A production master file is a file produced by processing the content in one or more archival master files, resulting in a new file or files with levels of quality that rival those of the archival master.
The first type of processing consists of the assembly of a set of segments into a unified reproduction of an item. For example, an image of a large map may be produced by stitching together a set of image files, each representing a portion of the original paper item. In the realm of sound recordings, a number of segments may be assembled to represent a long work, or a number of separate tracks may be mixed into a single, stereo presentation. If both the segmented and the assembled representations are retained by the digitizing organization, the originally captured segments are called archival master files and the assembled version is called a production master file.
The second process that may be applied consists of aesthetic or other technical corrections to the original file. When both the uncorrected and corrected representations are retained, the uncorrected files are archival master files and the corrected versions are production master files. For most preservation-oriented archives, aesthetic changes will be modest. For images, aesthetic changes may include such things as adjusting tonality; for sound recordings, such adjustments may reduce the audibility of clicks and pops, or correct drastic changes in level. Certain technical changes may be more significant. For example, the archival master version of a pictorial image may employ a linear representation of light intensity, while the production master may employ gamma correction. The transformation from linear to gamma-corrected is not reversible in a mathematically exact manner.
Master files of all types have permanent value for the digitizing organization and should be managed in an appropriate environment, e.g., one in which read and write executions are minimized and other preservation-oriented data management actions are applied. In contrast, derivative files are frequently accessed by end-users and are typically stored in systems that see repeated read and write executions.