Network Bulletin No. 05-25.
Date: May 20, 2005
Subject: Digital long-term planning group
Index term: Report from the September 2004 meeting
The Digital Long-Term Planning Group held its fifth meeting on September 22-24,
2004. A summary is enclosed. As Bob McDermott is no longer with NLS,
comments on the summary may be addressed either to Jean Moss or to Michael
Moodie, deputy director, who will coordinate future meetings.
Enclosure
For further information contact:
Jean Moss
Digital Projects Coordinator
(202) 707-9238
[log in to unmask]
____________________________________________________________________
Digital Long-Term Planning Group
September 22-24, 2004
Group Members Participating:
David Andrews, NFB
Lori Bell, Illinois
Kim Charlson, Massachusetts
Paul Edwards, ACB
Barbara Goral, Colorado
Karen Keninger, Iowa
Robert Maier, Massachusetts
Donna Jones Morris, Utah
Karen Odean, Illinois
Peggy Rudd, Texas
Guynell Williams, South Carolina
Michael York, New Hampshire
Bob Axtell, NLS (Recorder)
Kurt Cylke, NLS
Judy Dixon, NLS
Robert Fistick, NLS
Bob Jones, NLS
Brad Kormann, NLS
Bob McDermott, NLS (Chair)
Michael Moodie, NLS
Jean Moss, NLS
Steve Prine, NLS
Carolyn Sung, NLS
Deborah Toomey, NLS
David Whittall, NLS
Wednesday, September 22
9:00 - 10:30 Announcements and NLS reports
Announcements
Kurt Cylke welcomed the group, particularly the new members. A number of
COSLA members have joined the group. Robert Maier, Donna Jones Morris, and
Peggy Rudd were able to participate at this meeting. Sara Jones of Nevada, and
Doris Ott of North Dakota, are new members but were unable to attend this meeting.
David Andrews has joined the group from NFB. Lori Bell of the Mid-Illinois
Talking Book Center has also joined. Cylke remarked on the significant number of
state librarians present, as it reflected the number of voices representing the parent
agencies of regional libraries and it ensured that COSLA as a whole would get a
broader view of the proceedings. He also noted the large NLS staff presence and the
high interest that it signified. He mentioned that they would be participating in the
proceedings occasionally.
Robert McDermott covered the agreements on conduct for the meeting: to begin on
time, to be back on time, that there should be only one speaker at a time, and that
everyone is equal. Also be more forward than polite in one's remarks. He said that
previous meetings began with brainstorming and discussion of the prior meeting but
that, with the current meeting's full schedule, the format would be different.
Opinions are needed on many things related to the digital talking book and those
issues will dominate the meeting. He opened the table for comments and questions
and there were none.
NLS reports
NLS staff presented update sessions on NLS digital talking-book activities.
Strategic plan
Jean Moss gave a presentation which noted that the strategic plan calls for digital
talking books and players to become available in 2008.
Discussion began with a question about the newsletters and publicity. The three
primary targets of the newsletters are state libraries, network libraries, and patrons.
Patrons are also targeted through inserts of the newsletter content in TBT and BBR.
The newsletter content will also be on the Web and will be sent to the media. NLS
will ensure that network libraries receive the press release content before the media
receives it. It became evident at the last COSLA meeting that communication on the
digital program needs improvement. NLS is spending $100,000 to remedy the
problem.
Player design contract
Michael Moodie said that the player design contract is the major project of NLS on
which everything rests. A Request for Proposal (RFP) for the design of two major
components, player and cartridge, was released in September 2004. The RFP
includes a requirement for mailing container and duplication design concepts which
can be fed to other studies specifically on those subjects.
The RFP will go to seven hundred professional engineering companies on a GPO
schedule for contracting. The firms on the schedule work on or contribute to
complex projects of all kinds. Many will be qualified for this project. If a firm does
not have experience working with the blind, they are required to partner with a
company that does. That requirement is a cornerstone of the process so that all have
this capability. There is a basic requirement to validate with key stakeholders that
the requirements are correct, up-to-date, and complete.
There are many design parameters: that the user interface design is intuitive, easy to
use, and functional; that the design works for diverse users' needs; that the needs of
the most to least sophisticated users are accommodated; and that industrial design
principles are followed to make the product look and feel good. The player must be
designed for ease of manufacture. It must have good audio engineering and software.
It must be designed for player support: it must be either repairable, or robust but not
repairable. If repairable, the design must include how it is to be repaired and ensure
the ease of repair.
A key requirement of the RFP is testing for usability. The testing will be iterative
with users, librarians, and repair personnel, the three groups that will be working
with the player for years. The design effort for the cartridge and container must be
coordinated with the firm doing the distribution study. The overall system must be
streamlined, with less manual labor built into both machine and book efforts. The
preliminary estimate is eighteen months plus field testing and hand-off to the
manufacturing firm. The design firm must stay involved a year after the design is
complete to work with manufacturer, ensuring that the player is built according to the
design specification and that, if there are problems with the design, they are resolved.
The group discussion raised the issue of disposal for disposable machines. While the
quantity is expected to be small, NLS has seen the issue with C-1 machines and will
address it. Regional libraries are interested in helping with testing. A discussion of
price points for the player yielded $150 as the low end of NLS expectations, with
$200 being the more likely figure. Concerning the possibility of the price being too
high, NLS has prepared Congress for a substantial funding need. Congress seems
well disposed as we are doing well in our requests. If the price were $300, it would
not put the player out the picture, but it would take longer to come up to a full
inventory. Nothing so far indicates that the price will be higher than $200. The
player is less complex than the current C-1 machine.
Digital asset management
Robert McDermott summarized the continuing evolution of the systems needed to
support the production and future distribution of digital talking books at NLS.
Initially, digital recording was not perceived as an automation issue. The NLS studio
started using digital audio tape (DAT) for creating DTBs, in the same paradigm as
analog audio production. DAT proved to be cumbersome and a change was made to
recording on removable hard drives. These also had cumbersome aspects and were
prone to failure. It became clear that networked online storage was needed for
smooth studio operations. That storage could also be used for performing quality
assurance review on completed books. The requirements for networked storage were
analyzed and a storage area network (SAN) solution was chosen. The SAN system is
now up and running in the NLS Studio and will be used by the Quality Assurance
Section in the near future. It is designed to hold a year's worth of books and the
current workload of the Quality Assurance Section.
Beyond the immediate operational need for handling the workflow of books in NLS,
there is also a need for a system to electronically receive books from producers and
to store and manage the books once they are approved for distribution. The original
idea of storing the CDs on shelves as is now done with master tapes is not adequate.
NLS will occasionally need to update the entire collection to take advantage of
improvements in compression or digital rights management, or to include new NISO
features. To do so, NLS will need access to an electronic archive of the original
recordings. NLS estimates that the archive will require about thirteen terabytes of
storage per year. NLS is proposing a digital asset management system to track the
location and pertinent data about a book when it is in-process, in the archive, and in
distribution. We also have a requirement to accept books from producers
electronically rather than physically receiving CDs through the mail. For NLS, mail
is becoming increasingly unreliable due to security requirements imposed on the
legislative branch. Electronic transmission of books is not a simple task because the
uncompressed books are so large four gigabytes on average. NLS has assigned an
analysis of our requirements to a contractor. The analysis was begun in March 2004,
and NLS received a draft in August which is being reviewed for amendments and
additions. When the analysis is complete, an RFP for service will be issued.
Web-Magazines
Judy Dixon described a pilot that began in late 2003. Magazines were considered a
good candidate for testing Web delivery. They are smaller than books, are amenable
to partial downloads, and are released monthly. The test started with thirty-five beta
testers and interest has increased. Three titles are being posted, each from a different
producer. A contractor works with NLS to mount the books on the web site The
producer sends files, the contractor unzips them, checks the NCX for articles, and
automatically generates HTML for title names and files.
The user logs onto the site listing the magazines, picks one and goes to the issue
menu. He/she can then go to an article menu or download the whole issue. The users
were provided with Victor Readersoft and an unpack tool from Bookshare. At the
end of June NLS surveyed the thirty-five users and received a 65 percent response.
All but one had successfully logged on and downloaded a magazine. Positive
feedback included the ability to navigate from article to article, not having to wait for
mail delivery, and exceptional sound quality. Users were disappointed with being
tethered to a computer and waiting for magazines to post (a problem at our end).
They also did not like the amount of time download took. Two users said there were
too many steps in the process. Suggestions: seven users suggested integrating the
unpack tool with Readersoft so books load directly to the bookshelf; three suggested
more markup; and three suggested being able to download to a portable player and
decreasing the file size. NLS is looking at how to decrease the file size and how it
would affect sound quality. We are aware that the download process is somewhat
tedious. Neil Bernstein, a new member of the Engineering Section, is looking at the
site to get a fresh look at simplifying it.
In the discussion, four testers were among the members at the table. They had
numerous comments and suggestions. Many were concerned with making the
process easier and broadening the number of people who would be able to use it
effectively. One liked the system but suggested trying a lower sampling rate with
smaller file sizes to broaden the range of people who could use it. Another
commented that the download process had a lot of steps and required some computer
literacy. The success of the experiment depends on who the target is. The current
target is computer users who choose to read that way. Perhaps users would find it
helpful to know how many steps there are (for example knowing that it is step 1 of
5). It is believed that the demand will grow as more people move to broadband
access to the Internet. Growth in turn would require more support, which is a
concern, as only two people at NLS can give support now. Another trend that will
increase interest is the growing number and availability of portable devices that can
read digital files. The ability to use portable devices with NLS books would require
accommodating DRM and meeting NLS requirements for sales restricted to eligible
readers. One tester felt the Bookport needed more explanation and another said that
he used the material with a Telex Scholar but he found using more than one layer of
navigation levels confusing. There needs to be clearer feedback from the reader on
navigation levels and there is no consensus on how to do it yet.
Web-Books
Kurt Cylke noted that we plan to have a program in 2008 with books on flash
memory. At the same time we want to have books online. Our first step is Web-
Magazines. Web-Books are scheduled for implementation in May 2005. We
contacted Plextor, now doing a streaming book program in Tokyo. The proposal is to
do streaming for 1,500 books with testers and for full rollout in 2008. Later NLS
will design a downloadable books system. It will take longer due to DRM issues.
NLS may not be able to do a sole source contract with Plextor, but they will submit a
proposal in a week or so.
Oregon petition
Kurt Cylke called this an awkward situation but said it should be discussed. Since the
meeting in San Diego, a COSLA representative contacted the ACB and NFB
representatives on this group to discuss the situation behind the scenes. An article
has been submitted to Public Libraries, along with comments from two users to be
selected by the consumer organizations A petition was distributed to the group,
along with the Oregon article. Regarding the petition, the author was encouraging
support from whoever wanted wished to sign. Most signers (85%) appear not tobe
users or librarians for the blind. The articles will be discussed on Friday after
everyone has a chance to read them.
Following the COSLA meeting in San Diego, NLS has initiated a program to visit
each state library in one year and has made thirty-five visits thus far. NLS
determined earlier not to try to raise readership before the digital rollout. Because of
the level of anxiety expressed during the visits, a public relations firm has been
engaged to pursue a nationwide campaign to raise the number of readers by 70,000.
This is in direct response to COSLA concerns. In addition to the $100,000 for news
on the digital effort, funds will be expended on raising the readership.
The group discussion centered on readership. A Colorado group is encouraging
learning disabled readers to use the program. This is an area where numbers could
increase, but is tricky because most learning disabled do not fall under the federal
physical disability requirement. If the requirement is violated, NLS could lose postal
privileges, as almost happened to Recording for the Blind & Dyslexic (and to NLS)
when they tried to ignore it. The difficulty of adding this type of detail (the physical
eligibility requirement) to a TV spot was noted. The basic requirement is that, for
physical disability, a medical doctor must certify.
10:45 - 12:15 Digital medium: flash memory
Michael Moodie introduced Neil Bernstein, who is new to NLS. Neil has a B.S. in
electrical engineering from Brown and an M.S. from Berkeley. He has worked with
MCI in networking and in hardware and software for delivery of digital media. Neil
led a presentation and discussion on USB flash memory. The USB flash drive is
NLS' medium of choice for physical digital distribution. It consists of a flash-
memory chip, a controller, and a USB connector.
The arguments presented for choosing flash memory included: it has no moving
parts; it is reusable; it can be mass-duplicated; it is a widely available and mature
technology; it is continuing to get cheaper; and there is a multiple and growing
number of sources, as all chip makers are moving towards making them.
USB flash memory is particularly appropriate to the NLS program. It will allow
technology independence of the book because it allows changes later without
redesigning the player. It makes players more independent because the player's USB
port can be used to connect to other devices. It allows customization such as
allowing NLS to write the book to the chip, but not allowing the chip to be written by
the user. The book can be sized, labeled, and packaged to order, with braille and
large-print labeling. USB allows customization and is widely supported. The
network would be able to procure supplies for local production.
Concerns about flash media were addressed:
Cost. The July 2003 report of the Gartner Group, a premier IT consulting firm,
indicates that costs are going down and the price drop is projected to continue. A
chart showing cost projections to 2007 for 128MB and 256MB with costs declining
to $9.09 and $15.18 respectively was shown. NLS will continue communication
with the author of the report for updates and IDC for corroborating evidence, as well
as with manufacturers such as Sandisk. All think the numbers are conservative and
prices will probably drop faster. Earlier projections for 2004 bore out in quick test of
Amazon.com retails. Retail prices are cheaper than predicted wholesale.
Capacity floor: Capacity floor is the price and size of the smallest available chip.
As technology changes, small chips are no longer available, but the price of the
smallest available chip is expected to level out. This is not a problem since a bigger
chip can always be used.
Durability: USB flash is extremely durable. They have ten years data retention with
constant use. Sandisk says that, if stored, they can last fifty to sixty years. Humidity,
shock, bending, twisting, and electromagnetism are not issues. The BBC reported
tests of memory card formats dipped into Coca Cola and coffee; run through a
washing machine; run over with a skate board and a toy car; and given to a six-year-
old to play with. All survived with 100 percent of the data intact.
Write speed: It takes one to two minutes to write a book from a computer. The
write time will decrease, but not as much as price.
The cost of developing customized parts: The controller chip and shell will be
customized which will present a one-time cost for design.
New memory technology: The next big thing is magnetic RAM. It may replace
flash and hard drives. NLS could move to that, using the USB port, and still keep the
player. This technology is five to ten years away and not developed enough to cause a
change of choice.
Michael Moodie discussed the group's initial media requirements and how USB
flash measures up:
Not under three dollars but 128MB will be under seven.
The memory chip is in a sealed cartridge.
The cartridge is large enough for large-print and braille labels.
It is large enough to be picked up.
It can be inserted only one way for ease of explanation to patrons.
It will stand up to cleaning from household agents.
It facilitates handling by automated equipment.
It is stackable.
It will function after immersion.
It has 250 write cycles.
Data remain readable under normal conditions and occasional extremes: good for
ten, probably twenty, years.
It is playable after exposure to mailbox conditions.
It is not affected by humidity, shock, normal vibration, bending twisting, or
electrostatic discharge.
The connectors maintain reliable contact after 10,000 insertions.
They resist contamination (the actual connectors will be protected by a shell.)
They resist mechanical damage.
They resist corrosion.
Most cartridges are in 2MB/second range for write speeds.
The read speeds are as fast as write speeds.
It will have write protection from unauthorized parties.
It is not ignitable.
It will not outgas.
It will avoid single source procurements.
It should adapt to changes in memory technology.
The labels should stay on but will be removable--important for distribution-on-
demand system. The cartridge must make relabeling as automated as possible.
The presentation was enthusiastically received and there seemed to be satisfaction
with the progress on the media. There was interest in getting these media by non-
library agencies and groups. NLS plans to make them available to other groups
through a rider on the NLS production contracts, although NLS itself could not sell
them. Resellers could get involved later. There was also interest in being able to use
the player with other flash devices, which would be possible. NLS believes
commercial audiobooks will move to flash, and we want our player to be able to play
them as well. Other uses for the standard USB port would include connectivity to
other devices such as add-ons for physically disabled readers, for diagnostic devices,
or for upgrading the player software to the next version of the standard. This led to a
short discussion of upgrades, which RFB&D does by sending CDs to patrons. NLS
plans on handling the upgrades. We could even append the upgrades to books the
patron will receive so that it occurs transparently. One participant asked what would
happen if USB changed. NLS responded that we would not have to change in
response. NLS could procure enough spare parts to ensure a long production period
as we have done for recorded disc and cassette players. There was a question about
robotic equipment for burning flash units, as is done now with cassettes. At first
such equipment would need to be purpose-built but after that we would definitely
have it. This is not a difficult technology. We have it for cassette duplication, and
cartridge manipulation is easier than for CDs.
To a question concerning patrons copying books, NLS responded that it would not be
allowed. Our books are produced under law and written agreement, to be available
only to blind and physically handicapped readers. We have to be able to prevent
patrons from giving them to ineligible readers. NLS has not yet determined how to
do this, but we will. Otherwise, the program is in jeopardy. The books will be
downloadable, but for the same reason we will be doing streaming first until we
know how to prohibit use.
General discussion
The group returned to the topic of the ___Public Libraries__ articles. NLS
suggested that consumer agencies supply commentators. The articles will appear in
April. There was a suggestion that the members of the group should take the
initiative to answer these sorts of criticisms. NLS has communicated to NFB and
ACB conferences for twenty-five years in an effort to keep them abreast of
developments here, nevertheless we have dropped the ball on more extensive
communications. We are making amends. With respect to the article from Oregon,
the statistics cited do not consider the effect of the end of the disc program. The
information about the Netherlands books is incorrect. They are not using DAISY
technology. The CNIB has approximately 2,000 titles, which is not a complete
program. A COSLA representative said it was incumbent upon them to make sure
the facts were on the table at COSLA. This is why there are now several COSLA
members in this group to make sure one person will not dominate reporting to that
body. The group agreed that the author of the article in question was no doubt
sincere in his writing but that the facts and statistics used were misleading.
The discussion moved to reaction among the blind community. A consumer
representative said that there is a group of blind people aware of digital talking-book
developments and who have expectations that are not being met. The DAISY and
RFB&D people believe NLS is moving too slowly. He did not share this sentiment
but suggested that NLS maximize navigational markup on the books, and that they
should be marked to the page level, or there would be more dissatisfaction. This
issue has been extensively discussed at NLS. NLS will make every effort but some
books, especially the retrospectively converted ones, will probably not be as heavily
marked up as the most intellectual reader might want. NLS will learn from RFB&D
and others who are already releasing digital talking books but marking to the page
level could be problematic. NLS is trying to learn to make the technology work so
that this is feasible, but it is still not easy to mark up. The issue of synchronized text
and recordings was raised. NLS gave a lot of early attention to this possibility, but it
is much more complex than anticipated. A librarian noted that these types of books
were very promising and that she sometimes reads braille and recorded books in
tandem to help with foreign or unfamiliar names. Another added that the promise of
synchronized full text was a dream worth having and that NLS should continue to
examine ways of making this highly featured type of book feasible to produce.
There was discussion about the current range of players and their frustrating
inconsistencies. But these inconsistencies were viewed as natural at this early stage
and that the market would choose the best features.
Finally, a state librarian mentioned that she wished to do a performance-based study
of libraries for the blind, going to customers to see if their objectives were being met.
She would like to determine the impact of the program on people's lives. She asked
if other libraries had tried this. NLS noted that Deborah Toomey had assumed
responsibility for a similar project, initiated by the Librarian of Congress. Deborah
added that the first meeting of the committee will be next month and that she would
be happy to discuss it further. On the issue of how the program affects peoples lives,
there are the contrasting effects of lifelong users who don't know what life would be
like without the program and new users who may have unrealistic expectations based
on their experience of total access to the printed word. In past studies done by NLS,
interviewees were reluctant to criticize the program for fear of losing it. Perhaps
using specific questions would resolve the problem. One member suggested that
BPH libraries not be judged by the standards used for other libraries.
1:15 - 2:45 Player transition
Jerry Ducrest, program planning consultant and contractor to NLS, gave an overview
of the Player Transition Study. NLS needed to determine the requirements for future
cassette machine production, repair services, parts requirements, and policies
affecting supply and demand during the transition to DTB. The purpose of the study
is to plan to phase out cassette- machine production and to phase in digital talking-
book machine production up to 2008 and after. Alternative transition plans are to be
formulated.
One primary audience of the study is the Inspector General's Office at LC that has
proposed to discontinue current production of machines immediately and just do
repairs. NLS needs an objective study to demonstrate the effects of discontinuing
production (or for that matter repair, and spare parts) at various times.
The introduction of digital machines will be considered simultaneously with the
phase- out of cassette players. The study has produced a modeling tool for
evaluating alternative transition plans over a fourteen-year transition period. The
model allows one to concentrate on detail, allows changes of assumptions or
variables to run various scenarios, and forces objectivity when the outputs are
created. The focus is on NLS (i.e., not network) costs. C-2 cassette machines and
record players were not considered in the study.
The model considers: production of machines, attrition from damage or loss, repair
capacity (commercial and volunteer), and parts inventory needed for support. It
calculates parts needed and number of repairs (warranty, volunteer, commercial). It
accounts for differences in patron desire to switch from cassette to digital. The
model produces reports, tables, and graphs.
The discussion centered on assumptions of how long people would want to retain or
receive cassette players in addition to or instead of digital talking-book players.
Magazines were a key issue. Magazines will probably remain on cassette and not be
distributed on flash because patrons keep their magazines. Thirty percent of our
readers receive magazines, so the study assumes that thirty percent of current readers
would want to retain their cassette machines. Additionally, it was assumed that
another twenty percent would want them to play older, unconverted cassette books.
There was also a question of how and to whom the digital talking-book players were
to be distributed, as that might affect demand for cassette players. The assumption
for the model was that new patrons would be assigned DTBMs rather than CBMs
and that only fifty percent of existing patrons would want DTBMs. These
assumptions were widely challenged on a number of grounds.
Ducrest reviewed scenarios that had been run through the model and presented as
graphs. They predicted where shortfalls in CBM and DTBM supply, if any, would
fall. The plan assumes the completion of transition phase in 2012 (meaning everyone
who wants a DTBM will have one) and the completion of the residual phase by 2017.
Five years is a reasonable period to assume for cassettes to drop to a negligible
number. Parts obsolescence also a factor in this timing. The scenarios were:
1. CBM production would continue for one more year with 6,000 repairs per
year. Unacceptable shortfalls for CBMs in 2008 were predicted.
2. Repairs increased to 12,000 per year (allowing for a period to ramp-up to that
number) but this also hit inventory problems in 2007-2009.
3. CBM production would continue for two years with 6,000 repairs per year.
Availability in 2007 fell below the ten percent line which was considered dangerous.
4. CBM production would continue for two years with 12,000 repairs per year.
Fall below ten percent availability was postponed to 2008.
5. Assumes production for three years with 6,000 repairs a year. Inventory never
falls below eleven percent.
6. Assumes production for three years with 3,000 repairs a year. Inventory falls
to ten percent in 2008, but it could be feasible.
A cost comparison of the three most viable implementation scenarios (nos. 4-6) was
presented. The average annual costs for the three years ran from $15,371,938 for
scenario four to $16,183,854 for scenario five, with scenario six costing $16,045,284.
All scenarios would produce a surplus in the end but NLS needs to get past the
2008-2009 pressure point.
The group discussed the distribution of the DTBMs and how it would affect demand
for CBMs. Some thought that new readers, if given a DTBM, would have little
interest in cassettes but that existing readers would want to keep the CBM, possibly
until they were more confident in the digital program. NLS proposed that the
propensity of existing patrons to want the CBM is likely but perhaps not ninety-three
percent as assumed. Some thought that the decline in demand for CBMs would be
rapid. The group was interested in finding a way to migrate the magazine program to
digital form as a way of significantly reducing the demand for CBMs. Using a
cheaper cassette player was suggested, but NLS believes there is no realistic way to
introduce and maintain a new cassette machine it would require a solicitation and
award a new production item as well as bringing it up to speed. A disposable
cassette player was also suggested. Because many readers also have their own
portable cassette players, many may send back their CBMs. Still, a supply of NLS
machines is needed as an insurance policy. It was suggested that the new machine
would be easier to explain to new patrons. With the publicity effort, there will be a
large demand initially for the DTBM among new readers.
3:00 - 4:30 Reading at risk
Mark Bauerlein, director of Research National Endowment for the Arts, gave a
presentation on the report__ Reading at Risk,__ which has received widespread
media attention. He is presenting his study to different groups: educators, publishers,
writers, and librarians. The report is the first step in addressing a difficult problem.
It presents the results from the literary segment of the Survey of Public Participation
in the Arts, conducted by Bureau of Census and designed by NEA. __Reading at
Risk__ presents results from 1982-2002 and provides a detailed profile for the role
literature plays in the lives of Americans.
The study correlates literary reading with geographic region, gender, race/ethnicity,
education, age, other cultural and social activities. It profiles households through
interviews and follow-up phone calls to ensure a proper demographic representation
of the American population as well as correlating with other cultural events,
entertainment. Literary reading includes any of the following: novels or short stories
of any kind, poetry, and drama. It does not include non-fiction. No distinctions are
made for quality or length. One poem a year would count as literary reading. It
included any medium such as books, magazines, or the Internet.
The key finding is that literary reading has declined dramatically over twenty years:
56.9 percent in 1982 to 54 percent in 992 to 46.7 percent in 2002. The rate of decline
is increasing: less than half of population now reads literature. The decline in literary
readings parallels a decline in total book reading at about half the rate. Women read
more than men, but literary reading by both groups is declining. The gender gap has
existed since last century but is getting wider. The west has the highest rate of
reading, over fifty percent. The south has the lowest at forty percent. One-third of
men read literature. In a time-use study of Bureau of Labor Statistics, men spend two
hours watching television, one hour playing games, and eight minutes a day reading.
Dr. Bauerlein projected that the next study will be much worse. Reading skills and
rates on texts requiring some engagement are declining in the younger generation.
Literary reading correlates to education, but every education group dropped. The
college and graduate school rate dropped the most. All age groups have dropped.
Literary reading by 18-24 year olds has declined by seventeen percent, but there is an
asymmetrical drop among the three youngest groups. Formerly the most active
group, young readers are now the least active, a 13-17 percent decline vs. 2-4 percent
decline in the older groups.
Publishers say revenues are going up but that reflects inflation and rises in price. Unit
sales have been flat for the last ten years even though the population is growing. The
decline in reading foreshadows an erosion in cultural and civic participation. The
survey shows that literary readers perform more volunteer work and go to art
museums.
Dr. Bauerlein considered events in the last ten years that could have caused the
dramatic results. The decline in reading coincides with increased participation in a
variety of electronic media, including: the Internet, video games, portable digital
devices, DVDs, CDs, and cell phones. Television watching is constant while new
technologies have absorbed further leisure time. In 1999, the average American child
lived in a household with 2.9 televisions, 1.8 VCRs, 3.1 radios, 2.1 CD players, 1.4
video game players, and 1 computer.
In summary, American literacy (the inclination to read) is in steep decline. More
than reading is at stake: this decline has cultural, social, civic, and economic
consequences. The literary heritage is endangered. Cultural participation is shrinking
and civic engagement is eroding.
The discussion tended to focus on various failings of the education system and
standardized testing. The results seemed consistent with the NLS experience of
declining readership. The biggest declines, four percent over four years, in younger
readers seems to support the report.
Thursday, September 23
9:00 - 12:15 Digital book distribution
Distribution scenario
Michael Moodie introduced this morning's session as being a wide-ranging
discussion of how we move to the distribution of digital talking books. Following up
on yesterdays's machine transition session, Jerry Ducrest sought the group's input
on some assumptions to be used for the CBM transition model. The current
assumption is that fifty percent of new patrons will want CBMs (thirty percent for
magazines and twenty percent for older cassette books that had not been converted).
For existing patrons, when they receive DTBMs it is assumed that ninety percent will
want to keep their CBM because they are aware of older titles only available on
cassette.
In group discussion, a consensus was reached that ninety percent of patrons would
want to keep their CBMs. One group member suggested that libraries could shape the
response by encouraging or discouraging the use of CBMs. Another factor in
retention of CBMs would be how quickly regional libraries switched to digital
talking-book format for their locally produced books and magazines. Many are still
not sure and will not necessarily have to switch when NLS does. One way or
another, there will be a lot of residual material only on cassette for years to come.
In addressing the assumption of fifty percent CBM usage for new patrons, the
discussion moved to the effects and fairness of offering DTBMs to new patrons first.
Many librarians had problems with this plan. They were sure they would face
strident objections from the many existing patrons who would have heard about the
digital talking-book program by then. It was suggested that NLS increase DTBM
production in the first year. NLS feels that it is too risky to go straight to full
production because too many problems will only be found during the first year of
production. We need to correct initial problems before ramping up.
Another factor on demand will be the availability of non-NLS players. Many people
will buy them if they can. This is a possibility. The manufacturers will have to agree
to restrict sales to eligible NLS readers before we can enable them to use the DRM.
It is expected that there will be two classes of patrons who want the DTBM. There
are the early adopters who already have DAISY equipment. They should be among
the first group to receive DTBMs because they will be good testers and they will
complain otherwise. The second group is vaguely aware of digital equipment, if only
from their children. If they hear that new patrons are getting DTBMs they will
complain. But there will be a huge number who are not aware of digital players and
who do not care about them. Possibly sixty percent of readers would not care that
much. Also we certainly do not want to turn off young new readers who are used to
digital equipment. It seemed possible to balance distribution by being aware of these
differences within the groups. A recent listserv message suggested that DTBMs be
offered to the most active patrons.
Some librarians want guidance from NLS on how to allocate new DTBMs.
Guidelines will help prevent perceptions of favoritism, which was a problem during
the distribution of CBMs.
Some group members thought that most people were expecting full availability by
2008, and they do not realize that the transition could take years. NLS proposed that
the people who are aware of the date could be accommodated. Others disagreed
because they were already seeing a demand and expected that demand to increase.
Publicity needs to be clear that 2008 is the start of a transition and not full
availability. The problems of early demand argue for a hastening of the time line for
machine design, testing, and production. NLS proposed that the 2008 time line is the
best that can be done. It is not a resource issue but rather the number of steps and
tests that need to be done before full production can begin.
Moodie summarized the discussion so far. In response to the group he agreed that
the transition model would have to be revised in favor of giving ninety percent of
new DTBMs to current patrons from the existing fifty percent assumption. The ten
percent for new readers would go primarily to the younger new patrons who should
be encouraged to remain in the program. He thought there was more work needed on
defining the groups that would receive the DTBMs first.
Magazines
The discussion then turned to magazine distribution. This plays an important role in
continuing the demand for CBMs among DTBM recipients. Many of the network
librarians and others thought the assumption that patrons would resist the idea of
returning magazines should be examined, perhaps through a survey. If magazines
were returned then they could be converted to the flash memory cards as well, but if
they were "throwaways," it would be too expensive to consider. Many of the
librarians currently distribute magazines on cassette and expect them back. Patrons
are surprised when they hear they do not have to return NLS issues now.
Additionally, the usefulness of digital formatting and mark up for the magazines was
noted, since most people do not read magazines completely or straight through.
There is concern that even a ten percent nonreturn rate could be too costly for NLS.
Would it be feasible to cut off readers who did not return magazines? Some libraries
do it now.
Digital book distribution
Michael Moodie reviewed the upcoming study on digital book distribution that will
analyze the three options for distributing digital talking books to patrons: mass
duplication, in which copies of every title are sent to the network libraries;
duplication on demand, in which network libraries have nothing on their shelves and
all patron requests are filled by duplication centers; and the hybrid model, in which
the most popular titles are distributed from network libraries and the less active titles
are duplicated on demand at duplication centers. The hybrid model, the most likely,
was prompted by an early look at circulation figures that suggest twenty percent of
the titles account for eighty percent of circulation. The RFP for the study will be out
in October.
The first phase of the study includes building a cost model; suppling values for key
variables that affect costs such as loss rates, turnaround time, and the proportion of
cartridges being reused by libraries; creating screening and evaluation criteria for the
options under consideration; and evaluating the options and making a decision. NLS
sees it as a five-month study and needs a conclusion on the scenario so that the
circulation system vendors can make the appropriate changes to their systems. A
second phase of the study develops the design of the selected system. The study
includes an analysis of mailing containers and labels, which are crucial to an on-
demand type of circulation, because of the need to replace the label for every
circulation.
During discussion, the question was raised on network input to the study. The point
was made that input from this meeting is not the same as having the contractor
getting the input for the study. NLS has planned further coordination with the group
at the end of phase 1, presenting the selected scenario for discussion, along with
design issues. The contractor can meet with the group prior to phase 2. Midway into
the design of the DTBM, they will discuss circulation system issues with vendors.
Many librarians were concerned about network cost figures. NLS does not have the
time nor the resources to do the kind of in-depth study to recreate all of the network
costs. The figures for the network were to be taken from a 1989 study and adjusted
for inflation. The proportions should be close for facilities, staff, etc. Concern was
expressed that some network libraries have been changing the way they operate,
revising the kinds of teams that carry out the work, and that proportional adjustment
of 1989 figures will not give an accurate picture.
There was also concern about the cost and effort of circulating both cassettes and
DTBs during transition. The problem was considered, but NLS believes that the
circulation numbers will remain the same. NLS plans to design the new container
with the same footprint as the current container or smaller so they can be
intershelved. Gaps in book numbers may require random shelving. All agreed that
the shelving solution would vary depending on the situation of each library. There
will be a large one-time effort involved in initially getting the new materials shelved.
Beyond the initial setup, the costs of cassette and DTB service by libraries should not
be additive. There was also the question of how circulation systems will handle
requests for titles that are in both formats. This will be similar to when NLS had FDs
and RCs. Patrons will have the preference of receiving only one format or either,
perhaps with different choices under different circumstances.
The design phase of the distribution study contract will consider: operating
procedures, service standards, NLS policies, detailed resource specifications;
functional specifications for automated circulation systems, duplication process, and
a data distribution system for copies on demand. It will also consider the duplication
process for copies at network libraries, shelving issues, communication interfaces,
telecommunication needs, and packaging and labeling operations. It will include a
transition plan and the development of statements of work, if we choose an on-
demand or hybrid system.
In discussion, librarians were concerned with handling requests when a title might be
held locally or centrally in a distribution center. NLS envisioned that the circulation
system will know where the book is and route the request to the center when
appropriate. The center will not be a backup site. Requests will not be routed to the
center if a book is local but the library does not have a copy on the shelf. The center
(in a hybrid scheme) will not be designed to handle the massive numbers of requests
that would come in that case. Conceivably the library could duplicate it locally.
Libraries will use copy allotment to request an appropriate number of titles and
handle out-of-stock as they do today. Another librarian was dubious about the ability
of a center to get everything out on the day received. NLS is designing a first-in,
first-out service model. The network will not be charged if we can help it. The study
will examine the costs of duplication on demand in greater detail. The cost is
critical. If the cost was three dollars, NLS could not afford it, in which case NLS
would look at getting money from libraries but we realize that is also not realistic.
However, other options might be expensive to libraries as well. Concerning the
ability to prioritize a request, NLS plans on filling all requests in one mail card run,
in which case prioritization would only move it up on the mail run list and have no
effect on actual mailing. There was also concern about the effect centralized
distribution would have on circulation statistics. If figures were cut because they
only showed circulation of locally held books, funding agencies would want to know
why. Requests to the central facility could be counted much as if the book were in a
branch library. This led to discussion of counting circulation for downloadable
books. That issue will be put into the analysis.
Further discussion with the group is needed concerning the allocation of books in a
hybrid system. How are titles categorized, i.e. which twenty percent goes to libraries
and which eighty percent goes to the centers? How is it determined that a title is no
longer popular and should be moved from the libraries to the center? How does one
respond to a sudden high demand for a title? What will happen if a library decides to
keep all copies of books? These questions are appropriate to a hybrid system and
will be discussed after phase 1.
One issue in the RFP is the interface with library circulation systems. NLS will bring
in vendors and independents to talk with the contractor to hammer out the details of
what systems have to know and what data will go back and forth. Conceptually,
think of two printers connected to a system for printing mail cards. One printer is in
your building and one is in the center. Your system knows which printer gets a
particular mail card.
The issue was raised that some libraries may want to abolish their collections and act
as their own on-demand centers. The question was raised as to why we would not
center the on-demand distribution in regional libraries rather than contract centers.
NLS believes that the vast majority of libraries do not have the infrastructure needed.
On-demand centers require duplicating and labeling machines, huge storage servers,
and support staff. Most could not do it, although some will. Some libraries may be
supplemented by very large computing centers with lots of staff and automation
support, but we cannot base the national system on requiring regional libraries to
have such support.
The session concluded outlining the plan. Phase 1 of the study should be complete in
four to five months. Results can be presented at the next meeting. Design issues
may be discussed then, after a scenario is chosen. In the meantime, the group was
encouraged to discuss the issues with their colleagues. Midway through phase 2
there will be discussions with system vendors.
1:15 - 2:45 Analog to digital book transition issues
Michael Moodie and Steve Prine started the discussion with a brief explanation of
how the hybrid system would work. During production, each book will be
designated as being for mass duplication, for libraries, or for on-demand circulation
from the center. This cannot be customized for individual libraries, with some
receiving copies of a title and others using on-demand circulation. At some time, as
a book becomes less popular, it will be declared no longer mass circulation and
libraries will send all copies of the book to the on-demand center.
NLS analyzed circulation data for three large states and found that the top moving
books were mostly the same. When the data was examined by age category of the
book, the heaviest circulation was in the first two years, after which it dropped off.
The data implies that the highly circulated books might form a T-shaped distribution,
consisting mostly of the books produced in the previous three years along with the
small set of perennially popular titles.
The group discussed how titles in a series would fit into this model as readers will
want to read a complete set. Newer series titles tend to create demand for earlier
ones. All titles in a series should be converted to digital but not necessarily mass
duplicated. The less popular (earlier) ones would still be available on-demand. The
system will be designed so that all titles, whether mass duplicated or on-demand, are
equally available. Circulation systems will need to ensure that a preference for books
in series overrides any preference for locally held books over on-demand books.
Moodie solicited some particular input from the librarians. How could the system be
abused? Why couldn't a library get all circulation from the on-demand center and
have no collection to maintain? NLS will decide which titles are in on-demand
status. Only on-demand titles can be ordered from the on-demand center. Some
titles will be on-demand from the start. Others will become on-demand over time
possibly to be determined by monitoring circulation statistics in bellwether regionals.
Will a library have to return the cartridges when a title leaves mass duplication
status? Yes. NLS cannot afford to keep that many flash memory cartridges tied up
on the network library shelves. How do we deal with libraries that might try to
retain copies of books that are called back for on-demand distribution? What about
sleeper titles that unpredictably become popular? Just as now, they are tricky to plan
for. A digital process similar to the tape quota program might be necessary for cases
like this. NLS would need more power in this situation to move copies from one
regional to another. It will be difficult to centrally track. It is possible that the
holding patterns of subregional libraries would change very little. The hybrid system
will be similar to the systems used in states with large subregional libraries.
Returning to the circulation sample, even some recent titles do not circulate very
much and do not need to be mass duplicated. In the recent year's production, 600
titles were circulated ten or less times. The critical question for analysis is how to
define a cutoff for on-demand titles. It will be even harder to decide when to pull a
title, which is why we may need to use bellwether libraries to collect data on
circulation. While the duplication center can work as fast as a regional, it will be
costly for NLS to circulate popular titles on an acceptable scale. NLS can only afford
duplication of low-demand items because the processing costs are lower. And it
represents a small shift of burden from the states to NLS. The librarians reiterated a
point from earlier sessions that they needed a way to do duplication-on-demand for
walk-in patrons.
Could local libraries skew their selections towards the duplication-on-demand?
Patron interest in new materials, prompted by TBT and the annual catalogs would
prevent it.
What about the evaluation criteria for selecting a model? In earlier meetings the
group said that a duplication center should take the same time as a regional to get
material to the patrons. But that raised two issues: time to get into the mail stream
and time in the mail stream. It seemed to some that it would take longer in the mail
stream from a distribution center than from a regional library but some suggested that
this is not necessarily true either. We could stipulate that a book must be mailed
within twenty-four hours or by the next business day, since the pick up time is
determined by the postal service. It might be necessary to limit the number of on-
demand items a patron could request at one time.
Local systems will need to be updated when books are returned so that more books
can be sent. Vendors will discuss how this will be handled. Libraries will be
notified when a book is issued to a patron from the distribution center. Some
librarians were concerned about books getting into the wrong container and going to
the wrong place. NLS might want to add an ID with the library label for mass
duplicated books so the center will know where to forward misdirected titles. Some
suggested a tactile symbol to label a cartridge as belonging to a regional or to a
duplication center. Libraries should also be notified of wrong address returns.
The group returned to issues of the transition. If it is to be a hybrid system then the
libraries must be "frontloaded" with a subset of the 20,000 DTB titles that will be
available. Possibly, NLS would send the production of the most recent three years
minus those designated as probable low circulation. The number of copies per title
would be initially small, but would increase as the supply of DTBMs increased. At
some point, production of cassettes will stop. It may be difficult or costly for NLS to
continue producing cassettes until everyone who wants a DTBM has one. There
might be a period of two years when no new books were produced on cassette but not
everyone had a DTBM. The librarians were concerned about this. It will not be easy
to tell a patron that they can only have a CBM but no new books for it.
3:00 - 3:45 Illinois libraries' new technology initiatives
Lori Bell reported on a number of experimental projects at the Mid-Illinois Talking
Book Center and elsewhere. The eAudio and the Lobe Library projects focused on
using Audible.com titles that were loaded onto Audible Otis players and circulated.
The readers like the portability and sound quality of the Otis, and the availability of
popular titles. Many disliked the small controls and ear buds and missed the variable
speed control available on other players. Library staff found loading the players time
consuming. There is also no way to purchase a license to deliver content to library
patrons over the Internet. They have to be loaded onto a device. The Lobe Library
project expanded the test to five talking-book libraries who share an Audible.com
collection.
In the OverDrive/Adobe project, volunteers evaluated Adobe e-books with the Read-
Aloud function. They were checked out over the Web and downloaded. Nearly all
of the volunteers preferred their own screen reader to the built in Microsoft speech
synthesizers. Some titles are not enabled for read-aloud by the DRM. Adobe's e-
book reader was not initially easy to use and readers wanted better training and
documentation. The volunteers would prefer to use the Adobe books on the various
commercial talking-book players. Adobe needs a better speech engine for readers
who do not have their own screen reader.
The Mid-Illinois Digital Talking Book Project: 150 volunteers are testing a variety of
mainstream digital audio players and digital talking-book players with different types
of content. The experiment is ongoing but Bell discussed preliminary results from
testing various players. Several libraries share a collection and a set of Otis players
(small mp3 player not designed for visually disabled people). Users love the
portability and size but dislike the controls and lack of variable speed control. The
new Nouveau, a flash-memory device for mp3, is the premier player now. One
problem is that none of digital players play Windows Media Audio (WMA), but
Overdrive and Netlibrary are looking at that. People are requesting that Bookshare
enable WMA and other formats. Some commercial players are acceptable, but NLS
patrons prefer features on devices designed for visually impaired people and the
ability to handle multiple file types (text, audio, audible mp3, etc.)
The Unabridged Project: Four state talking-book libraries will share a digital
audiobook collection on the Web. The content is formatted for WMA and can be
played on the computer or device that plays WMA.
The purpose of these projects is to make people aware of what else is out there. They
are letting people try Bookshare by being an institutional member, a new account
type. The plan is to buy books in bulk and download them but they have to buy and
download for a specific customer. So if someone wants a book, we download it in
various formats. That has resulted in people joining Bookshare because of books and
formats available. A trial was done with Adobe e-book reader. People prefer human
voice, but are excited nonetheless. Overdrive has a system so we can download with
a library number. The read-aloud function was tested, but not all books are enabled
for read-aloud. But for those that are, the patron can use the book with their own
synthetic speech software. There is a steep learning curve, but they liked the books.
Bookcourier and Bookport are also creating excitement.
During the experiments, use of the iVocalize Talking Communities software to
conduct online orientations was successful. The sessions are recorded and available
to volunteers unable to attend the original presentations. Tip sheets on the players
and various content sources have also been posted on the Web.
Responding to questions from the group, Bell reported that most participants are
elderly, but that age did not make a difference. Interest in technology and new players
seemed to be more important than age. Adobe is somewhat accessible, but not easy.
The software takes some learning. One of the readers did a tip sheet that helped
people. They are hoping Adobe licenses a better text to voice speech. They have not
tried MS e-book reader. Users dislike Microsoft SAM. Overdrive books are human
audio, commercially narrated. They now have 800 to 1,000 titles. They are getting
demand from mainstream libraries as well. All books are unabridged. Demand is for
unabridged books. Technical support does take time.
3:45 - 4:30 Graphic audio: access to entertainment media
Jamie Cutting, president of Potomac Talking Book Services, introduced members of
his staff who discussed their new product line called Graphic Audio. The new
product line consists of unabridged recordings of adventure and science fiction
published by Harlequin Entertainment, with related Web content such as
downloadable objects, games, ring tones, and the cover art available as screen savers.
This approach to recording is seen as an entertainment experience more than a
straightforward book experience. It revives the spirit of the golden age of radio:
dramatic readings, multiple voices, music, and extensive use of sound effects.
The product will be available on compact disc, but Potomac plans on going to
Internet distribution as soon as possible. They would be downloaded in 600MB
sections, with usually about ten sections to a book. They use the WMA format. They
thought the product could readily be adapted to the DTB format.
While blind individuals are not being specifically targeted, they are seen as a
potential audience. Potomac Talking Books is interested in the group's reaction and
how the product might be received by and promoted to visually impaired individuals.
Would there be interest in the blind community for products like these? Distributers
will sell them for $14.99 on CD, cassette, or as downloadable WMA files. It was
suggested that this is analogous to descriptive videos, and they sell. There is a niche
within the blind community, but they are used to free books, so it may be harder to
sell. The Destroyer series is on Bookshare, so there is some experience. There are
some blind operated web sites in Canada with such material in various forms. So
there is an interest. Given the price range and the length of product, it is a good
value.
The librarians in the group said there might be interest in purchasing this product,
especially for younger readers both by libraries and by individuals. Lori Bell said she
would be interested in making this available through a project as there was a lot of
interest in old time radio. A promotional piece directed at librarians might be useful.
Friday, September 24
Kurt Cylke began the session by distributing a New York Times article on flash
memory. He also mentioned that the Plextor proposal on streaming and download of
DTBs was received. NLS may or may not work with Plextor, but we will have this
capability by 2008.
9:00 - 10:30 Network library role in Web delivery
Steve Prine led the discussion. He noted that NLS has a proposal for distribution
using streaming audio. NLS will have a complete program by 2008, but we hope to
have more than a pilot long before 2008 for students and early adopters. NLS will
also have Web distribution. NLS wants one site that is a portal for all books. The
question is: what is the network's role in this, especially in the area of technical
support? So far network involvement in Web-Braille is not high. They do not
provide much in the way of user support, other than answering basic questions about
passwords, etc. Part of the reason is that there is no consistency in the level of
knowledge at the network level. NLS is looking for guidance on this issue.
There was a suggestion that the ALA standards be revised to include technical
support. For several reasons, staff in network libraries can only be expected to do
very basic technical support. With just a handful of users, a library could not justify
having a position for technical support. It would be different if forty percent of
patrons were downloading books. There is also a recruiting and reasonable pay
problem. States are not going to increase the pay scale to support this. Network
libraries already had a problem providing minimally qualified readers advisors. The
ALA standards are just guidelines and the state libraries are in a fiscal crisis.
The technical support at the Mid-Illinois Talking Book Center was considered
relevant. They pay for a person to do technical support as well as doing some
locally, but they have only one person on staff who can help. Without such contract
support, a library could get inundated with questions and problems. Another
approach is for a gradual rollout so that the patrons themselves can help each other.
In the Mid-Illinois projects, some patrons who received the online training took on a
mentoring role and enjoyed helping others.
Most libraries have Web OPACs and are having to provide technical support for it,
but many librarians do not use assistive technology and cannot support it. Those on
the staff who use assistive technology have to be the ones who provide support. So
we will need to think about getting assistive technology specialists on board. One
library tried to train nonuser staff to provide help with assistive technology and it was
a disaster. Now the librarian, an assistive technology user, does it herself.
The idea of recruiting patrons to help others, possibly by maintaining a volunteer list
in the state, was supported. That might work, but people are not competent with all
tools, most only use a few, so they may not be able to answer questions about other
screen readers for example. And the combinations of equipment and software are
endless. It is bad enough for braille. With audio it will be exponentially more
complicated. And while training is important, you need a manual to get people
started.
With everything getting more technical, libraries have to ask what they can do,
otherwise people will say that they are adding nothing to the process as the
operations become more technical. This situation is similar to the mid-nineties,
when patrons wanted help with the basic software available using public access
computing. Libraries entered programs to bring up the level of knowledge with some
success. If the Gates Foundation committed to provide training for technical support
(not the technical support itself), they would probably do a wonderful job. They
could recruit blind users as trainers. But who would take the lead?
Perhaps refocusing resources on technical support expertise could be seen as part of
the move from manual workload to automated support. Since downloading takes
away some of the expense of mailing, you are shifting costs to technical support and
it may not be more expensive in the long run. Mid-Illinois had relevant experience
with a cooperative technical support program. They brought up their own experience
first with an electronic list and with an online chat room (iVocalize), then added
participating libraries. Each participating library staffs the chat room for ten hours a
week so users can go there with their questions.
The Massachusetts regional library is just finishing up a grant to do an online training
course for the Keystone OPACs around the country. It uses an iVocalize chat room.
Anyone using the catalog uses a computer and so they should be able to use the chat
room. The web OPAC course will be free to any patron of a KLAS library in the
country. There is a lot of potential for that kind of training. It minimizes staff
involvement but creates a resource for patrons to use. She recommended creating a
subcommittee on training, saying it could not be solved today but that we needed to
take some responsibility.
With the idea of a Gates Foundation support grant, a cooperatively staffed chat room
has potential. A cadre of staff available would make it more likely that even very
specialized questions could be answered. It could be like having a multilevel system
of support with the real experts only tapped occasionally. This could be a topic for
the next conference.
The group then turned to the subject of streaming audio. It was said few people
would want to be tied to a computer to listen to a whole book. Would there be a way
to bookmark where you left off reading? Yes, that can be done. Also, are there
many ways to rebroadcast a stream to other players? One user said that he was
comfortable with streaming especially since it was only the first step.
It was asked if libraries could promote the online books as a way to stem demand on
the digital talking-book players? This led to a consideration of the statistics on NLS
readers who have computers, whose numbers of course are growing. Libraries could
target promotion of the online books better if they knew which patrons had
computers. Some OPACs would tell you who was accessing the system. Local
OPACs will soon have MARC records with links to electronic resources. These
might provide information on patrons with computers. Another suggestion was to
use narrators to do promotional welcomes on web sites or using streaming media for
book samples.
10:45 - 12:00 Open discussion
Digital rights management (DRM)
Michael Moodie opened the discussion by announcing that three NLS staff members
would be attending a DAISY Consortium Group meeting on digital rights
management the following week. We will start work soon on a DRM system for
NLS books. We are looking to this group for ideas about the DRM system from the
perspective of users and librarians, things the system should do and not do.
Obviously, it has to protect copyrighted material from getting into the hands of
noneligible users. Can we get a clear statement of requirements?
There were more questions than suggestions. It was noted that the DRM should be
invisible to the end users. The group wanted to know more about the DRM system
and how it would work. It was explained that NLS machines will be able to decrypt
and play NLS books. It will be trickier to authorize after market devices, such as
Plextor and VisuAide that have more bells and whistles, to play our books and to
allow our machines to read locally produced network books.
Our player will play unprotected books in standard formats such as wav or mp3. For
a network library to produce material to be played on an NLS machine, it could
ignore protection altogether. But if protection is needed, it would have to have a
production tool that will add encryption. Such production tools are available and can
be bought from vendors. The tool needs to be easy for users and producers
(volunteers, etc.)
The system will be based in software, not hardware. It will probably work around a
key system with NLS having its own authorization key and network libraries doing
their own production having another. The players would work with both keys.
DTB/Cassette transition
Michael Moodie led a return to the discussion of the DTB transition. How many
titles would be sent to network libraries initially under a hybrid distribution system?
We thought it would be a subset of the 20,000 DTBs we expect to have in 2008. If
the 20/80 rule holds up, then it would be 4,000 titles and possibly seven hundred
titles a year while we continued retrospective conversion and perhaps four hundred a
year after that. Titles will be sent back to the distribution centers as circulation drops
off so a regional might only have 6,000 to 10,000 titles at any time. The 20/80 rule is
not fixed. Distribution will depend on our experience with circulation.
Will NLS convert more popular titles at first? The collection development section is
trying to select titles for conversion across a range of subjects and genres. They do
not have circulation statistics, but they have a sense of what is popular. Will libraries
have input into what titles are distributed? NLS is thinking that copy allotment
might make sense as a way to do that, with highly requested books going for mass
distribution. Is there a correlation between the copy allotment selections and
subsequent circulation? There seems to be, but there are always misses, like The Da
Vinci Code. But the 20/80 rule should not apply to a monthly basis. Some months of
copy allotment are loaded with popular books and other months have mostly low
circulation books.
The comment was made that the conversion to digital was similar to putting in an
integrated library system. Every process in the library has to be opened up,
examined, and put together again. It is a highly complex process.
Discussion returned to the problem of the one- to two-year period during which, as
was suggested the previous day, the DTBMs had not been completely distributed but
NLS would not be producing cassettes. That could be a significant fairness problem
and possibly a public relations problem. Those without DTBMs will not get new
material.
Looking at the three scenarios in the study, the DTBM ramp-up figures seem low.
While there are good reasons for a low run in the first year, NLS should look at
increasing production during the next years. That might resolve the fairness issue as
well as resolving some frustration over the rollout length. NLS responded to the
production question saying that the figures from the study were best guesses of what
might reasonably be done. We hope to get manufacturing done in the United States
so we can have more quality assurance visits. U.S. manufacturers are not geared to
the mass production they can do in the Far East. Doing a short run at high volume
for a year may not be realistic, but if we are able to, we will. We agree that a shorter
rollout would be better for public relations and an easier transition. Is it impossible
to get a U.S. manufacturer to make more than 150,000 a year? It took years of effort
for NLS to get a second CBM manufacturer. There is not a lot of capacity in the
United States for this type of electronics product. Further, we are prohibiting the
design contractor from participating in a production contract to lower the risk of a
vested interest. Companies that could do more might not be interested in our
relatively small scale production. Can the libraries turn over their patron base in a
few years? Librarians thought that it was possible. Colorado sends out about 100-
150 machines a year. A three-year turnover is about twice the normal rate of patron
turnover.
How do we deal with clients who cannot get new titles because we cannot give them
a DTBM? Perhaps libraries should target the message if they are not certain they can
fulfill the need generated by over-advertising the DTBM. Planning must occur at the
COSLA level. What they tell the legislature is told to the world. Perhaps we could
delay sending DTBMs to downloading patrons to reduce initial demand. That is
probably not possible since they would be the most interested in having a machine.
NLS agreed to revisit the topic of ending RC production. The decision was not solid.
A gap is unacceptable.
Other ideas for the transition included keeping advanced lists of patrons who want
the new machines, or putting people who request titles not available on cassette on
such a list. Perhaps we could try getting outside funding to ramp up initial
production, but outside funding tended to jeopardize our more routine funding and
are not worth that risk. Could NLS provide cassette masters for ILL distribution?
The meeting concluded on the positive note that the digital effort was going on as
planned. Members commented that these were some of the most thoughtful
discussions they have had so far. They find the information is useful in defending
NLS against unfair or shallow criticisms.
The next meeting will focus on the distribution study. The contractor will outline a
basis for making the distribution decision at the end of phase 1. NLS also plans on
having more information about the distribution ratio (how many books should be
held by regional libraries versus the distribution centers.)
|