National Computational Infrastructure
NCI National Facility
Making Video Animation Using the ANUSF Vizlab Facilities
The Vizlab uses a write once read many (WORM) video laserdisk to record single frame animation. If you want to produce an animation you will have to provide the original images in a format suitable for recording to disk.
Image specifications.
  • Format - The best format to give us images is probably SGI's .rgb/.sgi format, however most common image formats can be handled. Some good tools for producing SGI formatted files are the San Diego Supercomputer Center's Imtools. Imtools can be ftp'd in binary and source formats from SDSC's ftp server. XV also can read and write SGI format files.
  • Resolution - Full PAL resolution is 768x576. You may use smaller images, they can be centered in this framesize. Bear in mind that the amount of a the full PAL image seen on monitors varies and that you should leave approximately 10-15% blank borders. (For advanced users an even better format would be to give us the images with dimensions 720x576 and an aspect ratio of 1.0666 - ie 768x576 scaled to 720x576 eg
    imscale im.sgi -xsize 720 -ysize 576 newim.sgi
    where im.sgi is 768x576. This is the format of serial digital video CCIR-601).
  • Naming of images - the preferred format is name0000.sgi name0001.sgi ..., but name.1.sgi name.2.sgi ..., is reasonable (though doesn't list alphabetically in correct order which complicates things for some tools).
  • Where to put them - once you have your images ready to record contact the vizlab programmer to find out where the images can be transferred to recording. A recording typically takes a few hours or more, usually done overnight.
  • How to get the video onto VHS tape - once the raw video has been recorded to disk you will usually need to have a session in the vizlab with the programmer to make a small program (titles, repeats etc) and record it onto VHS tape (frequent and/or experienced users are taught how to do this for themselves).
  • Aesthetic issues to think about when making animation for video.
  • No single or double pixel lines or dots - video interlace will cause them to flicker badly.
  • Leave 10-15% border around the images main content to avoid clipping.

  • Make text and annotation twice as large as you think it needs to be, based on workstation monitor previewing. It still may not be large enough.
  • Avoid superfluous and distracting annotation. Consider moving static annotation to a title frame.
  • Avoid highly saturated colors especially bright reds. (i.e. don't use an RGB value (255 0 0). This is particularly important if you wish to make NTSC (the US video format) copies of your video.
  • For high quality results you will probably need to look into spatial antialising your graphics (commonly known as removing the jaggies). One simple method is to generate your graphics at twice the PAL resolution then shrink them down using a simple averaging filter (many image toolkits have this capability).
  • Be sparing with color, use the minimum necessary to highlight the features you're showing.
  • Use changes in brightness for 3d effect, not color.
  • Email problems, suggestions, questions to