BEFORE YOU BEGIN
Just jump right in.
77 Convert Between Image Formats
78 Change Photo Resolution and Size
106 About Size and Resolution
One of the key factors affecting the size of an image file is the maximum number of colors it can include. If a file is theoretically capable of including a large number of colorseven though it may actually contain very fewthe file's size will be large to ensure that capacity. If you're working on an image to be shared over the Internet, small file size is often a high priority. One way you can reduce a file's size is to change its color mode the number of colors an image can contain, even if it doesn't actually contain that many. However, reducing the number of colors an image can use might lead to striation and patchiness in areas you think should be a single, solid color.
A full-color image in RGB color mode generally does not actually contain all 16 million-plus colors that standard video cards support. (For RGB color mode, each color channelRed, Green, and Bluemust be capable of "counting" to 256 for each pixel. 256x256x256=16,777,216.) To reduce the file size in a color image, the Editor gives you the option of switching to an Indexed color mode similar to the encoding scheme for GIF images. With an Indexed color mode, the entire image uses only 256 colors, although these colors are selected from all the 16,777,216 hues the standard video card produces. If your image is black and white, or black and white plus gray, there are other modes you can use to make your image file even smaller. How perceptible the difference is, when changing to a lower color mode, depends on the image you're working with. For this reason, the Editor makes it possible for you to sample different color reduction modes, enabling you to choose the least detrimental mode for your image.
Because some commands are available only for images that use RGB or grayscale mode, you might sometimes find yourself temporarily increasing an image's color mode (from grayscale to RGB, for example). This won't, however, improve the resolution or quality of a low-resolution imageincreasing an image's color palette simply makes more colors available for use; it does not tell the Editor where to use them in an image to boost detail and clarity.
109. Change Color Mode
If file size is your main priority, you can compress an image into GIF, JPEG, or PNG format (for example), a process that also reduces its color palette a bit more scientifically than the method discussed here. You can also convert an image to grayscale to reduce its color palette.
Choose Image, Mode
In the Editor, open the image you want to convert, and save it in Photoshop ( *.psd ) format. Choose Image, Mode from the menu bar. Select the color mode you want to convert to from the submenu that appears:
Technically speaking, the number of bits (binary digits) required for an image to encode the color value for one pixel is the base-2 logarithm of the maximum number of colors. In other words, 2 raised to that power equals the maximum number. It takes 8 bits to encode up to 256 values, and 24 bits to encode up to 16,777,216 valuesthus the arithmetic behind the phrase 24-bit color .
If Indexing Colors, Choose Options and Click OK
If you're reducing colors in an image with Indexed Color mode, select how you want the Editor to choose the colors for the palette by choosing from various options in the dialog box that appears. Before you begin making selections, enable the Preview option so that you can see how your selections affect the actual image. From the Palette list, choose one of the following options:
If the image already uses 256 colors or fewer, the Palette option is automatically set to Exact , which means that all colors in the image are added to the palette. You do not have to make a selection.
The options in the Forced list instruct the Editor to override some or all of its palette color choices and to include specific color values, some of which you can choose yourself from the Forced Color dialog box that appears. These "forced" choices may or may not be represented in the actual image, but they are included in the image's palette:
If the image contains transparent pixels and you want to retain them, enable the Transparency option. If the image contains semitransparent pixels, open the Matte list and choose a color to blend with them to make them fully opaque.
To reduce the side effects caused by using a smaller number of colors than the original image contained, select the dither pattern you prefer from the Dither list:
Because it uses a mathematical formula, error diffusion (when used in images with a very limited color palette or large blocks of color, such as comics art) can sometimes generate artifacts in a color-blended area, more so than using an ordered dither method such as Pattern .
If Converting to Grayscale, Click OK
When you're converting a color image to various hues of gray (grayscale), click OK ; if the image has multiple layers, you'll be asked whether you want to flatten all layers before proceeding. Click Merge .
If the color layers currently in the image use blend modes other than Normal to create its current appearanceespecially if that appearance depends on how the color of one layer interacts with the colors of the layers beneath itthese effects will probably be completely lost if the image is flattened while converting it to grayscale. To preserve the layers and their blend modes, click Don't Merge in step 3.
If Converting to Bitmap, Choose Options and Click OK
When converting an image to pure black-and-white ( Bitmap mode), the Editor could simply make relatively dark pixels black and the relatively light ones white. However, the result might not be desirable, so you might want to apply dithering.
First, let the Editor convert your image to grayscale by clicking OK . It's easier for the Editor to convert grays to black-and-white than to convert colors directly to black and white. If there are multiple layers, the Editor warns you to flatten them first; click OK to have it do that and continue. In the Bitmap dialog box that appears, in the Resolution area, make sure that your image is set for the resolution of your output device. At first, this is set to the image's current resolution. To ensure best appearance, you might have to adjust resolutionand thus, sizeaccordingly. For onscreen use, choose 72ppi; for printing, choose 150300dpi. Altering this setting resizes the image, both in print and onscreen.
In the Method area, choose how you want the Editor to apply dithering. The 50% Threshold option applies no dithering whatsoeverlight pixels are made white, and dark ones are made black. The Pattern Dither option applies a geometric dithering pattern, which might be adequate if your original image is a simple drawingsuch as a corporate logorather than a photograph. Diffusion Dither applies an error diffusion pattern, distributing vast differences in brightness value over wider areaswhich is generally more appropriate for photographs.
To finalize your choices, click OK .
After changing the color mode of your image, make any other changes you want and then save the final image in JPEG or TIFF format, leaving your PSD image with its layers (if any) intact so that you can return at a later time and make different adjustments if you want.
If you're curious about what color mode your image is currently using, simply look at the title bar, after the image filename.