organization’s perspective on the future of libraries and archives. In an evolving environment, new operational modalities will be required even if the financial base remains constant. In the for-profit environment, new operational modalities can increase productivity as in the development of a new product line. The benefits (profit) in the non-profit sector are not as easy to measure. It is difficult enough for any administration to balance the exegesis of one function over another, let alone to address the introduction of a heretofore-unaddressed activity.
Few would argue that libraries and archives have not experienced a period of change in recent years. These changes have made many question the very relevance of libraries. Over twenty years ago James Thompson wrote, “In terms of size, arrangement and catalogues, the conventional library has reached an organizational and financial impasse. Coincidentally there has emerged a preemptive new technology for the storage, handling and transmission of information, potentially better suited to the convenience of users. Libraries may disappear like the dinosaurs; or they may, by returning to first principles, be able to adapt and successfully survive.”1 As to what those “first principles” might be, Thompson continues, “to provide online access to resources for individuals who do not have their own terminals; libraries to be centres in which trained personnel will be available to assist the user to exploit databases and databanks; libraries to take chief responsibility for materials of purely local and very specialized interest; libraries to provide a high-quality question-answering service based on a vast shared ‘electronic encyclopaedia’; and libraries to act as centres for important community information services.” 2
Clearly with the advent of low cost computers and wireless communications, the provision of workstations may not be relevant to libraries of the future. While libraries certainly have “trained personnel,” the skill sets have been changing. In many respects, libraries have deemphasized some aspects of their operations, which could be seen as core strengths. For example, with the advent of the outsourcing of selection via approval plans, the notion of the subject specialist, conversant in the vernacular of a discipline, is increasingly becoming a rarity. While there are questions of the quality of much of the information published on the Web, it is not uncommon for an individual to complete an undergraduate degree without ever having been in a library or relied upon those “trained personnel.”
As for that “preemptive new technology for the storage, handling and transmission of information,” the bulk of information is being created with that new technology, the computer. “The world produces between 1 and 2 exabytes of unique information per year, which is roughly 250 megabytes for every man, woman, and child on earth. An exabyte is a billion gigabytes, or 1018 bytes. Printed documents of all kinds comprise only .003% of the total. Magnetic storage is by far the largest medium for storing information and is the most rapidly growing, with shipped hard drive capacity doubling every year. Magnetic storage is rapidly becoming the universal medium for information storage.”3 Relative to the storage and retrieval of information, print is ephemeral, yet it remains central to the acquisitions budgets of most libraries. It would seem, based on casual observation, that libraries are having a difficult time defining their function.
A notion understood by archives but not often given a high priority in libraries can be found in the subheading of “Ethical Considerations,” in the technical committee report of the International Association of Archives: “There are four basic tasks that are performed by all archives—acquisition, documentation, access, preservation. The primary task, however, is to preserve the information placed in the care of the collection.”4
If one accepts some of the notions expressed by Thompson, the future of libraries perhaps resides in the core strengths of its own past as well as in the artifacts of the past, and in the unique materials placed in its trust, those “materials of purely local and very specialized interest.” As for the activities and services related to those materials that can cannot reasonably be expected to be automated, one could look to the preservation of unique materials and informed access to those collections, areas historically undervalued in libraries, but central to archives. Unique recordings are amongst the most significant artifacts of the past. To minimize the value of audio recordings is to minimize the importance of the information we receive through our sense of hearing.
Audio preservation is problematic for a variety of reasons. Unlike print, recorded audio cannot be auditioned without playback equipment. It is also necessary to have the technical background and listening skills to know what represents optimal retrieval of that audio. Machines have to be maintained in proper calibration, and often times, such technicalities as equalization require specialized listening skills. Even many well-informed technicians and musicians can make plausible arguments for differing perspectives on such fundamental issues as pitch in a recording. What size needle is needed to provide the least amount of noise in playback and is this desirable as the greater the noise level, the easier it can be identified by much of the noise reduction software? These are but a few of the variables that indicate that there cannot be an absolute in reformatting. This leaves us with the primary and central expense to any audio preservation program: technical expertise coupled with a trained ear and an in-depth knowledge of preservation techniques. Knowledge of the subject and what recordings may or may not be unique (what deserves reformatting) is similarly requisite. It is rare to find all of the requisite skills in a single individual.
This one component, informed personnel, is perhaps the most expensive long-term investment, the most needed, and the most difficult to acquire. Unfortunately, at this time, there is no certification in audio preservation and only one institution of higher learning in this country is offering classes in audio preservation, namely, Preservation and Conservation Studies at The University of Texas at Austin.
Audio exists in time. Digitizing analog audio requires real time playback and the full attention of the technician. This suggests the technician needs the opportunity to work uninterrupted in an acoustically isolated environment. Not only does the isolation provide the technician the opportunity to listen at amplitudes that are needed to adjust his or her work without disturbing adjacent work areas but, of equal importance, it keeps extraneous noise from disturbing the monitoring of the audio. Drop-in soundproof modules can be purchased for under $4,000.
There are no mutually agreed upon standards for audio storage. Any cost estimates will be subject to the level of resolution and target format for reformatting and your preservation of that target format. Even the fundamental question of what sampling rate should be used is debated. The audio CD features 44,100 samples per second with a resolution of 16 bits. Is this adequate? “To approximate a high-fidelity sound with a bandwidth of 15,000 Hz, we require 30,000 samples per second, or a sampling time of 1/30,000 per second.”5 Today our standards are higher. Digitizing a sine wave of 12,500 cycles per second, a pitch near or just beyond the top of the available range of hearing for those of us past the age of 50, will clearly demonstrate the limitations of a sampling rate of 44,100 per second. A graphic display of the wave will show significant distortion. The new audio standard calls for a resolution of 24 bits with 96,000 samples per second. Even higher resolution and sampling rates are considered desirable by some. Choosing an appropriate standard to accommodate the needs of a particular institution or group of materials is yet another aspect that requires informed staff assistance.
Then, what of the final storage format? “It must be stressed that coding schemes used for preservation purposes must be openly defined and not proprietary to a limited number of manufacturers.”6 Further, “it has become generally accepted that, when selecting a digital target format, formats employing data reduction (frequently mistakenly called data compression) based on perceptual coding (lossy codecs) must not be used.”7 Oddly enough the “aging” technology of the CD-R remains as the one universally accepted format which can accommodate these criteria.
One of the often spoken axioms in audio preservation is “it is better to do something than nothing.” A lacquer disc not transferred is likely to be lost. If your organization is not willing to devote the resources to “do it right,” low cost equipment can handle some of the more familiar formats. There are inexpensive systems that combine cassette playback with a CD recorder in a single unit costing less than $700. That system also includes inputs for a magnetic cartridge. There are also stand-alone systems that will partially automate the digitization of reel-to-reel tapes.
While the primary equipment needs for an audio reformatting lab will be for the preparation of a “straight” or unaltered transfer of a recording, it is important that consideration be given to restoration. In many instances, perhaps due to an excessively high level of background noise, or a significant amplitude difference between an interviewer and his subject, a “straight” copy will be of little value to the patron. Therefore, it is recommended that any facility include equipment and/or software for restoration. A reasonable argument can be made for preparing a restored copy along with the “straight” copy as part of the normal process of reformatting. The noise reduction technology of today is affordable, and when used judiciously by a trained ear, can produce remarkable results. While there are always new technologies on the horizon (such as imaging for the playback of discs) which hold the promise for better noise reduction, the initial monitoring of the original during the transfer will identify items in need of restoration, making it more effective to do the restoration at that time. Considering the potentially large collections of recordings, it may be unreasonable to assume one will have time to deal with a recording more than once.
A reasonably well-equipped lab, designed for a moderate level of resolution, could easily consume $60,000 in equipment costs. This cost estimate includes basic analog playback of most disc formats, cassette tape and reel to reel, monitoring equipment, digital to analog converters, a computer for editing, basic restoration software, editing software, and CDR burning capabilities. For cylinder playback, one needs to add approximately $12,000. These costs do not include the recommended acoustically isolated studio or the cost of staff and supplies.
Outsourcing audio reformatting can cost between $90-$100 an hour. This modality still requires that the quality control be monitored and may present security problems for rare materials, which could need to be transported. An employee working a 40-hour week with a two-week vacation, holidays, and sick leave can be expected to be available approximately 1,920 hours a year. Subtracting time spent on equipment maintenance, meetings, participation in professional organizations, professional development, etc. many work models estimate productivity of professional positions at 80%. That could leave a full time employee with approximately 1,440 hours a year to do restoration. A salary of $40,000 equates to approximately $28 an hour. That does not include the cost of supplies, a work place, or the equipment. Outsourcing for 1,440 hours at $90 an hour could cost $129,600 versus $40,000 in house, leaving a balance in a single year of approximately $89,000 which could be used for supplies and equipment. The major benefits of outsourcing are for those organizations that have a finite quantity of audio to reformat or formats that require esoteric and generally unavailable analog equipment for playback.
Libraries frequently outsource the preservation of their digital information to computation vendors, database managers, and technicians not directly under the control of the library. However, it is essential that at least one informed technician be available on the library staff to create and monitor the specifications of any contractual agreements with outside vendors. The same is true for audio reformatting. Even if the bulk of the work is to be outsourced, it is essential that informed staff draft specifications and monitor the quality of the work. This requires, at the very least, the equipment and expertise to check what is on the recordings and to monitor the work done by the outside vendor. In-house expertise is also likely to be needed to provide and maintain patron playback facilities.
As with any expense, decisions regarding audio preservation and restoration should not be viewed in isolation, but considered in the light of the operational or systemic functionalism of the organization. Cost recovery can be measured in many ways. It can be viewed in the context of one task providing ancillary support to another, thus providing a reduction in expenditures. Similarly an institution that provides training in preservation or audio engineering might consider internships, directed at the processing of materials, as being beneficial to both the collection and the educational process.
The audio technician can serve as a resource to many preservation activities such as the determination of all digital storage modalities and the preservation of all digital information. This person can serve as a resource for the emerging preservation technologies. The experienced technician can provide mentoring of staff and assist in the preparation of materials for publication on the web. There are also opportunities for lower level staff to perform reformatting operations, once an informed methodology is established, that can be monitored by a trained professional. When the copyrights allow, in rare circumstances, the publishing of donated recordings may help defray the costs of preservation.
Audio was the first information stored electromagnetically. In many ways, the study of the history of audio preservation is a guide for the future of the preservation of electromagnetic storage. It is a history filled with changing formats, lost information, all subject to quickly evolving market driven technologies. If the past is any indication, the future is not bright.
The uninformed often look to the digital technology as the savior of preservation. Unfortunately, “digital carriers must also be regarded as endangered by decay, especially if they have never been checked for their data integrity. Several formats already show obvious signs of chemical decomposition. Some are additionally threatened by obsolescence of hardware.”8 “To date, none of the digital recording systems developed specifically for audio has achieved a proven stability in the market place, let alone in an archive. The commercial lifetime of modern formats and systems is likely to become shorter and shorter. In the future this may increasingly lead to the obsolescence of hardware while the carriers are still in good condition…It is also likely that in some arenas there will be no physical carrier to distribute, in which event the issue of format obsolescence applies to the file format itself.”9
Audio archives owe their existence to those librarians and archivists whose vision of the future embraces both the technology of the present and the future and the history of the past. It is for those who value the substantive information found in sound. When that sound is music, it can be what the great American composer William Schuman once said to me, “amongst the most noble expressions of the human spirit.” It would seem when we lose any unique audio recording, we are losing a part of ourselves.
The ultimate solution to information preservation requires the development and/or ownership of proprietary software to access digital information, software that can be universally adopted, made available at low cost, and maintained indefinitely. It needs to be coupled with a storage technology with the permanence of stone. Only if such a utopian notion becomes a reality can we ever consider information preserved. As the quantity of information continues to grow exponentially, the knowledge of technology is a most obvious requisite skill. Less obvious a need, but of equal value, are those professionals with the subject expertise to make the informed decisions regarding what will be preserved. Ultimately, as the economic limitations and even the theoretical limits of information storage tell us, we can’t save it all.
Thompson, James, “The end of libraries,” The Electronic Library, Vol.1, No.4 (October 1983): 245.
Ibid., 254.
http://www.sims.berkeley.edu/research/projects/how-much-info/summary.html, c.2000 Regents of the University of California.
International Association of Sound Archives, Standards, Recommended Practices and Strategies, Version 2 (September 2001): 3. Also available online: http://www.iasa-web.org/iasa0013.htm (accessed July 2004).
Mathews, Max V. The Technology of Computer Music (Cambridge, M.I.T. Press, 1969): 5.
Standards, Recommended Practices and Strategies, 7.
Ibid., 8.
Ibid., 11.
Ibid., 5.
© Karl Miller 2004