How Digitization Affects Reference Services And Access To Collections

by Dan Hazen, Jeffrey Horrell, Jan Merrill-Oldham
August 1998

Contents

Foreword

Summary

Authors' Acknowledgments

Introduction

Copyright: The Place to Brainstorm

The Intellectual Nature of the Source Materials

Current and Potential Users

Bodily and Anticipated Nature of Utilize

The Format and Nature of the Digital Product

Describing, Delivering, and Retaining the Digital Product

Relationships to Other Digital Efforts

Costs and Benefits

Conclusion

Selection for Digitizing: A Decision-making Matrix

About the Authors

About the Commission on Preservation and Access

Digital Libraries

---

Foreword

Collection building in the digital era presents challenges that libraries and athenaeum accept never before faced. They vary from having to work within licensing agreements in order to acquire serial publications, to having new, non nonetheless well-defined options for providing service of analog items through digital conversion and dissemination. What role does the digitization of enquiry collections play in a library'southward efforts to provide resources to its patrons when, where, and how they prefer to apply them?

This newspaper proposes a model of the controlling process required of research libraries when they embark on digital conversion projects. It is ane of a series by CLIR dedicated to selection policy questions that have arisen in the digital information environment. The authors of the paper offering a serial of questions to be answered that will facilitate the controlling process for library managers. They place the questions of what and how to digitize into the larger framework of collection building by focusing, start, on the nature of the collections and their use, and, 2nd, on the realities of the institutional context in which these decisions are fabricated. Their method is, to a higher place all, about helpful in its pragmatic approach to the unsettling dynamism of the digital technology itself. They view technology every bit a tool to serve specific collections-related goals and assess the available technology for its power to aid or obstruct access and preservation.

Summary

Selection for digitization is a complicated process having much in common with option for purchase, microfilming, and withdrawal, and with other strategic decision-making that is integral to the work of libraries. The conversion of textual, visual, and numeric data to electronic form—from preparation and conversion to presentation and archiving—encompasses a range of procedures and technologies with widely varying implications and costs. The judgments nosotros must make in defining digital projects involve the following factors: the intellectual and physical nature of the source materials; the number and location of electric current and potential users; the current and potential nature of use; the format and nature of the proposed digital product and how information technology will be described, delivered, and archived; how the proposed production relates to other digitization efforts; and projections of costs in relation to benefits.

Copyright assessments play a defining role in digitization projects and must be addressed early in the selection procedure. If a proposed digitizing project involves materials that are not in the public domain, permissions must be secured and appropriate fees paid. If permissions are not forthcoming, the materials cannot be reproduced and the focus of the project must alter. We will be able to catechumen to electronic form only a pocket-sized percentage of existing scholarly materials, and to do even that volition require substantial investments. Therefore, the intellectual value of the original sources, together with the types and levels of apply, must shape priorities for conversion. Ideally, the electronic version of a source will permit new kinds of use and more sophisticated types of analysis. Decisions to digitize must also take into account the physical size, nature, and condition of source materials as they affect the characteristics of the desired production. Decisions must be based on the electric current state of technology, just they must too anticipate how changes in technology could raise or make obsolete an investment in digitization. One must also assess how the production will be described for users, delivered to them, and managed over time.

Digitization, similar other reformatting endeavors, takes place inside a context larger than a single institution, subject field, or land. Selection decisions should be informed past both duplicative and complementary efforts. This may show challenging, because information technology is difficult to determine whether an item has been already digitized and past what means. Cost-do good analysis for digital conversion may as well be hard to conduct reliably, because the costs of creating electronic resources vary considerably. File size, associated storage needs, and processing requirements account for role of the differences, though labor requirements are fifty-fifty more than important. Functions such as grooming of materials for scanning, indexing, bibliographic clarification, post-scan processing, and long-term file direction oft fail to exist factored into toll equations. Incomplete price analyses can impute benefits that are difficult to stand for on a project residual sail. Though digitizing projects must calculate the likely costs and benefits, our ability to predict either of them is as nonetheless rudimentary. Thus, the decision to digitize must begin with an inquiry into copyright and an assessment of the nature and importance of the original source materials, but information technology must then proceed to analyze the nature and quality of the digitizing process itself—how well relevant information is captured from the original, so how the digital data are organized, indexed, delivered to users, and maintained over fourth dimension.

Authors' Acknowledgments

This essay grew out of the work of a Harvard Academy Library task force appointed belatedly in 1995 and charged with drafting a broadbased white newspaper to help Harvard'due south librarians and curators plan digital projects. Rapid developments in digitizing and processing technologies, file naming and metadata cosmos, estimation of copyrights and management of permissions, archiving, and other technical and administrative issues resulted in the emergence of "pick" equally the topic that could be most effectively addressed at this time.

The essay owes much to Barbara Graham, Associate Director for Assistants and Programs in the Harvard University Library, who convened the task forcefulness. Special thanks are due to colleagues Stephen Chapman, Preservation Librarian for Digital Initiatives, Lee Anne George, formerly Librarian for Data and Certificate Delivery Services in the Harvard Higher Library and now Plan Planning Officer at the Clan for Research Libraries, and Robin McElheny, Associate Archivist for Programs in the Harvard University Athenaeum, who were members of the task strength and fabricated helpful comments on successive drafts. Stephen besides collaborated on development of the accompanying period chart.

Introduction

Electronic resources are immensely appealing to well-nigh everyone concerned with education and scholarship. The potential benefits of data in digital form—unfettered admission, flexibility, enhanced capabilities for analysis and manipulation—are profound. The widely held notion that existing collections of books, manuscripts, photographs, and other materials should (and volition) be digitized wholesale is not surprising. In reality, of class, the creation and maintenance of electronic resource crave funding, skill, and ongoing commitment. Those that are intended for permanent use, moreover, will almost certainly require repeated intervention to ensure that they remain feasible as technologies evolve. In creating digital products, libraries are chosen upon to residuum the competing worlds of boundless hope and express resource. Because hard choices are unavoidable, the decision-making procedure must exist well organized and its results fully consonant with the institution's goals and values.

Selection for digitization is a complicated process having much in common with selection for buy, microfilming, and withdrawal and with other strategic decision-making that is integral to the piece of work of librarians and curators. Conversion of textual, visual, and numeric data to electronic grade, nevertheless, involves boosted layers of complication. The digitization process, from preparation and conversion to presentation and archiving, encompasses a range of procedures and technologies with widely varying implications and costs. Digital reformatting of library collections is nevertheless in its infancy, at once limiting what tin can be accomplished now and forcing determination-makers to anticipate future improvements. Scanned images optimized for viewing on today's computer monitors, for example, will display poorly on tomorrow's high-resolution screens and will require reprocessing. The same may ultimately be truthful of bitmap texts, which, if they are not made give-and-take-searchable once conversion is affordable, may be underutilized past researchers who have come to rely on key-word search capability. Considerations such as these make choice for digitizing more than challenging than choice for purchase.

The judgments we must brand in defining digital projects require consideration of many factors, including: cess of the intellectual and physical nature of the source materials; the number and location of current and potential users; the electric current and potential nature of use; the format and nature of the proposed digital product and how information technology will be described, delivered, and archived; how the proposed product relates to other digitization efforts; and projections of costs in relation to benefits.

Copyright: The Place to Begin

In that location are many interdependent and interacting factors to exist weighed in selecting materials to digitize. The specific choices that result from the choice process will reflect subjective judgments, any of which may change over time. Nuanced assessments, ambiguity, and shades of gray are all to be expected.

Questions concerning copyright, however, are far more clear-cut. Only stated, if a proposed digitizing project involves materials in the public domain, the work can proceed. If the source materials are protected by copyright simply rights are held by the establishment or advisable permissions can be secured, the work can move ahead. If permissions are not forthcoming for copyrighted sources, even so, the materials cannot be reproduced and the focus of the projection must alter. Copyright assessments thus play a defining role with regard to digitizing projects. Since the impact of copyright is so decisive, we have given it pride of place in this discussion.

Copyright issues in the digital environment are yet very much in flux and have provoked ongoing international word. While the wide thrust of digital technology is toward enhanced access, diminished costs, and more versatile capabilities, information technology is far less articulate that copyright law will besides encourage wider use. The legal strictures applicable to a particular project volition vary depending on the land in which the project is based, the country in which the source materials were produced, and prevailing international agreements. Different kinds of materials, moreover, ordinarily pose dissimilar types of rights-management problems. The functioning rights associated with musical scores, for example, or exhibition rights for films, differ from rights for nonperformance materials such as electronic journals or documentary photographs. To complicate matters, all these rights are susceptible to alter over time.

Digital projects must be undertaken with a full understanding of ownership rights, difficult as they oftentimes are to ascertain, and with total recognition that permissions are essential to convert materials that are non in the public domain. Rights that must exist negotiated with the copyright holder often entail fees. The institution hosting a project may also have policies and procedures that inform intellectual belongings negotiations. The general counsel or legal office of most institutions can provide guidance. The Internet site IFLA: Copyright and Intellectual Belongings Resources (come across ) is a good resource for maintaining current sensation. It includes articles, reports and white papers, discussions, and data about organizations related to copyright problems, intellectual property in full general, and electronic distribution of intellectual belongings.

The Intellectual Nature of the Source Materials

The post-obit sections of this newspaper separately discuss the complement of considerations that bear on decisions to digitize. The elements are presented in a sequence that moves from relatively abstract assessments of intellectual value to basics-and-bolts problems concerning whether bachelor resources and applied science tin provide a production that meets expectations. In practice, the pieces interact in means that are frequently complex.

Decisions virtually what to digitize must commencement and foremost address the intellectual value of the original sources. We are likely to be able to convert merely a small percentage of existing scholarly materials to electronic grade, and doing even this will require substantial investments. We therefore need to determine what it is truly worthwhile to convert.

Questions to Ask

Does the intellectual quality of the source cloth warrant the level of access made possible by digitizing?

Materials with marginal scholarly value are best left in their original class or made accessible in a less costly manner. Scholarly value, of course, is a subjective assessment and even the most marginal materials can back up some kinds of research. Near users, nonetheless, would opt for electronic access to original monographs rather than to derivative works, or to the papers of a prominent scholar over the administrative records of a academy department. Bibliographers regularly make purchase decisions that reflect their evaluation of the intellectual quality of unmarried items or collections of materials. Similar judgments apply in choosing what to digitize.

Will digitization enhance the intellectual value of the cloth?

Scholarship can be facilitated when texts are made fully searchable by rekeying (retyping) them or by employing OCR software. Comparisons betwixt successive drafts of a text and the final published work, for instance, or with later on editions and translations, are vastly simplified when the words and phrases are searchable. A concordance or thesaurus is likewise most easily mined when information technology is in searchable form. Electronic texts can be moved readily from one environment to another (from the World Broad Web onto the hard drive of a personal calculator, and and then into a word processing programme, for example), shared with other users, and manipulated and reconfigured for multiple purposes. Digitized prints, drawings, and other visual resources tin can be viewed in groups at low resolution or inspected individually for very fine particular. Digital charts and tables, appropriately coded, tin be loaded directly into statistical software packages for additional analysis. Census results, for instance, are virtually easily used when the information have been formatted and imported into the Statistical Package for the Social Sciences.

Volition electronic access to a trunk of information add together significantly to its potential to enlighten, or are the original books, manuscripts, photographs, or paintings sufficient to the task?

A collection of thousands of portrait images, nonetheless promising a resources, might be nigh unapproachable because of its size and the condition and dimensions of private items. Well-indexed and in digitized class, still, the collection could exist searched with relative ease for images of a particular person or for some indexed feature (the country from which the portrait originates, for example). Likewise, the digitization of large-format architectural drawings could enable comparisons of small- and large-scale drawings, different views of the same architectural feature, or sequential phases of construction.

To what extent will the combination or aggregation of original sources increase their value?

Digitizing related scholarly monographs, like building a coherent collection of paper copies, tin strengthen the context within which each title is approached. Ephemera—leaflets from a political campaign, for example—are often near useful when studied in the amass, as are posters, broadsides, and pop literature. Harvard has digitized daguerreotypes from thirteen repositories to facilitate the combinations and comparisons that are otherwise precluded past the fragility, value, and dispersion of the original images.

Current and Potential Users

Some scholarly resources are heavily used; others are consulted infrequently. With only limited funds available for reformatting, types and levels of use can help to shape priorities.

Questions to Enquire

Are scholars now consulting the proposed source materials? Are the materials existence used every bit much equally they might be?

These are complicated questions. Intensive use does not automatically brand a collection a expert candidate for digitizing. If the principal audition is local, for case, and if competition for a particular resources is non a problem, admission may already be sufficient. Ephemera produced by a customs political arrangement may be of great involvement to local scholars and of limited value to a worldwide audience. On the other hand, if use is heavy and widespread, digitizing may at in one case guarantee convenient and reliable access, and brand it possible for some institutions to discard their original copies. The JSTOR project (see ), through which a large array of core scholarly journals is existence made attainable in digital form, is a prime example of an initiative focusing on high-use materials.

Is current access to the proposed materials so difficult that digitization will create a new audience?

Low use may signal that a collection has marginal intellectual value, but there are many other reasons for valuable materials to have generated little interest. A collection may be held in a remote location, for example, or exist endemic by an institution with highly restrictive admission policies. Bibliographic records may be poor, as is often the case with pamphlets. The value of digitizing such materials may become across the simple fact that the resulting files can be widely distributed. Broader access, every bit it creates a new community of users, tin as well facilitate more active scholarship.

Does the physical condition of the original materials limit their employ?

Some resources are likewise fragile to be consulted. Aging newspapers or palm-leaf manuscripts that break at the slightest flex simply cannot exist browsed. In such cases, a digital copy might be provided to improve access, and a microfilm or other photographic surrogate made to ensure long-term survival. (Film can be made from a digital file or vice versa.)

Sources may also be at risk because of loftier user demand or extraordinary monetary value. A nation'southward founding documents, glass-plate negatives of vanished architectural sites, or rare maps may do good from the creation of digital copies that satisfy the purposes of most users. These files do not necessarily need to come across archival standards. They are created to protect the originals from treatment.

Are related materials so widely dispersed that they cannot be studied in context?

Cooperative efforts to digitize disparate pieces of a greater whole tin create or restore a more usable collection. Papyrus fragments, a prominent private's far-flung correspondence, scattered photographs of a particular subject or by a specific lensman, and broken serial runs are amidst the many materials whose coherence, accessibility, and scholarly utility can be enhanced through digitization.

Will the proposed digital files be of manageable size and format?

Digital resource need to match users' technical capabilities and equipment. Most crave Internet admission and standard web browsers, or a CD-ROM drive. Images delivered to the Internet in formats other than JPEG or GIF require boosted software for viewing or press. Fifty-fifty when electronic resources are optimized for on-screen delivery, some network connections, especially those via modem, are nevertheless far too slow to back up browsing of digital collections at satisfactory speeds. And scholars in some locations may lack training opportunities or the ongoing technical support needed to have advantage of the electronic surround. These limitations, however, are not necessarily reasons to rule out digitizing. The worldwide tendency is toward greater capabilities. Moreover, the more important the resources available electronically, the greater the incentive to acquire the network, viewing, and press technology necessary to use those resources. Digitization may, in and of itself, stimulate improved access.

Volition digitization accost the needs of local students and scholars?

Immediate demand tin can inject a mensurate of practical reality into decisions to create electronic resources. An fine art historian might seek to scan art images and make them available to students as electronic reserves, equally an culling to slide-based classroom presentations and reviews. A historian may choose to teach from digitized images of manuscripts that would otherwise be unavailable to a large course. Considering set up access to shared electronic files can transform the classroom, proposals to digitize in back up of immediate instruction needs may garner faculty support.

Bodily and Predictable Nature of Utilize

A person reading a book, looking at a photograph, or consulting a manuscript encounters few barriers to use. One might take to handle an object advisedly, or utilise a magnifying drinking glass to read fine print, merely in general the work is immediately outgoing. The same resource, when digitized, should be equally accessible and outgoing. Ideally, the electronic version will also allow new kinds of utilise and more sophisticated types of analysis.

Questions to Inquire

How do scholars apply the existing source materials? What approach to digitization will facilitate their piece of work?

Different digitizing techniques consequence in electronic files with different characteristics. These in turn can correspond well or poorly with scholarly needs. If the goal is to provide an paradigm-based finding aid that helps users identify original materials of interest, for example, mounting slow-loading high-resolution images would be counterproductive. If, on the other hand, the intention is to reduce or eliminate treatment of original materials, an image that fails to convey all critical information embodied in the original will fail to serve its intended purpose.

The simplest approach to digitizing involves use of a scanner or digital camera to create electronic pictures (bitmap images) of original materials. Decisions concerning the number of dots recorded by the scanner (resolution), how many shades of gray or colors volition exist recorded (bit depth), and other factors related to scanning equipment and settings will decide how well the digital production replicates the original. High-quality bitmap images tin usually capture all the significant item in texts or graphics. Scanning rare and unique texts or visual resource can brand them accessible to users who would otherwise never see them. In such a case, merely reproducing the original in electronic grade represents an extraordinary enhancement.

For textual materials, post-browse processing tin support expanded capabilities. Scanned text tin exist processed with Optical Character Recognition software to produce searchable indexes. OCR software is at present simply occasionally employed in digitizing projects because it cannot even so interpret accurately all fonts and alphabets, and because it adds significantly to per-folio costs. Text tin can too be rekeyed to create ASCII files (very straightforward digital text files that permit searching by keywords or phrases). In some cases this enhancement is the primary justification for digitization. Directories, dictionaries, and indexes are all significantly easier to use when specific words tin be searched inside a well-designed digital file.

ASCII texts accommodate primal-word searching (e.g., searching for all instances of the word "temperance") and some kinds of analysis, merely they do not readily replicate the structure and format of an original document. Without special coding, researchers cannot directly consult the seventh paragraph of the third affiliate of a particular text. Nor tin can they search for all occurrences of "welcome" used every bit a verb rather than a noun. These capabilities become possible in marked-up texts, which are coded to highlight elements of structure, format, and syntax. The Standard Generalized Markup Language (SGML) is the emerging model. One SGML awarding, the Encoded Archival Clarification (EAD), is being used to create electronic versions of archival finding aids.

These and other approaches to digitizing carry very different costs, benefits, and resource requirements. While electronic versions can be more versatile than original materials, in some cases they hinder inquiry. A scholar studying bookbindings or papermaking, for case, is poorly served by a reproduction of whatsoever kind. So too is the scholar whose immediate access to a large and important collection of literary works is sacrificed in lodge to serve a worldwide constituency—mayhap because bound volumes have been disbound for scanning.

Volition digitization increase the utility of the source materials? Will information technology enable new kinds of teaching or enquiry? Practise scholars agree that the proposed product will be useful?

Digitization can raise original materials in many ways. Paradigm quality can exist improved past eliminating inapplicable stains and marks. Thumbnail images of visual resources (photographs, drawings, paintings) can be browsed to discover patterns, trends, and relationships amid individual items, and specific images can then be scrutinized at higher resolution. Likewise, patrons can review scanned images to identify needed materials before requesting that they be retrieved from storage.

Electronic transcriptions of texts, in ASCII format or marked-up files, can be linked to bitmap images of original documents. Readers tin can then decide for themselves whether "authoritative" transcriptions are in fact accurate. Comparisons of different versions of a text are also simplified. Related texts and images tin be assembled together inside a single, unified corpus. Examples such as the Dante Project mounted past Dartmouth Higher (run into ), which reproduces and links related texts and commentaries concerning the Divine Comedy, and Tufts Academy'southward Perseus Project (see ), an interactive, multimedia database on Archaic and Classical Greece, advise the potential of electronic texts.

Virtually all electronic products volition provide basic links that allow users to navigate them (to locate a item map within a printed text, for case). The degree to which a digitization project exploits electronic links will depend upon its intended use. For digital resources created every bit pedagogical tools, predetermined connections are office of the package. Products intended for research tend to be less aggressive in ordaining relationships among sources, since their creators assume that researchers will build their own structures of meaning.

Are there other scholars, librarians, and archivists who tin collaborate to create a useful product?

Colleagues and potential users tin analyze ideas, help select meaningful materials for conversion, improve project pattern, and stimulate early involvement. "User demand" reflects both the intrinsic utility of specific source materials besides equally a social context of participation and promotion.

The Format and Nature of the Digital Product

Decisions to digitize must take into account the physical size, nature, and condition of source materials as they touch on the characteristics of the desired product. They must likewise address whether available means of conversion can satisfy expectations for the result. Projects must also, from the very first, consider how users volition be guided through the electronic version.

Questions to Ask

What critical features of the source fabric must be captured in the digital product? Are very high resolution copies, authentic rendition of colors, a seamless combination of images and text, or other qualities considered essential?

The cost and nature of digitizing hardware and software continue to evolve, and preferred solutions are likely to shift too. It may sometimes make sense to defer certain digitizing projects so that engineering science tin can catch up to needs. The success of a project to digitize oversized maps at Columbia University, for example, depended partly on the ability of users to see detail and read place names. Every bit a result, the maps were scanned at relatively loftier resolution, thereby creating challenges for digital image delivery and presentation. File sizes were very large and initially outran the capacity of the library's computers and network. Greater bandwidth and more powerful machines take enhanced functionality.

If the original sources are to be retained, can they withstand the digitization process?

Automatic canvas feeders are fast and efficient, only they may destroy breakable paper. Digital cameras can minimize the manipulation of source materials, but subjecting certain media—watercolors, for example—to prolonged lighting is problematic.

What type of hardware should exist used for conversion?

Colour slides, for instance, cannot be fully represented by scanners that create only black-and-white images. Even a colour scanner with express capacity to reproduce tonalities will be inadequate when high-quality images are of import. Digitizing equipment can be expensive, and the costs may exist difficult to justify when utilise is sporadic. Some projects may thus exist done virtually economically if they are contracted out. Agreements with external vendors, in improver to specifying technical atmospheric condition, performance expectations, and treatment guidelines, must fully define ownership and distribution rights for all digital products.

Volition a digitized sample see users' needs? If then, how should the sample be constructed?

Many collections are too large to catechumen in their entirety. In the case of an creative person's drawings, 1 might select materials from each of the artist'south major periods, or representatives of the diverse media in which he or she worked, or detail subjects, such as cityscapes or portraits. Subsets of big collections can be defined in many ways and for many purposes. Collaboration with scholars and other experts is essential.

Volition the information resources upon which the project is based continue to grow?

Ongoing commitments and extended arrangements for copyrights may exist required when collections are still expanding, equally is the example with current journals and annual reports, or the papers of a living private. Consultations with scholars and other experts can be particularly useful, since the long-term value of current materials is often difficult to discern.

How will users navigate inside and amid digital collections?

Printed sources orient readers by means of tables of contents, capacity and sections, pagination, indexing, and formatting cues. Manuscript materials often rely on finding aids linked to the organization of file folders. Photographs may exist mounted in albums. At a minimum, electronic products need to provide the same kind of functionality. The procedure may require several steps. For a multi-volume piece of work that has been scanned page by page, for case, each page is a divide computer file that must exist individually labeled and stored. The files for critical pages of the piece of work—for case, the title folio, table of contents, and the starting time folio of every new affiliate—must and so be linked to electronic navigational tools so that they can be easily located.

Describing, Delivering, and Retaining the Digital Production

While libraries can betoken with pride to their collective achievements in organizing and describing an enormous number and variety of collections and material types, some perplexing issues have not nonetheless been resolved. At that place is still no consensus on how to handle ephemera that cannot realistically be cataloged past the piece and that are too insubstantial to shelve like most books and journals. Providing access to mixed media (a book accompanied by a floppy deejay or CD-ROM, for example) is likewise problematic. But these bug, complicated as they are, pale next to the challenges of making digital products available to users. Decisions as to what resource should be digitized must be informed by an understanding of how the product will exist described for users, delivered to them, and managed over time.

Questions to Inquire

How volition users know that the digital file exists?

Bibliographic records, finding aids, and indexes can all be adapted to include references to electronic resources. Still, our ability to determine what has been digitized remains well behind what nosotros know about materials that take been microfilmed or photocopied.

I of the principal challenges is to decide what data is essential in describing an electronic product. The "Dublin Cadre" (see http://purl.oclc.org/metadata/dublin_core/) and other special initiatives for structuring and standardizing descriptive data suggest to combine information about the technical characteristics of digital files (how they were created), their location, and a summary of their contents. The resulting records are known equally "metadata." Their office is to provide users with a standardized ways for intellectual admission to digitized materials. Despite these and other initiatives, projects to catalog digital files are but in the developmental stage. No system has however been widely adopted for tracking the digitizing activities of libraries, athenaeum, and museums, although new approaches go on to sally.

How tin can the digital production best be delivered to users?

Alternative modes of digital storage and delivery must exist considered from the showtime of a project. CD-ROMs, for instance, are distributed and used differently from information made attainable over the Internet. The differences are reflected in hardware requirements, software, and ease of utilize. CD-ROMs are sometimes bundled with software for searching and analysis that is superior to that more often than not provided for Cyberspace files. On the other hand, access to CD-ROMs is limited to individual workstations or small networks, while Internet files tin can be made available to a very broad audience. And Internet resources, past nature, tin can be updated or augmented without requiring users to supplant objects that accept become obsolete.

Internet products, however, generate questions of their own. How immediate must admission be? Files tin can exist mounted on a server so that they are instantaneously available on-line. They tin can be stored on disks in a jukebox and loaded on demand ("near-line" access), or kept off-site ("off-line") and retrieved and delivered on demand. Near-line and off-line access can save on server space and requirements, though there are countervailing staff costs associated with retrieving and mounting the files. Expected demand, file sizes, fee structures, and bachelor staffing and equipment must all exist considered.

Who will be authorized to use the digital resource, and under what circumstances?

Copyright holders may limit distribution rights, institutions may be unable or unwilling to provide the infrastructure needed to back up universal access, and cost-recovery enterprises cannot by definition make their products available without restriction. Digitizing projects must thus consider access policies and control, pricing mechanisms, and billing procedures. Access problems impinge upon selection decisions in a number of means. A university may mount high-resolution images of unique holdings for scholarly use (a medieval manuscript, an of import collection of drawings), simply would non allow unauthorized publication of those images. Moreover, electronic resources price money that must be secured through subsidies or fees. When neither internal budgets nor external subventions provide acceptable financial support, digitization volition require a paying audience.

Access, when it is not universal, must be managed. Current alternatives include passwords, direct user fees, and limitations according to organizational affiliation. Unlike capabilities for viewing, downloading, and press may be offered at unlike prices or to different sets of users. There are many options, each reflecting a different pathway toward a self-sustaining effort.

How volition the integrity of the digitized data exist ensured?

The malleability of electronic products makes them particularly useful for many kinds of scholarship. Digitized files must be embedded with detailed information apropos the methods used to create them. The aforementioned information should be included in external bibliographic or descriptive records. Users who are consulting or copying the sources must also exist able to confirm that the files they see or receive match the originals. Means to authenticate and protect digital products, long available in financial and industrial applications, are but beginning to take hold in the scholarly world.

Specially for digital products created to meet local need, is the existing technology infrastructure acceptable?

Robust reckoner systems and an appropriate number of work stations are perhaps more hands provided than such ancillary features as network press capabilities in the library and in offices, classrooms, and residences.

What are the long-term intentions for the digital file?

In the case of electronic certificate delivery systems such every bit ARIEL (a product of the Research Libraries Group, Inc.), the goal in virtually cases is to provide very rapid access to specific articles or chapters. While images must be legible, they demand not be perfect replicas; and copyright constraints, indexing complexities, and storage economies arrive simpler to rescan on need than to organize and retain random files. In other cases, the file may be kept for a longer, but notwithstanding express, menstruum and then discarded—a reserve reading listing or copyrighted images of artworks scanned to support classroom didactics, for example.

Is the long-term preservation of deteriorated materials a project goal?

Preserving documentary resource in electronic format presumes that, to the greatest extent possible, all the information independent in the original fabric has been captured completely and accurately. This requires careful attention to significant detail, whether the smallest text character on a page or all the shades and tones of blue and green in a seascape. Targets for resolution, grayscale, and rendition of colour either be or are existence adult to ensure the needed particular and fidelity.

Digital preservation also requires a supporting system and infrastructure dedicated to storing the electronic files and to migrating them to new formats and/or media as technologies change. Unless these capacities are all in place, digital files cannot be regarded as permanent. Creating an enduring digital preservation master file is a multidimensional task with long-term implications. Hybrid projects, in which digital files are complemented by copies on microfilm, alkaline paper, or some other stable medium, provide the insurance that exclusively electronic projects do not.

Digital processes see preservation objectives without pretending to permanence. In the case of Spain's Archivo de Indias, for instance, depression-resolution grayscale images were prepared so that frail original documents, some more than five hundred years sometime, could be spared the rigors of repeated consultation. The digital files, while they fall well short of capturing all the information in the originals, nonetheless fulfill a vital preservation function.

Relationships to Other Digital Efforts

Digitization, like other reformatting endeavors, takes place inside a context larger than a single institution, subject, or country. Pick decisions should be informed by both duplicative and complementary efforts.

Questions to Inquire

Accept the materials proposed for digitization already been converted to electronic form?

Every bit we take seen, it can be difficult to make up one's mind whether a specific detail has been digitized and past what ways. If an electronic copy does exist, is information technology authentic, satisfactorily functional, and accessible? Does it have advantage of the capabilities of electric current technologies? If the existing production does not serve the intended purposes of the proposed project, a new version may be warranted.

Can cooperative digitization efforts bring together a cohesive body of material that would otherwise remain disassociated?

Standardized descriptors and a common approach to indexing and storage can allow dispersed materials to exist combined in an confederate digital resource. The process involves institutional alliances as well as technological conventions. Different levels of participation and different expectations for returns may touch on the upshot. If i institution provides the majority of materials for a digital project, for example, with many others completing the whole, the "atomic number 82" institution may claim special consideration or returns, requiring actress negotiation.

Successful projects to combine digital resource through a mutual organisation for organization and delivery suggest a new kind of model for drove edifice. Even in preservation microfilming, cooperative efforts to preserve a single title typically involve assembling dispersed materials at a key location for filming, or bringing together film prepared at diverse locations for splicing and duplication. The workflow of digital collection development can remain radically decentralized provided a robust infrastructure for collaboration is in place. The Research Libraries Group projection, Studies in Scarlet: Union, Women, and the Police, 1815-1914, is a case in point (see http://www.rlg.org/blood-red/sister.html). Six U.S. libraries and one in Great United kingdom have scanned trial accounts, case police force, statutes, treatises, and other materials related to the theme expressed in the projection title. RLG established file naming conventions and other guidelines, designed the interface, and will serve the images—one of several models being explored for the creation of virtual collections. The conceptual kinship with traditional drove evolution is articulate.

Costs and Benefits

Cost-benefit analysis assesses the human relationship betwixt functionality, demand, and expense. Limited resources and competing demands on organizational time and energy hateful that the analysis must be rigorous and consummate. The costs of creating electronic resources vary considerably. File size (and the associated storage needs) and processing requirements account for part of the differences, though labor requirements are fifty-fifty more than of import. Bitmap images in black and white are relatively inexpensive to produce and shop. Grayscale images, currently capable of capturing upward to 254 shades of grayness plus black and white, are more costly; color images are the most costly of all. In each case, images with higher resolution upshot in larger digital files.

Accurate ASCII files of searchable text, even though occupying far less estimator retentiveness than any image file, are more expensive to produce than bitmaps of the same textile. The main reason is that OCR software is not even so fully reliable. Materials converted by machine must be painstakingly proofread, or the source documents must exist rekeyed in combination with careful attention to the detection and correction of errors. Costs ascent even more for marked-up text, which entails yet another level of analysis and intervention. Creating other kinds of special databases or enhanced capabilities, for image files or for text, likewise raises the costs.

Costs vary even inside specific approaches to digitization. All other things being equal, for example, it is less expensive to browse from single sheets than from bound volumes. Small sheets are less expensive to scan than oversized ones. Items in good condition are less costly to process than those that are deteriorated and thus require special handling.

Bachelor cost figures for digitizing projects are often misleading. Cost projections seek to pin downwards a apace moving target. Although the prices of computer storage and processing power, for example, continue to fall, most projections simply extrapolate from available information well-nigh electric current price structures. Analyses oftentimes fail to account for sure categories of effort that, were they included, would alter toll calculations significantly. Labor expenses, for instance, oftentimes reflect but a pro-rated price per page that overlooks the real cost of a full-time employee. Crucial pieces of the workflow are sometimes written off equally one-fourth dimension "research and development" expenses. Functions such every bit grooming of materials for scanning, indexing, bibliographic description, mail service-browse processing, and long-term file management may non be factored into cost equations. Incomplete price analyses can impute benefits that are hard to correspond on a project balance sheet. It may be true, for example, that prepare access to backfiles of digitized journals volition ultimately reduce or eliminate structure costs for new stack infinite. Unfortunately, money not spent on capital projects is unlikely to exist reflected in support for other library initiatives. Though digitizing projects must calculate the likely costs and benefits, our ability to predict either of them is as yet rudimentary.

Questions to Ask

Who will benefit from the proposed digital product?

Information technology is of import to consider whether the product will support better pedagogy or research and enable students to learn more, or in different ways—if, for example, texts or images are more than fully revealed. Digitization may let librarians to manage collections and provide services more effectively, or to provide traditional services such as copying or interlibrary loan at lower cost or at less risk to collections.

Is the intellectual value of the proposed product commensurate with the expense?

The limited resources available for digitization might have greater impact if they were directed at another project, or directed toward an entirely different arroyo to providing access—through exhaustive indexing maybe, or microfilming, or another type of reformatting that would prove in the cease more useful to scholars.

Could an acceptable product exist created at lower cost?

When materials are scanned to support curt-term class work, for example, careful (and expensive) post-scan processing to eliminate inapplicable marks and speckles or to deskew misaligned images may be a waste of time. Likewise, an adequate substitute for full-text scanning of little-used journals might be provided by linking scanned tables of contents and indexes to bibliographic records and relying on traditional forms of document commitment.

How will the proposed project address the long-term costs associated with digital files?

The accumulated body of digital products may enable savings elsewhere in the institution—for example, by reducing staff costs for reshelving jump journals, or by lowering the costs of storage, circulation, and preservation—and these savings could offset some or all of the expense of digitizing. Only such savings as may be realized are difficult to predict. It is essential to realize that the costs of digitization are just beginning at the time of initial capture. The programmatic capacity to distribute and maintain electronic resources, and to migrate them to new forms as original digital platforms fail and formats and software are superseded, is fundamental to long-term efforts. In addition, there are staff costs associated with grooming and user support. Finally, rising user expectations may require that existing digital files be reprocessed in new ways. When OCR software is perfected, for example, unsearchable bitmap images of texts could be thought unsatisfactory. Projects that do not plan for change may get obsolete, and therefore irrelevant.

Tin can external funding exist secured to support the proposed project?

Some foundations are particularly interested in electronic products, and specialized scholarly initiatives may attract their support.

Decision

Research libraries are eagerly embracing the digital world. They are acquiring admission to great quantities of electronic materials produced exterior their walls and are making digital versions of their own holdings. These projects, every bit they become more common, are bringing both the broad bug and the nuances of the digitizing procedure into sharper relief.

Projects based on conscientious review, analysis, and planning can yield electronic resources that are functional and faithful to the original sources, and that support new kinds of scholarship. A detailed plan of work, regular assessment of progress, closely documented adjustments and corrections, and the retention of other project-related data can strengthen the knowledge base for future efforts. Each success, as well as each failure, will bring us closer to fulfilling the promises of the electronic environment.

The process of deciding what to digitize anticipates all the major stages of project implementation. Digital resources depend on the nature and importance of the original source materials, only also on the nature and quality of the digitizing process itself—on how well relevant data is captured from the original, and and so on how the digital data are organized, indexed, delivered to users, and maintained over time. Disciplined efforts to address the themes and questions outlined in this essay will help ensure that new digitizing projects fulfill the expectations of libraries, students, and scholars.

About the Authors

Dan Hazen, Librarian for Latin America, Spain, and Portugal at the Harvard Higher Library, is an surface area specialist actively involved in preservation microfilming and digital conversion projects both in America and abroad. Jeffrey Horrell, Associate Librarian of Harvard Higher for Collections, has served in library positions at the Academy of Michigan, Dartmouth College, and Syracuse University; he has also pursued enquiry in the history of photography. January Merrill-Oldham, Malloy-Rabinowitz Preservation Librarian, directs the work of the Harvard University Library Preservation Center and the Harvard College Library Preservation Services Department, where appropriate means of integrating digital capabilities with established preservation and access strategies are being sought.

Committee on Preservation and Admission

The Commission on Preservation and Admission, a program of the Council on Library and Information Resources, supports the efforts of libraries and archives to save endangered portions of their paper-based collections and to encounter the new preservation challenges of the digital surroundings. Working with institutions around the world, the Commission disseminates knowledge of all-time preservation practices and promotes a coordinated approach to preservation action.

Digital Libraries

The Digital Libraries plan of the Council on Library and Data Resources is committed to helping libraries of all types and sizes understand the far-reaching implications of digitization. To that end, CLIR supports projects and publications whose purpose is to build confidence in, and increase understanding of, the digital component that libraries are now adding to their traditional print holdings.

Source: https://www.clir.org/pubs/reports/hazen/pub74/

Posted by: meyersnobbland.blogspot.com

Share this post