Staff Papers

Australian Cooperative Digitisation Project 1840-1845 – an update

This paper was prepared for the ALIA Rare Books and Special Collections Special Interest Group 1999 Conference which was held at the Conference Room, University of Queensland Library, St.Lucia, Brisbane on Wednesday 7 July 1999. The paper was read by Richard Overell, Rare Books Librarian, Monash University Library.


We have made two previous reports on this project to the ALIA Rare Books SIG Conferences in 1996 and 1997. I think on both occasions I have made the observation that digitisation projects seem to take twice as long and cost twice as much as one ever expects. Well, two years later I might want to extend that observation even further.

The project is at the very end of what has been a lengthy production phase. We are finishing off the last few titles and sorting out any re-work that is needed, so that we have a set of digital images ready to place on the website set up for the purpose.

In case there is anyone who is not familiar with the project – which is likely, even if surprising (how easy it is to assume that the project has been everyone’s constant companion over the past four years!) – I should remind you that the Australian Cooperative Digitisation Project, or ACDP, is a joint project between the University of Sydney Libraries, the State Library of NSW, the National Library of Australia, and in the early stages of the project, Monash University Library. The project uses funding from the participating partners plus an infrastructure grant from the Australian Research Council. Our aim was to make digital copies of all Australian serials and all Australian fiction monographs first published between 1840 and 1845, and to make them available over the World Wide Web. To improve preservation as well as accessibility we digitised from microfilm, in most cases requiring production of new microfilms that would give us a suitable digital image.

The fiction component, which ran to only a very few titles, was completed early in the project, so I don’t intend to say any more about that. The titles are available as searchable text from both the University of Sydney’s SETIS site ( http://setis.library.usyd.edu.au/acdp/) and the NLA site.

For the serials, our basic approach has been to create TIFF images at resolutions of 400 dots per inch, converting to PDF format for multi-page delivery. This means we get images of pages, rather than searchable text.

Of the 90-odd serial titles listed for the period in the Ferguson Bibliography (hence our alternative project family name: the Ferguson Project), there were 11 titles we were not able to track down anywhere in Australia, and another five titles that were withdrawn from the project because they proved to be impossible to digitise adequately with current techniques – mainly because of the page size and the poor quality text. This leaves us with about 74 titles, of which five that proved to be unsuitable for scanning from microfilm are now being filmed directly from the source using digital cameras. This is an experimental addition to the project, and looks to be quite successful.

Since our report of two years ago, the Project management team, staff in the participating libraries, and our project vendors have all struggled to achieve a consistent suitable image quality at a reasonable price, with correct file names and embedded metadata. However, despite anguish often felt and frequent frustration, I think there is much we can be pleased with – and by "we" I mean all parties involved.

We will have delivered a significant body of material into a much more accessible form for research use. Being images of pages, it is not as usable as we would wish it to be, but we have built a foundation on which improved technology may be able to improve accessibility further. We’re taking some small steps in that direction ourselves as a modest extension to the project. We’re attempting to apply optical character recognition software to a few titles to see whether we can produce useful searchable text. If it is successful – and of course that requires some definitions also – the National Library will look at the business case for trying to take that process further. That case will come down to an argument about the level of demand for searchable text, the level of searchable text we can deliver, and the costs involved. Of course, OCR’ing this kind of material is a nightmare because of the great variety of print fonts and the variable image quality.

We will achieve another primary goal of the project in learning a lot that has already been ploughed back into other digitisation projects, and into policies. We have pushed the boundaries quite a lot, and found some of them to be real constraints, not just lines on our mental maps of the world. As the project developed we found numerous other projects that had looked to be similar, turned out to have either much more modest goals, or to have experienced exactly the same difficulties as we’ve had. We had to learn this through personal contact – it is not the kind of information that is generally shared in public forums.

We’re committed to sharing our experience – in many ways we see that as more important than the images themselves. Our project deliverables include a detailed report, and we are working on that now. We also want to engage in widespread discussion about our experience, so we’re planning to hold a public forum on the project, probably in Sydney later in the year (and I mean this year – with this project we’ve found we need to specify what year we’re talking about!)

What else needs to happen? Well, we are working away at the National Library, checking the last batches of images from our vendors. We have had to make a business decision to limit the amount of image-by-image checking we do, so after a certain level of assessment we conduct ourselves, we’ll be inviting users to report faults to us. Any other approach meant we would not be able to make images available well into the next millennium.

When the images have been mounted on the Library’s website, they will be retrievable both through the navigational tools on the Project pages and at title level through the NLA catalogue. And they will be maintained and developed as part of the National Library’s growing collection of digital research materials.

(By the way, anyone interested in looking at how we plan to manage these and other digital resources should look at our Digital Services Project site, at http://www.nla.gov.au/dsp/).

We’re planning a public launch of the site in Canberra in August or September. However, you can look at the images currently on the site, by going to http://www.nla.gov.au/ferg/. This is still a site somewhat under construction, but it does give you access to some of the background information, and to a growing list of titles and their digital images. (There were 20 titles available when I last checked and it should be growing almost daily.) We know we need to do quite a lot of work on the site, improving the contextual information and the navigational tools, but at least you can see some of the serials we have been working on.

By the way, to view the images requires a PDF viewer. Please don’t be put off by this. It is easily downloaded to your PC, and the site provides the link for you to do it.

For anyone with an interest in the content of this project, please visit the site and use it, both now, and especially after its public launch. For anyone interested in the processes of the project – and I think it remains one of the most significant and ambitious digitisation projects in this part of the world – please feel free to contact one of the Project Committee members:

And please look out for the public forum to be held later this year.

Colin Webb
Director (Acting)
Preservation Services Branch
National Library of Australia

1 July 1999

Top

About This Site | Copyright | Privacy | Accessibility | Site Map | Site Search | Content A-Z | Contact Us