The rendering of document content using speech synthesis and sound effects
is fundamentally different from content rendered for visual consumption.
There is no concept of font size, intrinsic visual size to an element
or even a visual canvas where elements may exist in relation to one another.
Crude aural rendering of document content is currently possible through
serialization of text content by screen readers. Stripping down content to
this level loses most of a document's structure, and ignores many document
components outright. The Aural CSS properties introduce the concept of an
aural canvas, where elements are rendered over time, with positioning properties
existing in a three-dimensional soundstage around the user, and white-space
interpreted as silence. Styling and emphasis are applied to an element through
the use of volume, pitch, vocal stress and "auditory icons", which cue the
listener to content that is intended to stand apart from other content.
While these properties can augment visual rendering, it is only the tip
of an iceberg - there will be cases where documents can only be rendered
aurally. In these cases, the power allowed by these properties becomes obvious.
These properties allow powerful interfaces to information for a wide range of
needs: easier accessibility for special visual needs and impairments, educational
teaching tools and reading aids, as well as 'hands-off' information access
(eg: retrieving directions while driving a car.) The market for this sort of
rendered content is expected to grow considerably in the future.
These properties were incorporated into CSS2, but were first detailed in the W3C
note "Aural presentation with CSS style sheets"
(http://www.w3.org/Style/CSS/Speech/NOTE-ACSS,
January 7, 1997) edited by Chris Lilley. The properties and aural canvas of ACSS
detailed in this note were mainly adapted from proposals by T.V. Raman of Adobe.
CSS 2.1 - a change of feature status
A few major changes to aural CSS have occurred in CSS 2.1. All properties
relating to aural CSS, along with specific units relating to ACSS have all
been moved to a separate appendix (appendix A). Spec-wise, this means that
ACSS is not "officially" part of the basic CSS 2.1 feature set - a browser
that supports all features of CSS2.1 barring ACSS may still call itself
fully CSS 2.1 conformant. The "aural" media type is also deprecated in
favor of the "speech" media type.