Powered by
JSPWiki v2.8.2
g2gp 17-01-2009
View PDF
This is version . It is not the current version, and thus it cannot be edited.
[Back to current version]   [Restore this version]

Vector Images#

Edited by Kieron Niven#

Section 1: Introduction to Vector Images#

Section 2: Creating Vector Images#

Section 3: Archiving Vector Images#

Bibliography and Further Reading#

Section 2 Creating Vector Images

    • project planning and requirements
    • sources of data
    • file types (whilst creating, working with, processing data)
    • documenting data creation and processing

As highlighted in the previous sections, many vector images are derived from data held in other applications. In regard to creating vector images, it is advised that if an image is derived from e.g. a CAD dataset then the original CAD file should take precedence over the ?? vector images. Where images have been created natively in a vector graphics package, or where a derived file (e.g. an illustration) is seen as being worth preserving in its own right, then the guidance in this chapter should be followed.

Adobe IllustratoraiA 2D layer-based raster/vector format based on PostScript. Not suitable for preservation.
Computer Graphics Metafile.cgmA 2D raster/vector format of some vintage and a documented standard format (ISO/IEC 8632). Not suitable for preservation.
CorelDrawcdrA proprietary layer-based 2D format which can hold both raster and vector data. Not suitable for preservation.
Encapsulated PostScript.eps, .epsfEPS is a PostScript file embedded into another document. It may contain vector, bitmap and fonts. A preview image at 72 dpi is normally included. EPS headers often contain a version number. Some programmes reject the files if the version number is absent.Not suitable for preservation.
Hewlett-Packard Graphics Languageplt, prnA control language to drive Hewlett-Packard plotters but many programmes can render the contents to screen. Files are largely ASCII and thus human readable. Not suitable for preservation.
Macintosh PICTpictA meta-format developed by Apple in 1984 that can store both bitmap and vector images. The file contains all the QuickDraw commands used to draw the image. PICT files containing only one bitmap are supported under Windows using QuickTime for Windows. Not suitable for preservation.
Macromedia Flashfla, swf, swd, flv, swc, swt, flpA 2D vector-based animation format optimised for web delivery. The format specifications are freely available. Not suitable for preservation.
Macromedia FreehandafA popular illustration package that uses a bitmap vector hybrid file type. The file format has changed substantially, resulting in some problems when migrating early versions. Not suitable for preservation.
Micrografx Designerdrw, dsfA vector/bitmap-based programme aimed more at technical than artistic drawing. Not suitable for preservation.
Microsoft Windows Metafilewmf, emfA vector-based file format that uses Graphics Device Interface (GDI) commands to render an image. EMF is an enhanced version of the WMF standard for 32-bit architectures. Not suitable for preservation.
Portable Document FormatpdfA subset of Postscript. The content is ASCII text unless compressed (which is normally the case). See also Binary Text / Word Processor Documents Preservation Handbook. Not suitable for preservation.
PostScriptpsA programming language, developed by Adobe, to describe the appearance of text and images in a device-independent manner. Suitable for preservation.
Scalable Vector GraphicssvgAn emerging, open standard to describe 2D vector graphics in XML. Suitable for preservation.

WordPerfect Graphics MetafilewpgA 2D graphics meta format capable of storing bitmapped, vector graphics or EPS data. WordPerfect 5.0 and earlier can store either bitmap or vector image data, but not both at once. Not suitable for preservation.

.ai is a proprietary format but is in fact a minor variation of postscript. Change the extension to ps or eps and programmes such as GhostScript can display (at least partially) the content. Illustrator files will probably have been created as page illustrations and therefore a tiff (or png) export of the file will be an adequate format for preservation, and jpeg or png suitable for presentation. If the vector content is significant then the files should be converted to svg. Illustrator files can contain scripts in AppleScript/OSA, JavaScript or VBScript. These scripts are aids to editing the document and hence do not add to the meaning so should be discarded.

.cdr is a proprietary format and not suitable for preservation. CorelDraw files will probably have been created as page illustrations and therefore a tiff (or png) export of the file will be an adequate format for preservation, and jpeg or png suitable for presentation. Ideally the depositor should create these files. If the vector content is significant then the files should be converted to svg or, if the drawing is layered, dxf.

.svg is fine as is. Their may be additional functionality provided by Javascript, if the svg is specifically built to be viewed within a web page, or similar container. The container and svg must be treated as linked files and both preserved and/or presented together. See ADS Data Procedures: HTML (or markup, or whatever it is to be called) for web pages.

Future directions?#

STEP is being developed as a neutral and fully documented CAD and CAM exchange format standard (ISO 10303).

SVG is potentially a better option than DXF, however it does not yet adequately support layers, currently this must be done by including elements in groups and using Javascript to control their visibility, of course the whole must be in a web page.

DWF has progressed over the past few years and the latest incarnation can hold 3D data. Also the file format is compressed XML (much the same as OpenOffice documents), although it does to have binary elements, and is an open format http://usa.autodesk.com/adsk/servlet/index?siteID=123112&id=4358577

.dwg is a proprietary binary format that undergoes more or less significant changes on average every three years. In recent versions of AutoCAD (since AutoCAD 2004) the data in the file is compressed using an undocumented algorithm. The OpenDWG Alliance have reversed engineered .dwg in an attempt to make it an open format but AutoDesk have shown a degree negativity towards their endeavours, in part because the Alliance receive funding from Bently, a major competitor. A number of cheaper CAD packages use .dwg as either a native format or as an export option but these are built on reverse engineered AutoDesk files and may not be entirely accurate. It is thus difficult to recommend .dwg as either a preservation or a dissemination format because of the necessity to migrate files to newer versions on a regular basis to ensure the files can be read by current versions, particularly as this would lock out those who are still using older versions and perhaps those using other CAD programmes.

AutoCAD files can have associated, external, scripts in either AutoLisp or VB for Applications. These may be aids to editing the documents but could add additional meaning or functionality or meaning (e.g. animation). The scripts may be stored in either ASCII or compiled binary format. If the scripts add functionality or meaning and are in ASCII they should be archived with the drawing and also made available for downloading.

.dxf is generally considered to be the universal CAD exchange format but there are problems with it. The format of .dxf files has changed in parallel with the changes in .dwg format. Release 14 saw the last major change to the .dxf format. This added more entities and also the possibility of including a preview image as binary data. dxf suffers from some of the same drawbacks as dwg but is, so far, the only viable format for saving three dimensional CAD data.

3.1 Deciding what to archive

      • Selection and retention
      • preservation intervention points / file and data lifecycles (specific to guide, will also be covered generically)

3.2deciding how to archive

      • archiving strategies (migration (to new format, to 'basic' format), emulation, refreshment)?
      • significant properties
      • file types

JISC Digital Preservation Programme: Study on the Significant Properties of Vector Images

All from AHDS Guide

Significant Characteristics Vector graphics programmes are often used as a convenience for creating high quality line drawings or illustrations and represent a final product or the source for a hardcopy version. When this is the case then a bitmap version will probably be suitable for preservation and the depositor should be asked to export the graphic to a suitable format and size (see the Bitmap (raster) image Preservation Handbook). In many cases the image is unlikely to have been created in isolation but as an illustration associated with some text. It is important that the document(s) the image is associated with is recorded plus the location within that document to which it relates. The caption or a description of the contents and purpose of the image is also important. Any meaning incorporated into conventions used in the drawing must be explained, i.e. the significance of colours, layers (including invisible layers), object types, line styles, line weights, text styles, fill and hatch styles, dimension styles.

Technique Check the graphic for hidden information (invisible layers, etc.) and the relevance or suitability of this material for archiving, i.e. does it consist of construction lines, text paths, etc., does the depositor have copyright.

Ensure that any linked data (e.g. other image files, fonts) are present. These should be embedded into the graphic.

If necessary export the graphic to a suitable archive format(s). This may be dependant on the nature of original, whether pure vector or vector and raster data. Mixed formats may require the separation of the vector and raster data. DXF is generally a good preservation medium for the vector content. Use a suitable preservation format for the raster content (see the Bitmap (raster) image Preservation Handbook). It may be possible to archive mixed content as SVG. Vector data should be archived at a suitable level of precision, preferably that used when collecting the original data. Exporting to an excessively low level of precision will degrade the data. Exporting to an excessively high level of precision will cause file bloat and may imply a false degree of precision for the data.

Validation of Exported Data There is no objective way to compare the original file and an exported version. The best that can be done is to use programmes that can be trusted to give an accurate rendering of each format, either to paper or screen, and to compare the results by eye.

Problems and Issues File migration to newer formats or to different packages can result in significant degradation of the image quality. This can be exacerbated, particularly if the migration is across platforms (e.g. from Mac to MS Windows), where the availability of fonts may differ. Programmes will usually substitute a font if the originally specified font cannot be found but the substitute may be inadequate and the resulting image spoiled.

3.3 Metadata and Documentation

      • project level
      • file level
      • standards specific to #

Relationship to other documents, caption

Level 1 (Essential) ?? Title or caption describing the graphic ?? Creator ?? Purpose ?? Relationship to other documents ?? Externally referenced files must be present ?? No embedded material for which the depositor does not hold copyright. Level 2 (Preferred) ?? Software and version used to create image/model ?? Creation/completion date ?? Explanation of all conventions used in the document (colours, layers, line styles, line weights, etc.) ?? Scale (i.e. what distance does a unit length of 1 represent). ?? Externally referenced material should be imported (bound) where possible. ?? Relevance of hidden material (in frozen / off / invisible layers)

3.4Structuring your archive

  • Copyright
    • specific copyright considerations for each guide.

  • Case study/studies

  • Check for any dependencies (linked files, object libraries, bitmaps, custom linestyles; shapes)

PDF as a possible dissemination format for 2D vector data. Means the user won't need any plug ins apart from the Adobe Acrobat reader 8.1