Preservation of image formats

René van Horik

Issues covered in the presentation:

• Definition and description of image formats.
• Types of image formats.
• Overview of theories (and the assumptions on which the theories are based) on digital preservation relevant for preservation of image formats.
  • Metadata
  • File format standards
  • Role of XML
  • Registries
  • Etc.
• Overview of available digital preservation solutions relevant for the preservation of digital images.
  • Format registry
  • Format identification
  • Digital archiving
  • Distributed storage
  • Emulation
  • Etc.
• Importance of evolutionary approach. Digital data formats are relative young. Only the future can judge which assumptions were right.
• 'Building blocks' for the long term preservation of digital raster images
  • Based on following assumptions:
    • Graphics file format standards are durable
    • Digital data encoded in the XML data format is durable data
    • Metadata on digital objects is essential in order to understand and process digital images in the future
• Building block 1: Graphics file format standards.
  • Discussion of graphic file formats used in the period 1994 - 2004
  • TIFF format seems the best format for long-term access.
  • Discussion of TIFF file format
• Building block 2: XML data format for durable encoding of the bitstream of digital images.
  • How to express the bitstream of a digital image in XML?
    • Expression of content model in XML
    • Binary to XML conversion
    • XML to binary conversion (in the future)
  • Methods available to express a digital image in XML
    • Bitstream syntax description language (BSDL)
    • Universal Virtual Computer (UVC)
    • Format language for Audio-Visual object representation (Flavor/Xflavor)
  • Comparison of three methods
• Building block 3: Preservation metadata element sets for digital images.
  • Methods to create and store preservation metadata on digital images (e.g. "automatic metadata exposure" project of RLG).
  • Some important metadata element sets for digital images:
    • NISO Z39.87 (technical metadata for digital still images)
    • EXIF (for digital images created by digital camera's)
    • SepiaDES (for digital surrogates of historical photographs
    • Etc.
• Conclusions:
  • (Baseline) TIFF seems best format for long-term storage of digital master images
  • Expression of bitstream in XML: more research required
  • Preservation metadata: application profiles and registries help to 'discriminate exactly what we know vaguely'.