Categories World Wide Web

Continued from Media Types

Aural style sheets are designed to help Web publishers create presentations that can be heard rather than read. With these extensions in the style sheet specification, you will be able to hear a synthetic voice read a Web page without special equipment.

Aural style sheets provide easier access to the Web for people with disabilities and also make it easier to surf the Web when you're on the go. For example, your car's on-board computer could read Web pages to you during your morning commute. (Think computers in your car are a long way off? Think again. Quite a few manufacturers already sell devices for mobile computing.)

If you've ever worked with speech synthesis, aural style sheets may seem quite intuitive. On the other hand, if you've never worked with this type of technology, you'll need to spend some time experimenting with it. At the most basic level, aural style sheets are a new twist on an old technology--voice synthesis software. For the Web, the voice synthesis software could be implemented as a browser add-on.

You use voice synthesis software to simulate various types of voices by modifying the volume, frequency, and speed of computer-created speech. Aural style sheets build on this basic model by providing a rich set of controls over the speech synthesis. You can add pauses to the speech, for example, and auditory cues such as a bell. You can also mix in background sounds.

Beyond this, you can control the stress, richness, and spatial properties of the voice as well. To change the stress of an element in a Web page, you emphasize or de-emphasize it by controlling the amount of inflection associated with the element. To change the richness of an element, you add resonance so that the voice carries or reduce the resonance to make the voice softer. By manipulating the spatial qualities of voices, you can create 3-D sound or even sound that seems to come from different areas of the room. Can you imagine a Web page that could carry on a running conversation with many different voices?

Next: Visual Rendering Model

Published as Internet User in the 12/1/98 issue of PC Magazine.