The Professional Journal of the Special Interest Group on CD-ROM Applications & Technology Volume 8, Issue 5 July/August 1994 Copyright 1994 SIGCAT P.O. Box 3706 Reston, VA 22090 President: E. J. (Jerry) McFaul U.S. Geological Survey (703) 648-7126 (703) 648-6536 (Fax) Internet: JMCFAUL@ISDRES.ISD.USGS.GOV Executive Vice President: Duane Marquis Department of Commerce (202) 482-3055 (202) 482-4595 (Fax) Vice President: Mike Rubinfeld National Institute of Standards and Technology (301) 975-3064 (301) 590-0932 (Fax) Internet: MIKER@MML.NCSL.NIST.GOV Vice President: Reenie Prettyman National Institutes of Health (301) 496-1936 (301) 480-6183 (Fax) Treasurer: Carol Cini U.S. Government Printing Office (202) 512-1283 (202) 512-1255 (Fax) Secretary: Joan McKean National Oceanic and Atmospheric Administration (301) 713-1170 (301) 713-1179 (Fax) Editor: Yvonne Kidd Kidd and Company (703) 385-1323 (703) 385-1493 (Fax) Technical Editor: Kathie Fraser The Special Interest Group on CD-ROM Applications & Technology (SIGCAT) is a network of people who share a vision that CD-ROM will profoundly change the way in which they deal with information. Begun in 1986 and sponsored by the U.S. Geological Survey, SIGCAT provides an open forum for people from all walks of life to investigate CD-ROM technology and its myriad applications. SIGCAT now comprises a wide spectrum of individuals from government, industry, academia, and the interested public. With its global network of over 6,700 members representing more than 300 organizations, SIGCAT has grown into the world's largest CD-ROM user group. SIGCAT members benefit from the ongoing exchange of ideas, information, and experiences through SIGCAT meetings and the bimonthly SIGCAT DISCourse newsletter. Members receive substantial discounts on CD-ROM drives, media, software, and publications, plus free passes to technical expositions. Members can also participate in any of 15 working groups that address specific special interests. Membership in SIGCAT is free. Subscription rates for the SIGCAT DISCourse newsletter are $40 per year. SIGCAT INFORMATION LINE: (703) 648-4452 SIGCAT BULLETIN BOARD: (703) 648-4168 The SIGCAT Information Line provides an update on the latest SIGCAT events. You can also obtain SIGCAT information by dialing into the SIGCAT Bulletin Board. To register as a new SIGCAT member, simply call SIGCAT Central, (703) 435-5200, and leave your full name (please spell out), title, organization, address, and phone and fax numbers. To receive the SIGCAT DISCourse newsletter, please register as a SIGCAT subscribing member by filling out the subscription form on page 15 of this newsletter and sending it with your $40 check payable to SIGCAT Foundation, P.O. Box 3706, Reston, VA 22090.
The National Archives and Records Administration (NARA) took a step forward in recognizing CD-ROM as an archival medium with the announcement of a new CD-ROM policy on July 28, 1994. NARA Bulletin 94_4, "Use of Compact Disc-Read Only Memory (CD-ROM) Medium to Transfer Records to the National Archives," states that NARA will begin accepting transfers of permanent records stored on CD-ROM. Previously, NARA accepted electronic information only on magnetic tapes, 3480-class tape cartridges, certain analog videodiscs (which typically contain photographs), and audio compact discs. The new policy permits Federal agencies to transfer permanent records to NARA on CD-ROM if the information consists of fielded data or text files and if the CD-ROM meets the following requirements as specified in the bulletin: - Conforms to the International Standards Organization (ISO) 9660 standard. - Complies with the American Standard Code for Information Interchange (ASCII) standard as defined in the Federal Information Processing Standard 1_2 (11/14/84). - Is not dependent on control characters or codes that are not defined in the ASCII character set. - Is not compressed unless the software to decompress the files is provided. - Is individually addressable (meaning that the file can be easily located on the CD to facilitate subsequent transference of archivable content to 3480 magnetic tape). - Complies with the documentation requirements of 36 CFR 228.188. In cases where permanently valuable electronic records exist on both CD-ROM and other media, the agency and NARA will mutually agree on the most appropriate medium for transfer of the records. The texts of NARA Bulletin 94_4 and a companion Bulletin 94_5 on use of optical disks to store Federal records are available through Internet on the NARA gopher at GOPHER.NARA.GOV. Printed copies of the bulletins will be distributed in August to agency records managers. SIGCAT recommends that vendors and agency personnel download the complete bulletin and review it to determine whether the policy is in any way limiting or if there are problem areas that should be addressed. A representative from NARA's Center for Electronic Records will join SIGCAT at the October 18_19 regional meeting to discuss the policy in further detail and address questions. (See page 1 for meeting details.) For further information on the CD-ROM policy, contact the National Archives and Records Administration, Center for Electronic Records (NSX), 8601 Adelphi Road, College Park, MD 20740-6001, (301) 713-6630.
October 18_19, 1994 NOAA Conference Center 1305 East West Highway(Bldg. No. 4) Silver Spring, Maryland The next SIGCAT regional meeting will be a special 2-day event taking place on Tuesday and Wednesday, October 18_19, 1994, at the National Oceanic and Atmospheric Administration (NOAA) facility in Silver Spring, Md. The theme for this meeting will be "The State of CD-ROM Authoring Software" and, as usual, will be open to the general public with no admission charge. The technology of CD-ROM provides a storage medium that is ideally suited to electronic publishing. It's inexpensive to produce, it's durable, it's compact (many magazines are now including discs with each issue), and it's standardized throughout the world. But simply putting 680 megabytes of data on a small, round piece of plastic is of little use without some sort of access methodology. The means by which those millions of bytes of data are transformed into a page from an encyclopedia or a lifelike rendition of a Monet painting or a movie clip of Apollo 11 landing on the Moon is implemented through software. Generally referred to as "authoring software," this component of a CD-ROM publication is quite often the most challenging part of a CD-ROM project. Selecting the appropriate authoring package typically involves an in-depth analysis of the myriad variables that underlie the requirements of the CD-ROM product. These variables must be weighed against the capabilities of the authoring packages available in the marketplace. What is the best package for my fielded database? How effectively can I link my images to my text? Which product will allow me to run my disc across three different computing platforms? Which package has no runtime royalties so I can keep the price of my disc down? How much can I customize the user interface for my particular product? These and dozens of other questions and issues face today's CD-ROM producer. Recognizing the importance of CD-ROM authoring software, SIGCAT has organized a special 2-day regional meeting dedicated entirely to this topic. We have assembled a comprehensive cross section of the industry's leading producers of authoring software to provide a series of in-depth presentations on a dozen state-of-the-art products. The presentations will focus on the software packages included on the latest SIGCAT Software Showcase disc, which was released at the SIGCAT annual conference in April of this year. These "Showcase Presentations" will give the audience a look at vendors' behind-the-scenes efforts as they prepared their applications for this year's Showcase disc. Each presenter will explain and demonstrate the detailed steps involved in building the Showcase application with his or her authoring package. You'll see exactly how each product handles things such as importing text, linking images, accessing audio and video, building indexes, customizing user screens, and many other functions critical to building an application. We are structuring the Showcase Presentations to be highly interactive and will encourage a lively dialogue between the audience and the presenters. The consistent thread through all of the presentations will be the "benchmark" applications contained on the Showcase disc. These 12 applications are all based on the same set of input data consisting of text, images, audio, and video information. Thus, attendees should be able to gather comparative information in a consistent, comprehensive manner. For once, you should be able to truly compare apples and apples. In addition to the Showcase Presentations, there will be an exhibitors' area featuring several dozen companies offering CD-ROM authoring tools and related products and services. These "tabletop" exhibits will allow attendees to receive more detailed product information and discuss their software requirements on a one-on-one basis. A unique feature of this t2-day meeting will be the follow-on SIGCAT Software Workshops. Approximately 6 weeks after the 2-day event, a series of half-day workshops will be scheduled where individuals can bring samples of their own data and build test applications under the guidance of experts from the various software companies. These workshops will take place in SIGCAT's Training Center in Reston, Virginia, and provide a hands-on environment for up to 20 students at a time. A dozen 486-class, CD-ROM-equipped machines are networked to a high-capacity server, allowing for each pair of students to share a machine. Two classes will be scheduled each day, with ample breaks to encourage additional interaction. The particular vendors scheduled for the Software Workshops will, in large part, be determined based on the requests of the attendees at the 2-day October SIGCAT meeting. Request forms will be available for attendees to indicate which vendors they would like to continue evaluating in a workshop environment. These requests will be compiled into a formal schedule of subsequent workshops and made available to all interested parties. There will be a modest fee associated with these workshops to defray expenses. We feel that the 2-day "The State of CD-ROM Authoring Software" meeting combined with the follow-on SIGCAT Software Workshops will provide a structured, comprehensive method of understanding, testing, and evaluating the myriad software options available in today's marketplace. Selecting an authoring package is not an easy task and is arguably the single most important step in a CD-ROM project. After all, you and your users will have to live with your choice for quite a while, so it had better be the right choice. SIGCAT hopes we can help you in this effort. Directions: The NOAA Conference Center is located in Building No. 4 at 1305 East West Highway, Silver Spring, Md. Take the metro to the Silver Spring stop (the Red Line), or, if driving, take the Georgia Avenue South exit from the Beltway. Proceed to Colesville Road and turn right, go under the viaduct, and then turn left onto East West Highway. Public parking is available next door in Building No. 3.
Doculabs, in association with the University of Illinois,
Chicago (UIC), recently conducted a "technology assessment" of
various CD-ROM software products for the Association of
Information and Image Management (AIIM). (AIIM is a trade
association and professional society representing the "imaging"
or document management industry and profession. AIIM's members
are primarily microfilm, scanning, and optical storage companies
and users.)
For the AIIM community, CD-R represents a cost effective
alternative to microfilm and write-once read-many (WORM) disks.
Surprisingly, the document management industry and the CD-ROM
industry (audio, catalogs, multimedia) know very little about
each other. When document management vendors are asked if they
support CD-ROM they typically answer, "Sure, CD-ROM is just
another storage media." While this may be true, there are many
specific issues that need to be addressed when considering CD-R.
Graduate students from the university tested seven authoring
tools and seven premastering tools to determine the strengths and
weaknesses of each software package. Although other studies have
compared CD-ROM titles, this test focused on the process of
authoring and premastering CD-Recordable discs (in the DOS and
Windows environments only). For each product, the students went
through the entire process of authoring and recording discs,
using the same set of source data. The students were timed to see
how long the process took with each tool. The students were also
asked to rate each product in a number of different
categories.
The Data
The test data consisted of course packets produced by UIC's
College of Business Administration, and provided by UIC's Office
of Publication Services. We specifically selected data that would
be most representative of the types of information most business
users would want to "burn" onto CD-R: text, images, and
structured data, as well as graphics, color photo CD images,
sound, video, presentations, and spreadsheets. The data, which
totaled about 150 MB, included:
- Syllabus, scanned and OCR'd.
- News Articles: Wall Street, Business Week, scanned.
- Harvard Business Review, case study, scanned.
- Electronic spreadsheets.
- Book Chapter, scanned.
- Lecture notes, electronic and scanned.
- Annual report, color PhotoCD images, Baxter International.
- Audio Clip, 30 s, lecture taping.
- Internet download: research notes, Hitchhiker's Guide to the
INTERNET.
- Video clip, 5 sec., lecture taping.
- Powerpoint presentation, executive visit.
- Structured data for input into SAS or SPSS (statistical).
- Annual performance data for input into Lindo (Linear
Programming).
- AutoCAD drawing, electronic.
- SGUL tagged document, electronic.
- An 83-page course pack subset called "Xerox 900" (scanned
output from a Xerox Docutech).
A majority of collegiate course packets in the United States
are produced on high-speed electronic copiers (such as the Xerox
Docutech). One intent of using this sample data was to develop
alternatives to the excessive paper reproduction costs in the
university publishing environment.
The Testers
Three graduate students with different experience levels
performed the testing. The table below shows the specific
experience level of each of the testers.
Tester 1 Tester 2 Tester 3
Experience level (Low) (Medium) (High)
PC literate Yes Yes Yes
Windows experience Yes Yes Yes
Application software experience No Yes Yes
CD software experience No No Yes
Methodology
This evaluation measured four variables for each product: time,
data manipulation capabilities, usability, and price. The test
data, the individuals performing the evaluation, and the hardware
were held constant. A total of 1743 observations were taken for
the authoring products and 1365 for premastering. Time was
measured in minutes. We used a mean score of 10 = excellent and 1
= poor, averaging the results of the three different skill levels
of our testers when no significant variance among the testers
resulted. Cost is in U.S. dollars.
Although the reader's experience level or data may be different
from those of the study, we feel this evaluation (sample) is
representative of a typical business environment (population)
because the range of observations was so broad. For example, if
one product's authoring time was twice that of another product's
across all three experience levels, the reader could expect
similar results. If all three of the testers experienced
difficulty with a certain data type _ say images _ we can assume
that the reader could expect the same difficulty. In cases where
we did not find a direct relationship (linearity) _ for example,
between authoring times and experience levels _ the results of
the test should be considered inconclusive and may not apply to
the reader's particular application.
Final test results of the study are currently being compiled
and will be presented in future DISCourse issues. Meanwhile, some
general findings include the following:
Authoring
Performance has been one of the biggest issues with CD-ROM and
CD-R. All of the authoring packages convert various file types
into a proprietary format for burning onto CD-R discs to optimize
retrieval performance. Yet, because files no longer remain in
their native format, you must use filters to import and export.
Incompatibilities may occur here (vendors will have trouble
keeping up with new releases, new products, etc.). Today, with
high speed hardware (Pentium, 4x readers), performance is
becoming less of an issue, and CD-ROM will soon perform the way
magnetic tape did only a few years ago. Thus, authoring tools
might go the way of other products in the multimedia world, where
authoring tools act as "media glue." That is, the authoring tool
only keeps track of itself, its own data, and external references
to picture, sound, and video files. None of the external files
get compiled into the actual program code or executable file,
they are referenced only inside the program.
Premastering
Of the variables we studied, the premastering business places
the most emphasis on speed. Fast conversion rates often demand
enormous memory resources. This requirement becomes critical in a
CD-R "assembly line" or in high-production applications. Data are
no longer static, sitting on a premastering workstation waiting
to be processed, but more than likely are coming from a host or
network and require some postprocessing as well (such as parsing
or verification, etc.). In this case, memory allocation is
critical to ensure that other tasks can be performed concurrent
with the premastering.
From an end user's perspective, premastering may eventually be
incorporated into authoring (or other application) software.
Seamless integration will be an essential factor contributing to
the growth of the CD-R market. For application developers,
API-level communications will be critical to high-production
systems (those producing over 1000 CD-R's a month for example).
In addition, look to the premastering software companies to offer
additional functionality that will address high-production
environments such as media tracking, robotics control (such as an
autoloader), and quality assurance.
Ten Things Document Imaging Vendors Should Know About CD-R
Technology
Vendors and service providers with expertise in document
management are not necessarily knowledgeable, experienced, or
qualified to assist with CD-R projects. Before selecting a
company to assist with CD-R authoring and(or) premastering
requirements, users should check to see if the prospective vendor
or service provider is knowledgeable in the following areas:
1. Management of images on CD-Recordable media is not the
same as that on traditional optical media. Because access speed
and transfer rates are slower than those of traditional optical
storage, file types and sizes must be modified for increased
performance. Depending on the retrieval engine, using compressed
PCX formats instead of CCITT Group 4 TIFF formats may improve
display speeds.
2. Text-based systems and image based systems are not the
same. Image-intensive applications require extremely fast image
display and relatively simple search criteria. Text-based systems
require sophisticated search and retrieval capabilities, yet
relatively simple display capabilities. The marriage of these two
many times results in compromised performance on CD-Recordable
media. Vendors should be knowledgeable about these trade-offs.
3. Database applications are excellent with a central pool of
information (such as a jukebox of images) yet perform poorly with
distributed volumes of information (such as the multiple
CD-Recordable discs distributed throughout an organization). Few
databases are accustomed to being managed in a distributed mode.
4. Not all CD-ROM readers are alike. Compatibility between
CD-Recordable discs and CD-ROM readers is a relatively new field
and not well understood by most in the industry. Deciding which
type of reader to purchase depends on your application. For
example, bridge disc or multisession CD-Recordable discs do not
run well on NEC 3X readers, and PhotoCD's only run in CD-ROM
XA-compatible readers.
5. An additional step is required when writing files (images)
to CD-Recordable discs vs. optical disks (jukebox or standalone).
A vendor's or service provider's experience with premastering
software is key. In some cases, file drop-out can occur when a
large number of small files are transferred; buffer under-runs
can occur when large files are transferred too slowly.
6. Users can write to CD-Recordable discs multiple times.
Managing multiple volumes is critical to the efficient use of
CD-Recordable media. "CD bridge" technology gives the user the
ability to let each volume or session "get acquainted" with the
next, thus increasing the utility of the disc, since files
(images) can subsequently be added.
7. File-naming conventions need to be modified for
CD-Recordable discs. Excessive use of subdirectories slows
performance, and makes many traditional database products
inefficient for use with CD-ROM readers. 3X, 4X, and soon 6X
readers should reduce this problem.
8. Data transfer to CD-Recordable media can be performed in
two ways. The first, using an image file, causes the root
directory to drop out. The second, using a direct transfer, is
extremely sensitive to transfer speed, hardware configurations,
and communication drivers (ASPI, HASPI).
9. Security of data on CD-Recordable is different from that
on traditional optical storage. Because data are written in an
industry standard ISO 9660 format, additional steps need to be
taken to limit unwanted access to the CD-Recordable discs.
10. A disc refers to a compact disc or CD. A disk refers to an
optical or magnetic diskette. If your vendor or service provider
is spelling disc incorrectly, they probably do not know the
difference.
Doculabs can be reached at (312) 433-7793, fax (312) 433-7795,
Internet U27555@UICVM.UIC.EDU,
attention James Watson.
The General Services Administration recently awarded SIGCAT President E. J. (Jerry) McFaul one of its Federal IRM "Best Practices for the Future" awards, in recognition of his achievements in advancing access to Government information through CD-ROM technology. McFaul was honored during an August 2, 1994, awards ceremony for his leadership role in founding SIGCAT and for his vision of how CD-ROM could be used to improve access to the Government's vast data holdings. SIGCAT now has more than 7,000 members in 50 countries and is headed by a board composed of Federal executives who believe that CD-ROM technology is a key component in the Government's evolving information infrastructure. GSA awarded eight Best Practices awards. Other recipients included the Air Force Standard System Center, for managing standard systems as a corporate asset; Benita A. Cooper, NASA Associate Administrator for Management Systems and Facilities, for her achievements in involving senior management effectively in IRM; the Department of Veterans Affairs, Information Resources Management Department, for re-engineering across organizational lines; the Department of Army's Sustaining Base Information Services, for delivering measurable outcomes to customers; NASA's Automated Information Management Office for using corporate management to leverage legacy systems; the Internal Revenue Service's IRM Department and Agency for International Development's IRM Department, for leveraging systems development with business process improvement; and the Tennessee Valley Authority, Information Services, for profitable downsizing.
CD Error Characterization: Differences Between CD-ROM and Writable CD
Read Only Compact Discs (CD's), such as the CD-ROM and Digital Audio CD, are made by concurrently forming a plastic disc and a spiral track of pits representing data carried on the surface of the disc via an injection molding operation; subsequently, a reflective metal layer is applied in a vacuum coating step, and a protective lacquer layer is applied via spin or spray coating. The original data that are replicated on these "pressed" CD's are carefully written on a master disc (used to form the injection molding tool) using a precision master laser writer housed in a very clean environment. On the other hand, writable CD's, such as CD-R and PhotoCD, are made by first using injection molding to form a plastic disc that has a continuous spiral (tracking) groove on one surface and then sequentially coating that grooved surface with a vacuum-deposited dielectric layer, a spin-coated organic "pit-forming" layer, a vacuum-coated metal reflector layer. and a spin-coated organic protective layer. Although the spiral tracking groove that is replicated on the surface of the writable CD's molded plastic substrate is initially written on a master disc using essentially the same precision laser writer that was used to create the pressed CDs' master discs, the marks that correspond to the user digital data carried by the writable CD are written one at a time via an optical CD disc drive that is housed in an ordinary environment. Thus, read-only CD's and writable CD's are physically different. The means used to form the marks that ultimately carry digital information stored on these two types of CD medium are completely different as well. It is quite possible, therefore, that the physical nature and statistical characteristics of the error events that contaminate data carried by pressed and writable CD's will be distinct. In particular, short (< 4 or 5 byte) error bursts may be more prevalent (in both an absolute and relative sense) on writable CD's than on pressed CD's, especially in the case of writable CD's that are written after substantial user handling and(or) aging. The CIRC Error Control System The basic code used in the CD optical recording system to control errors is a short (codewords are ~ 30 bytes in length), relatively weak (no more than two unlocated errors in any codeword can be corrected) Reed Solomon error-correction code. In order to handle long burst errors, two such codes that have codewords of 32 bytes (the C1, or outer code) and 28 bytes (the C2, or inner code), are individually interleaved and orthogonally catenated. Owing to the interleaving, the data bytes that belong to a given C1 or C2 codeword are not sequentially recorded along the disc track, so that sequentially written/read data bytes belong to different codewords. Orthogonal catenation causes each of the 28 bytes contained in a given codeword of the C2 code to also belong to one of 28 different codewords of the C1 code; during playback, the C1 codewords are reconstructed from the serial data retrieved from the disc and decoded prior to forming and decoding the C2 codewords. This interleaving and catenation of the C1 and C2 codewords is referred to as cross-interleaved Reed Solomon coding (CIRC). A basic strength of the CIRC error-control system is that it effectively randomizes burst errors; i.e, due to the interleaving that is done, burst errors _ even those with lengths up to 100 bytes _ will cause at most only one erroneous byte to occur in any given C2 codeword. In addition, because the CIRC system causes each C2 codeword byte to also be protected by (i.e., encoded by) a C1 codeword, its error-correction power is significantly enhanced (relative to that of the native base C1 and C2 codes). Specifically, the CIRC system can correct all possible error patterns that consist of fewer than 13 unflagged erroneous bytes that may occur in the 28x32 = 896-byte block of data that is concurrently spanned by 28 (32-byte long) C1 and 28 (28-byte long) C2 codewords. (Note: Circuitry external to the CIRC decoder may identify bytes that are suspected to be erroneous and mark them by attaching "flag" bits to them; these flags will subsequently be used by the C1 and C2 decoders to determine the location of these unreliable bytes within a received codeword. Such flagged bytes are known as "erasures.") Many error patterns that consist of more than 12 unflagged errors in an 896-byte CIRC code block can be corrected by the CIRC system, but correction of such patterns is not guaranteed. For example, patterns that consist of tens of erroneous unflagged bytes can be corrected only if all the errors are confined to just a few C1 codewords. However, if errors are truly randomly distributed, only patterns that contain fewer than 13 errors per CIRC code block are guaranteed to be correctable. Finally, the error-control capability of the CIRC system can be considerably extended if a reliable external means of creating erasures is employed, because a decoder can correct twice as many erasures as unlocated errors (i.e., up to 24 erasures per 896-byte code block can be corrected provided no unlocated errors coexist in the block). Achievement of the full error-correction capability of the CIRC system requires the use of enhanced decoding; i.e., the C1 and C2 decoders must cooperate during the error-correction process and have the ability to process erasures. They should also be able to dynamically reconfigure the correction program they will carry out. For example, the C1 decoder may flag errors that it was unable to correct for subsequent use by the C2 decoder, and both decoders may determine the number of errors and erasures they will attempt to correct on the basis of the number of flagged errors contained in an input codeword. Early CIRC system implementations did not employ enhanced decoding because of the (then) high cost of the VLSI required to support the necessary processing (recall that the CD system was originally designed as a consumer audio product). Today, the VLSI needed to support enhanced decoding at a sustained data throughput rate of ~ 200-Kbyte/s is available at low cost. Thus, modern enhanced CIRC decoders will usually switch to less powerful correction programs when operated at speeds >_ 300-Kbyte/s; i.e., error-correction capability is decreased as CD system data transfer rate is increased. Although the exact values are dependent on the design of specific decoders, a rough estimate of the reliability increase possible with enhanced decoding can be obtained by assuming that a simple decoder (one that can process externally marked erasures) will correct up to 7 errors per 896-byte CIRC code block as opposed to the 12 that can be corrected by a fully enhanced decoder. Assuming randomly distributed (spatially on the disc) byte errors, the probability that at least one C2 codeword per CIRC cannot be corrected by the simple decoder is 1019PB8, whereas for the enhanced decoder it is 3.551028PB13, where PB is the probability that any byte is erroneous. If PB = 5510_4 (which corresponds to a "Red Book" block error rate, or BLER, of about 118 error-contaminated C1 blocks per second at the 15 CD system data transfer rate of 176.4 Kbyte/s), the probability that the simple decoder will then fail to correct at least one C2 word per CIRC block is 3.9510_8; the equivalent probability for the enhanced decoder is 4.3510_15. When PB = 10_3 (BLER ~~ 235/s), these probabilities degrade to 10_5 and 3.5510_11, respectively. The respective probabilities that a CD-ROM sector will contain an uncorrectable C2 word are ~45 these values (because there are about four CIRC code blocks in a CD-ROM sector). Pressed CD vs. Writable CD Error Characteristics Newly made, good-quality, pressed CD's contain only a small number of localized error bursts greater than 5 bytes in length and virtually no bursts that are longer than 100 bytes. The background random errors (which are, for the most part, uniformly distributed spatially on the disc) consist of short bursts (less than 6 bytes in length) that occur with a probability of <_ 10_4 (BLER ~~ 24/sec). Such random errors and bursts are easily handled by the CIRC coding system. Writable CD discs, on the other hand, may exhibit a higher random short error event rate, especially if the user's CD writer is not optimally configured for use with a particular writable CD medium or if it is not well maintained. Moreover, discs written after user handling may exhibit high short error burst levels in the vicinity of defects that were caused by user handling. If such a correlation exists between long bursts due to handling and shorter bursts due to the influence of handling related defects on the data-writing process, the spatial distribution of errors found on writable CD's will not be random. The considerable capabilities of the CIRC error control system may be taxed by such an error environment. Measurement of Errors Commercially available CD testers are capable of providing high-level estimates of the error characteristics "seen" by a specific CD player. In gathering this information, these testers monitor the output pins of the CD player's CIRC block decoder rather than observe the actual data that are read from the disc. For example, the CD tester may accumulate and report the number of times the CIRC block decoder performs either single- or double-error corrections on received C1 and C2 codewords during successive 1-second intervals. Although this process may provide adequate information for a pass/fail test of the disc (i.e., to determine whether the disc is in conformance with the CD error specifications contained in the "Red Book") it fails to detect the size and location of every error event on the disc. To provide a means for rapid (real-time), complete measurement of both local (small neighborhood) and global CD error statistics, we have designed and implemented PC-based hardware and supporting software that directly measures the errors present on a CD. This system directly monitors data that are serially read from a disc by observing those data at the input of the de-interleaver/de-scrambler portion of a CD player's CIRC block decoder. It is capable of identifying single erroneous data bytes as well as single-byte good data intervals that may occur between two arbitrarily sized error events (a single byte occupies a distance of approximately 5 microns along the CD's spiral track). This systems error visualization software will produce graphical maps that show the physical location of measured errors on the entire disc or, via a zooming and panning feature, on user selectable local disc regions. Error burst length distributions and interburst good data gap interval length distributions for the entire disc, individual disc tracks, or user-selected local disc regions are also provided. This system can provide the information needed to estimate the reliability (i.e., the corrected error rate) of user data retrieved from CD's and assist in conducting disc aging/degradation studies as well as disc manufacturing process development and process control. Our error-measurement system is designed to operate with an IBM-compatible 486 computer that has 8 Mbytes of RAM and a VGA monitor. A CD player that uses a Philips CD3A CIRC block decoder chip is also required (see below). All error measurements are done in real time (i.e., at a continuous data transfer rate of 176.4-Kbytes/s). Choice of Decoder Chip Direct measurement of retrieved (from a CD) raw data error statistics requires the observation of data read from the disc prior to any de-interleaving, de-scrambling or ECC decoding by the player's CIRC decoder circuitry. (See [1] for a complete description of the CIRC coding structure.) We have chosen to directly access the data stream after the bit slicer has converted the analog playback signal into a digital (channel data) sequence and eight-to-fourteen (EFM) demodulation has taken place. Error measurements on data that occur at this point in the playback channel will include those errors due to demodulator sync problems as well those due to the usual impairments such as noise, disc defects, etc. Thus, we monitor sequential bytes of successive EFM frames as they enter the CIRC block decoder. A Philips CD3A Compact Disc Decoder chip is used to enable this data monitoring, because its specific architecture and operational properties make it possible to intercept the EFM frame data at the appropriate place. This chip is used in many commercial products such as the Magnavox CDD-461 multisession CD-ROM drive, the Philips CD-I system, and the Eastman Kodak PhotoCD player. It is hoped that the error-measurement system developed here can be easily adapted to other CD playback channels. Error Measurement Concept In order to obtain single-byte resolution error measurements, every byte retrieved by the read channel of the CD player must be compared to the originally recorded bytes. Normally, since there can be about 1 GB of information on one side of a CD (including all ECC parity and time code information), this process would require a large storage space (in PC memory) and an expensive and complex hardware system. An alternative approach is to utilize the data structure that results from CIRC coding to drastically reduce the required memory and the overall measurement system cost. It can be shown that, if a 24-byte repetitive block of user data is recorded using the CD "audio" format, the lower 32 bytes of each recovered EFM frame, in the absence of any errors, will be identical [2,3]. Thus, by comparing every recovered EFM frame, byte by byte, to a copy of the uncorrupted EFM frame that corresponds to the 24-byte input data block, all erroneous bytes can be located. This scheme requires only a 32-byte comparator and enough memory to store the addresses of the erroneous bytes. The fact that a repetitive 24-byte input data pattern must be present on the disc is not a limitation, since we are primarily measuring errors on writable CD's that can have any specified data written on them (a pressed CD containing appropriate data patterns would be required to use this system to measure errors on that type of storage medium). Using the CD audio format makes it possible to record up to 99 different patterns on one disc (1 on each audio track) and reduces the hardware complexity by simplifying error measurement system/CD player synchronization and CD player control. Future Developments The tools that we have developed enable us to completely understand the statistical and underlying physical nature of errors on either CD-ROM or writable CD discs that have 24-byte repetitive patterns written on them in the CD audio format. The next generation of these tools, which currently are under development, will enable the measurement of CIRC error processing flags as well as byte errors; e.g., all bytes that are erased by the EFM demodulator will be measured, in addition to the bytes that are erroneous. The display software is being upgraded to provide a two-dimensional graphical image of the error and erasure content of the four CIRC code blocks that encode the data in the immediate vicinity of an arbitrary user-selected location on the disc. By examining sufficient quantities of both pressed and writable CD's, we should be able to obtain valuable information about the characteristics and growth of error events. For example, using our error-measurement system in conjunction with environmentally stressed CD's will enable statistical characterization of error/erasure growth rates. Furthermore, because the errors measured are independent of the decoding program used by a particular CIRC chip, the measured error/erasure data can be used to determine the actual recovered data reliability (e.g., the number of uncorrectable and(or) misdecoded C2 words) that can be achieved with the disc under study versus different CIRC decoding programs. Thus, the user will be able to evaluate the utility of various CIRC decoding strategies against the actual errors that have contaminated a specific disc. References: [1]ECMA/TC31, "Standard ECMA-130: Data Interchange on Read-Only 120mm Optical Data Disks (CD ROM)," European Computer Manufacturers Association, July 1988. [2]Howe, D., and Tehranchi, B, "Compact Disc Error Measurements," The University of Arizona Optical Data Storage Center Quarterly Report, March 1993. [3]Harriman, K., "A Measurement System for Characterization of Errors in Data Stored on Compact Discs," thesis submitted in partial fulfillment of the requirements for the M.S. degree in Optical Sciences, University of Arizona, July 1993. Dennis G. Howe and Babak Tehranchi work for the Optical Data Storage Center, Optical Sciences Center, University of Arizona, Tucson, AZ 85721. They can be reached at (602) 621-4995 (voice); (602) 621-4358 (fax);dghowe@burke.opt-s ci.arizona.edu ; babak@ccit.arizona.edu.
SIGCAT recently welcomed three new Russian members as a result of its participation in CRIMEA '94, a conference sponsored by the Russian National Library for Science and Technology. A total of 268 people representing 16 countries attended the conference, which took place in Eupatory, Republic of the Crimea, Ukraine, May 23_28, 1994. Most participants were from Eastern European countries. The three new Russian members are Dr. Andrei I. Zemskov, Director of the Russian National Library for Science and Technology, Moscow; Boris R. Loginov, Director of the State Central Scientific Medical Library, Moscow; and Victor P. Zakharov, Department Head, Russian Academy of Sciences Library, St. Petersburg. Russia now becomes the newest country to be represented among the expanding list of nations that participate in SIGCAT. The theme of the CRIMEA '94 conference was "New Technologies for Use in Libraries." The conference featured several 1-day sessions on library science; modern computer and telecommunications technology; and databases, electronic publishing and CD-ROM databases. Duane Marquis, SIGCAT Executive Vice President, chaired the session on CD-ROM and databases. He also delivered two presentations covering CD-ROM basics and SIGCAT's goals, accomplishments, and activities.
CD-ROM and on-line services will be among the key themes of the May 22_26, 1995, Second Central European Conference and Exhibition for Academic Libraries and Informatics. Co-sponsored by the Latvian Academic Library, the Scientific Library of Riga Technical University, the Patents and Technology Library of Latvia, and the Information Department of the Latvian Parliament, the conference is designed to meet the needs of librarians in the Baltic states who wish to know more about electronic forms of information. The main themes of the 1995 conference will include: - CD-ROM and on-line services: trade-offs, complimentary values, desktop CD-ROM publishing, and Internet user services. - Printed and electronic information and their interrelationships. - Problems of electronic information support: copyright, accounting systems, etc. - Library automation and electronic library management and services: how to introduce such services, staff training, and user assistance. - Electronic duplicates of printed documents: their creation, use and handling. The conference will be conducted in English and will be held in the Latvian Academic Library, Latvia's main center for scientific and technical information and the leader in new information technologies in use among libraries in Latvia. For further information, contact Lee Burchinal, (703) 525-9045 (voice), (703) 351-0782 (fax) or lburchinal@ gmu.edu (Internet) or Dr. Aigars Krauze, Latvian Academic Library, Rupniecibas Str. 10, Riga LV-1235, Latvia, e-mail: alibcon@mii.lu.lv (Internet).
In the May/June 1994 Career Corner, we described the varied career paths of four SIGCAT board members. This month, we describe how three other SIGCAT members developed professional careers involving CD-ROM technology. These profiles are intended to help SIGCAT members become acquainted with their peers and also to illustrate the diversity of possible career paths by which one can enter the CD-ROM field. If you are considering a career involving CD-ROM technology, you might find ideas and inspiration in the experiences of these three SIGCAT members. - Dr. Gabriel Ofiesh, a long-time member of SIGCAT and co-chair of SIGMULTEC (the Special Interest Group on Multimedia Technology), saw CD-ROM technology as having revolutionary potential to transform the field of education. For him, CD-ROM's offered the possibility of an entirely different, learner-centered approach to education. He became involved in CD-ROM technology because he was excited about its potential to expand and enhance the learning experience. - Dr. Michael Rubinfeld, who serves as a vice president of SIGCAT, was seeking a job at the National Institute of Standards and Technology (NIST) because it offered an opportunity to be at the leading edge of technology. The job he accepted involved CD-ROM technology and development of standards. His background in computers helped him make the transition smoothly into a new and challenging career. - Francie Mendelsohn, a co-chair of SIGCAT's CDOWG (CD-ROM Data Organization Working Group), became involved in CD-ROM by being "put on the spot." Knowing very little about CD-ROM technology, she was asked to speak about it at a professional meeting 3 years ago. In rising to the occasion, she became fascinated with the technology and continued to develop her expertise as time went on. Seeing the Potential Gabriel Ofiesh has been monitoring educational technology for more than 25 years, so it was natural for him to spot CD-ROM early on as a technology with considerable potential for educational applications. In fact, Ofiesh shared his vision of CD-ROM's potential nearly 10 years ago in an article contained in The New Papyrus (Microsoft, 1986). His article, "The Seamless Carpet of Knowledge and Learning," suggested that CD-ROM technology would revolutionize not only education but also the structure of the knowledge on which education is based. The quantity of information contained on CD-ROM's and the opportunity to form instant linkages to diverse information sources are key elements that underlie the potential for these dramatic changes. For many types of learning, a sequence of printed pages is not the optimal format. In particular, the linear nature of the print medium makes it unsuited for dealing with recursive concepts. These concepts are an important element in discovery learning, which transforms the learning process from an instructor-driven experience to one directed by the student. With vast amounts of information and the ability to initiate linkages between various pieces of information, the student using CD-ROM technology becomes the force that establishes the direction of the learning process. Ofiesh cautions that merely placing CD-ROM's in a classroom does not transform a passive learning experience into an empowered, creative learning experience. The classroom itself needs to be restructured into a learning center that places the student at center stage. Ofiesh's philosophy on this matter is visible in the workshops he offers to nontechnical professionals and executives who want to understand this emerging and sometimes mystifying technology. In what he terms a "promissory note," Ofiesh says he will "refrain from being the 'sage on the stage' and make every effort to be the 'guide on the side.'" With a background in psychology and education, Ofiesh does not claim to be an expert in the technical aspects of CD-ROM. As he points out, he does not need to know how his car, television, or telephone works in order to use it. Whatever your background, understanding the potential that CD-ROM technology can bring to applications your field can take you a long way. On the Leading Edge Michael Rubinfeld had spent most of his career in industry as a mathematician, aeronautical engineer, and computer scientist but was intrigued by the special atmosphere at the NIST. The opportunity to participate in decisions that would affect leading technologies and to make contributions that would benefit both the public and the private sector held a strong appeal for him. After hearing through a colleague that a position was available at NIST, he applied and in April 1990 was pleased to accept a position as a computer scientist in the Computer Systems Laboratory. Rubinfeld's responsibilities at NIST include development of CD-ROM standards and of a Portable Document Format (PDF) that allows bit-mapped viewing of a document across multiple platforms. His job also launched him into SIGCAT, because his predecessor at NIST, Jean Baronas, was one of SIGCAT's co-founders. Until he joined NIST, Rubinfeld had no involvement with CD-ROM technology. His first task was to get up to speed on the standards already defined, first in the Red and Yellow Books and later in the Green and Orange Books. (This series of documents describes the standards for CD audio, CD-ROM, and other formats.) Rubinfeld's technical background allowed him to readily overcome his initial lack of familiarity with the technology. Some of his other activities include serving as editor for the Multimedia/Hypermedia Model and Framework Document for ISO SC 18 and participating in the development of the ISO 9660. Rubinfeld advises those interested in CD-ROM technology to become familiar with various platforms and learn to use several authoring systems. He sees a shortage of people well versed in sophisticated authoring systems. In addition, he predicts more interactive multimedia training and education. Like Ofiesh, Rubinfeld believes that the opportunity to get the learner more involved in the subject matter is likely to have a significant impact on education from elementary school through graduate school. Wing It! Francie Mendelsohn faced a significant challenge when, as a technology analyst working for IDC Government, she was asked to speak at a conference for information managers sponsored by the Environmental Protection Agency (EPA). EPA wanted her to address three optical storage technologies: CD-ROM, WORM, and magneto-optical. At the time, she was only slightly familiar with these technologies. She quickly began to do research in preparation for her speech. One telephone call led to another, and, as she learned about CD-ROM technology, she also became fascinated with it. As part of the networking process through which she gained insights into the field, she was referred to Jerry McFaul and soon became involved in SIGCAT. Mendelsohn's history of enthusiastically seeking professional challenges began in the mid-1960's when she joined the National Security Agency. The job for which she was hired _ German translation _ did not appeal to her, so she asked for a chance to become a computer specialist. Although she had never used a computer, her move into that new and rapidly expanding field led to a dynamic career in both industry and government. At the Food and Drug Administration, she set up an Information Center at which agency personnel could try out various PC hardware and software products. She also used the programming skills she had acquired to write a complex software program that produced a listing of approved prescription drug products. Now, Mendelsohn is president of Summit Research Associates, a technology assessment firm that she founded this year. The firm specializes in evaluating various emerging technologies _ both their potential and the barriers to their success. In one of her current projects, as part of the National Performance Review, she is assessing kiosks as a communication medium. She also evaluates CD-ROM discs and guides people through the disc preparation process. Mendelsohn comments that the quality of available CD-ROM discs ranges from "brilliant to awful." Her least favorite experience: waiting 14 minutes for a search on a pharmaceutical reference product developed by industry. Multimedia Lotus 1-2-3, which includes an excellent tutorial and on-line help system, gets her vote as one of the most impressive new CD-ROM products. For those interested in learning more about CD-ROM technology, Mendelsohn encourages dedicated research and networking. Asking for referrals to experts can lead quickly to the person who can answer your question. In her experience, a considerate researcher is likely to receive an informed and gracious response. As the experiences of these three SIGCAT members illustrate, there are many different paths by which one can develop a career involving CD-ROM technology. Whether you are motivated by an interest in a specific application like Ofiesh, a working environment like Rubinfeld, or the study of the technology itself like Mendelsohn, there are many ways to carve a niche for yourself in the CD-ROM industry. Networking with your fellow SIGCAT members is an excellent place to start! Judith Lamont is a freelance researcher and writer specializing in various aspects of computer technology. She holds a doctorate in experimental psychology with a specialty in career development.
CDOWG Conference Session Addresses Pitfalls To Avoid
The CD-ROM Data Origination Working Group (CDOWG) sponsored a session on "Setting Up and Configuring Your Write-Once System: Pitfalls To Avoid," which took place on April 28, 1994, during the SIGCAT '94 conference.
During this informative session, participants raised a number of questions regarding formats and systems. Lively discussion ensued, and a variety of solutions were offered. Topics of discussion included multi-session drivers and the Orange Book standard; archiving from a mainframe; the range and availability of retrieval software, particularly for archival purposes; resolving hardware interrupt problems in SCSI configurations; and problems related to hard disks that do constant thermal calibration.
CDOWG provides a forum for learning more about the development of CD-ROM products, especially premastering requirements. For more information, contact Maureen Prettyman, National Library of Medicine, (301) 496-1936 (Internet address: reenie@nlm.nih.gov) or Frances Mendelsohn, Summit Research Associates, Inc., (301) 670-0980 (Internet address: fmendels@seas.gwu.edu).
SIGCLASS Works on Specification for Encrypted CD-ROM
The Special Interest Group on Classified Applications (SIGCLASS) made final changes to the second draft of a specification for encrypted CD-ROM during its July 21, 1994, meeting. The SECURE CD Specification defines file system extensions that support an automated process to identify an encrypted CD-ROM and its associated cryptology. The goal is to support the application of cryptology for classified, sensitive, commercial, and exportable CD-ROM media using an ISO 9660 conferment disc. The specification allows for user-definable implementations ranging from the encryption of a single file to the encryption of the entire data area.
At future meetings, SIGCLASS will be reviewing commercially available cryptographic products with respect to their application to CD-ROM. SIGCLASS' goal is to publish a list of products that work with CD-ROM and to provide information on how to support those products using the SECURE CD specification. For further information, call Duane Marquis, SIGCLASS co-chair, (202) 482-3055.
SIGLIT Helps Librarians Evaluate CD-ROM Products
With the recent growth and diversity of CD-ROM products, many librarians face difficult choices in selecting products and evaluating their usefulness compared to hardcopy equivalents. Therefore, SIGLIT's upcoming activities will focus on the theme "Decisionmaking With CD-ROM's." SIGLIT co-chairs Felice Sacks and Denise Davis are planning an informative and educational series of programs that will feature presentations by vendors, consultants, and peer librarians. These presentations should help SIGLIT members make more informed purchasing decisions about new CD-ROM products.
SIGLIT's October kick-off meeting will be held in conjunction with the October 18_19 regional meeting and will feature the following presentation:
Topic:Updating CD-ROM Products With On-line Access Place:NOAA Conference Center 1305 East-West HighwaySilver Spring, MD(near Silver Spring Metro stop) Date:October 19, 1994 Time:9:00 _ 11:00 a.m. Speaker:A vendor representative from CCH, Inc. CCH's online service, CCH Access, enables users to search for up-to-date information on CCH's complete line of CD-ROM products in the areas of tax and business law, securities, health care, human resources management, benefits and pension, and payroll.
We will also discuss ideas and suggestions for expanding SIGLIT's compilation of quick reference fact sheets on CD-ROM products. Bring your ideas and your own quick reference displays to this session!
Please RSVP to Felice Sacks by calling (202) 663-6763, sending an e-mail to fsacks@capcon.net, or sending a message through ALIX. Mark your calendars for October 19! Light refreshments will be provided.
SIGSGML Raises Awareness of SGML Capabilities
The Special Interest Group on Standard Generalized Markup Language (SIGSGML) sponsored several sessions during the SIGCAT '94 conference. The SGML sessions were designed to raise awareness of the significant potential of the Standard Generalized Markup Language (SGML) to "structure" large text products for intelligent search and retrieval based on this international standard (ISO-8879). A variety of industry experts introduced basic SGML concepts to a very receptive audience.
SIGSGML co-chair Walt Klaus notes that CD-ROM technology has become the media of choice for many publishing products because of its large storage capacity, low cost, and ease of distribution. However, access to information and controlled maintenance of the underlying data need to be improved. Proprietary full-text retrieval software offering rudimentary solutions has been available for many years. These solutions include partially or fully inverting text (words, names, etc.) and building indices for later access. Although a "string-based" search mechanism may be the answer for some, it unfortunately creates a dependency on proprietary software, which also entails significant index overhead and index updating (depending on the extent of changes in the underlying data).
SGML provides a "standard" model for structuring text-based data by using a Document Type Definition (DTD) and associated Mark-up (Tags) governed by a formal and public mechanism, much like fielded data in conventional database environments. SGML provides the basis for "intelligent" design, subsequent conceptual and context-sensitive search, and the significant potential for a "standard" retrieval engine (or, as a minimum, a generic retrieval interface).
Future SIGSGML sessions at SIGCAT '95 will reinforce and expand knowledge about this promising technology, which provides a natural and synergistic enhancement for many CD-ROM projects.
The SIGCAT Library Forum is now up and running on Compuserve. It is being managed by Gabe Ofiesh, co-chair of the SIGMULTEC Working Group.
To access the Forum from Compuserve, type "GO CDROM" and then browse Library. The SIGCAT Library Forum is the 11th Library (i.e., choose number 11 in the Library menu).
SIGCAT Working Group chairs are urged to provide information on Working Group activities for uploading onto the Forum. Please send your short articles and(or) meeting notices via e-mail to Gabe on Compuserve ID # 72662.230. Alternatively, you may send a floppy disk containing the information to Gabe at 4031 27th Road North, Arlington, VA 22210. Gabe can be reached by phone at (703) 243-4271.
This summer SIGCAT offered The Fundamentals of Producing a CD-ROM , the CD-Recordables Workshop, and Digital Video _ Capture and Compression. All SIGCAT courses feature learning by doing in a classroom with 11 workstations with CD-ROM readers and LANschool.
SIGCAT founded The Technical Education Center as a way of fulfilling its ongoing educational mission. This fall, a series of classes is being offered that should meet the needs of those just beginning to explore CD-ROM technology and those seeking to develop the specialized expertise required in multimedia authoring and development today. At a time when training monies are scarce, SIGCAT recognizes students' needs to acquire practical experience and expertise that they can immediately use in their areas of specialty.
If you are still undecided whether to take a class this fall, here are a few of the comments that we have received from students attending SIGCAT training courses. Anthony Ciaffoni, a CPA from the Philadelphia area who attended the CD-Recordables workshop, said that "It clarified issues that I had been reading about for months. I wish I had known about it 6 months earlier." John Ecklund, a curator at the American History Museum, described Jan Ozer's digital video class as "first rate" and said that "The instructor's explanations highlighting the trade-offs of using different codecs were especially enlightening." Regarding The Fundamentals of Producing a CD-ROM, Peter Shen remarked that "The Fundamentals is a very valuable course _ especially for people contemplating jumping into producing a CD-ROM."
For more information on the variety of classes to be offered this fall, please see the accompanying schedule.
The following courses are scheduled for the fall. To find out more information about the classes or to receive a course catalog, please call Reed MacMillan at the SIGCAT Technical Education Center at (703) 435-4661. Our Training Center is located at 1851 Alexander Bell Drive in Reston, Virginia. SIGCAT DISCourse subscribers and corporate sponsors receive a $25 discount on each class they take. Please note: Classes are contingent upon a minimum level of enrollment.
Basics Week September 8_9 The What & Why of CD-ROM _ Reed MacMillan & Michael Grainger Imaging Week September 12_13 Digital Video _ Capture and Compression _ Jan Ozer September 14 The ABC's of Imaging -Bill Hooton September 15 Advanced Imaging Applications _ Bill Hooton September 16 Spatial Compression _ Dr. Nicholas Beser Multimedia Week September 19_20 The What and Why of Multimedia _ Gabe Ofiesh September 21_22 Capture Your Conference on a Multimedia CD-ROM _ Barry Hudson September 23 Multimedia Training Applications _ Gabe Ofiesh Production Week September 26 Multimedia and Today's Market _ Jay Hoffmeir September 27 Technical Multimedia _ Jay Hoffmeir September 28_30 How to Produce a CD-ROM -Reed MacMillan & David Lind Software Week October 10_11 Focus on Text Retrieval Engines -Tom Tiedeman October 11_12 Premastering Software and CD-Recordables System _ Walid Achi October 18_19 The SIGCAT Meeting _ SIGCAT Software Showcase Series Multimedia Presentations Week October 24 Learn to Astonish with Astound _ Instructor to be announced October 25 Beyond Bullets _ Tips and Tricks for Creating Multimedia Presentations