NINCH guide home interview table of contents previous interview next interview
HATII interviewed Svein Arne Brygfjeld, Adviser for IT and Innovation at the National Library of Norway’s Office of the National Librarian, on October 27 2000. The Office of the National Librarian is a co-ordinating unit for all activities within the National Library. Among the programs run by the Office, is the National Broadcasting Corporation (NRK) Radio Archive Project. The NRK was formerly the only national broadcaster, and the project intends to preserve the material it produced. In this way, the resources that the project is able to archive and make accessible will be of significant educational, cultural and historical value.
The National Library of Norway is divided into three main structural components:
The question of a collection survey assumes that a formal digitization strategy exists, which is not the case. Rather, the National Library is in the process of learning from experience towards establishing a strategy. It now has several years of experience gained at its unit in Mo i Rana, to develop in the digital domain. It was in early recognition of the strategic importance of digital activities in large-scale multimedia information management that the substantial, specialist unit in Mo i Rana was established in 1989. As a testament to the forward-looking attitude of the National Library, the main argument for establishing the new unit was the relative maturity of IT at that time which has, of course, now been superseded. The unit at Mo i Rana consists of state-of-the-art IT activities, a Media Laboratory (c. 50 people working on restoration, preservation and digitization activities) and a sound and image archive (c. 20 information professionals).
Another issue which arises in connection with overall policy, particularly when considering audio and moving image material, is what is meant by digitization: do we include, for example, the copying of fragile, nitrate-based filmstock to digibeta format after restoration as a digitization activity? Overall, as far as collection surveys are concerned there is a mixed picture. There are several projects and activities across all media types being undertaken at Mo i Rana. For some projects surveys have been carried out, but the decision to do so is made on a project-by-project basis. For those projects that have had collection surveys, relevant people would be made aware of the survey and encouraged to participate. These decisions are made in a pragmatic way: those staff members who are available and have relevant expertise will be consulted.
In terms of passing on advice for good practice in this area, it is important that the needs of partner organizations are taken properly into account. The National Library carries out several partnership projects with other organizations, e.g. it is currently involved in a project for networked-based learning in primary schools. The partner organizations should also be encouraged to contribute to collection surveys.
The National Library’s view of its digitization operations is of general “activities within the digital domain”, rather than the narrower concept of just digitizing its own holdings. Many of its activities are carried out in co-operation with others in the community.
There is no formalized strategic policy statement in this area, rather the National Library’s priorities are, to some extent, derived from its funding remit. The National Library is funded by the government to serve the interests of “research and documentation” — this remit can be and is interpreted widely by the Mo i Rana unit. Activities within the digital domain tend to facilitate enhanced access to materials and digitization, and the dual aims of enhanced access and social inclusion in practice drive policies and thinking about digitization.
The organization is not monolithic — rather the impression is of a flexible and responsive modus operandi. The Mo i Rana unit investigates what technology and methods are available for different projects and what is possible in its different areas of operation. Many of the activities are research-led in order to gain experience of whether they are going to work on a large scale. Even for pilot, research-led projects, the tendency is always to digitize a fairly large chunk of the collection. There is considerable freedom, within the interests of research, to work in different domains and to pursue various methods and technologies.
The largest obstacles both to planning and building digital deliverables are perhaps organizational aspects within the larger institution, including the culture of the traditional analog library world. Although organizational aspects are most important here, technology is also relevant. Technology is always a little behind what is needed, especially for larger scale projects.
All of the criteria listed (with the exception of IPR) are relevant in the selection and prioritization of materials for digitization. Generally speaking the highest priorities would be:
IPR is not understood as a priority in this respect. The National Library does not necessarily own copyright on materials, especially those deposited under the fairly recent (1989) legal deposit legislation, and so it is not necessarily allowed to make digital materials available.
Other organizations with which the Library works may have other priorities, e.g. the National Broadcasting Corporation (NRK: http://www.nrk.no) over the radio archive project.
Commercial exploitation is almost always via partners, e.g. the radio archive, where the reason for digitizing is that the programs will be re-used by either NRK or other radio stations. In another case of working with partners, there is a database of Norwegian music which the company concerned wishes to make available for commercial use on a commercial basis, but which is deposited with the Library.
Selection and prioritization criteria have changed over time, in that preservation has become a higher priority.
There is a great deal of co-operation and partnership with others, with the National Library in the role of the provider of expert technical services across the board. All type of organizations (except foundations and charities) and all types of regional and international co-operation are undertaken. The National Library is proud of its achievements in this area and its record of co-operation with partners.
The best advice for those embarking on partnerships is to ensure that there is the best possible understanding of the roles and interests of each collaborator and that all involved are aware of these roles. There has to be a common understanding of the focus of the project. Personal relationships are also important — there must be attention to trust, respect and understanding of the professional skills and organizational aims of others. It is vital to try out working with partners and establish ways of working together via small-scale pilots.
The overall program is ongoing. It started with the setting up of the Mo i Rana unit in 1989 and is continuing indefinitely.
All of the items listed apply with teaching and learning, public access, preservation and research (or keeping a technological lead) as high priorities. The National Library is proud of its research-based record in choosing appropriate technology and looks upon this area as a successful one.
Revenue generation may be on the agenda of partner organizations, rather than the National Library’s. Response to previous demand for digital materials has not yet come from users, but there is a demand from partner organizations.
There is an overall statement of intent, or mission statement, on the National Library’s website. All of the individual points listed may apply to specific projects, but not necessarily to the overall program. The mission statement refers to the need to establish a framework for digital libraries and collections which make information available to everybody, wherever they are, in a democratic way. The emphasis, again, is on access and inclusion.
The thinking is to establish a long-term digital library with core functionality, and then provide different “windows” into this information.
At the moment, there are many different digital collections and media types, such as 100,000 still images, a large collection of newspapers, audio (including 400 radio programs), entire collections or radio programs on Norway in World War II from the US, “Spirit of the Vikings”, examples of poetry with poets reading their own works. Taking the example of the NRK radio archive, one “window” into the information would be for use by NRK staff for their professional activities; there might also be a public view; equally, there might also be a view for educational users from an authored, guided website where paths are more clearly defined. The way that the information is presented will vary a great deal but it is of greatest importance to get the core functionality right in order to facilitate different views of the materials.
A variety of source material is being or has been digitized with the exception of textiles and artifacts. Concentrating on audio-visual materials, all types listed are being digitized. The nature and format of the materials are similarly varied, and include TV and nitrate-based film stock. There are some entire collections, as well as some samples.
Digital deliverables are intended to be re-used and re-purposed, for example in the networked learning for schools project. The preference is for networked delivery rather than CD-ROM, but other delivery modes are not necessarily ruled out where they may be appropriate.
Tools for representing content are used as appropriate, e.g. for audio-visual materials, MPEG is used.
Dublin Core is used for describing content, while a variety of other standards are used from the original cataloging including non-standard systems, e.g. in broadcasting or film where institutions have often adopted their own house styles.
In relation to the use of standards in general, it is worth noting that rewriting metadata — even where it is desirable — is extremely time-consuming and will only be carried out when the collection is small. In most National Library projects it would not usually be done; rather existing cataloging is transferred to Dublin Core. For example, in the radio archive project the National Library had to rely on the previous cataloging metadata from NRK which was sometimes of poor quality, but resources were not available to improve it. In this case new metadata-creation would have to be done by actually listening to the material — ideal, but not realistic.
URN, as a persistent world-wide unique identifier scheme, is also used for identification in the NRK project. NRK has agreed to use Dublin Core. Agreement on standards makes it much easier to interoperate with “publishers” such as NRK.
Target audiences are primarily those in “research and documentation”, as in the National Library’s remit, but in practice they are very wide, including all the target groups mentioned. As noted above, the Library’s thinking is to provide different, tailored “windows” onto its repository of digital materials, and these windows would be constructed on a case-by-case basis, as appropriate to the particular needs of an audience.
Project management is done in-house. There was little previous culture of project management at the National Library before the Mo i Rana unit was set up. Mo i Rana staff came more from different backgrounds, including commercial backgrounds, and brought a culture of project management with them. The Mo i Rana unit now sees organizational/project management issues as a successful area of its operations.
The organization of the project management responsible for formally constituting projects, can be summarized as follows:
The main advice regarding project management is about inclusiveness and delegation of responsibility: it is important to ensure that all the relevant people are included and to place responsibility with the right people — project members should know who has responsibility.
Feasibility and pilot studies are carried out for individual projects. These tend to be informal, and based on experience, rather than formal benchmarking/time and motion studies. Decisions on delegation are made by relevant team leaders on the pragmatic basis of which staff members are available (there is little staff turnover).
Digitization is carried out in-house, with the National Library’s own installed equipment. The National Library acts for others as the expert provider of services and expertise.
The profile of staff in Mo i Rana tends to be quite different from that of the National Library in Oslo: they tend to be younger, technical, and IT-orientated. Approximately 10% of IT staff have Masters degrees, and there are few conventional librarians. Technicians in the Media Laboratory are unlikely to have had previous technical training as no training to that degree of specialization is readily available; rather, training is done in-house. The technical department has a very high degree of skill — e.g. they have recently constructed from scratch, a player to play wax cylinders more efficiently.
Training needs are assessed project-by-project and training is often carried out on the job. External training and work experience are also used: people have been sent to UK and US international corporations and laboratories, e.g. Kodak in Rochester, particularly for those working with film and audio.
Legal deposit was only introduced for the National Library of Norway in 1989. The legislation is focused on the information itself, not the information carrier and legal deposit covers all media types including web space, TV and radio. The National Library does not always own copyright and copyright does not necessarily loom large in the decision to digitize, due to this legislation.
User access to digital materials is varied: for example, researchers have free access, but for materials for schools, or otherwise re-purposed materials some payment may be made.
No electronic system is currently used to control copying. Methods of managing rights in an object-based way are being researched to control usage on the networked-based learning project. This research is investigating a model which “wraps” the digital object in a coating of rights-management layers.
Mo i Rana is very well equipped with high quality film and sound laboratories. The curatorial staff make decisions about what is digitized, taking into account conservation needs, preparation of material, risks and any need for special equipment and handling (such as with nitrate-based films).
Access would not normally be restricted after preservation/conservation.
A variety of cataloging or reference systems is used for original materials prior to digitization from standards such as MARC to in-house systems. As noted above, even where it is desirable, re-writing metadata is not normally undertaken, rather existing cataloging is transferred to Dublin Core.
Digitization is used to preserve materials (e.g. fragile sound or moving image material) and these types of materials would not normally be rejected for digitization on grounds of poor original quality. Rather, curators would decide that fragile and vulnerable materials would be prioritized for digitization. All the forms of reproduction or intermediaries for audio and moving images would be used.
The degree of sophistication of metadata for digital objects depends on resources and quality of the original cataloging. The records for digital deliverables and original digitized materials are stored independently of each other. Metadata records are kept separate from the main National Library Catalog and are available on the internet.
In general, audio material is captured and preserved in WAV format for high quality. For access and delivery (depending on circumstances) versions in MPEG layer 3, MPEG 2 (for digital radio broadcasts) and RealAudio are used. The audio sampling rate for capture and preservation is 48kHz and can be 11, 44 or 48kHz for delivery. The bit-rate at all three of these stages can be 16, 20 or 24. The sampling rate for old originals, for example from wax cylinders, is 16 bit stereo at 48 kHz. Where the originals are of better quality 24-bit stereo is used. The average file sizes are 0.5 GB for capturing and preservation and 25 MB for delivery. Bit-rate compression methods used are MPEG 1&2, layer3 and Real Audio for delivery only. The aim of compression is to reduce costs, improve access, enhance usability and decrease storage requirements. Post-capture editing is carried out.
Digitization from 35 mm film is done using a film scanner. The film scanner produces a digital video signal (270 mbit per second), which is processed to minimize noise and distortions. Then the film is stored as a high quality copy on digital beta and a “low quality” compressed copy made. The thinking is that digital formats are not yet good enough, and resources not big enough for huge storage requirements of disk copies. According to needs and circumstances, digital versions may be made in MPEG 2 at 6 mbit per second and MPEG1 at 1.5 mbit per second, but the digibeta copy is the main preservation medium.
Quality control procedures depend on context and the process and material in question, such as knowledge about the source of the original. Random checks are made on the quality of the end product compared to the quality of the original.
The catalog and certain materials are now available to users on open access, but this is not on a large scale and user statistics are not yet being collected. As noted above, search and browsing facilities will differ according to the user profile.
Availability of the radio archive is still under discussion. Some will not be made available, as it contains confidential material. Again, there is a mixed scenario. Some partners such as NRK may charge for access to their materials. Other partners, such as record companies selling digital copies of music on the net, will charge. Record companies have deposited audio material which can still earn money via CD distribution, but the National Library retains rights to make material available to core users under its remit of use for research and documentation.
Potential users are generally informed about the collections on the website, but also through special events for particular projects, for example a special launch of materials about Norwegian immigrants leaving for the US, which was also launched in the US.
No evaluation data is available for the overall National Library program, although individual projects may carry out evaluations.
The level of government funding is in the public domain but may be difficult to interpret generally. Investment in infrastructure is likely to be between NOK 7,000,000 and NOK 10,000,000. It is more difficult to calculate man-hours. Everything is more or less funded by the government, but there are situations where costs are shared, as in the NRK example where the digital radio costs are shared 50/50 by the library and NRK.
Preservation strategy for simple and single data types (i.e. discrete units not cross-referenced data) such as audio recordings or moving image recordings, is not to store on carriers like CD-ROM or DAT tapes. That is, not to store onto formatted information carriers, but rather to store on disk in the data repository, with migration as the key. The strongly held view is that migration is preferable to emulation, which has not proved successful and at the moment it is too difficult to maintain over time. Alternatively, it is possible to build up a technical museum, but this provides too many challenges when machines stop working and the knowledge of how to use and maintain them is lost.
There is no exit strategy — on the contrary, the digital activities at Mo i Rana are aiming towards building a sustainable digital library. Loss of the digital deliverables would be a matter of concern.