TECH NEWS--Digital Libraries--Part I: Choosing a Scanner
Calculating Scanned Image Size
Important End-of-Fiscal-Year Deadlines for OCLC Services
Staff Profile: Joan McCoy, Acting FEDLINK Section Head at C&L
Notify FEDLINK of Contact Information Changes
Save the Date! 1998 FLICC Forum--March 19
FEDLINK's annual report to the Library of Congress revealed that the program closed the fiscal
year with the highest service dollars total in its history. In FY97 FEDLINK provided its members
with $55.4 million in transfer pay services and $72.3 million in direct pay services saving federal agencies more than nine million dollars in cost
avoidance and millions more in vendor discounts.
To support federal libraries in their use of information technology, FEDLINK Network
Operations (FNO) staff conducted 113 OCLC and Internet training classes for more than 800
students in the US and abroad. Staff members also redesigned the FLICC/FEDLINK Web site to
expand members' access to financial records, program information, vendor services, and training
resources.
To improve the program further, FEDLINK contracted with Abacus Technology Corporation in FY97 to perform a strategic review of the following areas: program cost/benefit analysis; development of a formal five-year business plan; and program cost allocation/accounting for transfer pay and direct pay activity. FEDLINK program managers collectively reviewed and approved the Abacus reports; staff are working to implement a number of the Abacus recommendations in FY98.
Procurement Services
FEDLINK wholly revised its solicitation for electronic information retrieval services in FY97 to
reflect the changes in the information industry. The revised solicitation included several firsts for
FEDLINK: it specifically addressed electronic serials; it identified the federal community's
preferred terms for licensing of electronic publications; it allowed vendors to offer print
publications associated with their electronic offerings, thereby consolidating sources for
members; and it allowed vendors to make customer-specific offerings of their electronic
materials and associated research services.
FEDLINK also worked with LC's Office of General Counsel in FY97 to develop a model licensing agreement for electronic databases and publications and to identify specific points for negotiation regarding the licenses offered by the largest FEDLINK vendors (Knight-Ridder Dialog, Lexis-Nexis, Westlaw, Chemical Abstracts, etc.). The model licensing agreement provides baseline language that federal libraries can propose to vendors as an alternative to the standard commercial language if they wish to negotiate locally. The model agreement will be available to members in early 1998.
FEDLINK Fiscal Operations
To support FY97s record-breaking volume, FFO: processed 10,140 member service transaction
requests for current and prior years, representing $55.4 million in current year transfer pay, $8.8
million in prior year transfer pay, $72.3 million in current year direct pay, and $100,000 in prior
year direct pay service dollars; issued 54,866 invoices for payment of curren and prior year
orders; incurred virtually zero net interest expense for late payment of FEDLINK vendor
invoices; completed Fiscal Year 1992 member service dollar refunds to close out obligations for
expired appropriations; successfully completed work associated with Abacus strategic reviews of
the FEDLINK program; ensured that administrative expenditures/obligations did not exceed
program fee projections; and implemented plans to develop requirements for the successor
automated financial system (SYMIN) and improve the efficiency of FEDLINK's financial
processes.
OCLC Network Activity
OCLC Users Meetings in the Fall and Spring briefed members on enhancements to the online
systems, cost saving opportunities in interlibrary loan, and OCLC's new electronic collections
program. FNO developed new OCLC classes, including a basic overview of Windows to
complement the current Passport for Windows class and an advanced ILL class. FEDLINK staff also joined OCLC in site visits to FEDLINK member libraries throughout the year.
FEDLINK Internet/Technology Program
In FY97 FNO focused on Internet training, conducting 30 Internet classes in the DC area and
three in the field. FNO also participated in a number of special projects:
FEDLINK Training Program
All told, staff conducted 113 OCLC, Internet, and related classes for 820 students. FEDLINK
staff conducted training sessions in Korea, Japan, Okinawa, Guam, Alaska, and Hawaii. FNO
staff also worked closely with FLICC working group members to sponsor the first FLICC
Acquisition Institute in July 1997, the first FLICC Paraprofessional Institute in August 1997, and
a four-day subject cataloging workshop in September 1997.
By Jessica Clark
This is the first in a series on creating and managing digital documents. Part II will describe digital projects taking place in federal libraries and will address document management software systems.
Federal librarians around the globe are using scanners to generate metadata, provide accessible copies of brittle manuscripts, and create online exhibits of rare photos and documents. Digitization projects breathe new life into collections, but are expensive and can be tricky to manage.
To choose the appropriate scanner for both current and long-range projects, librarians have to weigh the relative importance of cost, scanning speed, and image quality. They also need to make sure that materials to be scanned are matched with the appropriate type of scanner and evaluate the benefits of image scanning vs. optical character recognition (OCR).
Like cameras, scanners use a lens to bounce light rays off of an object and onto a light-sensitive medium. While cameras traditionally use a chemical emulsion to capture light, scanners use a charged coupled device (CCD)--a circuit with a grid of light sensors which divide the scanning surface into pixels. These sensors are larger than photographic grains, so the resolution of scanned images tends to be lower than photographs.
When material is scanned, an analog-to-digital converter assigns a numerical value to the number of electrons in each gridded area. Each pixel is assigned a tonal value (black, white, shades of gray, or color) that is digitally represented in binary code. The binary digits (bits) for each pixel are stored in sequence, and then the computer interprets them to produce an analog representation for display or printing. The resulting images are known as "raster" or "bitmapped" images, and are different from "vector" images, which are generated by a mathematical formula rather than pixel mapping.
The term "bit depth" refers to the number of bits that a scanner includes in each pixel:
The size of a scanned image is related to its bit depth. Eight bits are equal to one byte, the unit generally used to express file size. Image hieght and width are measured in inches. The formula for determining file size is:
image height x image width x bit depth x dpi2 ________________________________________________________ 8 |
The quality of a scanned image depends on:
Different scanners may be used to process various materials.
Flatbed scanners resemble desktop photocopy machines; they allow the user to scan documents as large as 11 x 17 inches, one sheet at a time. Flatbed scanners work well for single-leaf, letter-sized documents and some transparent media. Right-angle, prism, and overhead flatbed scanners have been developed to improve scanning of bound materials. These scanners often include software with image enhancement tools, but allow no control over lighting. They may not be suitable for light or fragile material, as they include a cover which must be placed over the object to be scanned.
Sheetfeed scanners allow the user to load a sheaf of paper to be bulk scanned. Some sheetfeed scanners can also copy both sides of a page. These scanners work well for single-leaf letter-sized and oversized documents, and are significantly faster than flatbed scanners. They dont work well, however, with brittle documents, or documents of different sizes.
Drum scanners require the user to tape the object being scanned to a rotating cylinder. These scanners provide very high resolution images and are generally used in graphic arts studios. Drum scanners are slow, however, not good for brittle items, and may only be used on materials of a certain size. They are also very expensive, but may be worthwhile for imaging projects which include complex, oversized maps.
Film and slide scanners are used to reproduce images from transparencies. They provide high-resolution scans with better dynamic range, but are quite specialized.
Digital cameras allow direct scanning from an original object with no shape limitations. High resolution images may be created through "tiling," or taking close-ups of an image and then matching up the edges. Digital cameras allow control over lighting and don't require materials to be pressed or handled. Affordable, high-quality digital cameras are still in development, however, and operators require more training than those using other types of scanning equipment.
Most library projects require mid-range to high-end scanning equipment, with at least 300 dpi resolution, 30-bit color scanning, a scanning and image editing program, and an OCR program. The scanning program should incorporate "TWAIN" technology (the acronym purportedly stands for "Technology Without An Important Name")--this standard was developed to enable interoperability between imaging programs. Many scanners also require the installation of an Small Computer Systems Interface (SCSI) adapter card.
When choosing a scanner, librarians should also ask themselves:
Recommended scanner manufacturers include: Agfa, Bell and Howell, Fujitsu, Hewlett Packard, Kodak, Microtek, Minolta, Optronics, Ricoh, Scangraphics, UMAX, and Xerox, among others.
Scanning Guidelines | |||
---|---|---|---|
The following scanning guidelines are outlined in Digital Imaging for Libraries and Archives, written by Anne R. Kenney and Stephen Chapman for a workshop conducted by the Cornell University Library: | |||
Medium | Type of Scanning | Resolution | |
Published text/line art | bitonal scanning | 600 dpi | |
Illustrated text | bitonal scanning | 600 dpi with image enhancements | |
Color illustrations | color scanning | 24-bit | |
color oversize maps | color scanning | 24-bit, 200 dpi | |
Black and white photos | grayscale scanning | 8-bit, resolution depends on quality of original. | |
Color photos | color scanning | 24 bit | |
Halftones | bitonal scanning | 600 dpi | |
Color halftones | color scanning | 24-bit | |
Typewritten/laser printed | bitonal scanning | 300 dpi minimum | |
Ball point pen text | bitonal scanning | 300 dpi minimum | |
Pencil/quill/felt tip text | grayscale scanning | 300 dpi minimum | |
Damaged/stained documents | grayscale or color scanning | 300 dpi | |
Papyri | color scanning | 24-bit, 600 dpi |
OCR software translates scanned images into ASCII text. For accurate OCR translation, the resolution of the image must be crisp--300 dpi at the minimum, and 600 dpi for scanning small, italic, or old-fashioned typefaces. The bit depth should be 24-bit or 30-bit. OCR software choice is important; some programs may mistranslate, duplicate, or skip characters. The software should recognize different fonts and common symbols. Recommended OCR programs include OmniPage by the Caere Corporation, Textbridge by Xerox Corporation, and TypeReader by Expervision; all are available in both Limited Edition (LE) or professional versions. TypeReader is particularly recommended, as it claims to be able to read more than 2100 fonts.
The Adobe Acrobat Capture program lets you convert paper documents into electronic files in the Portable Document Format (PDF). These files are fully searchable by keywords, and the document creator can add annotations, cross-document links, and bookmarks. Documents created in this software still require proofreading (if the program doesn't recognize a character, it creates a graphic image of it), and the software is expensive.
Scanning Resources | |
---|---|
Links
Jeff Bone's Scanning FAQ: |
Books
Ashford, Janet and John Odam. 1996. Start With a Scan. Berkeley: Peachpit Press. Cornell University, Ithaca, NY 114853, 607-255-9841, [email protected]. Simone, Luisa. 1996. The Windows 95 Scanning Book. New York: John Wiley & Sons. |
The semi-annual FEDLINK OCLC Users Group meeting, October 29, 1997, highlighted new products and services and brought members up to speed on the technical and financial aspects of the network.
FEDLINK staff reminded members that their access to OCLC will be blocked unless signed IAGs are returned to FEDLINK by November 28. To assist customers for planning in the new fiscal year, staff highlighted a few other important service deadlines. (See chart below.)
Important End-of-Fiscal-Year Deadlines for OCLC Services | |
---|---|
June 15 | Last day to cancel asynchronous dial access authorization to avoid paying the annual fee of $220 charged each July. |
August 1 | Deadline to notify FNO OCLC team to cancel OCLC services for next fiscal year. |
August 30 | Last day to order Local Database Creation. (The date is subject to change, depending on OCLCs workload; verify that the order is still viable before submitting final paperwork.) |
Last week of fiscal year | Orders for products/services may still be laced by fax, but verify receipt and ability to process by calling FNO. |
Remember: | To request the removal or move of dedicated lines requires 25-45 working days. Allow time prior to the end of the fiscal year if billing needs to be stopped before a new fiscal year begins. Credits for dedicated lines are not available. |
OCLC has implemented Phase Two of PromptCat, supporting the provision of shelf-ready books by some participating book jobbers. Currently eight companies participate in PromptCat: Academic Book Center, Ambassador Book Service, Baker & Taylor, Blackwell's Book Services, DA Information Services, Majors Scientific Books, Rittenhouse Books, and Yankee Book Peddler. Two federal libraries are among the 36 libraries currently using PromptCat. Contact vendors to learn whether they provide shelf-ready books under PromptCat, and what their pricing for the materials and service is.
The first OCLC Institute will be "Knowledge Access Management: Tools & Concepts for Next-Generation Catalogers." The institute can be billed to your OCLC account: $495.00 OCLC Members ($550.00 non-members). Open dates as of 10/27/97 include February 16-18, March 6-8, and March 16-18. For more information, see the institute home page: http://www.oclc.org/institute.
OCLC continues to test the Local Data Record Batch Updating Service. It will allow a library's local system to send data on serials holdings to OCLC's Union List system to automatically create, delete, and modify union list holdings. The pilot is scheduled to conclude by the end of 1997. OCLC expects to introduce the product no later than Summer 1998.
The OCLC home page has specifications for "Batch Updating of Local Data Records." Visit http://www.oclc.org/oclc/specs/batch.htm for details. Records must conform to ANSI/NISO standards, and be MARC coded (MARC Format for Holdings Data). The specifications cited list all of the MARC elements.
Changes to the Interlibrary Loan system are outlined in OCLC's Technical Bulletin 222, Interlibrary Loan Enhancements (May 1997). These enhancements support the use of Direct ILL, which lets selected users send loan requests directly to selected libraries. ILL Management Statistics are described in Technical Bulletin 223 (May 1997). The statistics cost $420 a year, and are ordered via the OCLC Name Address Directory. Like other OCLC statistical reports, ILL Management Statistics are delivered via Product Services on the Web. The URL for the Product Services Menu is http://psw.oclc.org/psw. Call Georgette Harris at FEDLINK if you have questions about the ILL Enhancements or the ILL Management Statistics.
FEDLINK recommends Internet access for OCLC's cataloging and ILL as an economical access method for libraries that use the system less than about 70 hours per month, or about 3 hours a day. If Internet access is not available, consider using dial TCP/IP. TCP/IP uses Windows-based software and a high speed modem to provide faster access than asynchronous dial access. Also known as "X.25" from its ANSI designation as a standard telecommunications protocol, it has been available since the 1970s. In addition to faster, smoother access, OCLC offers a financial incentive for new TCP/IP access users: there is no $220 annual fee (billed in July for each asynchronous dial access account) plus the rate is one-half cent per minute cheaper, (i.e., $0.115/minute vs $0.12/minute. Internet access is $0.60/minute). FEDLINK can help you order a copy of the TCP/IP Dial Access Systems Administrators Guide. Technical Bulletin 224 (June 1997) also has specifics about setting up dial TCP/IP.
For those libraries with three or more dedicated line ports, also known as synchronous dedicated lines or as "multidrop" lines, OCLC is introducing a dedicated TCP/IP method. Currently undergoing tests in severl libraries, OCLC will release detailed specifications and order forms soon, so that installations can begin early in 1998. Some preliminary specifications are already available from FEDLINK.
For a consultation on your telecommunications options, please contact FEDLINK for assistance in coordinating access. Careful coordination is necessary to integrate various operating systems, hardware, and changes in OCLC software.
OCLC will not support Passport for DOS after December 31, 1997, requiring the user to install Passport for Windows. This $40 program runs under Windows 3.1, Windows 95, and Windows NT and works with all available OCLC telecommunications methods, making Internet access to OCLC systems very easy. To order a copy, request a form from FEDLINK or visit the OCLC home page at www.oclc.org (click on the Support and Documentation button, then Forms, and the forms are listed in alphabetical order). Orders via the home page will appear on your FEDLINK account.
Warning: Windows NT is not compatible with the multi-drop dedicated lines. This will not change. If installing Windows NT, you will need to find an alternative access method.
Cataloging MicroEnhancer Plus, the DOS version, is no longer available for purchase. The Windows version of CatME will be available by the end of 1997. OCLC is already accepting orders; please call FEDLINK for prices. Because Microsoft now only supports development for 32-bit processor applications, the software will not run under Windows 3.1 nor operate with multi-drop dedicated lines. For this reason, OCLC will continue to support the DOS version of the CatME.
The OCLC Cataloging Label Program is available free from OCLC's home page at http://www.purl.oclc.org/oclc/label. A new version will be released in early 1998 to correct some minor problems in the first version. It can be downloaded, and will be distributed with future versions of OCLC software in the OCLC Access Suite. It assists with printing labels both immediately and in batch modes, works with laser printers, and expedites editing and the formatting of print constants and ranges of copies or volumes, etc.
Immediately after the Users Meeting, OCLC announced that its software will now be available on a CD-ROM called the OCLC Access Suite. It will contain Passport for Windows, ILL and Cataloging MicroEnhancers for Windows, CJK software, and the Cataloging Label Program. Those who order the software packaged in this suite will pay an annual licensing fee instead of a one-time charge, but then will receive updates automatically without additional charges. The annual charges will be $129 for single workstations, or $1,290 for a site license.
Enhancements to the functionality of FirstSearch were outlined in the September issue of FEDLINK Technical Notes. FirstSearch Usage Statistics are now available via the OCLC home page at http://www.stats.oclc.org/, using a FirstSearch authorization number and administrative password. Statistics can be downloaded as a comma-delineated file for local manipulation. In addition, OCLC has loaded new databases--Union List of Periodicals and SIRS Researcher--and added information to other databases--Books-In-Print, publisher information added; Medline backfile (1965-1984) loaded; and brief and full files added to Disclosure and Worldscope. GeoRef access has changed so that per-search authorizations show only a brief record; subscriptions, which show the full record, are offered at discount through the end of 1997. BasicBiosis has also been redesigned to improve indexing and journal coverage.
The Library of Congress office of Contracts andLogistics (C&L) plays a crucial part in developing, evaluating, and negotiating FEDLINK members' contracts; seven C&L contracting officers are assigned to assist the FEDLINK program. In October, long-time FEDLINK Section Head Shirley Courtney left C&L for a job in the private sector. Contracting Officer Joan McCoy was subsequently appointed Acting FEDLINK Section Head. In this position, McCoy will oversee the development of FEDLINK procurement plans, preparation of RFPs, evaluation of proposals, negotiation of FEDLINK contracts, and preparation of awards. McCoy will share the Section Head position in four-month rotations with co-worker James Chajkowski until it is permanently filled.
McCoy began at C&L as a contract administrator in February of 1989. After a brief training period, she was assigned to work with Courtney on the FEDLINK program. During her tenure, she has completed a number of job-related courses at the US Department of Agriculture Graduate School and the Department of Defense, and is one course away from completing the General Service Administration's Senior-Level Contracting Officer's Warrant Program.
Before coming to the Library of Congress, McCoy earned a BS Degree in Hotel and Restaurant Management at the University of Maryland. While she was a student, she worked at the Department of the Interior, US Geological Survey computer center. After completing her BS, McCoy worked with the Marriot and Hyatt hotels, but decided she preferred working for the federal government, and applied for the C&L position.
Notify FEDLINK of Contact Information Changes When a name, address, phone or fax number changes, please be sure to notify FEDLINK. Changes that affect information about the IAG Point of Contact should be sent to the FEDLINK Fiscal Hotline: 202-707-4900 or FAX 202-707-4999. Remember to change information in the OCLC Name Address Directory record for your library: log on to either cataloging or ILL, search by entering a colon and then the OCLC symbol (e.g.:TPY); make edits and send each field and replace the record (rep). Changes to area codes and exchanges are especially troublesome. Please remember to notify FEDLINK of them, because in critical situations we count on contacting you by fax or phone. Also, if you are sending in a registration or IAG modification request and information has changed since last year, please call attention to the changes in a cover memo. FEDLINK Network Operations' OCLC team can help you make changes to OCLC's mailing lists and Name Address Directory. For other companies, send informatio concerning changes directly to the vendor. |
Save the Date! Plan to attend the 15th annual FLICC Forum on Federal Information Policies Adapting to Reinvention: Getting Results in Government Printing March 19, 1998 This year's Forum will take a closer look at how federal libraries and information centers respond to reinvention initiatives and the key issues of providing government information in this fast-paced environment. Watch your mail for details! |
FEDLINK Technical Notes is published by the Federal Library and Information Center Committee.
Send suggestions of areas for FLICC attention or for inclusion in FEDLINK Technical Notes to:
FEDLINK Technical Notes
Federal Library and Information Center Committee
Library of Congress, 101 Independence Avenue SE, Washington, DC 20540-4935
FLICC/FEDLINK:
FEDLINK Fiscal Operations:
Phone (202) 707-4800 Fax (202) 707-4818
Email: [email protected] Web Site: http://lcweb.loc.gov/flicc
Phone (202) 707-4900 Fax (202) 707-4999
Executive Director: Susan Tarr Editor-In-Chief: Robin Hatziyannis
Writer/Editor: Jessica Clark Editorial Assistant: Mitchell Harrison
FLICC was established in 1965 (as the Federal Library Committee) by the Library of Congress and the Bureau of the Budget for the purpose of concentrating the intellectual resources of the federal library and related information community. FLICC's goals are: To achieve better utilization of library and information center resources and facilities; to provide more effective planning development, and operation of federal libraries and information centers; to promote an optimum exchange of experience, skill, and resources; to promote more effective service to the nation at large; and to foster relevant educational opportunities.
Go to: Library of Congress Home Page