The Creation of an RNIB Centre for the Migration, Origination, Archiving and Production of Alternative Format Files

General Objectives

The following are the general objectives of a file creation and migration centre; to provide blind and visually impaired people with an alternative format production service which:

1.1 Meets individually defined needs within a defined turn-round time at an agreed level of quality

1.2 Publishes material which reflects the general print and media markets simultaneously with those markets (referred to in this Paper as equivalence)

1.3 Undertakes a classics publishing policy to make good the massive deficit in stock (referred to in this Paper as classic)

1.4 Makes international agreements for file exchange and copyright

1.1 Individual need

a) RNIB is the only significant general provider of material for individuals. The problem of meeting the demand of blind and visually impaired people for information in accessible formats is that it is a demand driven service which might lead to an excessive burden on the organisation, crippling it economically, overwhelming other publishing priorities and encouraging frivolous commissioning. Currently demand is partly rationed by turnaround time and this is an irrational and arbitrary device.

Production for individuals should be determined by such factors as:

Whether the requested item accords with our publishing policies (e.g., equivalence or classic)
Whether the item is required in a context where payment can reasonably be demanded (e.g., National Curriculum, DDA, formal study supported by a special needs grant)*

*Individuals must not be allowed to act as proxies for organisations to which they belong or from whom they are receiving services.

b) Factors which should enable the organisation better to meet demand are the determination with the end user of:

The essential textual requirement (e.g., is the request for a specific edition of a specific title or more general in either respect)
The essential publishing output format (file or hard copy; synthetic or real speech)
Accordance of the user requirement with existing templates and processes (diagram making, sound recording)
The quality of the output, including turnaround time as a factor in quality (e.g., sometimes there is a trade off between turnaround time and output quality)

If RNIB, on the basis of the above:

Establishes and publishes normative file format definitions (e.g. XML), standard template options and other standard processes (diagram making, sound recording)
Utilises a high level of librarianship skill
Minimises proof reading
Minimises printing (braille, modified print) and copying

it should substantially increase its capacity to create and migrate files and publish them in an accessible format.

c) Even with all these factors in place it may still be necessary to establish an individual information quota. This should be transparent, public and part of any campaign for public sector and fund raising support. It should be couched not in terms of entitlement to a quantity of data input (so many print pages etc) but in terms of a quantity of process capacity so that, for example, there is a trade-off between quantity of output and intensity of manipulation.

d) For the purposes of this Paper it is assumed that real sound recording will follow the same process as the creation of other file formats; a book may be 'provisionally recorded' for rapid use but will then be subject to additional processes which bring it up to RNIB's archiving and publishing standards. The current distinction between superior quality 'Talking books' and very ordinary quality student recording will simply be a matter of the distinction between an intermediate product for an individual and a final product for the archive. We must not continue to provide seriously sub-standard end products for students where quality is so important for comprehension. The quality for students should mirror the discussion of transcription error rate in 3.9 below.

1.2 Equivalence Publishing

a) The responsibility for publishing major 'fixed' documents in alternative formats (primarily books), is shared between RNIB, the NLB and CALIBRE.

With so few exceptions that these can be quoted, this form of publishing does not aspire to be Equivalence publishing, i.e. the same product at the same quality, price and time as the product for the general market. RNIB (e.g., Andrew Morton's PRINCESS DIANA) has occasionally attempted equivalence publishing which not only meets the price and quality but also, and exceptionally, the time criterion. The NLB, through its visionary Cheetah Project, published Booker Prize novels which met the time and price criteria for equivalence but not the quality criterion but in recognising the trade off between time on the one hand and quality and price on the other it was a landmark conceptual breakthrough. Nonetheless, as more than 80% of its consumers were pleased with the product (and those who were not were not bound to consume it) it is sad that the NLB apparently lost its nerve and did not pursue this product line. To say that Cheetah did not meet the quality criterion is not in itself a criticism; it might have done so had it subjected texts to much more rigorous proof reading and correction but this would have, as noted above, lengthened the turnaround time but it would also have radically raised the production costs.

It will not be possible in the foreseeable future (3-5 years) to reduce the cost of alternative format file preparation to anything close to that of a standard file for print output; and even with steadily improving automation there will always be a differential cost of origin for documents as we currently define and recognise them simply because there are more processes. There is always likely, too, to be an additional marginal cost for special materials for paper hard copy for braille and large print. Perhaps inevitably, however, we are likely to see a widening of the cost gap as conventional documents for the general market become more complex in the own rendition, requiring ever more interpretative intervention to render them for blind and visually impaired people.

b) Nonetheless, equivalence publishing, necessarily with public sector support, must be a primary policy objective of RNIB. This could immediately be given limited effect by setting aside all proposed new titles already scheduled to be published and defining these as items on the 'Classics' list. Equivalence publishing could then be launched on the basis of the simultaneous publication of new titles. This would cause inconvenience to some individual commissioners but the only alternative is a substantial use of overtime to reduce the planned backlog of commissioned work to zero to enable the new service to be launched.

c) Defining equivalence requirements in terms of the products to be published is not a very complex matter for popular fiction and non fiction as this can be based on projected book sales and library loans. Even here, there are always exceptions - books which become unexpected best sellers (EATS SHOOTS AND LEAVES) and those expected to do well which flop - but even then the recovery time should be minimal. A much more deliberative process is required for identifying books which will be regarded as classics, commanding smaller audiences than popular works but over much longer periods.

d) If RNIB determines that equivalence publishing should form part of its offer it must come to a proper understanding with the NLB in respect of digital/text publishing and CALIBRE in respect of sound recording. Its starting position must be that competitors are performing on sufferance and that RNIB will only refrain from publishing certain agreed titles if these are being produced according to its own standards of equivalence; NLB and CALIBRE cannot be given an unwritten, unquestioned monopoly of current title publishing, particularly when this falls below RNIB standards. If competitors fall short it will have to consider sub contracting competitors or competing directly with them. RNIB can only adopt this high ground when its own performance has reached its own defined standard.

After more than a decade of collaboration through Share the Vision (STV) consumers have seen very little benefit in production; there may have been a slight gain through a lack of duplication of production but the main gains have in the refinement of metadata systems.

e) In spite of its lower cost and greater flexibility, there has been very little interest in equivalence publishing in the area of digital information. RNIB has produced some audio described video and has been looking at audio described DVD (which must include accessible menus as well as additional description); but the visual impairment world in general has been content to condemn web inaccessibility without intervening directly in the market to render web information accessible. This topic requires some fundamental analysis:

How far is it good enough to condemn sites for being inaccessible? (on a legal, moral basis)
How sensible is it to provide consultancy on accessibility but not a repair or even construction service?
Is it necessary, while following the general line of imposing public obligations for accessibility, for RNIB to make parts of sites accessible to individuals or the blind and visual impairment sector as a whole, particularly if this reduces the need for origination?

Even though the answer to the third question should be positive, this is an extremely complex area because it puts pragmatism in opposition to principle; but RNIB would be well advised to adopt an intermediate position on this subject.

f) RNIB should establish a transparent and public process for selecting equivalence publishing product against a public sector subsidy and its own resources.

1.3 Classics Publishing

a) While alternative format publishers have failed to keep up with new publications there has been a steady deterioration in the stock of alternative format classics, particularly in braille and on tape (there has never been a great market for classics in modified print; whether there should have been or not is an interesting question).

b) Even if RNIB's current transfer of analogue titles to digital produces acceptable quality, new generations of hardware will require a basic, high standard of recording which our legacy stock does not meet. In respect of braille, NLB's manually produced single copies and RNIB's plates are deteriorating or have become obsolete.

c) It might be argued that all 'classics' should be subject to a new individual commission or to an 'equivalence' criterion but any basic service should possess a list of 'classics' files which can be produced on demand. Many of these may be rendered through combining classic text out of copyright with material such as commentary and notes still in copyright; but, for example, it cannot be acceptable for RNIB not to hold complete digital files of the canon of Shakespeare and the Authorised Version of the Bible.

d) RNIB should establish a transparent, public process for publishing classics material, the origination cost of which should be a public sector responsibility.

1.4 International File Exchange and Copyright

a) As more documents are initially produced in digital format, documents will become both more provisional and hybrid. Not only will they be much more amenable to authorial updating, they will also be open to critical upgrading. Both of these aspects will call into question national and international copyright agreements.

b) RNIB should continue to pursue copyright issues but it may be that in spite of its short term disadvantages, a long term approach based on a general right to information will be more effective than ad hoc publisher or national concessions.

c) One benefit of international agreements should be the exchange of files for multilingual productions; all alternative format agencies should endeavour to reduce their foreign language origination. They should also consider centralising specialist code origination - chess, braille music, maths and science - in specialist agencies, perhaps in Eastern Europe.