Web Design Primer

Chapter 6 – Image Optimization

This chapter provides a brief introduction to images for the web. More details are given in Chapter 14, Digital Photography for Web and Cross-media, and Chapter 15, Photoshop for Web and Cross-media.

The most recent versions of web browsers can read RGB images in JPEG, GIF, PNG-8 and -24, and SVG (Table 6-1). Images for web sites should be saved at final size and 72 ppi resolution (except SVG which is resolution-independent).

Table 6-1. Images Supported by Browsers
Format Compression Type Transparency Description
JPEG variable contone no Joint Photographic Experts Group, continuous tone, multiple compression levels
GIF variable indexed yes Graphic Interchange Format
PNG-8 none indexed yes Portable Network Graphic, 8-bit, indexed color
PNG-24 none contone yes Portable Network Graphic, 24-bit, continuous-tone
SVG none indexed yes Scalable Vector Graphic, line drawings and flat artwork

Web vs. Print

If you’re used to working with images for print, you probably know that their resolution should be twice the screen ruling, or 300 ppi for a 150 lpi halftone screen. For those without a printing background, presses can only print solid ink. To reproduce properly, images must be broken down into nearly invisible dots. The screen ruling describes how many dots there are per linear inch.

If you work with page-layout programs like Adobe InDesign, you also know that you should size images before placing them in pages; but you can also adjust the size in InDesign, as a percentage of the original size.

Images for the web should also be sized for correct reproduction. Although they could be resized in Dreamweaver or by using the height and width styles, resizing will not produce the optimum file size. That is, sizing them smaller will make the resolution too big. Resizing larger may reduce the quality by producing a pixelized or stair-stepped image. Optimum file size is important because it speeds up page loading.

Web Image Formats

Continuous-tone. Continuous-tone, or “contone”, refers to bitmapped images that are photographs or drawings with many shades and blends of colour. Contone images are best saved in JPEG or PNG-24 format (Figure 6-1).

Comparing three (3) identical images in different formats (TIFF,JPEG,PNG)
Figure 6-1. The same image is saved in TIFF, JPEG, and PNG-24 formats (then combined in JPEG format for this figure). Photo Richard Adams. Image licensed under CC BY 4.0. 

Flat colour. Flat, or indexed, colour refers to bitmapped images of drawings and line art that have a limited number of colours (Figure 6-2). Examples are comic strips and logos. Flat colour is best suited for GIF or PNG-8 format.

The same image is saved in GIF and PNG-8 format.
Figure 6-2. The same image is saved in GIF and PNG-8 format. Photo Richard Adams. Image licensed under CC BY 4.0. 

Vector graphics. Vector graphics include resolution-independent drawings and diagrams drawn in Adobe Illustrator, CorelDraw, and similar programs where points, lines, curves, and shapes, along with stroke and fill, are defined by mathematical equations (Figure 6-3). Vector graphics can be saved in SVG format for display on the web. SVG defines vector graphics in XML format. Linked SVGs are displayed in Dreamweaver as a separate tab with the XML code.

Browsers cannot display print-based vector graphic formats, including Adobe Illustrator (.ai) and encapsulated PostScript (EPS). Adobe portable document format (PDF) files can be viewed with browser plugins but display as separate documents.

SVG files can be imported into HTML pages using the same <img> tag as for bitmapped images.

Adobe Illustrator has an Export > Save for Screens dialog box similar to Photoshop’s and includes the option to save vector graphics in SVG, along with PNG, JPEG, and GIF bitmapped formats. Designers should consider whether scaling will be necessary, given scaling of the graphic will be limited by screen size.

A vector-graphic drawing of a bird Kiwi drawing saved in SVG format created in Adobe Illustrator
Figure 6-3. This Adobe Illustrator vector-graphic drawing was saved in SVG format and placed in an HTML page with the <img> tag. (Kiwi drawing courtesy of Chris Coyier, CSS-tricks.com.)

Working with Images

Compression. Depending upon the file format, bitmapped images can be compressed to varying degrees. Lossless compression reduces the file size while retaining image quality. Lossy compression, on the other hand, reduces file size further by discarding some information from the image, which lowers the quality.

Photoshop. If you have Photoshop, the best way to save a bitmapped image for the web is to use the File > Save for Web dialog box (Figure 6-4). It offers a choice of contone and indexed formats along with options for colour levels and compression.

Screenshot of Photoshop’s Save for Web dialog box
Figure 6-4. Photoshop’s Save for Web dialog box offers the opportunity to preview and save images in various file formats. Adobe® Photoshop® screenshot(s) reprinted with permission from Adobe Systems Incorporated.

Photoshop’s Slice tool is a convenient way to divide an image into multiple zones, each of which can be saved in the most suitable file format. This would be convenient if you created a site mockup in Photoshop and wanted to use parts of the mockup in the actual web site.

Apple Preview. Apple Preview can also be used to resize and crop images. They can be saved in TIFF, JPEG, and PNG formats.

Transparency. Transparency means that if part of a bitmapped image is masked, or a vector graphic does not contain a filled object as its background, the web page’s background image or colour will show through the image. Formats that support transparency include GIF, PNG-24, and SVG.


Icon for the Creative Commons Attribution 4.0 International License

Web Design Primer by Richard Adams and Ahmed Sagarwala is licensed under a Creative Commons Attribution 4.0 International License, except where otherwise noted.