The pros and cons of shapes
What good are object-oriented shapes inside Photoshop? Well, I'll tell you:
♦ Shapes are editable. Unlike pixels, you can change a shape by moving points and control handles. Likewise, you can scale, rotate, skew, or distort shapes, or even transform specific points and segments inside shapes. Nothing is ever set in stone.
♦ Shapes help to disguise low-resolution images. Sharply defined edges can add clarity to a printed image. The first example in Figure 14-1 shows a standard image printed at 75 pixels per inch. The second example shows that same 75-ppi image, but this time using an object-oriented shape outline. The low resolution works fine for the blurry fill and shadow, but where clarity is needed, the mathematical outline is there to serve.
♦ You can color a shape with a layer style. As we see later in this same chapter, layer effects such as drop shadows and beveled edges are equally applicable to shape layers as they are to standard image layers. And it's amazing what wild effects you can achieve with a shape, a style, and no talent whatsoever. To create Figure 14-1, for example, I drew a cloverleaf shape and applied the Striped Cone style from the Styles palette.
♦ Shapes result in smaller file sizes. As a rule, an object takes up less space on disk than an image. Expressed in PostScript code, a typical path outline consumes 8 bytes per anchor point, as compared with 3 bytes for a single RGB pixel. But while a shape may contain as few as 4 points in the case of a rectangle or ellipse, an image routinely contains hundreds of thousands of pixels. For example, the grayscale image shown in Figure 14-1 consumes about 132K on disk when saved as an uncompressed TIFF file. If you created an image with comparable resolution using exclusively pixels, the file would balloon to 16 times the size, or nearly 2MB.
♦ You can preview clipping paths directly inside Photoshop. In previous versions of Photoshop, you were never quite sure if you traced an image properly with a clipping path until you imported it into QuarkXPress, InDesign, or some other application. Now you can preview exactly what your clipping path will look like directly inside Photoshop.
4 Shapes expand with an image. In Chapter 3, I advise against using Image ^ Image Size to resample an image upward on the grounds that it adds pixels without adding meaningful detail (see the section "Resampling an image"). But with shapes, you can enlarge the image as much as you want. Because it's mathematically defined, the shape remains crystal clear no matter how big or small you make it. Layer styles likewise resize without problem. Suddenly, the old limitations of pixels are gone.
If vectors are so great, why not forsake pixels and start drawing entirely with shape layers instead? While a shape can clip a continuous-tone photograph, it can't replace one. Although there have been all kinds of experiments using objects and fractals, pixels are still the most viable medium for representing digital photographs. Because Photoshop's primary job is photo editing, pixels are (for the foreseeable future) the program's primary commodity.
Caution An even larger downside to shape layers is compatibility. Photoshop 6 has stretched the TIFF and PDF formats to accommodate any kind of layer — shape layers included — but that doesn't mean other programs have any idea what Photoshop is doing. As I write this, the most recent version of Adobe's own page-layout program, InDesign 1.5, doesn't recognize objects inside TIFF files. But it does support PDF files with objects. Of all the formats, PDF is the most likely to work with other programs as well. Just be sure to proof the document on a laser printer before taking it to a commercial printer. After all, when you create objects in Photoshop, you're working on the bleeding edge, so be prepared for the consequences.