Memory of the World Programme
Memory of the World Programme
The members of the Sub-Committee on Technology wish
to express their thanks to those people that have helped with
advice and comment on this document.The final conclusions and
wording of the Recommendations are, however, the work of the Sub-Committee
and our advisors should not share the blame for any errors that
may be contained within this document.
It should also be noted that the Recommendations
apply to today. They concentrate on digitisation and its potential
for improving the access to and preservation of documents of all
kinds including manuscripts, texts and sounds and images. The
aim is to encourage the exploration of the advantages offered
by digitisation whilst making it clear that digitisation is not
a panacea for all the problems facing collections.
In particular, the Sub-Committee on Technology wish
to make clear that the traditional conservation methods used to
keep documents in a useable condition will not be replaced by
the commencement of a programme of digitisation. To complete the
process of digitising a collection will frequently take many years
and the preservation programme for the original documents must
continue. The digitised copy will, in most cases, not be a replacement
for the original but will form an access copy acceptable to the
majority of the users thus helping to reduce the stress on the
original document.
Access copies provided by micro-films and photographs
may also still be required if the quality offered by currently
available digital technology is inadequate.
Technology moves on, however, and the standards of
digitisation are constantly improving. These Recommendations have,
therefore, to be seen as a draft which will require regular review
and updating.
Various scientific prefixes denoting magnitude
are used in this document. A short note on the hierarchy is given
to help those not familiar with the terms. Each prefix is 1000
times larger than the term above. The prefixes are combined with
Bit - a single binary digit - or a Byte - a group of eight binary
digits.
| Kilo | One Thousand | |
| Mega | One Thousand Kilo or One Million | |
| Giga | One Thousand Mega or One Billion | |
| Tera | One Thousand Giga or One Trillion | |
| Peta | One Thousand Tera | |
| Exa | 10 | One Thousand Peta |
1. Summary
The Sub-Committee on Technology to the International Advisory
Committee for the Memory of the World Programme:
Having considered the advantages offered by the various technologies
currently available;
Being aware of the increasing demand for access to documents;
But also being aware of the risk of damage to documents from over-use;
Being aware also of the need to provide preservation copies of
many endangered documents;
Taking account of the fact that many analogue data carriers have
a restricted life span and will require the information to be
copied in order to preserve it for the distant future;
But also taking into account the fact that copying in the analogue
domain inevitably leads to a loss of quality;
And taking account of the fact that not all documents are currently
suitable for copying to a new medium;
Suggests that a programme of digitisation of documents is the
best compromise between the conflicting demands for more access
to collections and for greater protection for the documents.
The Sub-Committee on Technology therefore recommends that the
Custodians of all collections examine closely the benefits for
their collection that may flow from the creation of digital copies
of the documents.
Up
2. Introductory Comments
a. Limitations in the Recommendations
The technical needs of a catalogue for a digitised collection
have not yet been agreed. It is clear that a catalogue for a digitised
collection can offer more than the traditional catalogue, particularly
with regard to images and sounds. It can, for example, also incorporate
an active system of pointers to related material. Discussions
by documentalists that lead to a definition of the catalogue requirements
would enable the Sub-Commitee on Technology to include the catalogue
needs in the technical recommendations.
b. Restrictions to Digitisation
The Sub-Committee on Technology recognises that some types of
documents are not appropriate for digitisation at present. The
reasons vary. For example motion picture films require more digital
storage space than technology can provide at an economic cost
at present and large maps and plans will not fit into existing
scanners.
In addition, there are documents that have a value beyond that
of the information that they hold. The artistry of illuminated
manuscripts and the historical significance of documents such
as the "Magna Carta" or the "Declaration of Independence"
take them beyond the realm of simple information carriers. Similarly,
more modern carriers such as art photographs, films and sound
recordings are valued objects in many collections. A digital copy
of these documents will not represent all aspects of their appearance
and character but will, however, reduce the level of stress caused
by over-use.
c. Urgency
A new storage system is required very urgently in parts of the
sound and moving image fields. Each day, more carriers decay beyond
the point of restoration. A slow "Fire of Alexandria"
is in progress. Books, manuscripts and prints are also decaying.
It is no longer a question of "If" information carriers
will be lost but a question of "When". Some carriers
will survive for thousands of years. Many are only lasting for
thousands of days. To preserve their information it has to be
copied.
d. Preservation of the Original Documents
The Sub-Committee stress that the introduction of a large-scale
digitisation programme should not be seen the pre-cursor of a
programme of scrapping the original carriers. For many collections,
digitisation will be primarily to improve access. The digital
copies will assist the preservation of the original document by
reducing the number of requests for access to it and, therefore,
reducing the stress on the document.
Not all documents can be copied to a digital format with sufficient
quality yet. Small inaccuracies and losses may occur. The original
documents will be required at a future date when technology offers
the necessary improvements in the quality of transfer and digitisation.
There may be documents, particularly in the sound and moving
image areas, where the disposal of the original material after
digitisation is sensible because the original carrier is too decayed
to be preservable. This is not, however, the norm.
e. Recommendation of Digitisation Technology
The Sub-Committee is not able to make a firm, single Recommendation
as to the type of digitisation equipment or the storage medium
to be employed. This is a matter for investigation by the collection
Custodians when assessing the requirements of the individual institutions.
Some of the points to be considered are given below, as are some
minimum technical standards that equipment and the transfer process
should meet. The members of the Sub-Committee on Technology are
prepared to assist with advice on specific questions and to recommend
specialist advisers for projects that require more in-depth assistance.
The technical situation is continually changing and the Sub-Committee
on Technology intends to keep the situation under review.
Up
3. Digitisation
a. Why Digitise?
The pressure for access to documents is increasing and this
demand, in turn, increases the risk of damage, both gradual and
catastrophic, to the document. Although many types of damage can
be repaired, the expense involved can be considerable. In cases
of severe damage, particularly with sound and moving image carriers,
the document is effectively destroyed.
By digitising a document, the access needs of the majority of
users can be met by using the digitised access copy. Many users
wish to examine the content of a document and do not need to examine
the original carrier.
In addition, for many documents it is possible to make a higher
quality digitised copy that, for many documents, can form a safety
back-up copy of the original.
b. The Advantages of Digitisation
The Sub-Committee on Technology believes that the main potential
advantages offered by digitisation are:
i. Copying - the potential to make copies of digitised information,
using the same storage format or another digital format, without
loss of quality.
ii. Automation - not only of the play-back of items requested
by researchers, by using robotic storage systems, but of the process
of making copies. As the document is represented by a string of
binary numbers, it is possible to automate the process of copying
and even to remove the need for human checking of the fidelity
of the copy.
iii. Auto-Checking and Repairing - the possibility of automating
the copying of digital information leads to the potential auto-checking
of the condition of the collection and, if excessive errors are
detected in a digital carrier, making a corrected copy of the
material without human intervention.
iv. Searching - digitization offers the possibility of being
able to search both local and remote catalogues and to create
a web of links or pointers from the document accessed by the user
to associated documents in the same or other collections. In addition,
full text searches can be made. Similar search techniques are
being devised for sound and image documents.
v. Access - the ability to send a digital signal over communications
networks without loss of quality. The use of large scale robotic
stores will enable the digitised collection to be accessible for
24 hours a day with the minimum of staffing.
vi. Speed of Copying - within the digital domain, it is possible
to accurately copy or transfer data at a very high speed. The
provision of a hard-copy to a researcher can be performed swiftly.
The future migration of collections to new carriers will be much
faster than the initial movement to the digital domain.
vii. Quality - the potential to digitise an information carrier
at a very high resolution or at a lower resolution as required.
It is also very easy to generate a lower quality copy from the
higher quality copy when required.
viii. Space Requirements - the density of storage of digitised
information on the carrier can result in a major reduction in
the shelf space required. This in turn reduces the space that
requires archival climatic control with a consequent reduction
in running costs.
ix. Future Migrations of the Collection - if a collection is
copied to an analogue carrier, future migrations of the collection
will cost (subject to inflation) the same as the current migration.
If the collection is copied to a digital carrier, future migrations
can take advantage of the possibilities for automated transfers
inherent in a digital format (see 3.b.i. and ii. above).
Migration of collections of digital data is not a new phenomenon.
The banking world, for example, has successfully migrated its
data banks several times. The experience of banks can give useful
guidance when faced with the migration of textual, visual and
aural information.
c. Disadvantages of Digitisation
The major disadvantage perceived by many people is the cost.
This can be summarised under four main headings:
i. Capital Costs - the equipment required to carry out the digitisation
process can be expensive to buy and often requires skilled operators
if the best results are to be achieved.
ii. Storage Requirements - it is often considered that the carriers used to hold digitised information will require a very clean and climatically stable environment with a consequent increase in the energy consumption of the collection. There would also be a capital cost in the creation of any such controlled storage area.
iii. Running Costs - there are fears that a digitised collection
may itself require frequent copying with the consequent labour,
energy and new carrier costs. A safe life of only two to three
years for digital information stored on magnetic tape and of three
to five years for optical discs is feared.
iv. Preparation Costs - before capturing the document, it is
necessary to prepare the material. In addition to any physical
preparations that may be required, the preparation includes the
ordering and indexing of the original material and the entry of
the textual references into the data base by specialist staff.
This can be an expensive process. The manpower requirement to
copy the existing carriers to a new carrier is seen as being very
high.
The Sub-Committee on Technology discussed these points and made
the following comments and suggestions:
i. It may be possible for a group of smaller archives and libraries
to share a set of digitisation equipment. The process of digitisation
may, after the initial major capture process, be a more intermittent
operation for smaller institutions. For larger collections with
an active acquisitions programme, the possession of their own
equipment will be a necessity.
The capital cost of the equipment required for a programme of
digitisation has also to be set against the potential savings
in preservation costs. Savings can be made by concentrating the
preservation resources of the collection on the documents in the
collection that have a value beyond that of simple information
carriers. For many decaying carriers of information, only digital
storage provides the means of preserving the information in an
undistorted form for millenia.
ii. The optical discs and magnetic tapes used by the currently
available digitisation systems do not require extremes of cleanliness
or climatic control. If an area constructed to high standards
of cleanliness and climatic control is required, it will be very
small in comparison to the area of traditional shelving.
iii. Early computer tape storage formats lacked the sophisticated
error protection systems used on modern formats. In addition,
the tapes were not in a cassette or cartridge format and consequently
suffered damage if carelessly handled. The banks and other institutions
using these early magnetic tape formats copied the tapes at frequent
intervals to avoid errors overwhelming the simple correction systems
available at the time.
Many television broadcasters are currently copying their old
video-tapes, dating back to the late 1950s, to digital carriers.
This copying is being done because of machine obsolescence and,
in some cases, because of carrier decay. Instability of the magnetic
particles containing the information is not a factor.
The broadcasters have carried out much research into the long-term economics of copying their collections and have concluded that a modern, digital video-tape format offers the best solution for the preservation of their programme material. The same tapes and error correction systems can be used for the storage of other forms of information including texts, still images and sounds.
Factors such as frequency of use and the storage environment
will affect the life-expectancy of tapes. Frequently played tapes
will have shorter life than the norm but these are expected to
be relatively few in number. For very popular items, duplicate
copies can be held.
Optical discs in the form of commercial CDs have been in existence
for over ten years. The number of cases of information loss have
been very small. Those cases that have been reported were mainly
due to a corrosive ink being used to print the label information
on the discs. This led to contamination of the digital data. The
problem was quickly detected and solved.
If a robotic storage system is employed, the costs of checking
and copying the carriers can be greatly reduced. The facility
of auto-checking for errors without human involvement is already
available on some robotic systems. Auto-copying without human
intervention is also feasible. The only cost then is the new carrier
required for the copy.
iv. Material to be captured on micro-film also requires preparation
and restoration before being photographed. The material to be
digitised often does not require physical restoration before digitisation.
In many cases the "restoration" can be performed more
effectively, and cheaply, in the electronic domain when the information
is accessed. This also permits the use of future improvements
in forensic techniques of document examination which may be impossible
to use on a physically restored document. Degradations such as
lack of contrast, ink bleeding, blotches, stains etc. that have
occurred during the life of the document are part of its history.
The "restoration" of these degradations in the digital
domain rather than by permanent, physical methods is highly recommended.
The skills required by the operators of the digitisation process
are of the same level as the skills required by operators of a
micro-filming process. Some re-training will be required but the
staff-cost element should not change greatly. A degree of automation
of the capture process may also be possible.
d. Cost and Quality Factors in Digitisation
To copy from a traditional analogue carrier to a new carrier
- either analogue or digital - is expensive. Trained staff, with
the ability to monitor and adjust the process, are required to
ensure that the best transfer quality is achieved. The technical
quality of an analogue copy made from an analogue original is
not as good as the original and the quality steadily deteriorates
with each succeeding generation of copying.
This deterioration may not be obvious with some types of information
carrier because the intrinsically high level of redundancy in
the information. This is particularly true of printed texts where
very large errors are needed before the text becomes unreadable.
Other types of information carrier suffer obvious quality losses
after minor deterioration.
To copy from a digital carrier to another digital carrier offers
the possibility of automation for both the mechanical parts of
the process and the monitoring. The technical quality of a digital
copy made from a digital original is identical to the quality
of the original.
e. Access and Preservation Copies
During the discussions of the Sub-Committee on Technology, it
became clear that a fundamental difference existed between the
texts and still images working group and the sound and moving
image working group. The first group sees digitisation as, primarily,
an access tool. The second group is seeking a tool to make high
quality facsimiles for preservation purposes. This does not negate
the Recommendations of the Sub-Committee; on the contrary it strengthens
the need for the dialogue. Improvements in both access and document
preservation can be provided by the same systems producing digital
copies at different levels of fidelity.
For some collections, a mass produced storage medium such as CD-ROM, which can, if desired, be physically distributed to the users, will be the best solution. CD-ROM and CD-R offer a reasonably inexpensive entry to digitisation and the discs can, as the size of the digitised collection increases, also be used in a robotic jukebox to improve the speed of service to users.
For other collections, a large capacity, robotic store offering
the possibility of remote access via networks is the best solution.
The decision as to what is the best solution will vary from collection
to collection. In some cases, both the mass-produced carrier and
the robotic store will be required.
f. Limits of Digitisation
As previously discussed, there are two main limitations to the
digitisation process:
i. The very large digital storage requirements of some types
of documents. In particular, moving image documents and large
collections of high quality photographic stills. Storage technology
is still improving and the unit cost of digital storage is dropping
while, simultaneously, the capacity of stores is increasing.
ii. The definition of a digital copy made using current technology
may not be adequate to provide the high level of fidelity required
by some types of documents.
g. Dangers of Digitisation
From the above comments it can be seen that digitisation is
not currently viewed as a universal preservation tool. The transfer
of large numbers of documents will, however, aid preservation
by reducing the pressure on the original documents.
The creation of a digital access copy does not mean that the
original document can be neglected or disposed of. The funding
bodies of institutions must be made clearly aware that digitisation
is, for most documents, a method of granting more people access
to the information (by means of a high quality copy) without increasing
the dangers of damage to the original document, with the consequent
high restoration costs.
Up
4. Standards
a. Technical Standards
It is possible to make a digital transfer at several different
standards of quality. A decision as to the optimum technical
standard of digitisation for a document must be made as part of
the process of examining and preparing a document prior to digitisation.
For documents that can be stored as a still image, the following
levels can be identified by examining the information to be digitised.
1. Many modern text documents require only a basic bi-tonal storage
standard - a digital photo-copy. This would include many books
and typed documents.
2. Documents that include mono-chrome images and photographic
stills require a grey-scale storage standard - the equivalent
of a black-white photograph - to adequately digitise the document.
3. Documents that include colours, such as illuminated manuscripts
and maps, and items that may require forensic type examination
require full-colour digitisation.
Sounds and moving images, by their nature, require different
standards definitions but in many cases these are already defined
by recognised international Technical Standards.
b. Usage Standards
Custodians of collections must decide what purpose any digitised
copies are to fulfil.
Basic Access - available for access by all with the original
document available for examination only by scholars and researchers.
This is a copy that will satisfy the needs of 95% of users.
Intermediate Access - may be required by some collections,
for example to provide more protection to an important or popular
document by restricting access to the original document even
for researchers and scholars. The Intermediate Access copy will
satisfy the needs of 99% of users.
Facsimile - the replacement of a severely decayed document will
require a preservation standard - as near a facsimile of the
original as possible. Accepted as a duplicate of the original
document and providing a safety or preservation copy of the document.
In addition, a "Catalogue" or "Browse Mode"
standard may also be required. This would, for example, enable
sounds, images or illustrations to be accessed by using low quality
copies as a search tool.
c. Optical Character Recognition
The use of Optical Character Recognition (OCR) systems in place
of bi-tonal images for texts was considered. OCR systems have
several potential advantages. The main advantages are a decrease
in the digital storage space required and the possibility to search
the text directly without the need of a key-word catalogue. At
present, the reliability and usefulness of OCR software is limited.
Work is still proceeding and commercial developments, such as
systems that can recognise a small range of handwriting and convert
it to type, are signs of the increasing commercial interest in
this area. If the systems become sufficiently reliable, the use
of OCRs would create a fourth technical standard for the capture
of textual documents.
d. Data Compression and Data Reduction
Data compression (loss-less data compression) is acceptable
as a part of the capture process if a lack of storage space makes
its use essential. Data compression is a reversible process: the
signal is fully restored after de-compression.
Data reduction, however, is not reversible. The disadvantages
of a data reduced copy may not be realised until some future date
when manipulation of the digital copy is attempted for special
purposes eg. to enhance certain colours to help detect watermarks
or writing that has been partially erased. Data reduction is only
acceptable if no loss of critical information that may be required
by users can be detected. Data reduction can be a powerful and
essential tool in the provision of access and dissemination copies.
It is, however, unacceptable in all processes related to preservation.
Up
5. Conclusion
The Sub-Committee on Technology recommends that the digitisation
of documents be considered by all Custodians of collections as
a method of providing improved access to the documents in their
care without simultaneously increasing the risk of damage to the
collection. Digitisation should also be considered, where necessary
and/or possible, for the creation of preservation copies of the
information.
The advantages of large scale digitisation programmes greatly
outweigh the disadvantages. Those groups of documents that are
not suitable for digitisation because of shortcomings in the equipment
and processes currently available will be monitored by the Sub-Committee
on Technology and further recommendations made at intervals.
Up
ANNEX A
Technical Standards
The Sub-Committee on Technology has discussed at length the technical
standards that should apply to digitisation of documents. This
is a complex subject because of the number of official and de-facto
standards currently in use. The figures quoted are given with
the knowledge that they will not meet with universal approval.
It is hoped that the following suggestions will generate a wider
debate, that will, in turn, lead to a general consensus.
The figures quoted are an attempt to tread a path between what
is in common use and what is desirable. In some cases a technical
standard that is at the leading-edge of current technology is
suggested. This is done with the experience of previous "leading-edge"
technologies that have become industry standards.
1. Texts and Still Images
For textual documents and still images, digital copies will normally
be made to improve access. Copies of endangered documents may
also be made to a higher standard to provide a safety or preservation
copy. There will, however, always be a residual group of documents
that can never be considered as suitable for preservation by digital
methods. Any digital copy of these documents will be an access
copy only. The reasons for their omission are not normally technical
but usually cultural and historical.
The sub-committee has proposed standards which it considers the
minimum required for Intermediate Access copies. However, where
circumstances require custodians of collections to treat digitisation
as a preservation process for endangered material, image capture
must be carried out to full facsimile standard, using the assistance
of curatorial and technical experts to judge the appropriate technical
standard of image capture, which will frequently be much higher
than the minimum standards quoted here.
The main parameters that affect the quality of digitised copies
of books, manuscripts and other paper-based material are the resolution
of the image and the range of the grey-scale or colour-scale for
each pixel. The figures that follow are the minimum value for
each parameter for each group of materials. Some documents may
require higher values of resolution and grey/colour-scale. Examples
include texts with very fine detail or small characters. The figures
given are for Intermediate Access copies. Copies being made to
Facsimile standards for safety or preservation reasons will require
higher values.
Flat documents may be captured by flat bed scanners. Documents
that are not flat and old books with bindings that cannot be opened
fully should be captured by a digital camera.
a. Modern Texts
i. Printed Materials, Typed Documents etc
Capture parameters : 200 dpi, bi-tonal (one bit per pixel)
Compression : CCITT G IV
File Format : TIFF 6.0
ii. Documents Containing Black and White Photographic Plates
Capture parameters : 100 dpi, 8 bit grey-scale
Compression : JPEG loss-less Compression
File Format : TIFF 6.0
b. Manuscripts and Early Printed Materials
i. Monochrome Material
Capture parameters : 100 dpi, 4 bit grey-scale
Compression : JPEG loss-less Compression
File Format : TIFF 6.0
ii. Colour Material
Capture parameters : 100 dpi, 24 bits/pixel
Compression : JPEG loss-less Compression
JPEG Lossy for non-critical images
File Format : TIFF 6.0
An alternative de-facto standard is Photo-CD.
c. Photographic Stills
For monochrome material, capture at 8 bits/pixel and colour
material capture at 24 bits/pixel.
i. Opaque
Capture parameters : 100 dpi
Compression : JPEG loss-less Compression, JPEG Lossy for non-critical images
File Format : TIFF 6.0
ii. Transparencies
(8 x 10 inches, 35mm and Microfilm)
Capture parameters : 200 to 2000 dpi
(depending on the reduction factor)
Compression : JPEG loss-less Compression, JPEG Lossy for non-critical images
File Format : TIFF 6.0
If a microfilm does not include gray scale information, then
it can be classed as part of Group i.a. Modern Texts
An alternative de-facto standard is Photo-CD.
d. Maps
For monochrome material, capture at 8 bits/pixel and colour
material capture at 24 bits/pixel.
Capture parameters : 100 dpi
Compression : JPEG loss-less Compression
File Format : TIFF 6.0
Maps larger in size than A3 should be first photographed and
the resulting image digitised.
The merits of using a digital camera in place of a photographic
camera followed by the digitisation process for large documents
were discussed by the Sub-Committee on Technology. Concern was
expressed about the limited number of lines and pixels per line
available with current digital cameras. It was acknowledged that
the latest production systems for feature movies were using digital
equipment of 3000 lines vertical resolution and 2500 pixels horizontal
resolution. These systems were not, however, manufactured in a
form suitable for the digitisation of large paper documents. The
resolution of CCD cameras is limited mainly by the size of the
chips. Larger chips can, in theory, offer higher resolution. It
is also possible to capture a large document in sections and join
the sections in software when viewing it.
Whenever possible, documents should be captured at a higher
standard and lower standard copies produced from this as required.
The above statements are what is realistically possible today.
The ideal standard may not be achievable today but it should be
kept in mind as the long-term target. If a document is in good
condition but technology does not yet permit its capture at the
ideal standard, it should not be digitised but left until technology
does permit its capture at the ideal standard. If, however, the
document is in poor condition or access to it is frequently required,
then digitisation at the best possible standard currently available
is recommended. To overcome the limits imposed by the capability
of current digitisation equipment, it may be advisable to consider
the creation of high quality photographic copies. It must be remembered,
however, that photographs are themselves subject to decay.
2. Sounds and Moving Images
a. Sound
In general, the standards quoted provide an Intermediate Access
standard. The standards are suitable for transmission using the
AES/EBU interconnection standard.
i. Historical Material (Cylinders, 78s)
| Sampling | Bits/ | Bytes/ |
| Standard | Second | Hour |
| 48 kHz/20 bit | 1920.0 kbits/s | 864.0 MB/h |
| two channel |
The use of a stereo pick-up and digitising each output separately
when copying grooved originals enables the signal from each groove
wall to be recorded independently, even though the two outputs
are nominally identical. This may assist if scratch and click
removal software is used when playing the digital copy.
ii. Modern Analogue (LPs, Tapes)
| Sampling | Bits/ | Bytes/ |
| Standard | Second | Hour |
| 48 kHz/20 bit | 1920.0 kbits/s | 864.0 MB/h |
| Stereo |
iii. Digital
Store at the same standard as the original material.
The commercial CD standard of 44.1 kHz/16 bit per channel sampling
standard can be used as a Basic Access and dissemination standard.
For access by networks and for multimedia applications, the
maximum acceptable level of reduction of data is that given by
ISO/MPEG-1 (Audio) Layer III at a data rate of 64 kbits/s per
channel. Only browse or catalogue copies should be digitised to
a lower standard.
b. Analogue Video
In general, the standards quoted provide an Intermediate Access
standard. There are two methods of storing and distributing colour
video signals. The video signal received off-air or by cable (NTSC
or PAL) or viewed from tape at home (VHS, BetaMax etc) are composite
signals coded to include the colour and luminance information
in one signal.
The Sub-Committee recommend the use, wherever possible, of a
component format which stores a luminance signal (equivalent to
a monochrome image) and two lower quality colour difference signals
separately . This uses more storage space but offers advantages
when copying or converting the signals. The standards quoted are
based upon CCIR Recommendation 601 for composite digital video.
The CCIR 601 standard gives two word sizes for storage - 8 bit words or 10 bit words. The early digital video recorders (D1 to D3) used 8 bit words but the trend is to use 10 bit words. The latest digital video recorders are 10 bit capable. The higher standard is chosen to give increased fidelity to the stored images.
| Sampling: | Bits per: | Bytes per: |
| Standard | Second | Hour |
| Luminance (Y) | 13.50 MHz/10 bit135.0 Mbits/s | 60.75 |
| Colour Diff. (U) | 75 MHz/10 bit67.5 Mbits/s | 30.38 |
| Signals (V) | 75 MHz/10 bit67.5 Mbits/s | 30.38 |
| Totals | 270.0 Mbits/s | 121.50 GB/h |
In practice, the most common access and dissemination medium
in current use is the VHS tape. The quality of VHS relative to
the above video standard is inferior to the quality of the CD
relative to the audio standard. The proposed new digital format
for domestic video tape may offer a better quality access format.
For access by networks and for multimedia applications, the maximum
acceptable level of reduction of data is that given by MPEG-2
(Video) at a data rate of 2 Mbits/s. Only browse or catalogue
copies should be digitised to a lower standard.
c. Digital Video
Store at the same standard as the original material.
d. Film
The minimum acceptable standard for an Access standard for copies
of 35mm films would be a High Definition Television (HDTV) standard.
At the moment of writing, however, the HDTV recording scene is
still unsettled. In addition to the differences between the format
advocated by North America and Japan (1125 lines, 60 fields per
second) and that advocated by Europe (1250 lines, 50 fields per
second) there are a number of other unresolved questions with
regard to recording formats.
Either basic HDTV format will, in principle, provide a suitable
Access standard copy. Both formats are being used for the shooting
and production of programme material for television transmission
or for transfer to film for public viewing. The use is, however,
often on a slightly experimental basis and using a range of potentially
incompatable equipment. Until the uncertainties in the technical
specifications are resolved, the Sub-Committee are unable to make
any meaningful comments.
An Intermediate Access standard for 35mm film - one which meets
the requirements of 99% of viewers - would require a storage capacity
beyond that which is economically practical today. Specialist
restoration and processing systems which exceed the quality of
35mm film are already being used. The work can, however, only
be performed in short sections. Each completed section has to
be copied back to film as the storage capacity is insufficient
to hold an entire 90 minute feature film.
Up
ANNEX B
Obsolescence of Machine Readable Formats
The obsolescence of machine readable formats and the problems
that this is creating for video collections in particular was
discussed.
A machine-readable format, by definition, requires both the carrier
and the machine required to access it to be in good condition.
It is one of technologies quirks that later formats are often
more at risk from obsolescence than earlier formats because of
their increased complexity. For example, to construct an Edison
cylinder player is within the capability of any reasonably equipped
workshop. The construction of machines to access modern formats
is much more demanding.
The video world is currently tackling the problem of the earliest
commercial recording format - the Ampex 2" quadruplex tape
format. This format was introduced in the late '50s and a crisis
is looming. Many tapes are showing signs of decay; the machines
needed to copy the material are old and spare parts are no longer
made; the experienced operators are nearing or have reached retirement
age; the quantity of tapes to be examined and copied is very large
in the order of hundreds of thousands of hours; the documentation
available for many tapes is inadequate to permit the selection
of material for copying without first viewing the tape. To preserve
the material, it has to be copied. The work of copying the tapes
to a digital format has been started by many broadcasters and
archives but it is a race against time.
The formats that followed the 2" tape format - the 1"
tape formats - are also approaching a similar crisis. These formats
came into use in the late '70s and were used as production formats
until the early '90s.
There are at least seven other analogue recording formats that
have been widely used by amateur and professional cameramen. The
problem is compounded world-wide by the different video standards
used in various regions of the globe. Material on any of these
formats may be offered to archives and libraries. Only the largest
and best equipped video archives can be expected to deal with
this range of formats and standards.
Since the first digital video format (D1) was introduced in the
mid '80s, at least four others have been launched. None of them
is in a dominant position in the market. D1 itself is now virtually
obsolete.
The problem requires technicians who understand the operation
of a range of formats. The collection also needs to own working
examples of obsolete machines. This increases the need for highly
trained, specialist technicians. There is currently no training
for archive technicians anywhere in the world. There are some
establishments that can, potentially, provide such training but
finance will be required to enable them to organise and run suitable
courses.
The training of technicians is not a purely technical matter.
Modern technology provides many ways of making changes when copying
a document. The ethical considerations of archival copying need
to be taught as well as the technical considerations.
Sound and moving image archives and libraries require a storage
format that is not linked, solely, to the demands of the broadcasters,
publishers or manufacturers. It must be able to contain documents
of several types so that, for example, the digitised sound of
an LP record may be stored with the image of its cover photograph
and the text of the sleeve notes. A storage system that does not
restrict the digital data-stream - a free-formatted store - is
preferred. Most of the commercial, mass-produced systems available
at present require the information to be in a specific formation
before recording.
Stores that are fully automated, self-checking and self-renewing
are becoming available from a number of manufacturers. Because
of their ability to copy and check carriers without constant human
intervention or monitoring, it is possible to migrate collections
to new stores whenever the need arises without incurring the very
large labour costs demanded by the migration of analogue or shelf-stored
digital carriers.
This is not science-fiction. Such stores are starting to be installed
in some audio archives and libraries. They are also in use in
a number of other areas requiring the storage of very large amounts
of digital information such as banking and scientific centres.
They will also be the backbone of new consumer services such as
"video-on-demand".
Up
ANNEX C
LIST OF PARTICIPANTS
Members of the Sub-Committee on Technology:
| Abdelaziz ABID General Information Programme (PGI) UNESCO 1 Rue Miollis 75015 Paris France |
E-Mail:A.ABID@UNESCO.ORG Tel: +33 (1) 4568 4496 FAX: +33 (1) 4449 0058 |
| Michael ALEXANDER The British Library Computing and Telecomms Boston Spa Wetherby West Yorkshire LS23 7BQ United Kingdom |
Tel: +44 (1937) 546883 FAX: +33 (1937) 546872 E-Mail: michael.alexander@bl.uk |
| Julian BESCOS Co-Director Archivo de Indias Project Co-Director Society "Archivos y Bibliotecas" created by IBM Spain and El Corte Ingles Apartado Correos 179 28080 Madrid Spain |
Tel: +34 (1) 397 9206 FAX: +34 (1) 735 1440 |
| George BOSTON Rapporteur 14 Dulverton Drive Furzton Milton Keynes MK4 1DE United Kingdom |
Tel: +44 (1908) 502610 FAX: +44 (1908) 502520 E-Mail: gboston@gn.apc.org |
|
Dietrich SCHÜLLER Chairman
Chairman Director Phonogrammarchiv Österreichischen Akademie der Wissenschaften Liebiggasse 5 A-1010 Wien Austria |
Tel: +43 (1) 40103 2734 FAX: +43 (1) 403 0465 E-Mail: pharchiv@kfs.oeaw.ac.at |
The following people have also taken part in discussions or have offered advice and comments at various times: |
|
|
Bill BULFORD
BBC Television - Television Centre Wood Lane London W12 7RJ United Kingdom |
Tel: +44 (181) 576 1566 FAX: +44 (181) 576 1639 |
| Michael CROLL BBC Research Department Kingswood Warren Tadworth Surrey United Kingdom |
Tel: +44 (1737) 836612 FAX: +44 (1737) 836665 |
| Brian JENKINSON National Film and Television Archive John Paul Getty III Conservation Centre Kingshill Way Berkhamsted Hertfordshire United Kingdom |
Tel: +44 (1442) 876301 FAX: +44 (1442) 862085 |
| Jacques KLOSSA General Director TRIBUN-VTCOM (France TELECOM) 40 Rue Gabriel Crie 92240 Malakoff France |
Tel: +33 (1) 4612 6971 FAX: +33 (1) 4612 6980 |
| Franz LECHLEITNER Phonogrammarchiv Liebiggasse 5 A-1010 Wien Austria |
Tel: +43 (1) 40103 2740 FAX: +43 (1) 4030465 E-Mail: pharchiv@kfs.oeaw.ac.at |
| John NUNN BBC Radio - Room 4102 Broadcasting House Portland Place London W1A 1AA United Kingdom |
Tel: +44 (171) 765 3522 FAX: +44 (171) 765 3822 |
Support has also been received from: | |
| Harald GARDOS Secretary General Austrian Commission for UNESCO Mentergasse 11 A-1070 Wien Austria |
Tel: +43 (1) 523 64 21 or +43 (1) 526 36 80 FAX: +43 (1) 526 13 01 |
