XML by design was created to help manage the same content with many different formatting or publication options. This leads to a number of situations in which XML could help various people with handicaps, not least among them people with visual impairments, for whom large print, Braille, audio and/or a combination of the three, might be the only way to access the content of a printed book. The Dutch Library for the Blind approached Dutch prepress technology firm Mediasystemen with the need to support enhanced formats through a single-source, XML based system. Custom development involving the Digital Talking Book followed, creating a solution that other, similar libraries could employ to increase the range and efficiency of their content transformation activities.

XML by design was created to help manage the same content with many different formatting or publication options. This leads to a number of situations in which XML could help various people with handicaps, not least among them people with visual impairments, for whom large print, Braille, audio and/or a combination of the three, might be the only way to access the content of a printed book.

The Dutch Library for the Blind (NLBB) in Den Haag was established to convert plain text into enhanced text as a service to Dutch-speaking visually handicapped people. The NLBB had a problem, though: It was taking too much editor time to transform plain printed matter into the different augmented formats needed to serve blind and visually impaired people. In addition to that problem, each enhancement, from large print to Braille to narration for audio book, was a separate step whose result was stored and distributed separately, depending on the reader’s particular needs. Also, Braille output itself represented unique challenges to efficient operation.

Developments in content management as well as a need to upgrade some publishing processes made it clear to the NLBB that the future of enhanced publishing for the visually impaired could, like other multi-output publishing projects, benefit from a single source in XML. Such a source could add multiple enhancement options to a single file that could be accessed in Braille, large print or audio versions, at once.

With the addition of a Digital Talking Book standard, it became possible for the NLBB to serve an even broader clientele eventually through the Internet and HTML. So the NLBB decided to analyze its overall transformation and editing process towards adopting a single source system. The advantages of single-source for the NLBB would be the same as for any organization: no need to re-edit or re-format content for a new output channel. Additionally, due to the synergy between XML and HTML, publication to the Internet now became a real, within-budget additional possibility.

Before adopting an XML solution, the NLBB was already publishing enhanced content in the form of Braille, audio books, large print editions and through a proprietary electronic access option. About 60 editors and 80 other employees produced, per year, about 8,000 enhanced editions per year, including large print, audio books and ink print orders such as transforming menu cards into Braille. These editions serve about 11,000 clients worldwide. Clients are asked to pay a voluntary fee for the service but about 90% of the cost is absorbed by Dutch government sponsorship.

NLBB editors already worked in Microsoft® Word. With the Digital Talking Book DTD to govern XML, what was left to establish in the overall workflow were a content management system, an XML editor that worked with Word, and a way to translate XML to certain unusual outputs (a Braille printer, for instance). Mediasystemen in Haarlem was called in on the project for its expertise in publishing workflow and content management and for its knowledge of the XML editor WorX for Word.

The following presentation outlines the Lexus Studio solution proposed by Mediasystemen, a publishing solution that other, similar libraries for the blind have already begun to inquire into. The system reduces the editing process time by reusing single sources for each output channel. The sources are automatically streamed through the channels in such way that editor intervention is supported for each step in the production process. This semi-automatic approach is a strong point that profits by the qualities of both automation and human interpretation, thus saving time while assuring accuracy.

The most important innovation of Lexus Studio is its support for Braille. At the start of the Braille production process, an ink print book is scanned using Omnipage. The result is a picture of the book that is interpreted with OCR functionality. To allow intelligent ink print interpretation later on the production process, a formatted output (HTML) is chosen to stream into Lexus Studio. Then the system is able to determine what the formatted text could mean. At this stage of development, chapters at all levels and page numbers as well are recognized. In the future, this will be expanded to support tables, footnotes, etc. Once the books is interpreted, a Digital Talking Book (XML) can be generated in which the logical meaning of the text parts is embodied.

At this point, the editor can choose one or more of the output channels to enter. At this stage of development, there are two output channels developed; Braille print and web publishing. Style sheets (XSL) are used to allow customizing the format per channel. In the future also support for audio, speech and large print will be added. On top of that, books can be enriched with multiple resources. For instance, an enriched picture can be shown, described, enlarged, read and printed with large font.

Between all these production steps the editor is allowed to make changes. Therefore powerful third party products are integrated with Lexus Studio, such as Word and WorX. Because WorX is embedded in Word, editors enjoy the powerful and mature functionality of Word that they are familiar with and they can use WorX’s XML functionality on top of that. In the meantime, editors’ decisions and work are protected by the workflow system, since Lexus Studio is built with features that prevent loss of information, allow re-editing at almost any stage and provides guides to the editor for completing the production process.

Since developing the NLBB solution, Mediasystemen has been contacted by libraries elsewhere in the Netherlands and in Europe who are interested in authoring for multiple enhanced outputs from Word. The technology and methods employed in Lexus Studio could find an even broader application in the United States where recent law requires that government-funded software projects and the content published by their developers/suppliers be made accessible to employees with a number of disabilities, not least vision impairment.