This document defines version 2.02 of the DAISY Digital Talking Book (DTB) format. The DAISY format is based on the W3C defined SGML (ISO 8879) applications XHTML 1.0 and SMIL 1.0. Using this framework, a talking book format is presented that enables navigation within a sequential and hierarchical structure consisting of (marked-up) text synchronized with audio.

This document can be found at: www.daisy.org/products/menupps.htm

This document can be found at: www.daisy.org/z3986/specifications/daisy_202.html

If you have any questions, comments, or suggestions please contact Markus Gylling at markus.gylling@tpb.se

This specification extends and revises the DAISY 2.01 specification. These revisions are are intended to bring this specification more in line with the DAISY 3/NISO DTB specification, and to clarify ambiguities in the DAISY 2.01 specification.

The major revisions made in this version of the specification are:

The DAISY 2.02 specification is technically backwards compatible with the DAISY 2.0 specification.

Definition: The key words must, must not, required, shall, shall not, should, should not, recommended, may, and optional in this specification are to be interpreted as described in IETF RFC 2119.

This specification uses the XHTML 1.0 and SMIL 1.0 specifications defined by the World Wide Web Consortium. Bibliographic and document metadata is based on the Dublin Core Metadata Initiative element set.

As specified in the DAISY structure guidelines, the DAISY 2.02 standard supports the following types of DTB (Digital Talking Book).

To comply with the DAISY 2.02 standard, a DTB must contain exactly one NCC.HTML document and one or more SMIL documents. Depending on the type of DTB made, the DTB may also contain one or more audio files, and one or more text content documents (XHTML). Finally, the DTB may also contain an optional Master SMIL document.

The structure of these document and file types, and the DTB functionality they provide, is defined below in sections 2.1 to 2.5.

The NCC also implicitly represents the continuous playback order of all the media objects that make up the DTB. This is sometimes referred to as "the flow" of narration and/or text.

The NCC is not necessarily identical to the table of contents (TOC) of the print source. It will often contain more elements than a print source TOC, that is, the NCC may be an expanded version of the TOC based on the content and structure of the body of the book.

The NCC should be a XHTML 1.0 transitional DTD compliant document. Use of HTML 4.01 is deprecated in this version of the Daisy DTB specification.

The NCC document must be named "NCC.HTML" or "ncc.html".

For bibliographic metadata, the DAISY 2.02 specification uses the Dublin Core (DC) Metadata Initiative element set, which is an internationally approved and broadly accepted tool comprised of 15 data categories, and the rules necessary for the description of document resources. Although the DC element set covers a wide range of bibliographic description for digital talking books, there is some vital information that is not adequately covered in those 15 data categories. Additional elements specific to DAISY DTB´s have therefore been developed. These additional elements are designated as "ncc:"-prefixed elements.

The "dc:" and "ncc:" prefixes should be lower case. However, playback systems must not be case sensitive when reading these attributes.

Please note that there is one exception. Information about the edition of the print source is included by using the metaname "ncc:sourceEdition". DC does not yet have a solution for the inclusion of data regarding editions.

The scheme attribute contains references as to how the value of the content attribute has to be interpreted. Such a reference may consist of a simple syntax model, but normally it is the name of a standard or an authority list. It is not meaningful to use a scheme name that does not refer to a file or a standard. In many cases no scheme is needed.

In addition to the above definition list, it is allowed to use arbitrary metadata elements to support producer-specific metadata issues. Such meta elements shall carry a "prod:" prefix and shall be ignored by playback systems.

(All mandatory and some optional elements are included in this example)

Class attributes may be used to provide additional semantic information. Typical values for class attributes occurring on the heading elements are:

title, jacket, front, title-page, copyright-page, acknowledgments, prolog, introduction, dedication, foreword, preface, print-toc, part, chapter, section, sub-section, minor-head, bibliography, glossary, appendix, index, index-category.

The value of the id attribute must begin with a letter ([A-Za-z]) and may be followed by any number of letters, digits ([0-9]), hyphens ("-"), underscores ("_"), colons (":"), and periods (".").