9.1.1.1 The production and long term preservation of digital audio objects incorporates a number of interrelated parts, many quite complex. These guidelines define the tasks as: Extraction of audio content to create archival digital audio objects; Ingest of the content into a digital storage system including the creation of necessary metadata: Administration and management of the data and system: Archival storage: Preservation planning: and Access.
9.1.1.2 Some institutions have the facility to undertake all the tasks, as well as a collection whose size justifies the expenditure. The alternative is to negotiate partnerships to manage some or all of the tasks on behalf of the collection owners. These partnerships may be with other, larger institutions, could include partnerships with like minded institutions, or could represent a commercial relationship with a supplier.
9.1.1.3 This section of the Guidelines examines the resources required to create and preserve digital audio objects according to technical requirements described in this document. It considers the issues related to size of collections and scale of work, recognising that the professional fulfilment of the requirements as described herein can only be met when the size of the collection held by the respective institutions reaches a critical mass that make autonomous preservation viable. Many institutions, collections or archives have particular expertise and resources in core areas which they can deploy to facilitate the necessary processes. It is recommended that they maximise the benefit from their core business area while carefully examining the areas where services may be better sought elsewhere.
9.1.2.1 The first decision to be made is whether an institution should engage in digital audio preservation at all. Often audio or audiovisual collections came about in institutions with a variety of other aims which may not include professional preservation of audio materials. The ever increasing problems related to the physical preservation of an audio collection, the obsolescence of dedicated replay equipment, and digital long term preservation may suggest a rethinking of the collection and preservation policy. Where appropriate alternatives exist, audio collections could be handed over to more specialised institutions. This would not necessarily mean fully relinquishing ownership a collection; the receiving archive could be asked to produce, in return, listening copies that could be held – without significant costs – for further in-house use. Various possibilities of retaining, partly or fully passing on the right of ownership as well as user rights could be applied.
9.1.3.1 Should an institution like to maintain archival responsibility for their collection a number of different scenarios may apply which do not require relinquishing the collection.
9.1.3.2 Producing digital audio objects in-house but entrusting digital preservation to another is one possibility. There are a number of ways this scenario could be enacted. One way, which seems most appropriate in academic institutions and universities, is where several units are engaged in the production and use of digital audio (and audiovisual) documents. Generally, such institutions have a central computer facility, very often with an existing responsibility for managing various digital objects. The data storage facility could take responsibility for the long term preservation of the created audio content. It is important, however, that the central unit would be fully acquainted with the specific of long term preservation of digital audio objects, and develops well defined rules for the production of archival files. The central unit would prescribe recording formats, resolution, annotation procedures, and other archival issues to be followed. Also, long term preservation tasks of that kind could also be fulfilled by private entrepreneurs. This concept would work for newly produced materials, specifically field recordings in various disciplines like anthropology, linguistics, ethnomusicology, and oral history.
9.1.3.3 Another way in which this scenario might be enacted is where a large collection exists with appropriate storage, transfer facilities and technical expertise, but where the infrastructure to support a digital storage facility is not developed enough to build a trustworthy digital repository. Under these circumstances the local facility may undertake the signal extraction and dispatch the resultant digital audio objects to the selected archive.
9.1.3.4 Should institutions, however, already have accumulated, though dispersed, analogue and historical digital originals, signal extraction from these originals to produce digital preservation files could be concentrated in one professionally equipped unit which could also be appended to the central computer unit. If the entire institution does not reach a critical amount of carriers, signal extraction should better be outsourced. The same is true if the institution does not have in-house expertise or equipment for professional digitisation.
9.1.3.5 In any of these scenarios, where a third party archive is to take responsibility for the ingest, management and preservation of the digital audio objects, it is imperative that there is a clear understanding of the roles and responsibilities of the various partners in the work. The ISO 20652:2006 “Data and information transfer systems -- Producer-archive interface -- Methodology abstract standard” identifies, defines and provides structure to the relationships and interactions between an information producer and an archive. It defines the methodology for the structure of actions that are required from the initial time of contact between the producer and the archive until the objects of information are received and validated by the archive. These actions cover the first stage of the ingest process as defined in the open archival information system (OAIS) reference model (see ISO 14721). http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?c...
9.1.4.1 Critical mass in the field of sound preservation is where the size of collection is sufficient to justify the expenditure to undertake all the tasks in house. It is difficult to quote concrete numbers when defining critical mass; the more professional institutions there are available in a country or region, the higher the critical mass will be. If, however, there are only few institutions engaged in professional audio archiving, or if there is none available at all, then the critical mass would be lower. Critical mass should always relate to specific media formats; coarse grooved discs, microgroove discs, open reel magnetic tape, etc. In fairly developed countries or regions critical mass would be at least several thousands of items, but often institutions with tens of thousands of carriers of one type make a rational decision to outsource signal extraction. Under less developed circumstances the autonomous transfer of few thousand items/hours only can be carried out successfully.
9.1.4.2 The critical mass will also depend on the homogeneity of the material within the respective format. Homogeneous collections can be transferred with some degree of automation. The cost associated with fully automated systems generally suggests outsourcing to institutions or service providers that offer computer controlled parallel transfers. Collections consisting of many different carriers or standard of recording – as often found in research collections – demands reliable manual transfer, which may be available at lower cost in house, provided the specialised expertise is available.
9.1.4.3 Even large, professional sound archives may consider sending parts of their collection to specialised institutions or service providers for the purposes of transfer. This may especially be so for some historical analogue and digital carriers.
9.1.5.1 Whenever material is outsourced for the purpose of signal extraction, especially to private entrepreneurs, it is important to accurately define the tasks to be fulfilled. This is best achieved by specifying the standards provided by IASA in these Guidelines as part of the contract.
9.1.5.2 When outsourcing any audio processes it is essential to establish a quality control system that provides a high level of assurance that all contracted work has been carried out appropriately. Such measures should be based on stringent delivery of preservation metadata, accompanied by tests of randomly chosen samples, including unannounced visits at service providers and testing of the transfer equipment. Specific attention should be given to test the automated and manual quality control systems established by the supplier, their capacity to manage long term contracts through the use of project management methodology, experience with similar contracts and with specific carriers, equipment maintenance, and finally the balance between cost and quality. Before the start of the production level digitisation phase specific small tests should be performed to ensure all aspects of the process meet criteria before commencing to process on a larger scale.
9.1.5.3 It is a responsibility of a sound archive to manage and control access to its collections in accordance with any legal, moral or ethical constraints which are associated with the content: Outsourcing the processes does not allow the archive to abrogate its responsibilities in this matter. When archival material is given to a third party to undertake any audio processes it is necessary to define in contract the restrictions under which the service provider must operate. For commercial copyright material the legal limitations are probably described in law and can be referred to. Where privacy or other ethical rights are of concern, these should be defined and the service provider should acknowledge their agreement to comply. It is also important to specify how and when copies will be eliminated from the contractor’s storage system when their responsibility comes to an end and the material and content is returned to the owners or archive.
9.1.6.1 Whether preservation is carried out autonomously in-house, or partly or fully outsourced, an indispensable pre-condition for seriously planning preservation is the quantitative assessment of the project. Serious and costly mistakes are often made by underestimating the amount of work needed for the optimal signal extraction from original carriers. Therefore, the first step is to count the numbers of carriers and their playing time.With mechanical carriers, compact cassettes, and optical carriers there is a fairly clear relation between the number of carriers and their respective playing times. This may be more complex in the case of magnetic open reel collections as the playing time is dependant on the length of the tape, the speed of recording and the numbers of tracks. However, with good knowledge of the specific collection, some well founded assumptions can often be made which lead to reasonably accurate estimates. In poorly documented or undocumented collections, a situation often encountered in the estates of prominent persons, this assessment can be extremely time consuming.
9.1.6.2 Once the duration of the carriers to be transferred is assessed, a second important factor is their physical condition. The time factors given in the respective parts of Chapter 5, Signal Extraction from Original Carriers, relate to well preserved items. Any cleaning and restoration measures required may add substantially to transfer times, and must be included in the calculations accordingly.
9.1.7.1 Paragraph 16 of IASA-TC 03 “The Safeguarding of the Audio Heritage: Ethics, Principles and Preservation Strategy”, describes that, except for lacquer discs, which may fail at any moment without pre- warning, the sequence of transfer within a specific collection is a multi-faceted decision based on the requirement for access to documents, their physical condition and, with ever increasing importance, the availability of equipment, spare parts and professional service support. The project “Sound Directions” has developed “FACET”1 , a tool to asses the respective parameters of a collection to assist in making a decision on a fairly objective and traceable basis. It must be noted, however, that obsolescence of formats and related problems like withdrawal of professional service support, e.g. for R-DAT machines, change rapidly, which calls for constant monitoring of the situation and re-assessment in regular intervals.
1. FACET was developed within project Sound Directions by the Archive of Traditional Music, Indiana University Bloomington, USA. http://www.dlib.indiana.edu/projects/sounddirections/facet/
9.1.8.1 It is quite common that, when commencing digital preservation, the costs of long term storage of digital audio objects are permanently and persistently underestimated. At the time of writing professional storage costs are considered to be in the order of a minimum of $US 5/GB/year2 for medium to large scale storage (over 5 TB) Although the hardware cost price has been permanently declining, the cost of the management of the storage, the continuous migration to new generation storage, the hosting in adequate premises (clean room, etc) are always underestimated. As a political target UNESCO has challenged the IT industry to arrive at $US 1/GB/year in the short term, a target which seems far from being met. Some figures detailed in a PrestoSpace study show a trend to stabilize long term storage costs at $US 9/GB/year. As digital audio objects on an average require 2 GB/hour, even future lower preservation costs will still be too high for many cultural institutions.
9.1.8.2 Lower digital storage costs can only be achieved for smaller quantities if the labour costs involved in small scale manual approaches are not incorporated. The systematic use of open source software may make autonomous, not fully automated processes viable in near future also for medium amounts (10-20TB) of storage requirements. The involvement of specialised staff to guarantee the permanent availability of archival files in manual or semi-automatic operation must not be underestimated.
9.1.8.3 Some service providers have recently developed adequate outsourced preservation strategies based on the mutualisation of the use of professional large scale mass storage systems with specific access schemes to users. Their fee is usually based on the size of the digital archives to be stored, the duration of the contract, and the associated services. For small and medium scale archives it can be an attractive solution, as well as for large scale archives before deciding to invest in their own storage solution.
2. Despite the current difference in monetary value USD and EURO are approximately same in the IT world
9.1.9.1 Perhaps the most crucial point when making these decisions is calculation of costs. Unfortunately, no generally applicable concrete figures can be given in this context. In-house costs are difficult to assess as many institutions holding audiovisual collections have infrastructures available (rooms, air conditioning, intranet) the cost of which is incorporated into the general budget, which makes it difficult to calculate overall costs for transfer and/or permanent digital preservation. Labour costs differ significantly even in developed neighbouring countries, which weakens any general conclusions as regard price. Finally, services offered by professional vendors vary considerably, depending on the amount of items per carriers, their state of preservations and hence the possibility to automate the process. Cost of staff, equipment and other resources generally rise over time, while it is possible prices for some automated processes may reduce.
9.1.9.2 Because of the many factors related to a specific preservation project these guidelines refrain from any quotation of price ranges for transfer. These guidelines suggest that holders of collections acquaint themselves thoroughly with the specific situation in their countries or regions and observe the market situation on a constant basis.
9.1.9.3 When seeking prices for audio preservation services, tenders must be well prepared and defined in detail, and any subsequent offers carefully examined. Bids that offer the same service for a fraction of other vendors should be examined with extreme scepticism. Finally, outsourcing can only be successfully managed if a stringent quality assurance system, as described, is established, and any substandard work is rigorously rejected.
9.1.10.1 In summarising preservation planning, it is strongly recommended that holders of audiovisual collections take the present need for preserving their holdings as an occasion to rethink their overall strategy: All scenarios, from total withdrawal from preservation responsibility,through cooperating in or outsourcing of signal extraction and digital long-term preservation, to taking full autonomous responsibility, should be examined. Each collection is different and institutions are embedded in a variety of environments. These multiple scenarios, which also change over time according to technical development, will make it difficult to decide on a purely economic basis. Generally, all holders of audiovisual collections, specifically of small collections, are strongly encouraged to seek cooperative relationships to manage their preservation requirements. The extent to which responsibility for signal extraction and digital long-term preservation is accepted in-house, should be linked to the general mission of that institution or collection. Memory institutions may decide differently from research collections, which have a strong interest in the availability of audio documents but whose core business does not necessarily include the processes that guarantee their further survival.