The cataloging for this portal has used the following guidelines: each file includes a descriptive record for a document, its origin, type of writing, binding, materials, contents, colophons, annotations, etc. The file thus provides a very detailed description which aims to give researchers and librarians all relevant information. This process is based on the working model that has been used by the IRHT (Institut de Recherche et d'Histoire des Textes) in France in recent years. First, a unified database was created for all manuscripts, regardless of their language, and was included in the Union Catalogue of the CSIC Library Network (CIRBIC), run by the ALEPH 500 system. The records were then incorporated into the CIRBIC in MARC 21 format and their descriptions adapted conform to the Cataloging Rules (1999) of the Ministry of Culture. For the principal and secondary access points such as author, entity, uniform titles, materials and places, the Authority Catalogue of the CSIC Library Network was used, whenever an authority record had been defined for this purpose. For cases in which none was available, one has been created by following the same guidelines used in the CIRBIC to produce the different types of headings.

The first step in the process corresponds to the photographic documentation and the preliminary analyses required for each document (evaluation of ink solubility, pH of the paper etc.), diagnosis of its state of conservation, and identification of the main kind of deterioration. Such deterioration can take the form of mechanical damage such as folds, holes and tears, which are especially visible at the edges; chemical changes due to the acidity of the cellulitic support; or general dehydration, a particular problem in parchment documents, which are made from a protein material that is highly sensitive to changes in humidity and temperature because of its high hygroscopic capacity.

In the paper manuscripts, texts are written with highly soluble inks and pigments, making it impossible to carry out any treatment involving the use of water. In these cases, the restoration process is based on mechanical cleaning, using erasers of variable hardness and fine brushes; the smoothing out of folds and creases plus the joining of cracks and tears; reintegration of holes in the support using paper inserts; and, finally, lamination of the inverse (using a vacuum laminator), in order to strengthen the support, which is greatly weakened by acidic components.

The parchment manuscripts were found folded or rolled up, and their dehydrated condition had resulted in tears in the support or the cracking of the supporting elements. In this case, the restoration procedure focused on cleaning the surface, hydrating and smoothing out the parchment, fixing the pigments and restoring lacunae using inserts of similar characteristics to the original material.

The Institute of Spanish Cultural Heritage has also devised a set of guidelines for the preventive preservation of manuscripts to be implemented in libraries storing the reserves which will help to guarantee the durability of these documents.

Since 2009, the Unit of Information Resources for Research has been promoting a digitalization policy for the CSIC Library Network, in an attempt to implement uniform technical criteria and standards in order to improve co-operation with national and international institutions and projects.

For this purpose, a set of documents was drawn up to support the creation and development of a common methodology for all libraries in the network.

At the start of 2010, the Master Plan for the Digitalization of CSIC Resources was outlined to consolidate the digitalization policy for CSIC heritage resources. This marked the beginning of the institution's digitalization plan.

Following the line of action then laid out, the first steps were taken towards digitalizing the Arabic, Hebrew, Persian and Turkish manuscripts now included in the Manuscript@CSIC Project.

Digitalizating this material has been an especially challenging task, due the typology of the material, its state of preservation, and the very limited cataloging of the resources.

In the first stage, specialists and researchers examined the material, taking care to identify the following:

• Documents in very poor condition not suitable for being digitalized.
• Documents that could only partially be digitalized. In many cases the beginnings and ends of the works were much more highly damaged than their central sections.
• Loose leaves and/or documents inside the manuscripts.
• Poorly bound or incorrectly inserted sheets.
• Undetected factitious elements.
• Láminas o dibujos de más tamaño.
• Larger prints or drawings, and uncatalogued documents (loose leaves).

Finally, for digitalization purposes, all the documents had to be precatalogued/catalogued and studied in order to reduce the number of unexpected incidents once the scanning process had begun.

In addition to all this, a timetable was devised to provide a daily control of the scanning process.

Digitalization was carried out at CSIC premises in Madrid and Granada. A librarian was appointed at each site to supervise the movement of manuscripts and the up-dating and maintenance of the digitalization table, and to record any problems arising and matters remaining unresolved. Technically, it was decided to make the masters in Tiff format in color, with a resolution of 600 ppp, on a Zeutschel scanner. The files are one-sided for their preservation and bear the name of the image embedded in the Tiff header tags.
Subsequently, jpg images of low resolution but verified legibility, were generated to be placed on the website and pdfs in color for downloading.

Due to the large size of the master images, it was difficult to create pdf files. However, it was eventually possible to make most files smaller than 50 Mb sufficiently legible, apart from a few with sections which were difficult to read owing to defects in the original manuscript.

A METS metadata file was also created for each bibliographic record. In addition to the standard MIX for the technical part of the image, this profile also includes a selection of PREMIS standard fields and a structural map on the page. For the generation of metadata files, the digitalizing company was provided with the bibliographic records in Marcxml format, which were added to the descriptive section of the file.

The metadata are generated according to the "CSIC METS Profile for monographs v. 1.2".

Digitalization produced 42,171 Tiff files corresponding to 3,200 GB.