool options can fix that (there are also options to keep height while varying width, and vice versa). It also offers Interpolation, Clipping, and Preview options similar to those for the Rotate tool.

For improving digital photos, you probably won't want to use the Scale tool. But keep it in mind for later when you're using lots of layers and doing crazy things to them. It might be just the ticket then, as it does have the advantage of giving a live preview of what the image will look like when scaled. Sometimes that's awfully helpful.

For sharing full-color photographs, the best format to use is usually JPEG. Pronounced "jay-peg," the name comes from the group that defined the format, the Joint Photographic Experts Group. Usually, a JPEG file will have an extension of ".jpg", or occasionally ".jpeg". JPEG images are highly compressed and are encoded in full color, so it's a very efficient format for photographs or other images with a wide range of colors.

The drawback to this format is that the compression is lossy. That means that every time you read a JPEG file, make a change (however minor), and write the file back to disk, the image quality degrades slightly. Don't use JPEG for images you plan to edit over and over. But as a format for exchanging photographs with other people, JPEG excels. (There's no quality loss when you copy the file from one place to another, only when you edit it and write it back to disk.)

You may be wondering, "My camera stores images as JPEGs. Does that mean I'm losing quality?"

The answer is yes. That's why many cameras provide a non-lossy raw format (discussed in the section, "Other Formats") as well as JPEG. However, JPEG files are so efficient that you can store many, many more images than is possible with raw mode on the same memory card. Use raw format if you're worried about making sure you preserve every detail, or if you're bothered by the image quality of a JPEG; but for most people, the slight quality loss caused by using JPEG on a camera, plus editing an image once or twice and saving it again in JPEG format, will not be noticed. The space savings are worth it.

The Graphics Interchange Format, pronounced "giff" with a hard "g" as in "graphics," is an indexed format. This means that it uses a fixed list of colors instead of encoding every color separately. This is very efficient for images with a small number of colors, like a five-color corporate logo.

GIF can represent up to 256 colors (256 is 2 to the 8th power, so this is also called 8-bit color ). Typical photographs have many more colors than that (typically tens or hundreds of thousands), so saving a photograph in GIF format will usually result in very poor quality. Even worse, with 256 colors, the file will usually be larger than a full-color JPEG version of the same image! The lesson is clear: don't use GIF for photographs; use it only for simple icons and logos.

So what is GIF good for? The GIF format offers two very useful features: transparency and animation. With transparency, you can make an icon with a clear background. If you display it on aweb page, or on a button in a program's user interface, whatever's behind the icon (a color, or even another image) will show through. GIF doesn't allow for partial transparency; a pixel is either fully transparent or not transparent at all.

GIF animation allows you to create images that move. Most web browsers support animated GIF images. (Whether this is a good idea is unclear; some users dislike animated images on web pages, and will avoid websites that use them. But animations have their uses.) To learn how to create GIF animation, see "Making Simple GIF Animations" in Chapter 3.

PNG, pronounced "ping," once stood for "PNG's Not GIF," though it's now sometimes referred to as the Portable Network Graphics format. It's a relatively new format, originally intended as a replacement for GIF because of legal issues regarding the GIF format.

PNG offers two modes: it can be used for full-color images, like JPEG, or for indexed images, like GIF. In full-color mode, it's not nearly as efficient as JPEG: a PNG will be much larger than a JPEG of the same image. However, it's not lossy, so if you read a full-color PNG, make a modification, and write it back out, the quality will be just as good as it was before. This makes full-color PNG a good format for storing original copies of your own images if you might want to edit them later.

Indexed PNG is just as efficient as GIF (sometimes more efficient) for images such as logos and icons with a small number of colors, and it can support more than 256 colors. Nearly all web browsers support basic PNG, so it's safe to use PNG images on web pages. PNG supports transparency, like GIF, and it also supports partial transparency so you can have translucent areas. However, some browsers (such as older versions of Internet Explorer) don't support PNG transparency, so transparent PNG images on a web page may not always display properly.

PNG doesn't support animation. There's a format called MNG that adds animation to PNG images, but no web browsers support it yet. For animated images, GIF is pretty much the only choice.

This is GIMP's own format. When you're editing an image and you have a lot of layers and paths and other information you want to save, this is the format to use. Otherwise, don't. XCF files are usually quite large and can only be read by GIMP, not by other programs. Since XCF files are so large, you can compress them using the GZIP or BZIP2 formats—save to a file name ending in xcf.gz or xcf.bz2, and GIMP will handle the compression and decompression when it reads or writes the file.

Tagged Image File Format (with a file extension of either .tif or .tiff) is another full-color, non-lossy format. Like full-color PNG, it's not very compact (don't use it on web pages—many web browsers can't display it anyway), but it's fine for keeping originals of images you might want to edit again. The reason I recommend using PNG instead of TIFF, aside from web browser compatibility, is that TIFF isn't a single standard, but many different standards with different interpretations. A TIFF written by one program may not always read correctly in another program. GIMP will read most TIFF files, but sometimes it will "have a tiff" about certain types of TIFFs.

One advantage to TIFF over other full-color formats is that it handles a wider range of colors. For the technically minded, it can handle 16 bits per color channel. This is important to some professional artists and graphic designers. However, GIMP doesn't currently handle 16-bit color anyway (that's planned for a future release), so if you're editing in GIMP, you don't get any benefit from using TIFF.

There are more image formats. Lots more. Too many more. Here are some of the most common:

JPEG's quality adjustments come by varying the amount of image compression. JPEG's compression takes advantage of tricks of human perception; it tries to degrade quality in ways that won't be very noticeable. By experimenting with the Quality setting in the Save as JPEG dialog (Figure 2-5), which appears after you specify a file name ending in .jpg, you can find a balance that gives you a very small file size but still a pretty good-looking image.

Figure 2-5. JPEG save options

Begin by making sure the checkbox for Show Preview in image window is checked. This lets you see the effect of changing the Quality setting. The file size will also be shown in the dialog. GIMP's default Quality setting is 85 when saving to new files (when editing existing files, it uses whatever Quality setting is already there). The value 85 is a good choice for photographic images stored locally, but you may find that you can get away with settings as low as 50 with surprisingly little degradation. Figure 2-6 shows an example of how the quality can change with different compression settings, but I encourage you to try this on one of your own images to see the differences in quality and file size.

Figure 2-6. Comparing JPEG quality settings. The original (at Quality=85) was 16kb.

You may be tempted to move the slider to 100% for images you store locally. Surely that would be best for images you're storing, and would prevent loss of data from JPEG compression?

Alas, even at 100%, JPEG still loses some data every time you save. Using a setting of 100% will produce a file two or three times larger than a setting of 95%, and the quality won't be any better. There's no advantage in using quality settings above 95%. If you want to save without any loss, you're better off using formats such as PNG or TIFF.

Finally, GIMP has some other JPEG settings in the Advanced area of the Save as JPEG dialog (Figure 2-7).

Optimize gives you an additional reduction in file size without any further reduction in image quality. It's on by default, and there's no reason to change that.

Progressive is a useful setting for images that will be uploaded to the web. It makes the image load in a different way, so people viewing the image will see a poor-quality version right away, which gradually improves, instead of seeing the image load line by line starting from the top.

Smoothing, Use restart markers, Force baseline JPEG, Subsampling, and DCT method control details of the JPEG format. You shouldn't need to change these unless you want very fine control over the way the JPEG file is saved.

Figure 2-7. Advanced JPEG settings

Save EXIF data is a good option to know. JPEG includes a way of holding information (called metadata) about the file using something called Exchangeable Image File Format, or EXIF. Most digital cameras add EXIF information about the date the photo was taken, the resolution, and the camera's settings, such as lens focal length and whether a flash was used. GIMP cannot show or edit this EXIF information (that will probably be available in some future release), but it can preserve the information so that you can view it with other programs. If you don't want the EXIF information preserved, deselecting this box will make the file slightly smaller.(You can see exactly how much smaller by watching the File size value at the top of the dialog.)

EXIF can embed a thumbnail of the image in the file. Many cameras include this thumbnail, which adds several kilobytes, but most programs don't do anything useful with it.

Deselecting Save thumbnail prevents saving this extra information, making the resulting filesmaller.

XMP data is another metadata format (from Adobe). Some of the same tools that can read EXIF can also handle XMP, but it's not as widely used as EXIF.

Comment is a place to put any text you might want to add to an image. You can use this to include your name, a copyright notice, or details about where the photo was taken or what it portrays.

Load Defaults and Save Defaults let you save settings you want to use most of the time. For instance, if you always want to preserve EXIF but don't want the thumbnail, and you find an 83% quality setting is enough for you most of the time, you can set those values in the dialog and click Save Defaults. GIMP still won't use them for every image—if you load a JPEG file that uses different settings, GIMP will use the image's settings unless you change them—but the defaults are helpful for JPEG files you create or convert from other formats.

GIF and indexed PNG quality settings are quite different from JPEG. Since these indexed formats use a fixed number of colors, you can make the files much smaller by reducing the number of colors they use.

To do this, you need to know about GIMP's image modes. Most images edited in GIMP are full-color images, denoted as RGB in the title bar of the image window (for instance, Figure 2-3).But images that are destined to be saved as GIF or indexed PNG should be converted first using the Image

Figure 2-8. Convert Image to Indexed Colors dialog

When converting an RGB image to indexed mode, you first need to choose a palette, the set of colors that will be represented in the final image. You can choose a predefined palette, such as a web-optimized palette, or one from the custom palette menu; you can choose a black-and-white palette, if you know that the image contains only thosetwo colors; or, the most common case, you can let GIMP create a palette based on the colors actually in the image. You specify the number of colors, up to a maximum of 256, and GIMP will try to approximate any other colors that exist in the image by using combinations of the colors in the palette (Figure 2-9).

Figure 2-9. Different levels of indexing, showing file sizes when saved as GIF. Note that all the GIF versions of this photo, even the smallest, result in a bigger file than the best JPEG.

This process of approximation is called dithering : it involves combining pixels of several different colors. You can choose from several types of dithering that create different effects, or you can specify no dithering, which may create a smaller file size and a cleaner appearance for images with sharp lines and only a few colors. See Chapter 8 for a detailed discussion of optimizing indexed images.

The Convert Image to Indexed Colors dialog does not include a Preview button. So you'll probably need to run it several times: choose some settings, click OK, and then use Undo to go back to RGB mode and try again with different settings. It's worth spending some time on this if you're trying to squeeze an indexed GIF or PNG image as small as possible.

When saving an RGB image as GIF, if you skip the step of converting it to indexed mode, GIMP will offer to do the indexing for you. In this case, GIMP will choose a palette intended to represent as many of the image's colors as possible. This helps preserve image quality, but it doesn't do much to reduce the file size. You may even end up with a larger, yet poorer quality file than if you'd used JPEG. You're much better off doing your own conversion.

The Save as GIF dialog offers a few additional options:

Saving an RGB image as PNG will save it as a full-color PNG. If you want indexed PNG, you must convert the image to indexed mode before saving it. That said, the Save as PNG dialog offers yet another set of options:

Finally, PNG offers a settable Compression level. Unlike JPEG, PNG's compression is not lossy. The only disadvantage to using high compression levels is that it will take slightly longer to save the file. Most of the time difference happens when saving; it doesn't take a viewer much longer to uncompress a highly compressed PNG image. So usually it's best to use the largest amount of compression possible.

The simplest level adjustment tool is Brightness-Contrast (Figure 2-12).

Figure 2-12. The Brightness-Contrast dialog. The original image is too dark.

The basic controls in the dialog are simple: two sliders, one for brightness and one for contrast. You can also type in numbers (the values go from −127 to 127, with 0 as the current image), but most of the time you'll probably just drag the sliders until the image looks best.

In addition, the dialog has a checkbox marked Preview. With Preview on, any changes you make to the sliders will be shown in the image window, so you can see exactly how your changes will look. No actual changes will be made until you click OK. Many GIMP dialogs have a Preview option; it's a useful feature that you'll want to leave enabled unless you have a very slow machine.

If you get the settings too far off, the Reset button at the bottom of the dialog will set them back to zero so you can start over. Cancel will dismiss the dialog without making any changes to the image, while OK will take the current settings and change the image accordingly.

Figure 2-13, 2-14, and 2-15 show some examples of the changes brightness and contrast can make. Play around with the sliders to get a feel for what they do. You'll probably find that when you make an image brighter, you'll want to increase contrast a bit, or it will look washed out. You may also be surprised at how well a slight contrast increase can bring out brighter colors in an image. Be careful of pushing contrast too far, though, or you can make the image look grainy.

Figure 2-13. Just brightness increased. The image looks washed out.

Figure 2-14. Just contrast increased. Notice how colors seem brighter, even though the image is still too dark.

Figure 2-15. Both brightness and contrast have been increased.

The Brightness-Contrast tool is straightforward and useful. But GIMP has two other brightness tools that are much more powerful: Levels and Curves. As with Brightness-Contrast, make sure that Preview is checked so that you can see the effects as you play with settings.

Levels (Figure 2-16) has two sets of controls: Input Levels and Output Levels. Input Levels shows a histogram representing the current brightness, or tonal range of the image: this is a graph where the horizontal axis represents brightness, and the vertical axis represents the number of pixels that have that brightness. If the image is fairly well balanced, the histogram should be spread fairly evenly across the image. Such an image is said to have full tonal range. If an image is way under- or overexposed, the histogram is probably bunched up at one end, or shows a spike over to the left or the right. Such an image has a reduced tonal range—it is not using all the possible pixel values that can be displayed.

Figure 2-16. The Levels dialog. Notice the values bunched up at the right (light) end: that's the white sky in Figure 2-12.

Both input and output have sliders that you can play with to make the image brighter or darker, or give it more or less contrast. The three input sliders are referred to as shadow, midtone, and highlight, representing the darkest points in the image, a midpoint, and the brightest points.

Sometimes you can use the histogram for guidance: drag the left input slider to where it lines up with the left edge of the thick part of the histogram, and line up the right slider with the right side of the histogram (Figure 2-17). Don't neglect to experiment with the middle input level slider, which is also called gamma value—you can fine-tune the contrast of the image with it.

Figure 2-17. Lining up the input sliders with the histogram. The bits of the histogram bunched at the far right are because of the bright sky.

You can also use the eyedropper buttons under the input levels sliderto pick points in your image that should be all black or all white: click on the left eyedropper button, and then click on the darkest point in the image. Then do the same to match the right eyedropper to the lightest point in the image.

At the dialog's upper right are two buttons, Linear and Logarithmic. These control the way the histogram is drawn. They have no effect on the operation of Levels, but switching to Logarithmic can make the histogram easier to see by making it taller.

One common point of confusion is that the sliders on input and output seem to work in opposite directions. If you move the leftmost input slider (at the dark end) to the right, the image gets darker; if you move the leftmost output slider (also at the dark end) to the right, the image gets lighter. What's going on here, and how can you remember which slider does what?

Think about it this way: the input sliders let you expand the tonal range of an image that has a restricted range. The leftmost shadow slider controls the black level: anything in the original image that was darker than the position of this slider will be mapped toblack. Similarly, anything to the right of (brighter than) the rightmost highlight slider will end up entirely white.

Once the image has been remapped according to the positions of the input sliders, the output sliders do the opposite: they let you restrict the tonal range of the final image. Moving the leftmost output slider right means that nothing in the final image will be fully black: anything that would have been black will now be a little bit less dark. Conversely, moving the rightmost output slider to the left means that anything that would have been white will now be a little less bright.

So input expands tonal range; then output restricts it. The leftmost input slider, which represents the blackest pixels in the image, gets mapped to wherever the black output slider is positioned; and likewise the right input slider gets mapped to the right output slider (Figure 2-18).

Figure 2-18. The input sliders, after the image's tonal range has been expanded, map to the output sliders to compress the final tonal range.

In most cases, full tonal range is a good thing; once you've adjusted the input sliders to expand your image's tonal range, you wouldn't want to restrict it again. Most of the time, you won't need the output sliders at all. They're there for the rare cases when an image needs to have a restricted range, or for images that are legitimately lacking some part of their tonal range (no highlights or no shadows).

HISTOGRAM GAPS

If you open the Levels dialog on an image when you've previously made a brightness, levels, or curves adjustment, you may notice gaps in the new histogram. What do the gaps mean? The following image shows gaps in the histogram after correcting brightness. Logarithmic mode is used to make the histogram easier to see.

The gaps represent lost information. When GIMP uses the input sliders to expand the tonal range of an image, some of the fine differences between pixels at the extremes of the image (in the very bright or very dark parts) may be lost. "Lost information" sounds bad, but most of the time, it wasn't anything you cared about, or you would have set the sliders differently. But you can see the result in the histogram when you look at the image. This also means that you can sometimes use GIMP as a forensic tool, to tell whether someone else's image has been edited in this way.

Another thing you may notice about the Levels dialog is the Auto button. This is theoretically comparable to the Auto Levels in some other photo-editing programs, but in practice, I find that it seldom helps photographs very much. (This may say more about which photos I try it on than it does about GIMP's Auto Levels.) It's always worth a try; you can fine-tune from there if it doesn't do quite the right thing, and you always have the Reset and Cancel buttons if you don't like the effect.

The Curves dialog (Figure 2-19) adjusts many of the same values as Levels, but in a different way. Some people find it more intuitive than Levels; others less so. You'll probably find that you prefer one or the other; which one you use is purely a matter of preference.

Figure 2-19. The Curves dialog

In the Curves dialog, the horizontal bar at the bottom of the graph represents input values, just like the input graph in the Levels dialog. The vertical bar represents output levels. But you don't need to know that to use Curves. To begin using it, just click on the line and drag it up a little (to make the image brighter) or down a little (to make it darker).

As soon as you click on the line, GIMP creates a control point : a handle you can use to grab the curve and drag it higher (brighter) or lower (darker). You can go back and click again on an existing control point and slide it up or down along the curve, or you can leave existing control points where they are and click somewhere else on the line to create more control points (Figure 2-20). If you need to get rid of a control point you've made, just slide it all the way left or right to another control point or to one of the endpoints of the curve.

Figure 2-20. Control points in the Curves dialog

The advantage of the Curves tool (besides being fun to play around with) is that you can easily control which parts of the image are adjusted. The upper-right part of the curve corresponds to parts of the image that are already bright, while the lower-left part corresponds to darker areas. If most of the image is fine, but there are some bright areas that look blown out and you just want to tone those down, you can make a curve that's straight except at the upper right, where it curves below the original line to make the bright areas darker. Then if the dark parts are too dark, you can grab the lower-left corner and drag it upward to brighten just those parts. You can make as many control points as you wish, and as complicated a curve as you desire, to fine-tune the correction you will make. (The catch is that such complex curves seldom accomplish much more than a single control point moved to the right place on the curve. But try it yourself. Experiment and see what you can do!)

In addition to the three general tools for adjusting exposure, GIMP's Colors menu has several other useful tools.

Threshold (Figure 2-21) lets you map an image to black and white, adjusting the threshold point between the two. Like Levels and Curves, Threshold shows a histogram of the image's brightness. The area between the black-and-white sliders, colored blue, represents the range that will be white in the final image. Everything else will be black. You can drag the sliders directly, or mouse down in the histogram area and drag out the area that should be white.

Figure 2-21. The Threshold dialog. Notice the black sky, corresponding to the white area at the far right of the dialog.

Posterize (Figure 2-22) reduces the number of colors in the image, creating a gaudy display like you might see on an artistic poster. You can achieve a similar effect by converting an RGB image to indexed mode, but posterize is a quick way to achieve this effect, and it still leaves you with an RGB image in case you want to do further editing. The effect is most obvious when you use a small number of colors.

Figure 2-22. A posterized image

Desaturate (Figure 2-23) removes the colors from an image, changing it to a grayscale image without requiring you to convert to grayscale mode. The three choices in the dialog let you choose the gray level in three ways, giving subtly different results. (GIMP 2.2 and earlier didn't offer a choice of methods.) Desaturate has no preview, so you'll have to try it and use Undo if you don't like the results.

Figure 2-23. A desaturated image

Invert reverses every color in the image, turning it into its own photographic negative. This is most useful with black-and-white images but can have an interesting effect on color images(Figure 2-24).

Figure 2-24. An image with the colors inverted

The Auto menu includes a set of automated functions useful for correcting photographs with poor exposure. They're worth trying if you have a difficult image you're trying to save; if you don't like the effect, there's always Undo.

I shoot lots of verticals : photos where I rotate the camera by 90 degrees to the left or right in order to make an image that's taller than it is wide.

That's fine, but when I put it on a website, or load it into GIMP, everything's sideways! The camera doesn't know that the image should be rotated, so it just saves it in the normal way, wider than it is tall. (Some cameras are smart enough to notice when they're rotated, and save the information in the file. Starting with 2.4, GIMP detects this and asks whether you want the image rotated.)

But fixing a rotated image in GIMP is incredibly easy, using the Image

The final option in the menu, Guillotine, is a different operation (not a rotation) and will be discussed in Chapter 7.

Do you ever shoot a photo of a landscape, and notice later that you didn't hold the camera perfectly horizontal?

Okay, maybe you don't, but I have to confess I make that mistake from time to time. Fortunately, GIMP comes to the rescue with the Rotate tool (Figure 2-25).

Select the Rotate tool in the Toolbox. There are several other tools that look similar, but remember, you can always hold the mouse over a tool (technically: hover ) and read the tooltip to remind yourself which tool is which. Now look at the tool options—in particular, the Transform Direction option. For correcting a photo with a non-level horizon, set Transform Direction to Backward, and set Preview to Grid.

Figure 2-25. The Rotate tool

With the Rotate tool selected, click in the image to begin the rotation process. The Rotate dialog appears. You can use it to type in an explicit angle in degrees, or to change the center of rotation. Most of the time, though, you won't need to do that.

If you have Preview set to Grid, you'll see a grid of lines drawn on the image.

Now drag in the image and watch the grid change. The goal is to line the grid up to where the horizon should be in the image (Figure 2-26). When you're happy with how the grid lines up, clicking the Rotate button will rotate the image (Figure 2-27). Rotating a large image takes a while: GIMP's progress bar, at the lower right of the image window, gives you an idea of how long the operation is taking (there's a Cancel button next to it, if you change your mind).

Figure 2-26. The Rotate tool in action: lining up the grid lines

Notice that rotating makes the picture larger: the corners hang off the edges of the image. Behind the image, you see a gray checkerboard pattern. This pattern isn't really part of the image; it's how GIMP represents transparency. Your rotated image is now on a transparent background. If you would prefer the image cropped to its original dimensions, with no transparency added, thecheckbox for Clip Result in the Rotate tool options will do the crop for you. Alternately, you can increase the image size to show the corners using Image

Figure 2-27. A rotated image (not clipped)

The Rotate tool has quite a few options: