Comparison of Methods and Costs of Digital Preservation

Citation

Turner, J.R. (1999), "Comparison of Methods and Costs of Digital Preservation", Library Management, Vol. 20 No. 3, pp. 196-199. https://doi.org/10.1108/lm.1999.20.3.196.3

Stephen Ostrow’s pamphlet is about how to convert visual images (paintings, engravings, photographic prints and negatives, pictures of all kinds, but not written or printed documents, or any other form of information) into electronic digital signals. He is not concerned with the technology or what the computer is doing, but with all the practical managerial questions and problems. He has been personally involved with digitising images at least since 1984 as chief of the Prints and Photographs Division of the Library of Congress. His comments are clearly argued and well worth the attention of librarians, curators and archivists.

The two fundamental questions are: is digitisation worth doing, and if so, how do we do it. It will come as no surprise to find that the answer to the first question is complicated, with arguments both for and against. Among the drawbacks are that in any collection there are often vast amounts of material to be processed and therefore it is probably never going to be possible to digitise every item. As well as existing in large numbers, many images, by their very nature, will be unique. Choosing which items to digitise will therefore always be difficult. Incomplete digitisation could give an unbalanced impression of the collection; all the drawings of a famous architect would be digitised, but what about the drawings of her almost unknown teacher? A related factor to incompleteness is that digital images are viewed one at a time on a VDU so the user is prevented from seeing a group of images. For example, all the original shots a news photographer took at an event, and from which one was selected by the newspaper for publication, could be laid out on the desk top, but not viewed together on a VDU.

Digitisation also needs a great deal of documentation, sometimes much more than already exists in the library or archive. For example, uncatalogued information which a user of the original may find on a slip of paper pasted to the back of the picture will not be included in the electronic copy unless it is all transcribed alongside the digital image. Furthermore, since the digitisation will probably not cover the whole collection, some description of the extent of the collection should be included in the documentation.

Digitisation always loses something of the original for the end user and even the highest resolutions do not produce exact details on screen. There are also the added drawbacks with high resolutions of high costs, slow electronic transmission times, and large amounts of computer storage space. Worse still is that a digital image can be downloaded and changed in almost any way the user wants. The image can be cropped, distorted, details can be removed, colours changed, and so on, and after the changes have been made a secondary user of the changed image would have no means of knowing that anything at all had been altered. Perhaps worst of all is that, being simply electronic impulses, all the records at once are vulnerable to sudden and catastrophic loss. Consequently, digitisation (for text as well as images) is still not accepted as being “archival” and at present is not recommended for long‐term preservation. From all the different types of pictorial originals Ostrow agrees that archival standards can be achieved, with the very highest resolutions, only for black and white continuous tone photographic negatives.

In other words, digitisation can never replace the original. But despite all the disadvantages, there are arguments in favour. The images are obviously made much more widely available than ever before, particularly to scholars around the world who could never visit the collection. Digitised images can also be used by groups other than scholars, for example, classes of school children, who before were denied access. Even for regular users of the collection there are advantages. Without digitisation, a user very often has to look at a selection of originals in order to decide which image contains the best information. Such decisions can be taken, usually more efficiently, from digitised images, and an advantage for the curator here is that the handling of originals is greatly reduced and handling, of course, is the cause of most deterioration. However, a drawback not mentioned by Ostrow, is that those selected originals which have been digitised are the ones which most users request, so that digitisation increases the handling of the originals chosen for digitisation.

An Investigation into Digital Preservation Needs looks at a different side of the problem. The report is not concerned with the transfer of non‐electronic information to a digitised format, but with the preservation of information which was created and exists only in digitised format. The investigation was attempting to find out if universities and research funders already had preservation policies, and if they were aware of the nature and implications of preservation of digital information.

Questionnaires were sent out to selected research funders, university vice‐chancellors, and researchers likely to be producing large amounts of digital information, and the questionnaires were followed up with interviews. It is not stated exactly how many questionnaires were sent out, but the total seems to have been at least 800. The report admits that the 109 replies, for all kinds of understandable reasons, was a low response. It also admits that the questions were difficult and the people who completed the forms were likely to be those who already understood something about the problems. People who had not thought about digital preservation were probably defeated by the questions, did not have preservation policies, and did not complete the questionnaires.

In fact very few organisations had preservation policies but all who replied accepted the need for them. They also accepted the importance of preservation, and the need for more information about all aspects of the long‐term storage of digital records. Other related concerns were the difficulties of protecting the ownership of electronic information and, despite the acknowledged importance of preservation, worries about the costs of carrying out the work properly.

An Investigation into Digital Preservation Needs is a good, clear report which maintains impartiality throughout. The overall finding is that there is general uncertainty about digitisation and a need for more information and guidance. The report concludes with three recommendations, with which nobody surely could disagree:

1. 1.

   (1) the development of national guidelines covering key areas of concern;

2. 2.

   (2) the development of standards to allow kitemarking of centres where research data can be managed and preserved in the long term; and

3. 3.

   (3) the development of a national policy on research data and dissemination of information about this national policy.

These two publications will be of use to managers looking for help on a particular problem or wishing to keep up with recent developments. They are written in a straight‐forward style, easily accessible to non‐specialists. The third report, Comparison of Methods and Costs of Digital Preservation, is aimed at readers with more than a basic understanding of computers and digitisation. Part of the problem for non‐specialists is the profusion of available electronic documents and the potentially complicated methods of producing them. For example, the report classifies six data types (alphanumeric, raster graphic, vector graphic, etc), which can be used to create ten kinds of digital documents (structured texts, design data, speech and sound recordings, etc) and these documents can be produced by at least 19 application programs (word processing, spread‐sheet packages, SGML editors, etc). In fact, the report admits:

It has always been the view of the consultancy that the people best qualified to make the final decisions on preservation options are the people who understand the specific digital resources and data types to be preserved.

and one of the final recommendations is that a group of at least 20 “specialists with expertise in preserving each of the [ten] categories of digital resources” should take over the next stage. It is good to know that the experts have begun working on the problem and, for the moment, it seems all that library and information managers can do is await further advice.

In any event, what other choice is there but to go along with digitisation? Whatever the present lack of archival reliability of digitisation, the system in future will probably be modified to become acceptable for preservation, or its shortcomings will be brushed aside and we will have to live with it. There seems to be no other option because of the impetus from the continuing development of computers, the increasing output of digital information which has to be preserved, the increasing use of original manuscript, printed and electronic records, the increasing rate of deterioration of these originals, and the inability of present non‐electronic preservation systems to stop the rot. At least some thought is being given to the problem, as these publications show.