Niels Thogersen

There are very, very few people in the DAISY community who have been in some way involved with DAISY since before the formation of the DAISY Consortium. Niels Thogersen is one of that small group of people who remember the days before DAISY, who experienced both the excitement and the sometimes discouraging moments of those very early days. He was one of the small group of people acknowledged in the February 2003 announcement of the Official Release of LpStudio/Pro, the first DAISY production software developed by the DAISY Consortium for its Members. The first part of his story takes us back to before the days of DAISY, to the days of manually produced braille and audio books on cassette.

(This is the first part of a two-part story)

PART 1

Niels Thogersen working at his laptop in his xml-tekst officeI'm Niels Thogersen from Denmark in Scandinavia, and I've been working with the production of educational books and materials in accessible formats for 20 years now – about 5 years before the DAISY Consortium was established.

The 'Not so Good' Old Days

Things were very different then – personal computers were new to most people, internal networks were as yet unknown, and the Internet did not even closely resemble what it is today. We were not aware of the possibilities and opportunities that would arrive with DAISY and with XML. We had not yet seen the advantages of applying open standards to our tools; we were working independently and the idea of "sharing" and of joint developments was not really there in 1990. We all had our own requirements and standards that often differed, even between the various organisations in a small country like Denmark. The concept of "outsourcing" parts of the production process was not even thought of.

Provision of educational books was already a public service in Denmark for blind students and for students with visual impairments when I began my work as the Head of the Student's Library at the Danish Institute for the Blind (IBOS) in 1991, and with new legislation in the year 2000, students with dyslexia and other print disabilities were also included in this public service. We served students from high school through university level education.

At that time the students were given their books in braille (on paper) or as audio cassette books on tape.

For study purposes, all cassette books had page numbers with tone indexing that allowed the user to find pages and chapters on a tape using fast forward/rewind, and we made separate tapes of the Table of Contents (TOC) and of the notes in the books. Using 2 or 3 playback units at one time, the students could read the TOC which would point them to the tape number of the required section, and by counting the "beeps" they would be able to find the right place in the book. A lot of students wanted us to print the TOC in braille to allow them to find the text faster.

To produce braille books, we were still retyping the entire text of the books directly into contracted braille. Even if this work was done on computers, it was still very costly, very slow and time-consuming, and it required that the staff who keyed in the text were experts in the contraction and formatting rules.

And even worse – the files were stored in proprietary formats, formatted for only one type of embosser – it was not possible to re-use the text in other formats. We tried almost everything. One project involved scanning of braille pages (each braille dot creates a small shade when the page is scanned in greyscale), followed by recognition of these shades back into letters and text, and then into standard text after applying a de-contraction process. The system was developed as a project by two engineering students and it actually worked very well! But it did not scale for real production – we had to hand feed each page and had to use large scanners because the braille pages were larger than standard letter paper. These scanners were extremely costly at that time. In the end the project was abandoned.

Photograph of the Braille Note Apex BT32 braille notetaker from HumanWare; current braille notetakers are not much larger than an adult's hand About that same time a Danish-made braille note-taking and reading device was beginning to be issued to the blind students in Denmark. By today's standards it was a huge box (about the size of two modern 17" laptops on top of each other) with just a 30-character braille display and a 30-character LED display which allowed a teacher to see in print what a student was reading/writing in braille. The machine had 8-dot braille cells, giving it a large character set that could be mapped directly to the ASCII character set. It also made it possible to print both an ink print of the text for the teacher and a braille printout for the student, and with more characters, it was easier to construct at least simple mathematics directly, thus avoiding Nemeth code and other complicated math editors. That machine became very popular, and many students wanted us to produce their books in the format for this machine.

The Early Single Source Concept

We realized that there was an urgent need to be able to provide the same text in different formats and versions. To meet this requirement we started working with single-source-mastering into SGML (Standardised General Markup Language) and by 1992 we had begun working in SGML.

At that time SGML was only used by large companies, primarily in the pharmaceutical industry and by aircraft producers. SGML editors were costly and would only run on absolute high-end computers, and SGML itself was also very complicated. We decided to use Word 2 (!) – the first fairly reliable Windows version of Word that had a good programming language (WordBasic) and good features for import and export to and from the most common formats. Our FrameMaker editor for SGML could import and export Word files, and most of our staff members were able to use Word for the text processing part.

Soren, one of Niels Thogersen's team, scanning and permforming quality control Instead of retyping everything, we started to experiment with scanning. We were using the very first versions of OmniPage for OCR (Optical Character Recognition), and an HP scanner with a sheet feeder, that allowed us to scan an entire book: first all the front pages and then the backside pages, and if you were lucky, these two scans were combined nicely into a book with all of the pages in the right place! If it didn't work for you on the first try, you simply had to scan from the beginning all over again. The OCR software knew nothing about our special Danish characters, and all of these had to be corrected afterwards along with all the other "normal" errors that always occurred.

We realised that it was faster to make these corrections than to retype, and from OmniPage 2 at least a rudimentary international character set was accepted and recognised.

Then XML came out as a new standard; it simplified our work dramatically and reduced our costs for software and high-end equipment. We were able to apply computer-based contractions to our files in Danish and English, first with our own contraction engine, then with an engine used in most of the Nordic countries, and finally with Sensus Braille (which became RoboBraille). Even if we needed to do a lot of proof reading and formatting, we were finally able to produce more braille with the same staff and at a lower cost.

Then Came DAISY

Cassette books were still recorded in the same way – but then came DAISY!

The first notions of DAISY came to us at a meeting attended by representatives of the Nordic countries, where some of the sessions were focused on different digital issues. This was very exiting, and for the first time ever we heard terms like "compression", "CD-ROM" and "Mini Disc". But most importantly we heard Kjell Hansson from the Swedish Library of Talking Books and Braille (TPB) inform us about a test project they were running that included digital recording of audio material and synchronisation between text and audio, and, with much better navigational features than any existing alternate formats. This was simply thrilling to hear.

These were the earliest days of DAISY and the beginning of some life-changing experiences for Niels. His association with DAISY has fostered friendships and important business opportunities (including the establishment of his own company, xml-tekst – a Friend of the DAISY Consortium). It has brought travel to places he might otherwise never have visited, and experiences that have changed his life. Please read Part 2 of Niels Thogersen's story "…but then came DAISY!" which will be published with the June 2011 DAISY Planet. Niels can be contacted by email at niels[at]xml-tekst[dot]dk.